Title: Florida Entomologist
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098813/00036
 Material Information
Title: Florida Entomologist
Physical Description: Serial
Creator: Florida Entomological Society
Publisher: Florida Entomological Society
Place of Publication: Winter Haven, Fla.
Publication Date: 2000
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: VID00036
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: Open Access

Full Text

Lloyd: On Quantifying Mate Search in a Perfect Insect 211


Department of Entomology and Nematology, University of Florida, Gainesville, FL 32611


Male Photinus collustrans LeConte fireflies fly over their grassland habitats flash-
ing and seeking their flightless females. I followed individual males, measured, and
took note of various aspects of their behavior. Then, from a sample of 255 male runs,
with a total distance of 13.9 miles and 10,306 flashes, various sets of these males,
those seemingly directed by other than search flight-plans, were removed to leave a
sample to characterize "pure" search flight. Fireflies are good subjects for students to
study foraging ecology and sexual selection, and from studies of common grassland
fireflies it will be clear to students that even simple behavior by males of a single spe-
cies, under seemingly uncomplicated and homogeneous conditions, can be complex,
but provide opportunity for theoretical and empirical exploration. Among factors
identified here as influencing male mate-seeking behavior were ambient tempera-
ture, ambient light level, and time of night. Other influencing factors, enigmas, and
student explorations are indicated.

Key Words: Lampyridae, Photinus, mate search, sexual selection, foraging, teaching


Las luciernagas machos de la especie Photinus collustrans LeConte vuelan sobre
los pastizales destellando su luz y buscando a las hembras que no pueden volar. Seguif
a los machos, los media y tome notas de various aspects de su comportamiento. Luego,
de una muestra de 255 vuelos de los machos, con una distancia total de 13,9 millas y
de 10.306 destellos, various grupos de estos machos, esos dirigidos aparentemente por
alguna otra raz6n que la de un vuelo de busqueda, fueron removidos para former una
muestra que caracterice el vuelo de busqueda "puro". Las luci6rnagas son buenos ele-
mentos de studio para los estudiantes de ecologia del forraje y selecci6n sexual. Del
studio de las luci6rnagas comunes de los pastizales quedara claro para los estudian-
tes que, incluso el comportamiento simple de los machos de una sola especie, bajo con-
diciones aparentemente sencillas y homog6neas, puede ser complejo, pero proporciona
la oportunidad para la exploraci6n te6rica y empirica. Entre los factors identificados
aquf que influyen el comportamiento de busqueda de la hembra, estan la tempera-
tura, el nivel de luz ambiental y la hora de la noche. Otros factors influyentes, enig-
mas y exploraciones de los estudiantes son incluidos.

In this symposium series I have passed along notes on the natural history of fireflies
I have met in the field while exploring their species-level taxonomy, in the form of writ-
ten lectures (Letters) to an introductory biology and natural history class. On the face
of it, this Letter is an attempt to quantify the mate search of flashing males over a pas-
ture, and apply this information toward understanding mate competition. In actuality
it reveals the biological complexity of this seemingly simple behavior, and finds a num-

Florida Entomologist 83(3) September, 2000

her of difficulties that students can appreciate and avoid in their turn. The study was
stimulated by papers in a symposium moderated by Dan Otte in the mid 1970s, which
surely introduced many naturalist/systematists to a different view of insect signal-
ing and associated behavior. The symposium was a timely event in my life with fireflies,
for it offered new perspectives that fit in naturally with what I had learned from read-
ing papers of pioneer fireflyers F. A. McDermott and H. S. Barber, and discussions with
systematists T. H. Hubbell and W. L. Brown, and especially, R. D. Alexander.
A pictorial moment of firefly search flight and its variation among species is seen
in the illustration orthopterist Otte created as Frontispiece for a Photinus behavioral-
taxonomy paper (Fig. 1; note his surreptitious acridid), a graphic used not long ago as
cover illustration for a mathematical (and for some naturalists an abstruse and mys-
tical) treatise entitled "Quantitative Analysis of Movements" (Turchin 1998). This
book has a number of useful considerations even for such an elementary study as re-
ported here, and views individuals to develop models for understanding populations.
In "the present study" individual fireflies were watched closely to learn something of
their (adaptive) programs for mate search flight. The two views, individual and pop-
ulation, overlap, then merge when mate-seeking fireflies leave or enter local popula-
tions and these "migrations" influence population vigor, independence, and fate.
Other relevant books and teachers that students may wish to consult are: "Foraging
Theory" (Stephens and Krebs 1986), "Spatial Ecology" (Tilman and Kareiva 1997),
and "The Ecological Detective" (Hilborn and Mangel 1997).
My leaves-of-grass-top project began with the naive notion that it would be easy to
collect some few data on males of each of several low-flying species, and make a ready
comparison of species that have different flash (signal) patterns and fly in somewhat
different ecological situations. Innocent at inception, results were reminder that raw
nature is not as it is often condensed for textbook generalizations, theoretical model-
ing, and taxonomy, a discovery students will make when they try to quantify mate-
seeking behavior. It was great fun to follow individual fireflies through a twinkling of
their nocturnal careers, and to see them in another dimension for the first time, and
so closely that I saw individual-but nevertheless yet nameless-males quit search-
ing and in climbing flight enter the boughs of the scattered pines in the study site. I
even saw some that crashed into herbs and shrubs fall to the ground, lights burning,
marking meteor-like descents. Decades ago when spinning tackle was introduced into
the U.S., an expert noted that spin-fishing was a soothing, meditative experience, to
be compared with making thread at a spinning wheel (though I wondered how he
knew this). Chasing fireflies across a meadow while pushing a measuring wheel is
also good for contemplation-and data just roll in.
It is of course obvious to almost anyone, today, that there is variation among the
phenotypes of individuals even within local populations, due to a number of "innate,"
experiential, and of-the-moment inputs to each firefly's central processing system. Af-
ter miles in pursuit of males of a near-perfect species, I was reminded that simple
variables in method, equipment, and assistants can also mess-up tidy results. In the
end, the chase provided previously unknown and eye-opening facts and basic statis-
tics about firefly search behavior, and it suggested interesting, short-term studies that
students can do in night-time labs (Appendix). Best of all, it incidentally dramatized
an important source of selection pressure that "must" often have led to divergence of
local populations, toward and, maybe, even to ... speciation.
In summer any teacher in the range of a grassland firefly, such as the widely dis-
tributed and very common "All American Firefly" Photinus pyralis (L.) (Fig. 1: flash-
path #8), will find that with stopwatches and foresters' measuring wheels students
can acquire a new understanding of sexual reproduction in the animal kingdom, and
discover that males often share a lonesome misery. They will learn to focus on details

Lloyd: On Quantifying Mate Search in a Perfect Insect 213

Fig. 1. Flashes and flight paths of males of several different Photinus species as
they would appear in a time-lapse photograph. Arrowheads indicate direction of
flight. The species illustrated are not all sympatric; number 4, Photinus collustrans,
with its low arcing flight is the species of focus here. (1) P. consimilis complex (slow
pulse sp.), (2) P. brimleyi Green, (3) P. consimilis complex (fast pulse sp.) and P. car-
olinus Green, (4) P. collustrans, (5) P. marginellus LeConte, (6) P. consanguineus Le-
Conte, (7), P. ignitus Fall, (8) P. pyralis (L.), and (9) P. granulatus Fall.

of behavior and quantify elements against an array of variables and distractions, and
discover sources of variation that they themselves introduce into their data. In my
own use, this Letter serves as an introduction to a field problem for a "biology with
fireflies" class-a companion guide has instructions for observing and quantifying
firefly search behavior, analyzing data, dealing with ecological variables, and devel-

Florida Entomologist 83(3) September, 2000

oping experiments for new-found questions-it provides a challenge for engineering
majors to conceive of a system of aiming devices, servos, and a portable computer to
quantify the 3-dimensional search paths of arboreal, crown-cruising fireflies.
The Internet (electronic) publication of this paper has additional figures as Info-
Link attachments to illustrate the text. They are cited here by their number as ILR
figures and their legends are included in the End Notes section. These copyrighted il-
lustrations may be used freely with the citation: J. Lloyd, Univ. of Florida.


In Search Of "The Pure-Seeker" Male-Following Sex-Driven Fireflies
Through Pastures For the Mind (Lampyridae: Photinus collustrans)

... many a requisite we see must be fulfilled in living things
ere they avail to propagate their kind ..."
(Lucretius, 95-52 B.P.E.)

"You can observe a lot just by watching."
(Yogi Berra, 1925-P.E.)

Dear Fireflyers, Luminescent fireflies are good subjects for the study and quantifica-
tion of mate-seeking behavior. I realized this somewhat belatedly when attending a sci-
entific gathering on sexual selection and mate competition in insects. It came to me, with
some embarrassment, that all the while I had been viewing firefly flashes exclusively
from the standpoint of a biological-species-seeking taxonomist, merely as signals that
were involved in mate recognition and reproductive "isolation," they were intimately
connected with mate competition among conspecific males. Suddenly it was potently
clear, the competitive mating context had probably been a major force of selection pres-
sures on signaling behavior, and if I were to understand firefly signals and their evolu-
tion, and use them taxonomically, I would need to get some sense of how sexual selection
might have influenced them. With information on several species, comparisons might
even give clues to adaptations in signaling behavior that are tuned to specific features of
the mating arena. Could there be a general theory of flash patterns, perhaps even a de-
scriptive and predictive equation that a mathematical modeler could formulate?
Some of our commonest species in North America occur in grassland habitats
where mating flights of males occur in a thin volume of air, sometimes an almost 2-
dimensional space, low over the ground, and individuals can be followed and watched
closely with relative ease. It was not difficult to select a prime subject that occurs in
north central Florida that has characteristics that would especially lend themselves
to such study. Photinus collustrans LeConte is a twilight firefly of fields, pastures,
lawns, and savannas (Figs. 1: flash-path #4, and 2; ILR 2000, Figs. 1 and 2). The male
flash pattern is a single, short, yellow flash that is emitted about each 2 seconds of
flight, with flash rate varying predictably with ambient temperature (Figs. 3 and 4).
Male flight paths are diagnostic, for during each flash males typically arc to the right
or left (ILR 2000, Figs. 3 and 4). These arcs may enable them to see the species-typical
flash-then-glow responses they have just elicited from females, and perhaps also to
scan more broadly for female responses that nearby, rival males have stimulated.
Female P. collustrans are flightless and found on the ground or very low on grass
stems near their burrow's entrance (ILR 2000, Fig. 5); their sedentary, virtually
sessile "search" for passing airborne mates, and other aspects of their biology has been
studied in intimate detail by Steve Wing). They respond to male flashes with a single

Lloyd: On Quantifying Mate Search in a Perfect Insect 215

Fig. 2. Habitus of Photinus tanytoxus Lloyd, a sibling species of P. collustrans, dif-
fering conspicuously in external appearance only in the black rather than pale elytral
sutural bead in the apical third or so of its length. Length, ca. 7 mm. Searching flight
in this species will make an interesting comparison with that of P. collustrans, be-
cause it begins about 5 minutes after P. collustrans ends, at full darkness, and contin-
ues for one-half hour or more. This photo-like illustration is actually a carbon-dust
drawing by Laura Line.

Florida Entomologist 83(3) September, 2000

5.0 -


S. .

2.0 *

16 18 20 22 24 26 28 30
Temperature (C)

Fig. 3. Flash period duration (in sec) as a function of ambient temperature (in Co).
A chart such as this can be useful for species identification in the field. However, flash
rate (= 1/period) is better for comparing species because its regression is approxi-
mately a straight line, and can be plotted with fewer data points, see Figure 4.

flash that begins about a half second after the male flash, and ends in a briefly persis-
tent, tapering glow. Males of each local, conterminous population (patch, deme) are
active for about 20 crepuscle (twilight) minutes each evening, though each individual
may only fly during a part of this already narrow window (note circles in Fig. 6). Such
flight "compaction" in time is great for fireflyers because it makes each night's flight
window a discrete and comfortably-managed sampling unit, reducing the number of
environmental, behavioral, and human variables that must be considered. Also, P. col-
lustrans' flight straddles one especially significant ecological event, the rapid de-
crease of ambient light known as twilight ("tween" light).
Is it possible to characterize, to describe the ideal, the optimum mate-seeking
flight for male P. collustrans? That is, can we sample the behavior of males, and then
describe a "pure" search flight that will have evolved because it is the most appropri-
ate and efficient (competitive) for males under the ecological conditions they live
with? Not likely, and the fallacy behind this notion is that there actually could be a
most efficient or most appropriate search flight. Males, their genetic constitutions and
the individual circumstances they experience through their lives are not identical.
And, on any given day some are older and have less search time remaining before they
die-males may be programmed to take more risks as they age, say, risks in where,
when, and how fast they fly; also, sometimes males experience high levels of compe-
tition, with rivals always within "eyeshot"-males may be programmed, upon detec-
tion of such conditions, to employ different, more concealing flight and flash tactics,
and such flight would certainly reduce their measured efficiency as "ideal searchers."

Lloyd: On Quantifying Mate Search in a Perfect Insect 217


t con ed f *r to 128i4 N u N L last
t s Lo T T T tree
0.3- a 368 T *B 3T B both
B NNa B P N none
0.1- -- --- i --
16 18 20 22 24 26 28 30
Temperature (C)

Fig. 4. Flash rate (1/period) as a function of ambient temperature (in C); data
points were converted from those in Figure 4. A "hurry up" flash period regression for
field use can be approximated by converting mean flash periods from two remote tem-
peratures to rates, drawing/extending a line through their plots, converting several
points along this straight line to period, and replotting. See text for an explanation of
letter and number dot tags.

Also, we intuitively understand that what might well be the best flight program for
the discovery of females in an ideal grassland could also be more dangerous when oc-
casional tall spikes of wing-tearing thistles are present, and an unfortunate flight ac-
cident could, with some calculable level of statistical probability, reduce the total
number of days a male has to search. Thus, we expect "imperfections" in search adap-
tation to arise from various trade-offs, and for other reasons, too. Nevertheless, we
might determine the overall search-flight characteristics of"our average male" across
the (mostly unspecified and little understood) range of conditions that prevail during
the evenings that data are recorded. Parameters of such a search plan, when visual-
ized as axes of XY charts will frame a cloud of "slightly-off" as well as more deviant
searches clustered'round a mean. These are the (dots of) males with different pheno-
typic characteristics and whose control systems are processing different inputs. Per-
haps we can find clues for some of the observed differences among males, and
speculative notions that can be developed into formal hypotheses for testing by care-
fully focused research. (How could one construct, and then place in a meadow, 100
fake thistle spikes, and quantify male responses and losses to them?)

On Firefly Trails
To collect data I followed individual P. collustrans males as they flew and flashed
over a pasture grassland, pushing a measuring wheel through each subject's curves
and turns-I trailed a few meters behind them and never saw any indication that my

Florida Entomologist 83(3) September, 2000

S 1 g o 0 0 .

0.9 ,

.* 0.8 .


0.5 :

0.4 .. *

o 0.3 '* *

*- 0.2

| 0 .* 6

0 4 8 12 16 20 24
Minutes After Mate Search Began (=F4)

Fig. 5. Height of flight flashes of 74 male P collustrans, given as the proportion of
flashes (Y axis) that were emitted above one meter altitude. The ratio (above-l-meter/
all altitudes) for each male is marked. The X axis shows the number of minutes after
evening flight began, "marked" when four males had begun flashing flight (F4). This
convention was adopted as an attempt to avoid biasing by idiosyncratic males, such as
those embarking on their evening flight from especially shady places. Other conven-
tions are possible, perhaps even better? Samples were made during 38 evenings.

delayed presence influenced their flight. A 2-channel event (blood-cell) counter
mounted on the handlebar was operated by thumb and stored certain "digitized" data
(ILR 2000, Fig. 6). For example, I had noticed that males fly at somewhat different al-
titudes over the ground, and altitude seemed to be connected with the time of evening,
so in each of 74 individual followings (i.e., runs; during 38 evenings) my thumb kept
track of the flashes emitted above and below 1 meter altitude. I found that during the
first few minutes each evening males generally flew lower than 1 meter, but gradually
through their twilight window, ever more of their flashes were emitted from above 1
meter (Figs. 5 and 6). That this change is in response to diminishing ambient light is
suggested by the observation that in early evening when males fly into shady spaces
beneath trees they fly conspicuously higher, and then fly lower again as they move out
under open sky. The scatter seen in the plot of individual records (Fig. 5) occurs in part
because males flew in or through the shade of trees.

Lloyd: On Quantifying Mate Search in a Perfect Insect 219

0.9- * * * 0.9 :

0.8 .4 0.8

0.7 .."' 0.7
10.6- 0 6

0.5- ..0 0.5

e 0.4- .'* 0.4 s

S 0.3 .. "- 0.3

2 0.2- 0.2

0.1 .. b 0.1

0 0' , 1 0
0 4 8 12 16 20 24
Minutes After Mate Search Began (=F4)

Fig. 6. Mean height ratio of males for each minute (left Y axis). With increasing
darkness males tend to fly higher, possibly to avoid unseen spikes of herbs, or to see fur-
ther, or to spread their light further?; can these notions be demonstrated, or be differ-
entiated experimentally? The figure also shows (open circles, right Y axis), the
proportion of followed males that quit search during consecutive brackets of time, be-
ginning 10 minutes after evening flight began (n = 138, 24 males quitting). Lines were
fitted by the graphing program, linear in dots of 5 and 6, and exponential in circles of 6.

Among the categories of data recorded for each run were: duration in seconds, from
the first flash to the terminating one; number of flash periods per run, with the starting
flash counted as zero and the count at the run's terminating flash giving the number of
flash periods; distance flown, as read from an automatic counter clicking feet at the rim
of the wheel; and time of day at the end of a run. I also made written notes of unusual
flight features and the reason for ending the run; for examples, runs were terminated
because males flew to a perch apparently ending their evening flight (Fig. 6, circles),
because followed males intersected paths with another and I was uncertain which male
I had been following, and because the run was sufficiently long (>30 flashes).
First and simpler questions to be answered by the wheel-pushing data are, what
are the flash-rates and flight speeds of mate-seeking males? Next, do these parame-
ters vary through the activity window (period)? Then, when interactions of these pa-
rameters are viewed, are there any generalizations or predictions to be made, any
puzzlements or surprises? The total data set from the pasture is from 255 males, fol-
lowed on 43 evenings, during the years 1976-7, for a distance of 22,356 meters (13.9
miles); this required 5.25 hours of actual following, during which time the males emit-
ted 10,306 flashes. The longest run was 0.49 miles, and the fastest, 4.8 mph. I should

Florida Entomologist 83(3) September, 2000

note that of all of my firefly field studies only this one seemed to demand professional,
knee-length snake boots; but they were not called into their intended service.
On nine occasions males flew into vegetation and fell with lights burning to the
ground, each losing a few seconds of search time (= 1 crash per 1.54 miles); males
found two females and coupled with them (= 1 per 157.5 min, per 7.0 miles!). Statis-
tically speaking, if each male should actually search 20 minutes per evening, on aver-
age, only one in eight of them would find a female? More realistically, but still
statistically, if individual males average only 16 minutes of search per evening, only
one in 10 will find a female! (How many males find none in their entire adult lives (7+
days?), and how many find two or three-and, beyond some degree of good fortune, do
exceptionally successful (super) males share an uncommon set of mate-finding fea-
tures? How could you determine this, and how big a sample would it require?)
Flash rate. The period of consecutive flashes (period = time between beginning of
one event to the beginning of the next such event) emitted by flying searching males
decreases with increasing ambient temperature, and ranges between 3 seconds at 160
and 1.5 seconds at 29C (Fig. 3). There is a curvilinear relationship between flash pe-
riod and temperature, but between flash rate (1/period) and temperature the relation
can be considered linear for our purposes (Fig. 4). Observe the slower, individually la-
beled runs in Figure 4. When field note (written) descriptions of these slower runs
were examined it was obvious that several of these males may have been in an "activ-
ity-terminating mode" rather than a "mate-seeking mode," because they flew up into
the scattered pines or down to the ground, apparently quitting flight for the evening.
To characterize "pure-seeking" behavior, runs that represent alternative or mixed be-
havior modes, such as this set would appear to represent, must be culled from the
data set. (But, note that the data set is shown in its entirety before any purging of
"bad" data; Figs. 3 and 4; n = 255 runs).
Selecting pure seekers. To get statistics for presumptive "pure-seeking" flight-pa-
rameters, I culled runs when field notes suggested the males might be following
mixed or other flight plan directives. Thus, I removed: all flight terminating runs
(when followed males flew up into tree boughs or down into the grass and stopped
flashing, "T" in Fig. 4); all last-of-the-evening runs (such males were always the last
one or two that could be seen flying and flashing in the population, "L" in Fig. 4-the
rationale being that they may already have been slowing down); runs during which
males behaved "erratically" (flying up then down, conspicuously slowing and then
speeding up); runs when the males crashed into vegetation (males can be seen glow-
ing brightly as they fall down on the approached-side of herbs and bushes); and runs
when males noticeably paused in flight, as they do occasionally over small sandy
patches and pale flowers. A few very short runs were also removed, those comprising
fewer than 8 flash patterns.
As examples of purged sets, note labeled run #368 (Fig. 4), in which the male
paused in flight and hovered over one place for 8-9 sec; and in run 128, the male ran
into a twig, fell down and lost about 2 seconds on the ground. The male of run 552 rap-
idly approached a 50-meter wooded sinkhole, momentarily stopped flashing at the
wall of foliage at the edge, then abruptly flew back over my head. His headlong dash-
ing flight makes me suspect that his measured parameters were not those of even an
extreme "pure" search flight plan. Finally, note that there are several runs at the slow
edge of the clustering marked "N", meaning that I made no relevant verbal notes at
the end of the run. I must interpret these as representing the slow tail of a "normal"
distribution of a pureseeker flash rate, and retain them in the hopeful set. In every
case, all of the males of a given rejected category were culled, and not merely those
that were conspicuous outliers on the chart. Figure 7 shows the flash-rate/tempera-
ture regression of the now much reduced, remaining sample of 123 runs. The correla-

Lloyd: On Quantifying Mate Search in a Perfect Insect


0.7 *

C* 0.6 so U"


0.3 -

0.2 --. -. -. .
16 18 20 22 24 26 28 30
Temperature (C)

Fig. 7. "Purified" flash rate regression, with all males in categories mentioned in
text removed (see outliers and key in Figure 4), as a function of ambient temperature
(in C); compare with Figure 4.

tion coefficient (r) for the complete data set in Figure 4 is 0.69, and for the selected set
in Figure 7 it is 0.80, though the slope remains about the same.
Flash-rate revisited. The slope of the "purified" data in Figure 7, can be used to ad-
just ("correct") the flash-rates of the selected runs to what they would be at 25C. The
temperature-adjusted rates can then be plotted as a function of the time-of-night (i.e.,
sun-time of run midpoint) they were measured (Fig. 8). This regression slope (of flash-
rate/midpoint), reveals no time-of-night effect, and nearly zero correlation (i.e., t =
1.25, slope not significantly different from flat; r = 0.12, on a scale of 0 to 1). In other
words, twilight time with its ever increasing darkness, and other changes such as a
(probable?) reduction in the availability of unmated females and perhaps even the
number/proximity of rival males, does not appear to influence the rate of male flash
pattern repetition).
Flight-speed of this "purified" set of mate-seekers increases slightly with temper-
ature (Fig. 9; t = 2.975), significantly different from flat, though the correlation is
pretty weak (r = 0.26). Then, when speed data are adjusted to 25C and plotted as a
function of the time of night, flight-speed is seen to increase through the activity pe-
riod (Fig. 10: t = 5.285; r = 0.43). But, to the naked eye and common sense, the chart's
dots do not really lie along a straight-line. A linear regression was imposed on the
data by the graphing program, but the data points clearly increase linearly to about
1.3 creps, remain flat to about 1.5 creps, and then decline slightly.
Before addressing this, I want to whack some more outliers. Note the several
faster runs especially after 1.6 creps (Fig. 10); might they indicate the presence of yet
another search tactic or influence? Field notes reveal that the data for four of them
were taken by a young family member (unpaid assistant) who had used a measuring

Florida Entomologist 83(3) September, 2000



o U *

C 0 *
40 0 *e .
** *
S * 0 0*

0.5- *

0.4 *

0.8 1.0 1.2 1.4 1.6 1.8 2.0
Mid-Point (Creps)

Fig. 8. Flash rate, adjusted to 25C, as a function of time of night (in crep units,
where 0 is sunset; a crep unit, which varies with date and locality, is the duration of
Civil Twilight).

wheel with a smaller diameter (ILR 2000, Fig. 6-one that would probably roll more
tightly over the surfaces of smaller humps and bumps, going further per linear dis-
tance and unit time. These data points thus suffered from two additional variables;
consequently, all data collected by this apprentice fireflyer were purged. One outlier
was a last run that escaped notice during the first culling, and another was a penul-
timate run that ended three minutes before the end of evening activity. These two
were also removed from the data set, though purging the last mentioned point was not
legitimate, not statistically proper-unless I were to remove all penultimate runs
(and replot, beginning with Figure 3). The data set is now reduced another 7 percent
(n = 114, down 55 percent from the original 255!). This refinement hopefully, theoret-
ically in a loose sense of the word, should reveal the best glimpse of search behavior
yet achieved, and show speed adjustments made by males through their twilight win-
dow about as well as I can now sketch them (Fig. 11). A verbal summary of what the
graph suggests would be: flight speed increases at first, while twilight darkness deep-
ens rapidly (causation?), and then after 1.5 creps there appears to be a slight down-
turn, indicating that males that remain active fly (flew) a bit more slowly.
Distance-flown per flash-period (flash rate melded with flight speed), decreases with
temperature (Fig. 12; t = 6.630, r = 0.52). This is to say, the cooler it is, the greater the dis-
tance along the flight track, out in front of the male, the flash is "expected/required(?)" to
cover ("to stimulate?" Tom asked, illuminatingly). Though flight speed appears to de-
crease slightly as temperature decreases (Fig. 9), flash rate decreases considerably (i.e.,
period lengthens, Fig. 7), and this results in a longer distance flown between flashes. This

Lloyd: On Quantifying Mate Search in a Perfect Insect 223

1.6- *,



0.2-8 -- -*
16 18 20 22 24 26 28 30
Temperature (C)

Fig. 9. Flight speed (in meters/sec) as a function of ambient temperature (in C).

would seem to indicate that with decreasing temperature, flashes "are required to" stim-
ulate progressively larger areas. There are too many unknowns to interpret this: we do
not know and cannot easily determine whether the intensity (photons emitted) of flashes
remains constant at lower temperatures, or, whether the angular spread of the light
leaving the lantern remains constant-some luminescent click beetles apparently con-
trol the spread of the light beams emitted by their ventral lanterns!
Predator Pressure and Deme Divergence. Females of many, seemingly most Nearc-
tic species of the genus Photuris prey upon mate seeking males of other species. They
perch in the activity spaces of their prey and mimic the flash responses of the males'
own females, attract and eat them. No such deceivers were seen at the pasture site,
and this makes for a very interesting contrast with data from another site. The latter
was a narrow, roadside, pine savanna with a nearby mesic hammock and marshy
catch-basin and drainage ditch. These habitats produce the predators Photuris versi-
color (Florida form) and Photuris bethaniensis (Florida form). Females of these two
species occurred in the P. collustrans site and flashed responses to P collustrans
males I was following. Thirteen such predators were observed to answer them, and
though some males landed and lost search time, none were caught by these femmes
fatales. What makes these statistics especially noteworthy, was that the savanna
sample of measured runs was a great deal smaller than that from the pasture-40
males were followed for a total of only 3,550 meters (2.2 miles). The score card: pas-
ture, 2 mates and no predators; savanna, no mates and 13 predators! Converting this
to a numerical comparison: if, during the next instant such a predator had been found
at the pasture (I need one predator to avoid the unworkable zero), mimicking the re-

Florida Entomologist 83(3) September, 2000


2.0 -

. 1.8 6

S1.6 -.
1 4 3

S1.2- .

1.0- ** t
S0.8- *



0.2 ---1
0.8 1.0 1.2 1.4 1.6 1.8 2.0
Mid-Point (Creps)

Fig. 10. Flight speed (in meters/sec, adjusted to 25C), as a function of time of night
(in crep units), with a forced linear regression line that even to the eye is not a good
fit; note the rise then fall of data points.

sponses of P. collustrans females, the hunter-exposure rate in the savanna would have
been nearly 69 times that of the pasture site (3.10 vs. 0.045 predators/kilometer). In
other words, in this predator connection, this savanna was "immeasurably", virtually
"infinitely" more dangerous than the pasture for male P. collustrans. This really
brings home the fact that local populations (demes) can be under grossly different se-
lective pressures, pressures that must be expected to have a significant influence
upon mate seeking and code-recognition behavior of males-and this is a phenome-
non that can be measured and compared!
Other Conclusions and Explanations. At the outset of this study, it seemed that for
a number of reasons mentioned in the introduction, it might be fairly simple to de-
scribe the mate-seeking flight of P. collustrans, and identify characteristics that made
it well-tuned, "optimal" for finding mates in a competitive environment. Each of the
sets of male runs that was distinguished and then removed at some point in the anal-
ysis, including quitting, crashing, and last-of-the-night runs, and runs whose mea-
surements may have had technical inconsistencies, not only produced a tidier sample,
but more importantly identified sources of influence and variation for future refer-
ence. Some of the culled sets revealed something about P. collustrans' biology, whether
tentative explanations were correct or not. But, it now is clear that to characterize the
simple search flight of P. collustrans for comparison with that of other species will re-
quire a great deal more intraspecific comparative study. Studies must first be compa-
rable at the technical and deme levels, and find and identify the influences of: (1) the
mechanical aspects of sampling techniques, including different wheel sizes, different
wheel operators, and different bump and hummock sizes (with consideration given to

Lloyd: On Quantifying Mate Search in a Perfect Insect


2.0 O


i 1.6 .

U 1.4

1.0- V-
0.8 /

0.4 y = -1.6497 + 3.9693x-1.3142x2 Ra = 0.378

0.2 --,
0.8 1.0 1.2 1.4 1.6 1.8 2.0
Mid-Point (Creps)

Fig. 11. Flight speed (in meters/sec, adjusted to 25C), as a function of time of night
(in crep units), with a regression line and equation for a second-order polynomial plot-
ted and calculated by the graphing program.

varying flight speed) on over-the-ground measuring accuracy and precision; (2) differ-
ent ecological conditions, including vegetation, predator and rival-density, ambient
light, and season; and (3) (presumptive) population genetic differences-for example,
by comparing contiguous pairs of demes, pairs of more remote demes such as those in
different drainage systems, and pairs of demes from extreme ends of the species' geo-
graphic range. The fact that females of this species are flightless and burrowing
means that any gene flow that occurs must be either through male flight, female or
larval dislocation/dispersion through flood waters, or flight by as yet unknown mac-
ropterous females, which I have unsuccessfully sought for many years. The fact that
sex determination in fireflies does not involve a Y (male) chromosome may mean that
we do not have a simple male versus female genetic marker to use to differentiate the
mechanics of gene flow among local populations.
The proper study and characterization of but one aspect of the biology of but one
presumptively simple, ground level firefly species could take a person a lifetime; or
the efforts of an insect behavioral ecology course for generations of students. It is for-
tunate that wheel pushing is a rewarding and a relaxing thing to do in itself, espe-
cially in high-stress times.


I thank Cindy Weldon Lasley, for making the hundreds of data sortings and ma-
nipulations that were necessary for this analysis, and Steven Lasley, Department
computer specialist, who gave us considerable instruction and wrote essential analy-

Florida Entomologist 83(3)

September, 2000

16 18 20 22 24 26 28 30
Temperature (C)

Fig. 12. Meters flown per flash period as a function of ambient temperature (in Co).
But what of actual area illuminated in front of the male by each flash? See text.

sis programs. The late Harry A. Lloyd made modifications on the wheel to make it
more suitable for the special application required for this study; Robert S. Lloyd fol-
lowed and measured some males. I also thank Tim Forrest, John Sivinsky, Steve
Wing, M. I. Montenegro, and Jade Williams for comments on the manuscript at vari-
ous stages of its development. Flora MacColl and the late Barbara Hollien provided
considerable technical expertise and assistance. Jenny Gavilanez-Sloan for translat-
ing the abstract. Florida Agric. Exp. Station Journal Series No. R-07406.
Measuring wheels were Rolatape Models; Model 623 was the larger and used for
all but the few exceptional runs mentioned in the text. The event counter was a Clay
Adams channel. For analysis, the midpoint time of each run was calculated by halv-
ing run duration and subtracting this quotient from the day-time of the run's end.
This time, expressed in minutes after sunset, was then converted to Crep Units by di-
viding by the duration of Civil Twilight for that date and place (Nielsen 1961). Sunset
times and C.T. durations were determined from a computer program, "Sunset", writ-
ten by J. P. Oliver with modifications by T. Forrest. Early and partial summaries of
certain data have previously been reported (Lloyd 1979).
The following enumerated statements are figure legends for color illustrations
(photographic slides) that appear as InfoLink attachments to this article in the elec-
tronic publication of this issue of the Florida Entomologist, and are cited in text here
as ILR 2000, Fig.#: 1. Copulating pair of Photinus tanytoxus Lloyd, a sibling species
of P. collustrans, which except for the dark coloration of the elytral bead is morpholog-
ically indistinguishable. The female was perched just off the ground near her burrow.
2. A copulating pair of P. tanytoxus on the ground, with attentions from a second (top)







b 2.2


Lloyd: On Quantifying Mate Search in a Perfect Insect 227

male. Wing (1984) made several interesting discoveries concerning this situation, and
the competitive and defensive tactics of males. 3. Twilight on the lawn of a lakeside
house near Gainesville, Florida, with several P. collustrans males searching for fe-
males. Note their arcing, slowing flight while flashing, and the thin slice of space they
use over the ground. 4. A firefly student observing the flashing flight of a single male
P. collustrans, along the grassy roadside on the west side of Newnans Lake, Gaines-
ville. 5. A female P. collustrans, showing her short elytra and large and thinly cuticled
abdomen, with her burrow's entrance at the tip of her abdomen. 6. Large and small
measuring wheels. The wheel at the right has the "event counter" mounted on the
handle bar; the smaller one was used by an assistant and seems to have been respon-
sible for somewhat different values. Another wheel model has a solid rather than
spoked wheel and would work better in brushy areas.


ADAMS, G. 1981. Search paths of fireflies in two dimensions. Florida Entomologist
HILBORN, R. AND M. MANGEL. 1997. The ecological detective. Princeton Univ. Press,
NJ. 315 pp.
LLOYD, J. E. 1966. Studies on the flash communication system in Photinus fireflies.
Univ. Michigan Museum of Zoology Misc. Publ. 130:1-93.
LLOYD, J. E. 1979. Sexual selection in luminescent beetles. Pp. 293-342 in M. S. and
N. A. Blum (eds.), Sexual selection and reproductive competition in insects. Ac-
ademic Press, NY. 463 pp.
LLOYD, J. E. 1980. Photuris fireflies mimic signals of their females' prey. Science
LLOYD, J. E. 1984. Occurrence of aggressive mimicry in fireflies. Florida Entomologist
LLOYD, J. E. 1990. Firefly semiosystematics and predation: A history. Florida Ento-
mologist 73:51-66.
LLOYD, J. E., AND S. R. WING. 1983. Nocturnal aerial predation of fireflies by light-
seeking fireflies. Science 222:634-635.
NIELSEN, E. T. 1963. Illumination at twilight. Oikos 14:9-21.
OTTE, D. AND SMILEY, J. 1977. Synchrony in Texas fireflies with a consideration of
male interaction models. Biol. of Behavior. 2:143-158.
STEPHENS, D. W. AND J. R. KREBS. 1986. Foraging Theory. Princeton Univ. Press. NJ.
247 pp.
TILMAN, D. AND P. KAREIVA [ed]. 1997. Spatial ecology. Princeton Univ. Press. NJ. 368 pp.
TURCHIN, P. 1998. Quantitative Analysis of Movements. Sinauer, Sunderland MA. 396
WALKER, T. J. 1975. Effects of temperature on rates on poikilotherm nervous systems:
Evidence from the calling songs of meadow katydids (Orthoptera: Tettigoni-
idae: Orchelimum) and re analysis of published data. Journ. Comparative
Physiol. 101:57-69.
WING, S. R. 1984. Female monogamy and male competition in Photinus collustrans
(Coleoptera: Lampyridae). Psyche 91:153-160.
WING, S. R. 1988. Cost of mating for female insects: Risk of predation in Photinus col-
lustrans (Coleoptera: Lampyridae). American Nat. 131:139-142.
WING, S. R. 1989. Energetic costs of mating in a flightless female firefly, Photinus col-
lustrans (Coleoptera: Lampyridae). Journ. Insect Behav. 2:841-847.

Florida Entomologist 83(3) September, 2000


Field studies for consideration.

1. How long should a sample run be (number of flashes) to adequately sample the be-
havior of an individual male?

2. How much agreement is there between consecutive samples (e.g., 30 flashes) of an
individual's search behavior-i.e., between a 30-flash sample and the next consec-
utive 30-flash sample?

3. Do older males take more risks in their flight? Do they fly longer each evening,
earlier or later, or faster and lower? Is it possible to sample an individual's search
behavior when he is young and then again when he is elderly?

4. Is flight altitude smoothly proportional to ambient light level, and what (physio-
logical) mechanism of altitude determination is involved?

5. From what distance are males able to competitively approach female flashes
given in response to a rival male's flashes? (see Otte and Smiley 1975?)

6. Is the direction of the arcing flash-trajectory of males influenced by the presence
of nearby males or other aspects of the natural environment, such as large or dark
herbs, shrubs, or the brighter (western) horizon?

7. Are local activity patches demee sites) as delineated by observing the flights of
males, constant in space occupied, or do "hot spots" change from moment to mo-
ment, night to night, or with the number of males active in the site?

8. Can patch-entering and -leaving males be detected by their flashes, or do they fly
flashless, and can be found only with flight-interception traps (e.g., window-pane

9. Do crashing males always light up, or do the data reported here need to be ad-
justed (calibrated)?

10. Do run-terminating males as observed in this study here actually end flight for
the evening or do they return to aerial search?

11. When males fly into a space just previously searched by another male in their pre-
sumptive vision-space do they adjust their direction of flight?

12. What is the adaptive significance (function) of the higher flight altitude observed
in later-flying males? That is, when it is darker will they see and broadcast fur-
ther? Or, is it that they cannot see to avoid tall plants and higher flight is safer?
Which of these, either or both, is the actual explanation, and can these two be
tested and separated simultaneously?

13. What is the adaptive significance of the (apparently) smaller area coverage of
flashes at higher temperatures? For example, at higher temperatures might
neighboring males react to female response flashes and interlope faster, so (the
apparent) smaller flash coverage is a defensive tactic?

Vulinec: Dung Beetles, Monkeys, and Conservation


Entomology and Nematology Department, University of Florida
Gainesville, FL 32611

1Current Address: Department of Zoology, University of Florida
Gainesville, FL 32611


Dung beetles are important in several ecological processes, including nutrient re-
cycling, soil aeration, the transport of other organisms, and the burial of vertebrate
dispersed and defecated seeds. Dung beetle species vary widely in their abilities as
seed dispersers. The biomass of beetle species that bury no seeds, bury small seeds
only, or bury small and large seeds, is significantly different among sites along the
Amazon River. The abundance of monkeys that act as high quality seed dispersers
also varies at different sites. Implications of these differences in seed dispersal dy-
namics are discussed. Recent clearing and disturbance of primary forest is having an
effect on the populations of primary and secondary seed dispersers, and suggestions
for conservation of these critical faunas are presented.

Key Words: dung beetles, Scarabaeidae, monkeys, primates, seed dispersal, Amazon,


Os cole6pteros copr6fagos sao importantes em varios processes ecol6gicos, in-
cluindo reciclagem de nutrients, aeragao do solo, transport de outros organismos e
enterramento de sementes dispersadas e defecadas por vertebrados. As esp6cias de
besouros copr6fagos variam imensamente quanto a sua habilidade em dispersar se-
mentes. A biomassa das esp6cies de besouros que nao enterram sementes, a biomassa
daqueles que enterram somente sementes pequenas, e a biomassa daqueles que ente-
rram tanto sementes grandes como pequenas 6 significativamente diferente entire va-
rios locals ao long do Rio Amazonas. A abundancia de primatas que agem como bons
dispersores de sementes tamb6m varia grandemente entire diferentes locals. As impli-
cao6es dessas diferecas na dinamica de dispersaa de sementes sao discutidas neste
trabalho. As recentes derrubadas de avores e as perturbao6es de dispersores primario
e secundario, e sugest6es para a conservagao da fauna sao apresentadas.

Dung beetles, an omnipresent component of tropical biotas, perform important eco-
system functions. Several characteristics make this group of beetles ecologically signif-
icant (Hanski & Cambefort 1991; Halffter & Matthews 1966). They are particularly
vulnerable to deforestation and other changes in habitat and fauna, and this sensitivity
makes them useful as indicators of ecosystem health (Halffter et al. 1992; Klein 1989).
As they are primarily associated with mammals, they are also indicators of mammalian
abundance and possibly diversity. Nevertheless, dung beetles' functions in ecological
systems go far beyond the status of an indicator. They contribute services including, re-
cycling of nutrients, aerating the soil, serving as transport for predatory mites, and

Florida Entomologist 83(3) September, 2000

burying of seeds in dung (Hanski & Cambefort 1991; Halffter & Matthews 1966; Es-
trada et al. 1991). Because of these roles, the decline in dung beetle abundance and di-
versity may have cascading effects on the environment. In this paper, I discuss the
results of some of my studies of deforestation, rainforest disturbance, and dung beetle
abundance and diversity. I also present my research on the seed burying characteristics
of different species of dung beetles, and how these differences may effect seed dispersal
in tropical rainforests. Finally, I will discuss rainforest survival associated with seed
dispersal and some suggestions for accelerating recovery from disturbed habitat.
Dung beetles are generally grouped into dwellers, which live and nest within a
dung pat, burrowers, and rollers (Hanski & Cambefort 1991; Halffter & Edmonds
1982; Halffter & Matthews 1966). Dung beetles of the burrower guild dig nests under
dung pats, and those of the roller guild roll the dung away from the site of deposition
to be buried later. Members of both of these guilds pack these nests with dung, usually
formed into balls, and lay eggs in the balls. The larvae develop and pupate within the
brood balls, and emerge as adults (Halffter & Edmonds 1982; Halffter & Matthews
1966). Additionally, most species of rollers and burrowers also make feeding balls that
are buried, may be abandoned uneaten, or if eaten, contaminants (from the beetles
point of view), such as seeds, are often moved underground (Vulinec, 1999). These be-
haviors hasten the decomposition of waste, aid in nutrient recycling, and contribute
to aeration of the soil (Halffter & Matthews 1966). They also reduce pests in dung,
such as dung breeding flies (Bornemissza 1970; Fincher 1981).
Additionally, dung beetles have an important role in seed dispersal. Seeds swal-
lowed by frugivorous mammals are often defecated intact and viable (Garber 1986; de
Figueiredo 1993). Nevertheless, as much as 90% of seeds defecated onto the surface of
the soil may be destroyed by rodents or other seed predators unless buried; this burial
is accomplished almost entirely by dung beetles (Estrada & Coates-Estrada 1991).
However, dung beetles are especially susceptible to ecosystem change (Mor6n
1987; Klein 1989; Halffter et al. 1992; Estrada et al. 1998). Disturbance of tropical
rainforest may affect dung beetles directly by altering temperature, humidity, or soil
characteristics, or indirectly by reduction in mammal faunas. Disturbance may have
a number of different outcomes. Intensive land use in tropical areas, such as bulldoz-
ing, may result in degradation that will never be reversed (Buschbacher et al. 1992).
On the other hand, less disturbed areas may have the potential to reforest. This re-
bound may be very important in developing management practices for revolving agri-
culture in tropical forests (Fearnside 1993). Studies done in Brazil and Costa Rica
suggest that the most important factor in reforesting land is getting the seeds to the
sites (Young et al. 1987; Pannell 1989; Nepstead et al. 1991; Holl 1999). Primary seed
dispersers such as birds, bats, and monkeys are instrumental in this process (Chap-
man & Chapman 1995. As rodent density can be very high in secondary growth (Chap-
man & Chapman 1999), secondary dispersal and burial by dung beetles may also be
an essential element in reforestation. Habitat characteristics that encourage use by
these dispersers will improve an area's chance of regenerating natural rainforest.
To determine the effect of land disturbance on populations of dung beetles in rainfor-
ests, I investigated beetle abundance and diversity in three habitats at three sites in the
states of Rond6nia, Amazonas, and Para, Brazil; more specifically, I censused beetle pop-
ulations at the three sites, and in three habitats, primary terra firma forest, secondary
growth, and clear-cuts. I also examined the most common beetles in these habitats for
seed burial capabilities. I asked: What beetles are best at seed burial? Are there limits
to the type of seeds that are buried by different species of beetles? What communities of
beetles occur at each site? Are there differences in the communities of monkeys? What
are the implications of community differences for forest regeneration? Finally, what ef-
fect will development of the Brazilian rainforest have on seed dispersal dynamics?

Vulinec: Dung Beetles, Monkeys, and Conservation



I examined dung beetle and monkey communities in three widely separated sites on
tributaries of the Amazon River in Brazil. The land near Caucalandia in the state of
Rond6nia, Brazil, is mostly cattle ranches, secondary growth, and some patches of pri-
mary rainforest. Mean annual rainfall is 2300 mm, and mean annual temperature is
27C (Landowner, pers. comm.). The state of Rondonia was undeveloped until the late
1960s, when the government brought in people from the overcrowded cities in the north-
east (Page 1995). This area currently has one of the highest rates of deforestation in the
world. With clearing, erosion has increased, and soil quality has decreased; laterization
of the soil is common. By 1991, over 37,000 km2 were cleared, more than 16% of the total
state; a good quarter of the state is probably cleared by now (Stone et al. 1991).
My research site here was a small section, 250 ha of primary forest, bordered by
overgrown banana and cacau plantations, and some areas of capoiera, natural re-
growth. Hunting is discouraged on the private land; however, poachers were encoun-
tered in the primary forest area. Even so, peccaries, agoutis, jaguarundi, deer, and
sloths, were often seen. Additionally, monkeys were easily observed. The work at this
site was done between October 1996 and March 1997.
The city of Manaus is located in the state of Amazonas at the confluence of the Rio
Negro and Rio Solimoes (Amazon). Rainfall in the area averages 2200 mm per year, and
average temperature is 27C (Salati 1985). The biological station ReservaAdolfo Ducke
(10,000 ha) is 25 km northeast of the city, and is surrounded by land that is rapidly be-
ing developed. Poaching is common, and most of the large mammals have been extir-
pated. Agoutis, which are seed predators, and coati were common, but no tapir or
peccaries have been seen for possibly more than 15 years. Monkeys are the most com-
mon large mammal, but are also hunted regularly within the reserve by people in the
surrounding communities. As these communities often have a higher standard of living
than many parts of the Amazon, this hunting pressure is particularly unfortunate. Two
dead howlers that had been shot and escaped to die later were found during my time
at this site. I worked at Reserva Ducke between November 1997 and December 1998.
In Caxiuana, Para, is a recently created biological station (established 1990), the
Ferreira Penna Scientific Research Station. It is located off the Rio Pard at the south-
west side of the Ilha do Maraj6. This station includes 33,000 ha, 80% of which is terra
firma forest, and 20% is blackwater floodplain forest (Lisboa 1997). Annual rainfall is
around 3000 mm (Lisboa 1997), however, October is the driest month, the month of
my collection, and the year of my work, 1998, El Nino increased dryness and fire po-
tential all over the Amazon. Surrounding communities of the reserve are sparsely
populated, and hunting pressure is low. Monkeys, especially howlers and the silvery
marmosets, have been habituated to humans by previous research. The upland area
at this site is surrounded by swampy seasonally flooded habitat.

Seed Burial Experiments

The number of dung beetle species was high at some sites (> 50); I tested the most
common species that are large enough to bury at least 5 cc of dung. Beetles not tested
in burial experiments were assigned the same categories as similar sized congeners.
To determine what proportion of seeds imbedded in dung particular species of
dung beetles bury, I used natural seeds or plastic beads that were small (<5 mm) or
large (>10 mm). I placed beetles in mesh-covered buckets (40 cm diameter x 36 cm
depth) filled with 150 mm sandy soil, and placed 50 cc of fresh cow dung with embed-

Florida Entomologist 83(3) September, 2000

ded seeds in the center of each bucket. Buckets were left up for 72 hrs. The number of
individuals of each species varied depending on the size of the beetles, such that bio-
mass of beetles remained approximately equal (Vulinec 1999). Sample sizes varied de-
pending on the number of beetles captured alive.
For smaller species (e.g., Sybalocanthon sp.), I placed beetles in 1-liter containers
with 100 mm sandy soil, and10 cc fresh cow dung containing only small seeds. After
72 hrs., I excavated the burrows, 10 mm at a time. I recorded seeds buried and depth
of burial.



Using baited pitfall traps set along the same route as the monkey transect, I cen-
sused beetle community composition and species abundance in three habitat types at
each site: primary rain forest, secondary growth (edges of primary forest, old planta-
tions, and previously cleared areas), and clear-cuts. Beetle traps were 1-liter plastic
cups buried to the rim in soil, and topped with a 3 cm aperture funnel. A 50 ml cup
baited with human dung was suspended by wire above the trap (Howden & Nealis
1975; Gill 1991). The volume of dung in the traps attracted even the largest beetles
(Peck & Howden 1984). I set a total of 27 traps each census, which were left up for 24
hrs, then collected. Three traps were set at 20 m intervals at 9 different trapping sta-
tions (3 sites in primary forest, 3 in secondary growth, and 3 in clear-cut). This trap-
ping regime yielded 1242 trap-days (sensu Klein 1989) for the three locations. Each
site was surveyed for nearly the same number of trap-days.
Beetle biomass was measured as dry weight; the beetles were dried at room tem-
perature for 1 week prior to weighing. Ten individuals of selected species, those most
common, were weighed on an Ohaus balance at the USDA-ARS laboratory in Gaines-
ville, Florida to determine mean biomass. For beetle species that were not weighed,
biomass was assumed to be similar to beetles of the same size. Extrapolation was nec-
essary only with small, less common beetles.
Beetle biomass at all sites was summed over time, and the differences between pri-
mary, secondary, and clear-cut habitats were tested with a one-way ANOVA. I used
Scheffes test for multiple comparisons, and Students t-test to compare primary and
secondary growth at Caxiuana (beetle biomass in the clear-cut at this site totaled only
0.006 grams, and was excluded from the analysis).
Beetles were categorized as those that buried almost no seeds, those that buried at
least 20% of small seeds, and those that buried at least 20% of both large and small
seeds. These groups were then compared for their proportion of total dung beetle bio-
mass at each site using one-way ANOVA.


To categorize primate primary seed dispersers, I censused monkey populations at
the three sites: 36 times in Rond6nia and Ducke, and 27 times in Caxiuana. Monkey
transects were stratified by the proportion of secondary and primary growth at each
site. Transects were walked by 2 people, and transect pace was about 1 km/hr. Monkey
species and number of individuals were recorded.
I used methods described in Chapman et al. (2000) to calculate monkey density,
using a 50% cut-off rule to select the sighting distance. Observer to animal distance,
as opposed to perpendicular distance was used, as quantitative comparisons suggest

Vulinec: Dung Beetles, Monkeys, and Conservation

perpendicular distance underestimates transect width for forest primates (Chapman
et al. 1988). Densities of groups were then calculated as the number of groups sighted
within the truncated sighting distance divided by the area sampled, that is, the
length of the transect times the truncated distance (Chapman et al. 2000). Densities
were calculated for groups per square kilometer and number of individuals per square
The proportion of monkeys that are seed predators as opposed to seed dispersers
was evaluated using a Chi-square contingency table test for differences among the
three sites.


Seed Burial

Some beetles, for example all species of Euysternus, bury almost no seeds. The
quality of other species as seed buriers depends on several factors, including seed bur-
ier size. Larger beetles bury more seeds than smaller beetles simply because they
bury more dung. Additionally, large species are capable of burying larger seeds (Table
1). Burrower guild beetles bury small seeds fairly well, while rollers are less capable.
Seeds over 10 mm are only buried by the largest burrowers (Figs. 1 and 2).

Biomass of Seed-Burying Beetles In Different Sites

Beetle biomass differed among sites (ANOVA, F = 7.58, P < 0.03); Rond6nia had
nearly four times higher biomass than Reserva Ducke. The mean biomass in primary
forest was significantly higher than secondary forest at Rond6nia and Caxiuana, but
not at Reserva Ducke (Table 2). Biomass at all sites was highly significant between
both primary and secondary forest and clear-cuts (only two beetles were collected in
the clear-cut at Caxiuana).
Biomass was significantly different among the three seed burial categories at
Reserva Ducke (ANOVA, F = 4.97, P < 0.02), and Rond6nia (ANOVA, F = 5.79, P <
0.005), but was not significant at Caxiuana. Biomass of small seed buriers was low at
Reserva Ducke, while beetles that buried large and small seeds dominated the bio-
mass (Fig. 3). At Caxiuana, beetles of all size categories were somewhat equal in bio-
mass, and at Rond6nia, beetles that buried large and small seeds comprised the
largest amount of biomass, but other categories were also abundant.

Proportion of Seed Predating Monkeys at Each Site

Monkey densities varied significantly among the three sites (ANOVA, F = 4.5, P <
0.01). Total densities for all species at each site were 38.55 individuals per square kilo-
meter for Caxiuana, 8.94 for Reserva Ducke, and 28.42 for Rond6nia. Differences in
the proportions of seed predators versus seed dispersers in monkey abundance were
significant among the sites (Z2: Caxiuana vs. Ducke = 44.7, p < 0.001; Caxiuana vs.
Rond6nia = 10.4, P < 0.005; Ducke vs. Rond6nia = 101.0, P < 0.001; Fig. 4).


In rainforests, seedlings have a higher survival rate when not directly under the
parent tree (Howe et al. 1985; DeSteven and Putz 1984; Augspurger 1983). This sur-
vival is thought to occur because the seedling escapes from predators and pathogens

Florida Entomologist 83(3) September, 2000


Type of dung Activity Mean biomass
Species Abb. manipulation period gms (SE)

Canthon aequinoctialis
Canthon triangularis
Coprophanaeus lancifer
Dichotomius nr. batesi
Dichotomius boreus
Dichotomius lucasi (Har.)
Dichotomius podalirius
Eurysternus caribaeus
Oxysternon conspicilla-
tum (Web.)
Phanaeus cambeforti
Phanaeus chalcomelas
Scybalocanthon nr.

Ca Roller

Ct Roller

Nocturnal 0.130 (0.045)


0.038 (0.010)

Cl Burrower Crepuscular 3.260 (1.021)

Dbat Burrower Diurnal

0.073 (0.007)

Dbor Burrower Nocturnal 0.452 (0.161)


Burrower Nocturnal 0.099 (0.020)

Dpod Burrower Nocturnal 0.452 (0.161)

Ec Neither Nocturnal 0.102 (0.028)

Oc Burrower Diurnal

Pcam Burrower Diurnal

Pch Burrower Diurnal

Sp Roller


0.798 (0.244)

0.136 (0.042)

0.146 (0.042)

0.027 (0.005)

of the parent (Howe & Smallwood 1982; Connell 1971; Janzen 1970), and because dis-
persal reduces competition for resources (Stiles 1989).
Trees have evolved strategies to disperse their offspring to other locations. Fruit
and the seeds within generally travel away from their parent inside animal guts.
Monkeys are considered by many authors to be high quality seed dispersers (Castro
1991; Estrada et al. 1991; Rowell & Mitchell 1991; Chapman 1989; Howe 1989; Gar-
ber 1986). However, some monkeys are better seed dispersers than others, for exam-
ple, some eat leaves or nectar in addition to fruit, some travel more, some defecate
seeds out singly, or may defecate them in a mass (Veracini 1996; Zhang & Wang 1995;
Castro 1991; Chapman 1989; Garber 1986). Most importantly, some monkeys, prima-
rily the sakis, are generally seed predators (Norconk 1998; Peres 1993; Van Roose-
malen et al. 1988). The proportion of seed predators in monkey communities varies
greatly at each locality. These findings may reflect a very different potential in differ-
ent areas for forest regeneration.
Secondary seed dispersal and burial by dung beetles may also have an impact on
rainforest regeneration (Feer 1999; Andresen 1998; Estrada et al. 1991). The highly
significant difference between beetle biomass at Rond6nia and the other two sites
might have a number of causes. It could reflect the shorter history of disturbance, the
abundance of mammals, climatic factors, or vicariance. Nevertheless, the overall bio-

Vulinec: Dung Beetles, Monkeys, and Conservation 22

Ec Ct Ca Sp

Dbat Dluc Dbor Dpod Oc Pcam Pch

Beetle Species

Fig. 1. The total proportion of small seeds (< 5 mm) left on the soil surface by some
beetle species. See Table 1 for species abbreviations and characteristics.










Ec Ct Ca Sp Dbat Dluc Dbor Dpod Oc Peam Pch CI
Beetle Species

Fig. 2. The total proportion of large seeds (> 10 mm) left on the soil surface by some
beetle species. See Table 1 for species abbreviations and characteristics.


Florida Entomologist 83(3) September, 2000


Site Primary forest Secondary growth Clear-cut

Caxiuana 5.07 (2.65)' 2.49 (1.66)b *
Ducke 9.85 (1.71)' 7.03 (1.15)' 0.41 (0.11)b
Rond6nia 36.55 (4.16)' 19.78 (3.10)b 0.58 (1.10)'

mass of beetles and the proportion of those that bury seeds may be highly important
to the local ecology. Reserva Ducke is dominated by beetles that bury both large and
small seeds. The proportion of the biomass that buries only small seeds is very low.
Because beetles that bury large seeds are mostly nocturnal (Vulinec 1999), the parti-
tioning of beetle biomass may have an effect on seeds that are deposited during the
day. Seeds deposited by monkeys in transit may not be buried immediately, and could
be left on the ground until discovered by rodents, ants, or weevils. It would be ex-
pected that at this site, seeds may suffer higher mortality from seed predation, first
by saki monkeys, and then by seed predators on the forest floor.
In regeneration of secondary growth areas, dung beetles may prove even more im-
portant than in primary forest. Primary seed dispersers, such as birds and monkeys
will enter secondary growth, especially if there are trees such as bananas and Cecro-
pia spp. (Holl 1999). Therefore, seeds from primary forest can get to these areas (Dun-
can & Chapman 1999). However, rodents are often more prevelant in secondary


1000 -

0 Burial of Small and Large Seeds
600 | Burial of Small Seeds Only
1l No Burial


Caxiuana Ducke Rondonia


Fig. 3. Total beetle biomass at each site stratified by three categories of seed burial

Vulinec: Dung Beetles, Monkeys, and Conservation

i 40--


14U Seed Predators
i --g.____5___ is_ ENon-predators

Caxiuana Ducke Rondonia

Fig. 4. Density of monkeys at each site and proportion of seed predators versus
seed dispersers.

growth, so predation may also be greater (Asquith 1997). It would be advantageous to
a seed to be buried quickly in these areas. Dung beetles are far more abundant in sec-
ondary growth than clear-cuts (Vulinec 1999; Estrada et al. 1998). Additionally, sec-
ondary growth has more similar microclimates and soil characteristics to primary
forest than pastureland, where soil compaction and low humidity may keep many
dung beetles out. Secondary growth has the potential to reforest relatively quickly
given the primary and secondary dispersers that enter it. Clear-cuts have much less
chance of rapid reforestation, and in some cases, no chance.
Beetles that roll dung, rather than burying it directly under the dung pat, domi-
nate clear-cuts (Vulinec 1999; Halffter et al. 1992). Rollers, despite relocating seeds
from the source, often do not bury brood balls deeply enough to protect the seeds
within from rodent predation. Additionally, the generally smaller size of these beetles
would make them less effective as processors of larger seeds, or a quantity of seeds.
This factor suggests that savanna adapted species would probably not contribute sig-
nificantly to forest regeneration in secondary growth (Vulinec 1999).
Beetles that frequent secondary forest are, like those in primary forest, more likely
to be good seed "planters", due to beetle size, depth of burrowing activity, and abun-
dance (Vulinec 1999). Nevertheless, in two of the three sites, beetle biomass was sig-
nificantly greater in primary growth than secondary growth. Many beetle species are
very sensitive to changes in habitat, while other species, such as Dichotomius, are
more flexible in habitat preferences (Vulinec 1999). The factors that encourage high-
quality seed buriers to enter secondary forest should be investigated. Manipulation of
these factors could increase regeneration of primary forest from secondary growth ar-
eas. For example, beetles may respond to humidity when foraging or mate seeking.
Planting broad leafed trees (such as Cercropia) in abandoned cut areas may increase
ground level humidity and recruit more large dung beetles (and monkeys) from pri-
mary forest to these areas, resulting in quicker primary forest regeneration.

Florida Entomologist 83(3) September, 2000

Amazon rainforest is disappearing at a distressing rate; in the Brazilian Amazon
2,554,000 ha of forest are cleared yearly, a number that does not include clearing due
to selective logging or destruction by fires (FAO 1999). Deforestation in the Neotropics
has profound effects on carbon cycling, the hydrological cycle, and soil and water qual-
ity (Salati 1985). For example, 50-75% of precipitation in the Amazon is returned to
the atmosphere in the form of water vapor through evaporation of the water retained
by leaves, and through transpiration of the plants (Salati 1985). Cutting the forest
can radically change the water cycle.
The majority of the clear-cutting in the Amazon is for cattle ranching, and the
greatest clearing is by large agribusiness corporations. According to Fearnside (1993),
beef productivity on Amazon soils is low, and is unsustainable. Available phosphorous
limits grass yields on oxisols and ultisols. Inedible weeds that are more adapted to the
poor soils quickly invade pastureland. Massive government programs that subsidize
pasture have claimed that pasture improves the soil, and is indefinitely sustainable.
However, further studies have shown that maintained pasture productivity is not pos-
sible without the addition of phosphorous fertilizer. But adding fertilizer will still not
solve the problem of pasture degradation. Soils become compacted through the expo-
sure to sun and the trampling of cattle. My own measurements of soil density in forest
and clear-cuts showed a 2 to 4 fold increase in relative density in clear-cuts, even with-
out cattle (Vulinec, unpub.).
In a recent study, Holl (1999) suggests that although regeneration of forest in pas-
tures is limited by colonization, establishment, growth, and survival, the major lim-
iting factor in seedling establishment is lack of seed availability. She suggests pasture
restoration to forest by planting native tree seedlings to increase canopy architecture,
installing bird perching structures, and planting rapidly growing shrubs that quickly
produce fruit and attract seed dispersers. I also suggest that there must be a nearby
refuge of protected species to provide the necessary colonizers, including plants, pol-
linators, and dispersers, and secondary dispersers such as dung beetles.
Logging is another large-scale use of tropical forests. Currently, decisions about
logging in the Amazon have no central planning or coordination. Without this plan-
ning, logging industries could easily log all of the forest in the state of Pard (Verissimo
et al. 1998). Verissimo et al. (1998) conclude that logging could be a sustainable indus-
try that would preserve diversity and indigenous rights and suggest several ways to
do this. Logging would have to be highly monitored, however. Nepsted et al. (1998b)
maintain that even selective logging diminishes forest cover, allows the drying of un-
derstory vegetation, and sets the stage for devastating fires as were seen in 1998.
Even one fire in an area significantly increases the chances of future fires. The more
fires in an area, the more brushy secondary growth and open canopies invite more
fires in a positive feedback loop (Nepstad et al. 1998a).
Perhaps most importantly, loss of tropical rainforest leads to loss of biodiversity.
There may be as many as 5-30 million species of plants and animals still undescribed,
and unknown, the vast majority in tropical forests (Erwin 1982). We have yet to un-
derstand the interrelationships among the flora and fauna living in these critical ar-
eas. What we don't understand may do more than just prevent us from finding a cure
for cancer. The loss of pollinators and dispersers will affect uncounted species of
plants, many of potential economic and ecological importance. My research shows
how even seemingly insignificant organisms may have important roles in ecosystem
function. If dung beetles were gone, the buildup of feces, the increase in dung breeding
pathogens, the loss of some soil turnover, would become rapidly apparent (Vulinec
1999; Klein 1989). Predatory mites that hitch rides on dung beetles would disappear.
Potential antifungal and antibacterial chemicals may never be discovered. And very
importantly, seeds deposited in dung would remain on the surface vulnerable to ro-

Vulinec: Dung Beetles, Monkeys, and Conservation

dents, fungi, and granivorous insects. Conversely, if vertebrate seed dispersers disap-
peared, the 60% of tree species, and almost 100% of understory plants that depend on
vertebrate seed dispersal would also decline. Dung beetles would vanish with them.


I thank the following organizations for their support of my research: The Fulbright
Commission for International Studies, The Charles A. and Anne Morrow Lindbergh
Foundation, the Florida Center for Systematic Entomology, the Dickinson Award for
Tropical Agriculture at the University of Florida, The University of Florida's Women
in Agriculture Club, and the J. D. Turner Foundation.
In Brazil, I am grateful to 0 Conselho Nacional de Desenvolvimento Cientifico e
Tecnol6gico (CNPq), Instituto National de Pesquisas da Amaz6nia (INPA), and Museu
Paraense Emilio Goeldi (MPEG). I am especially thankful to Claudio R. V. da Fonseca
of INPA and Pedro L. B. Lisboa of the Museu Paranese Emilio Goeldi.
Special thanks are due David Almquist and Freida Ansoanuur for sorting insects,
Bruce Gill, Dave Edmonds, and Fernando Vaz de Mello for help with identifications. I
am indebted to Colin Chapman and John Sivinski for comments on this manuscript.
I also thank Coleman and Corey Kane for their assistance in Brazil and the States,
and Dave Mellow for everything.


ANDRESEN, E. 1999. Seed dispersal by monkeys and the fate of dispersed seeds in a
Peruvian rain forest. Biotropica 31: 145-158.
ASQUITH, N. M., S. J. WRIGHT, AND M. CLASS. 1997. Does mammal community com-
position control recruitment in neotropical forests? Evidence from Panama.
Ecology: 941-946.
AUGSPERGER, C. K. 1983. Seed dispersal of a tropical tree, Platypodium elegans, and
the escape of its seedlings from fungal pathogens. J. Ecol. 71: 759-771.
BORNEMISSZA, G. F. 1970. Insectary studies on the control of dung breeding flies by
the activity of the dung beetle, Onthophagus gazella F. (Coleoptera: Scara-
baeinae). J. Aust. Entomol. Soc. 9: 31-41.
BUSCHBACHER, R., C. UHL, AND E. A. S. SERRAO. 1992. Reforestation of degraded Am-
azon pasture lands. Pp. 257-274 in Ecosystem rehabilitation Vol. 2: ecosystem
analysis and synthesis. M. K. Wali (ed.). SPB Academic Publishing, The Hague,
CASTRO, N. R. 1991. Behavioral ecology of two coexisting tamarin species (Saguinus
fusicollis nigrifrons and Saguinus mystax mystax) in Amazonian Peru. Ph. D.
Washington University.
CHAPMAN, C. A. 1989. Primate seed dispersal: the fate of dispersed seeds. Biotropica
21: 148-154.
CHAPMAN, C. A., AND L. J. CHAPMAN. 1999. Forest restoration in abandoned agricul-
tural land: a case study from east Africa. Conserv. Biol. 13: 1301-1311.
CHAPMAN, C. A., AND L. J. CHAPMAN. 1995. Survival without dispersal?: Seedling re-
cruitment under parents. Conserv. Biol. 9:675-678.
SAKER. 2000. Long-term effects of logging on African Primate Communities: a 28
year comparison from Kibale National Park, Uganda. Conserv. Bio. 14: 207-217.
CHAPMAN, C. A., L. M. FEDIGAN, AND L. FEDIGAN. 1988. A comparison of transect
methods of estimating population densities of Costa Rican primates. Brenesia
30: 67-80.
CONNELL, J. H. 1971. On the role of natural enemies in preventing competitive exclu-
sion in some marine animals and in rain forest trees. Pp. 298-312 in Dynamics
of populations. P. J. Den Boer and G. Gradwell (eds.). PUDOC, Wageningen.

Florida Entomologist 83(3) September, 2000

DE FIGUEIREDO, R. A. 1993. Ingestion of Ficus enormous seeds by howler monkeys (Al-
ouatta fusca) in Brazil: effects on seed germination. J. Trop. Ecol. 9: 541-543.
DESTEVEN, D., AND F. E. PUTZ. 1984. Impact of mammals on early recruitment of a
tropical canopy tree, Dipteryx panamensis, in Panama. Oikos 43: 207-216.
DUNCAN, R. S., AND C. A. CHAPMAN. 1999. Seed dispersal and potential forest succes-
sion in abandoned agriculture in tropical Africa. Ecol. Applic. 9: 998-1008.
ERWIN, T. 1982. Tropical forests: their richness in Coleoptera and other arthropod spe-
cies. Coleop. Bull. 36: 74-75.
ESTRADA, A., A. ANZURES D., AND R. COATES-ESTRADA. 1999. Tropical rain forest frag-
mentation, howler monkeys (Alouatta palliata), and dung beetles at Los Tuxt-
las, Mexico. Am. J. Primat. 48: 253-262.
ESTRADA, A., AND R. COATES-ESTRADA. 1991. Howler monkeys (Alouatta palliata),
dung beetles (Scarabaeidae) and seed dispersal: ecological interactions in the
tropical rain forest of Los Tuxtlas, Mexico. J. Trop. Ecol. 7: 459-474.
Dung and carrion beetles in tropical rain forest fragments and agricultural
habitats at Los Tuxtlas, Mexico. J. Trop. Ecol. 14: 577-593.
FAO. 1999. State of world's forests. Food and Agricultural Organization of the United
FEARNSIDE, P. M. 1993. Deforestation in the Brazilian Amazonia: The effect of popu-
lation and land tenure. Ambio. 22: 537-545.
FEER, F. 1999. Effects of dung beetles (Scarabaeidae) on seeds dispersed by howler
monkeys (Alouatta seniculus) in the French Guianan rain forest. J. Trop. Ecol.
15: 129-142.
FINCHER, G. T. 1981. The potential value of dung beetles in pasture ecosystems. J. Ga.
Entomol. Soc. 16: 316-333.
GARBER, P. A. 1986. The ecology of seed dispersal in two species of Callitrichid pri-
mates (Sanguinus mystax and Sanguinus fuscicollis). Amer. J. Primatol. 10:
GILL, B. D. 1991. Dung beetles in tropical American forests. Pp. 261-229 in Dung Beetle
Ecology. I. Hanski and Y. Cambefort (eds.). Princeton U. Press, Princeton, N.J.
HALFFTER, G., AND W. D. EDMONDS. 1982. The nesting behaviour of dung beetles
(Scarabaeinae): an ecological and evolutive approach. Publicaciones del Insti-
tuto de Ecologia, Mexico City, 175 pp.
HALFFTER, G., M. E. FAVILA, AND V. HALFFTER. 1992. A comparative study of the
structure of the scarab guild in Mexican tropical rain forests and derived eco-
systems. Folia Entomol. Mex. 84: 131-156.
HALFFTER, G., AND E. G. MATTHEWS. 1966. The natural history of dung beetles of the
subfamily Scarabaeinae (Col.: Scarabaeidae). Folia Entomol. Mex. 12-14: 1-312.
HANSKI, I., AND Y. CAMBEFORT. 1991. Dung beetle ecology. Princeton University
Press, Princeton. 481 pp.
HOLL, K. D. 1999. Factors limiting tropical rain forest regeneration in abandoned pas-
ture: seed rain, seed germination, microclimate, and soil. Biotropica 31: 229-
HOWE, H. F. 1989. Scatter- and clump-dispersal and seedling demography: hypothesis
and implications. Oecologia 79: 417-426.
HOWE, H. F., E. W. SCHUPP, L. C. WESTLEY. 1985. Early consequences of seed dis-
persal for a Neotropical tree (Virola surinamensis). Ecology 66: 781-791.
HOWDEN, H. F., AND V. G. NEALIS. 1975. Effects of clearing in a tropical rain forest on
the composition of the coprophagus scarab beetle fauna (Coleoptera). Biotro-
pica 7: 77-83.
HOWE, H. F., AND J. SMALLWOOD. 1982. Ecology of seed dispersal. Ann. Rev. Ecol. Sys-
tem. 13: 201-228.
JANZEN, D. H. 1970. Herbivores and the number of tree species in tropical forests. Am.
Nat. 104: 501-528.
KLEIN, B. C. 1989. Effects of forest fragmentation on dung and carrion beetle commu-
nities in Central Amazonia. Ecology 70: 1715-1725.

Vulinec: Dung Beetles, Monkeys, and Conservation

LISBOA, P. L. B. 1997. Estacao Cientifica Ferreira Penna. Pp. 23-52 in Caxiuana. P. L.
B. Lisboa (ed.). Museu Paraense Emilio Goeldi.
MORON, M. A. 1987. The necrophagus Scarabaeinae beetles (Coleoptera: Scara-
baeidae) from a coffee plantation in Chiapas, Mexico: habits and phenology. Co-
leop. Bull. 41: 225-232.
NEPSTAD, D C., A. A. ALENCAR, AND A. G. MOREIRA. 1998a. Flames in the rain forest:
origins, impacts and alternatives to Amazonian fires. PP-G7 Publication Se-
ries, Brasilia, Brazil.
1998b. Forest fire prediction and prevention in the Brazilian Amazon. Conserv.
Biol. 12: 951-953.
NEPSTAD, D. C., C. UHL, AND E. A. S. SERRAO. 1991. Recuperation of a degraded Am-
azonian landscape: Forest recovery and agricultural restoration. Ambio: 20:
Digesta retention time in white-faced sakis, Pithecia pithecia. Am. J. Primatol.
PAGE, J. A. 1995. The Brazilians. Addison-Wesley, Reading, Massachusetts. 540 pp.
PANNELL, C. M. 1989. The role of animals in natural regeneration and the manage-
ment of equatorial rainforests for conservation and timber production. Com-
monw. For. Rev. 68: 309-313.
PECK, S. B. AND H. F. HOWDEN. 1984. Response of a dung beetles guild to different
sizes of dung bait in a Panamanian rainforest. Biotropica 16: 235-238.
PEREZ, C. A. 1993. Notes on the ecology of buffy saki monkeys (Pithecia albicans, Gray
1860): a canopy seed-predator. Am. J. of Primatol. 31: 129-140.
ROWELL, T. E., AND B. J. MITCHELL. 1991. Comparison of seed dispersal by guenons
in Kenya and capuchins in Panama. J. Trop. Ecol. 7: 269-274.
SALATI, E. 1985. The climatology and hydrology of Amazonia. Pp. 18-48 in Key envi-
ronments: Amazonia. G. H. Prance and T. E. Lovejoy (eds.). Pergamon Press,
STILES, E. W. 1989. Fruits, seeds and dispersal agents. Pp. 87-122 in Plant-animal in-
teractions. W. G. Abrahamson (ed.). McGraw-Hill, New York.
STONE, T. A., I. F. BROWN, AND G. M. WOODWELL. 1991. Estimation, by remote sens-
ing, of deforestation in central Rond6nia, Brazil. For. Ecol. Manage. 38: 291-
northern bearded saki (Chirpotes satanas chirpotes): a neotropical seed preda-
tor. Am. J. Primatol. 14: 11-35.
VERACINI, C. 1996. Ecologia e comportamento sobre Callithrix argentata e Saguinus
midas niger, duas esp6cies simpatricas de Callitrichinae, Primates. In
Relat6rio de Actividades, 1995, ECFP, Pedro Luiz Braga Lisboa (ed.).
VERISSIMO, A., C. S. JUNIO, S. STONE, AND C. UHL. 1998. Zoning of timber extraction
in the Brazilian Amazon. Conserv. Biol. 12: 128-136.
VULINEC, K. 1999. Dung beetles, monkeys and seed dispersal on the Brazilian Ama-
zon. Ph.D. Dissertation, University of Florida.
YOUNG, K. R., J. J. EWEL, AND B. J. BROWN. 1987. Seed dynamics during forest suc-
cession in Costa Rica. Vegetatio 71: 157-173.
ZHANG, S. Y., AND L. K. WANG. Fruit consumption and seed dispersal ofZiziphus cin-
namomum (Rhamnaceae) by two sympatric primates (Cebus apella andAteles
paniscus) in French Guiana. Biotropica 27: 397-401.

Florida Entomologist 83(3) September, 2000


1Department of Biology, University of North Carolina at Asheville
Asheville, NC 28804

2Forestry Sciences Laboratory, Rocky Mountain Research Station
USDA Forest Service, Montana State University, Bozeman, MT 59717-2780


Scotch broom, Cytisus scoparius (L.), a weed in the Pacific Northwest, is rapidly in-
vading open areas and ecologically sensitive dunes along the coast. Scotch broom pop-
ulations also exist in the eastern United States, but are apparently stable and not
expanding. The eastern Scotch broom populations may be kept in check by the broom
weevil, Bruchidius villosus (F.), a bruchid found in eastern populations of broom but
absent from those in the Northwest (Bottimer 1968). We studied the natural history
and biology of the broom weevil in North Carolina. Our purpose was to relate the
bruchid's life history to the phenology of the host plant and to quantify oviposition and
seed destruction by the bruchids. Adult weevils were active around the plant from the
first flowering in early spring until dehiscence of the seedpods in summer. The sex
ratio of the beetles was nearly 1:1 throughout the adult activity season. The number
of weevil eggs laid on the pods was correlated to the length of the pod and to the num-
ber of seeds in the pod. The larvae develop in and destroy the seeds of the broom plant.
Seed destruction at two sites in North Carolina was more than 80%; a field experi-
ment showed that seed destruction was dependent on the density of beetles in cages
on the plants. Because of its impact on seed production, the broom weevil may be a
viable candidate for biological control of broom in the Northwest.

Key Words: broom weevil, oviposition, seed destruction, Scotch broom, Cytisus sco-


La escoba escocesa, Cytisus scoparius (L.), una maleza del Noroeste Pacifico, esta
invadiendo rdpidamente areas abiertas y dunas ecol6gicamente delicadas a lo largo
de la costa. Poblaciones de escobas existen tambien en el este de los Estados Unidos,
pero aparentemente son estables y no se estan expandiendo. Es possible que las pobla-
ciones orientales de escobas est6n bajo control gracias al gorgojo de escoba, Bruchi-
dius villosus (F.), que se encuentra en poblaciones orientales de escoba pero esta
ausente en el noroeste. Estudiamos la historic natural y la biologia del gorgojo de es-
coba en Carolina del norte. Nuestro prop6sito fue relacionar el historical de vida del
gorgojo a la fenologia de la plant huesped y cuantificar oviposici6n y destrucci6n de
las semillas por los gorgojos. Los gorgojos adults eran activos alrededor de la plant
desde la fluoraci6n primera al comienzo de la primavera hasta la dehiscencia de las
capsulas de semillas en el verano. La proporci6n sexual de los escarabajos fue casi 1:1
durante la temporada de actividad adulta. El numero de huevos de gorgojo colocados
en las capsulas fue correlacionado al lo largo de la capsula y al numero de semillas en
la capsula. La larva se desarrolla y destruye las semillas de la plant escoba. La des-
trucci6n de semillas en dos lugares en Carolina del Norte fue mas de 80%; un experi-
mento de campo demostr6 que la destrucci6n de semillas dependia en la densidad de

Redmon et al.: Bruchidius villosus in North Carolina

las jaulas de escarabajos en las plants. Debido a su impact en la producci6n de se-
millas, el gorgojo de escoba puede ser un candidate viable para el control biol6gico de
la escoba en el noroeste.

Scotch broom, Cytisus scoparius (Fabaceae: Leguminosae), was introduced from
western Europe into the Pacific Northwest as an ornamental and as a soil and coastal
dune stabilizer. This woody shrub is well adapted to dry, disturbed habitats. Since its
introduction, C. scoparius has become a pest by spreading and displacing valued for-
age and native plants (Andres & Coombs 1995). In addition, Scotch broom has made
reforestation difficult in many areas in British Columbia, California, Oregon and
Washington (Balneaves 1992). Historically, C. scoparius has been difficult to control
because of its large and long-lasting seed bank (Bossard & Rejmanek 1994).
In western North Carolina the distribution of feral C. scoparius is restricted to iso-
lated patches and its populations do not seem to be expanding (Syrett et al. 1999). One
possible factor contributing to the limited spread of Scotch broom in this area is the
weevil, Bruchidius villosus. Adult weevils lay eggs on the broom seedpods and the lar-
vae feed on the seeds within the pod. Parnell (1966) studied the life history of
Bruchidius on broom in England. However, little is known about the life history and
biology of this potentially beneficial bruchid in North America. Because the broom
populations in North Carolina are stable, the ecological agent that limits population
growth in North Carolina could potentially be introduced into the Northwest to con-
trol this problem plant.
Bruchidius villosus shows potential as a biological control agent for C. scoparius
due to its direct effect on the host's seed production. This study's purpose was to im-
prove our understanding of B. villosus' biology as it relates to its host plant in western
North Carolina. We compared the biology of the bruchid in North America with that
in England to see how consistent the behavior and ecology of this species are in widely
separated populations. Our studies focused on the overwintering of the weevils, on the
relationship between beetle activity and host plant phenology, and on measurement
of oviposition and destruction of seeds. An understanding of these basic aspects of this
insect's biology is useful if this beetle is to be introduced as a biological control agent
for Scotch broom in the Pacific Northwest.


Site Descriptions

Site 1. The larger of the two stands of Cytisus scoparius used for this study was lo-
cated on a hillside approximately 11 km east of downtown Asheville, North Carolina.
The 69 x 24 m area has a northeastern exposure and contains 36 shrubs. The majority
of the shrubs are 1-2.5 m in height. In addition to C. scoparius, the hillside is densely
covered with two species of Solidago, Lespedeza cuneata, Lathyrus latifolius, and Eu-
patorium hyssopifolium. Several small Robinia pseudo-acacia and Acer rubrum also
grow at the site. The C. scoparius shrubs at this site are exposed to full sun through-
out the day.
Site 2. The second study site was located in a residential area on the edge of the
Bent Creek Research and Recreational area approximately 16 km southwest of
Asheville, North Carolina. The broom plants are old ornamentals that are no longer

Florida Entomologist 83(3) September, 2000

being maintained. The area is approximately 19 x 12 m and it contains 12 shrubs that
are 1-2 m in height. The site has a dense ground cover of Hedera helix. Also growing
on the site are Lespedeza sp., Solidago sp., Clematis terniflora, Aster pilosus, Lonicera
japonica, Celastrus orbiculatus, Quercus alba, Tsuga canadensis, and Prunus sero-
tina. The centrally located shrubs at this site receive little sun most of the day,
whereas the shrubs at the edges of the site receive sun for about half of the day.
At each site, continuous temperature readings were made using a StoAway Tid-
biT@ temperature logger. The loggers were placed at the base of mature C. scoparius
shrubs with the temperature probes in the leaf litter. The temperature data were
downloaded each week and transferred to computer.


Field Sampling. To determine if the bruchids overwinter near the broom plant, the
soil and leaf litter around the plants at each site were sampled by two separate meth-
ods. In January, three soil samples (100 cm2 x 2.5 cm deep) were collected at 30, 60 and
90 cm from the base of a mature broom plant at each site. Soil samples were placed
in Berlese funnels over jars of alcohol. After the soil dried for two weeks, the alcohol
and dried soil were examined for bruchids.
Litter samples were taken weekly from 26 Jan to 20 Apr 1998. Each litter sample
was approximately 100 cm2 x 1 cm deep. From each site a sample was taken at 30, 60
and 90 cm from the base of a mature plant. Each of the six weekly samples were emp-
tied into a large tub and examined thoroughly with a hand lens for B. villosus.
Laboratory Observations. To observe activity of adult weevils during the winter, we
set up eight 10 x 10 x 8 cm3 clear plastic observation cages. Each cage was filled with
soil to a depth of about 5 cm and contained 8 to 10 weevils. Four of the containers were
kept indoors at room temperature and four were kept outdoors under natural temper-
ature conditions. Because the group kept indoors was near a window, both groups of
beetles received the same natural photoperiod. Moisture was applied weekly to each
container. If the bruchids were active during warm winter days, individuals kept in-
doors could be observed on the lid and walls of their cages more than individuals in
the cages kept under natural temperatures. We checked the cages each week (13 Feb
through 30 Mar) and recorded the number of bruchids located on the lid or side of the
cage or on the soil. The number of beetles not visible was also recorded. We used a sign
test for paired-sample data to test for differences between the activity of bruchids
kept indoors and those kept outdoors.

Seasonal Abundance and Host Phenology

Weekly beat and sweep samples were made at each site to quantify seasonal
changes in the adult bruchid population. Samples were standardized throughout.
Beat samples consisted of beating a branch of a broom plant over a white tray (0.5 x
0.4 m). Beetles falling into the tray were counted and collected using an aspirator.
Three plants at each site were sampled by beating branches at the top, middle and
bottom of each plant. Sweep samples consisted of 10 uniform sweeps of the vegetation
surrounding the broom plants. The contents of the sweep nets were emptied into a
white tray and beetles were collected using an aspirator and counted. Each week, the
bruchids collected at the two sites were examined individually under a dissecting mi-
croscope and sexed (see Parnell 1964). In addition to sampling for beetles, the pheno-
logical stage of the broom plants was quantified. We paid particular attention to the
stage of leaf, flower, and seedpod development.

Redmon et al.: Bruchidius villosus in North Carolina

Oviposition and Seed Destruction

We dissected females collected on three dates (N = 7, 4 May-before pod set; N = 10,
19 May-just after pod set; N = 9, 1 Jun-when pods begin turning brown) and deter-
mined the development of their oocytes and checked for sperm in their spermathecae.
Once oviposition was occurring in the field, we collected three to five seedpods from
different heights and directions on four bushes at each of the sites. Using a dissecting
microscope, we determined the number of bruchid eggs on each of the pods. We also
measured the length of each pod (calyx to tip) and counted the number of seeds in each
pod. We tested for a significant correlation between the number of bruchid eggs laid
per pod and the number of seeds per pod. We also correlated the number of eggs with
the number of infested seeds.
We also conducted a field-cage experiment to determine the effect of bruchid den-
sity on the seed destruction. The cages were (1 mm) mesh bags about 30 cm in diam-
eter. We used four densities of female bruchids: 0 (control), 1, 4 and 8 female bruchids
per bag. We tied and taped each bag around the end of a broom branch with 8 to 18
young (green) seedpods. Because the seedpods had been exposed to natural oviposi-
tion, we removed bruchid eggs from the seedpod exterior by hand prior to introducing
female bruchids into the cages. To control for the effect of plant and location, each of
the four densities were placed on branches of a single broom plant and the density
treatments were randomly assigned to each cage. Four blocks (different broom plants)
of the four treatment densities were left in the field. At the end of four weeks (24 May
to 22 Jun 1998), the cages and branches were collected and examined in the labora-
tory to determine the number and size of the seedpods, the number of seeds, and the
infestation rates. We used an ANOVA to test for differences in seed destruction among
the treatment densities. Because the number of pods and seeds in each cage varied,
we also used a regression analysis to determine the relationship between the density
of weevils per seed and seed destruction.


Once the seedpods had started to dehisce (late June through August) and the wee-
vils were maturing, we collected five to ten pods from different heights and directions
from four bushes at each site. The length and number of seeds were recorded for each
pod. The pods were opened and we recorded the number of live and parasitized



No B. villosus were found in any of the soil or litter samples from either of the sites
during January and February. Except those insects that were in pods in the leaf litter
sample, all beetles found in the litter samples were dead. Some pods from the previous
year were collected in the litter samples and these contained a few live beetles (N = 7).
The adults kept in indoor cages were significantly more likely to be active (on cage
lid and walls) compared with those in cages kept outdoors (P = 0.04). In only two of
twelve observations were a greater number of active weevils found in outdoor cages.
Both of these were later observations (30 Mar and 6 Apr) when the beetles were be-
coming active in the field. Based on our sampling and cage experiment, it seems likely
that the beetles overwinter away from the plants and may become active during the
winter when temperatures are warm.

Florida Entomologist 83(3) September, 2000

Seasonal Abundance and Host Phenology

Adult weevils can be found on the plants from early April to the end of August
(Figs. 1 and 2). Arrival of the weevils is closely correlated with the first bloom of the
plant. We found beetles at each site during the weekly sample where flowers first ap-
peared. At both sites we found a peak in the population during the flowering stage of
the plant (beat samples Figs. 1 and 2). Another peak in the population from late June
to August was observed in beat samples of the broom and in sweep samples of sur-
rounding vegetation. The second peak corresponds to the darkening and dehiscence of
the seedpods and probably represents the emergence of the new generation of weevils.
Adults were active over a wide range of temperatures (Figs. 1 and 2). The overall sex
ratio sampled was slightly male biased 542m:500f. The sex ratio of weekly samples
fluctuated around 1m: if throughout the adult activity (10 weeks with slight male bi-
ases and 9 weeks with slight female biases).

Oviposition and Seed Destruction

None of the females collected 4 May, prior to pod formation on the plants, had ma-
ture oocytes although some (2 of 7) had mated. These females noticeably lacked fat de-
posits. Once pods were set, all females (N = 19, 19 May and 1 Jun) had some large
mature oocytes and most (13 of 19) were mated. The mature oocytes were large com-
pared with the size of the females and typically filled the abdominal cavity. The aver-
age number of mature oocytes found in gravid females was 10 (range 4 to 14).
For field collected seedpods (N = 100), the overall seed destruction was 82% (492/
600). Eighty-six percent of the seedpods sampled had more than 60% of their seeds de-
stroyed by weevils (Fig. 3). Seed destruction was similar at both sites [site 1: 85%
(225/266); site 2: 80% (267/334)]. Not surprisingly, the number of seeds per pod in-
creased significantly with the length of the seedpod (Fig. 4: N = 100; P < 0.001; r2 =
0.53). There was a significant relationship between the number ofbruchid eggs laid on
the seedpod and the length of the pod (Fig. 5: N = 59; P < 0.001; r2 = 0.23), with more
eggs being laid on longer pods. Beetles also laid more eggs on pods with more seeds
(Fig. 6: N = 59; P < 0.001; r2 = 0.18), although the number of seeds was a less reliable
indicator of the number eggs than was the size of the seedpod.
In the field cage experiment, there was a significant difference among the treat-
ments with respect to the proportions of seeds infested with larvae (Fig. 7: ANOVA;
F3, = 5.67; P < 0.02). There was also a significant relationship between the number of
beetles per seed in the cage and the proportion of seeds damaged (P < 0.02; r2 = 0.35).


Parasitization occurred at relatively low rates at both sites, 14% (14/100) at site 1
and about 6% (6/102) at site 2. The parasites have tentatively been identified as Di-
narmus sp. (Pteromalidae), known parasites of bruchids.



The weevils apparently overwinter away from the plant. We did not find any live
B. villosus in the soil, litter or sweep samples during January and February, although
a few live weevils were found in seedpods in litter samples and in seedpods still on the

Redmon et al.: Bruchidius villosus in North Carolina



ves N l




23-Feb 30-Mar 27-Apr



22-Jun 20-Jul 17-Aug


0.00' '
26-Jan 23-Feb 30-Mar

27-Apr 25-May 22-Jun 20-Jul 17-Aug

Fig. 1. Field sampling and temperature data for site 1, Asheville, NC. Bar graphs
show the proportion of beetles in weekly beat samples (lower bar graph, N = 233 wee-
vils) and sweep samples (upper bar graph, N = 121 weevils). Maximum and minimum
temperatures (C) for each week are also shown. Shaded horizontal bars at the top
of the plot show the phenology of the broom plant at the site. Upper bar shows the
development of seedpods, with the darker portion showing the dried dark pods. The
middle hatched bar represents the presence of flowers. Lower dotted bar shows devel-
opment of leaves.


25.0 r

- 15.0


S 0.30

C 0.20


1 1 1


000 L''

1 0.30

| 0.20
2 0.10



Florida Entomologist 83(3) September, 2000




25.0 r


1 ' I






I 0,10

0.40 r

0.00-Jan 23-Feb 30 r
26-Jan 23-Feb 30-Mar

27-Apr 25-May

22-Jun 20-Jul 17-Aug

Fig. 2. Field sampling and temperature data for site 2, Asheville, NC. Bar graphs
show the proportion of beetles in weekly beat samples (lower bar graph, N = 435 weevils)
and sweep samples (upper bar graph, N = 35 weevils). Maximum and minimum temper-
atures (C) for each week are shown. Shaded horizontal bars at the top of the plot show
the phenology of the broom plant at the site. Upper bar shows the development of seed-
pods, with the darker portion showing the dried dark pods. The middle hatched bar rep-
resents the presence of flowers. Lower dotted bar shows development of leaves.

23-Feb 30-Mar 27-Apr 25-May 22-Jun 20-Jul 17-Aug
Site 2: BEAT SAMPLES N=435

0.20 L

- - - - - - -

S-- I -- ,


Redmon et al.: Bruchidius villosus in North Carolina


60 N=100
- 5 .


E 20 I
z 'I
10 "

0.00-0.19 0.20-0.39 0.40-0.59 0.60-0.79 0.80-1.00
Proportion of Seeds Destroyed per Pod

Fig. 3. Frequency distribution of seed destruction in broom seedpods by Bruchid-
ius villosus in North Carolina. Each bar shows the number of seedpods found with
proportions of their seeds infested with bruchid larvae or pupae. Of the seedpods ex-
amined, 86% had more than 60% of their seeds destroyed by weevils.


010 9



0 I I
20 25 30 35 40 45 50 55 60

Pod Length (mm)

Fig. 4. Correlation between the length of the seedpod and the number of seeds
within the pod for Scotch Broom. Pods ranged in length from 26 to 56 mm and larger
pods tend to have more seeds (P < 0.0001; r2 = 0.53). Trend line through the points is
the least-squares linear fit to the data (Y = 0.32X 6.79).
the least-squares linear fit to the data (Y = 0.32X 6.79).

Florida Entomologist 83(3) September, 2000

20 25 30 35 40
Pod Length (mm)

45 50 55

Fig. 5. Correlation between the numbers of Bruchidius villosus eggs laid on the ex-
terior of seedpods and the length of the pod. Beetles laid significantly more eggs on
longer seedpods (P < 0.0001; r2 = 0.23). The least-squares linear fit to the data is plot-
ted (Y = 0.37X 8.72).




a 14
L 12

Z 6




0 2 4 6 8
Number of Seeds

10 12 14

Fig. 6. Correlation between the numbers of Bruchidius villosus eggs on seedpods
and the number of seeds within the pod. Beetles laid significantly more eggs on pods
having more seeds (P < 0.0001; r2 = 0.18). The trend line (Y = 0.69X + 1.82) is the least-
squares fit to the data.









0 S


* 0 0 0
p p p

Redmon et al.: Bruchidius villosus in North Carolina





_0.3 A A A

0 1 4 8
Number of Females

Fig. 7. Bars are the average proportions of seeds destroyed by Bruchidius villosus
larvae when different densities of females were placed in field cages. There is a signif-
icant effect of female density on the proportion of seeds infested with bruchid larvae
(ANOVA; F,,1 = 5.67; P < 0.02). Bars with different letters represent means that are
significantly different (post hoc Tukey's studentized range test).

broom bushes in November and December the previous year. In a two-year study of
Bruchidius in England (Parnell 1966), no beetles were discovered in large leaf litter
samples. In our cage experiments more beetles were found on lids and sides of the
cages under warm conditions (inside) compared with cages kept at colder tempera-
tures (outside), suggesting that bruchids spend cold temperatures in the soil and be-
come active if the temperature warms. Parnell (1966) found some beetles on flowering
gorse early in the spring prior to the flowering of broom. The only time we collected
weevils away from broom plants was later in the season. Although site 1 was densely
covered in other legumes, B. villosus was not found in any beat samples of the vege-
tation surrounding the broom plants prior to or during the bloom of the broom plants.
There was an increase in the numbers collected in sweep samples from vegetation
around the broom plants in late June, July and August (Fig. 1). The beetles in these
samples were probably newly emerged adults. Large numbers of bruchids apparently
leave the host plant in late summer once they emerge from the seedpods. We still do
not know where they go and how the bruchids overwinter.

Seasonal Abundance and Host Phenology

The phenology of the broom in North Carolina is nearly identical to that described
for broom in England (Parnell 1966). Leaves bud in late January or early February.
The first flowers appear in late March or early April with the first pods forming a few
weeks later. The green pods begin to blacken in mid-May, and most of the pods have
split and fallen from the broom plants by the end of August (Figs. 1 and 2). The wee-
vils arrive simultaneously with the first flowering of the plant. Females have no visi-

Florida Entomologist 83(3) September, 2000

ble fat deposits at this time and probably have used all their reserves during the
overwintering period. Both sexes can be collected in flowers where they feed on pollen
and nectar (see Parnell 1964). It seems likely that pollen would be used as a nitrogen
(protein) source for egg yolk deposition. However, when we dissected females and ex-
amined their digestive tract, the tract was simple with no distended crop and we did
not find pollen. Individual plants at our sites differed in flower set by a week or two.
As the more advanced plants lose their flowers (begin to set pods), the beetles move
to plants with newly opened flowers. When the seedpods are green, females deposit
eggs on the outer surface.

Oviposition and Seed Destruction

In North Carolina, egg development within females occurs soon after females ar-
rive at the plant. Prior to the set of pods in mid-May, ovaries were not developed and
did not contain mature oocytes. Once the first pods appeared (19 May), nearly all the
females collected had mature oocytes and were mated. Parnell (1966) found that all
females were mature and had mated by the end of May. The number of mature oocytes
we found in dissected females agreed with Parnell's (1966) description of 14 ovarioles
in the ovaries. We never found more than 14 mature eggs in the abdomen of females
(mean = 10), and all the mature eggs seemed to be in the same stage of development.
We found a significant relationship between the number of eggs laid on a pod and
the size of the pod (length or seed number). Typically one larva requires a single seed
to develop, and Parnell (1966) found that late-arriving larvae died. The first larva to
reach the seed apparently prevents others from reaching and feeding on the cotyle-
dons. Females probably can detect oviposition by other females, and the number of
eggs laid is adjusted to minimize the loss of offspring due to competition within the
pods. The larva hatches from the bottom of the egg where it is attached to the pod and
excavates a small tunnel into the pod. Some of the tunnels can be 1 cm long before the
tunnel enters the seed cavity. The destruction of the seeds by the weevils is high, over
80% at both sites. In California broom populations where Bruchidius does not occur,
seed damage by another beetle (Apion) is variable and rises from 5 to 10% early in the
season to 22 to 80% late in the season (Bossard & Rejmanek 1994). It seems likely
that B. villosus could have a significant impact on seed production. However, seed de-
struction is dependent on the density of weevils. In cages without females, the aver-
age seed destruction was about 40% (Fig. 7), indicating that a number of eggs had
hatched in some of the pods prior to our removing them for the experiment. The aver-
age increased to about 70% when there were 8 females in the cages. When we normal-
ized the data and correlated females per seed with seed destruction, our data
indicated that 1 female for every 1 to 2 seeds would be required to reach about 60 to
70% seed damage.


We do not know what stage of the bruchid is parasitized by the Dinarmus we
found. However, the parasites apparently do not affect the bruchid until later larval
stages because most of the parasitized bruchids had already consumed much of the
seeds they occupied. Thus, the damage to the seeds by the bruchid will not be influ-
enced by the activity of this parasite.
Our work shows that the biology of the North Carolina population is synchronized
to the phenology of the plant and is similar to that found in Europe. Because B. villo-
sus typically reduces the seed production by about 80%, this beetle is likely to have a

Redmon et al.: Bruchidius villosus in North Carolina

significant impact on broom populations. The number of eggs laid on seedpods is cor-
related with the size of the pods and number of seeds in the pods. Females probably
distribute their eggs in an ideal manner (Fretwell 1972) to reduce competition be-
tween their larvae and those of other females. Based on cage experiments, the density
of females must be fairly high to have 80% seed destruction seen in North Carolina
broom stands. However, once populations of B. villosus reach sufficient densities they
should make an important contribution to the biological control efforts directed at
Scotch broom in the Pacific Northwest.


We would like to thank Hazel Creasman for allowing us to use her lawn as a re-
search site. We would also like to thank Jim Petranka for advice on statistical analysis
and Jim Perry for help in identifying plants on the two field sites. We also want to
thank E. Eric Grissell of the Systematic Entomology Laboratory for identification of
Dinarmus specimens. The United States Forest Service Rocky Mountain Research
Station provided funding for this research in cooperative agreement with the Univer-
sity of North Carolina at Asheville.


ANDRES, L. A., AND COOMBS, E. M. 1995. Scotch broom, pp. 303-305 in J. R. Nechols,
L. A. Andres, J. W. Beardsly, R. D. Goeden and C. G. Jackson [eds.]. Biological
Control in the Western United States: Accomplishments and Benefits of Re-
gional Research Project W-84. University of California, Division of Agricultural
and Natural Resources Publication No. 3361.
BALNEAVES, J. M. 1992. A comparison of surfactants to aid control of gorse and Scotch
broom with herbicides. Plant Prot. Quart. 7: 174-177.
BOSSARD, C. C., AND REJMANEK, M. 1994. Herbivory, growth, seed production, and re-
sprouting of an exotic invasive shrub, Cytisus scoparius. Biol. Cons. 67: 193-
BOTTIMER, L. J. 1968. On the two species of Bruchidius (Coleoptera: Bruchidae) estab-
lished in North America. Canadian Entomol. 100: 139-145.
FRETWELL, S. D. 1972. Populations in a Seasonal Environment. Princeton Univ. Press,
Princeton, N.J.
PARNELL, J. R. 1964. The structure and development of the reproductive organs of the
two seed beetles Apion fuscirostre F. (Col., Curculionidae) and Bruchidius ater
(Marsh.) (Col., Bruchidae). Entomol. Mon. Mag. 100: 265-272.
PARNELL, J. R. 1966. Observations on the population fluctuations and life histories of
the beetles Bruchidius ater (Bruchidae) and Apion fuscirostre (Curculionidae)
on broom (Sarothamnus scoparius). J Anim. Ecol. 35: 157-188.
TER, AND A. W. SHEPPARD. 1999. The potential for biological control of Scotch
broom (Cytisus scoparius) (Fabaceae) and related weedy species. Biocontrol
News and Information 20: 17N-34N.

Florida Entomologist 83(3) September, 2000


1TREC-IFAS, University of Florida, 18905 SW 280 St. Homestead, FL 33031

2Department of Entomology and Nematology, University of Florida
Gainesville, FL 32601


The toxicity of fourteen different pesticides used in 'Tahiti' lime, Citrus aurantifo-
lia (Christman) Swingle, to the spider, Hibana velox (Becker) was tested under labo-
ratory conditions. Among the nine pesticides tested using a coated glass vial method,
the five broad-spectrum insecticides (azinphos-methyl, chlorpyrifos, ethion, carbaryl,
dicofol) were all highly toxic to H. velox, causing 100% mortality even at the lowest
concentration. Avermectin and Provado (a.i., imidacloprid) applied as sprays had
moderate toxicity; whereas, Admire (a.i., imidacloprid) applied as a drench and Tri-
Basic@ (copper fungicide) caused the lowest percent mortality (10-30%) even at the
highest concentration. With a leaf-dip method, petroleum oil exhibited a low toxicity
to H. velox. However, when combining petroleum oil with avermectin, a synergistic ef-
fect elevated the toxicity to moderate. Azadirachtin, Bacillus thuringiensis, and di-
flubenzuron showed low impact on H. velox. Less than 20% mortality was recorded at
the highest concentrations for all of these products.

Key Words: toxicity test, Hibana velox, Phyllocnistis citrella, predatory spider, citrus


El efecto t6xico de catorce pesticides comunmente usados en lim6n 'Tahiti', Citrus
aurantifolia (Christman) Swingle, se evalu6 en la aranfa, Hibana velox (Becker) bajo
condiciones de laboratorio. Entre los pesticides evaluados usando un m6todo de frasco
de cristal cubierto, los cinco pesticides de amplio espectro ([azinphos-methyl, chlor-
pyrifos, ethion, carbaryl, dicofol]) fueron altamente t6xicos hacia H. velox, causando
100% de mortalidad aun a la concentraci6n mas baja. Los insecticides, Avermectina y
Provado [(i.a., imidacloprid)] aplicados en forma atomizada demostraron tener baja
toxicidad; mientras que aplicaciones hasta empape de Admire [(i.a., imidacloprid)]
y Tri-Basic (funguicida de cobre) causaron el porcentaje de mortalidad mas bajo (10-
30%) hasta en las concentraciones mas altas. Al usar el m6todo de sumergir las hojas
del lim6n en la soluci6n con insecticide, el aceite de petr61leo mostr6 baja toxicidad ha-
cia H. velox. Sin embargo, al combinar el aceite de petr61leo con Avermectina, el efecto
de sinergismo elev6 la toxicidad de baja a moderada. Las concentraciones mas altas
de Azadirachtina, Bacillus Thuringiensis, y diflurobenzuron mostraron bajo impact,
al causar una mortalidad menor del 20% de los especimenes de H. velox.

Amalin et al.: Toxicity of Pesticides to H. velox

The citrus leafminer (CLM), Phyllocnistis citrella Stainton (Lepidoptera: Gracilla-
riidae), is a widely distributed and major pest of Citrus spp. as well as other species in
the family Rutaceae (Heppner 1993). Insecticides provide a rapid means of suppress-
ing CLM populations, especially during heavy infestations. However, chemical control
of CLM is difficult to achieve due to the development of resistance by CLM (Tan &
Huang 1996), harmful effects on natural enemies (Huang & Li 1989) and due to CLM
adults' prolonged and overlapping emergence which may require multiple sprays
(Pena & Duncan 1993). Nevertheless, chemical control is still considered an adjunct to
nonchemical control (cultural and biological control), particularly for the protection of
the main leaf flushes that are important in tree growth and fruit production.
Several species of spiders are found commonly inhabiting lime, Citrus aurantifolia
(Christman) Swingle, orchards in south Florida. A preliminary survey revealed that
three species of hunting spiders, Chiracanthium inclusum Hentz, Hibana velox
(Becker), and Trachelas volutus (Gertsch) and one species of jumping spider, Hentzia
palmarum (Hentz), fed on the larvae and prepupae of CLM (Amalin et al. 1996). In
addition to insecticides used for CLM control (Beattie et al. 1995, Heppner 1993), cit-
rus trees, particularly in nurseries, are regularly treated with pesticides to protect
them from other arthropod pests and diseases (Villanueva-Jimenez & Hoy 1997,
Knapp 1996). These materials could threaten survival and sustainability of native
and introduced natural enemies (Browning 1994, Pena & Duncan 1993, Villanueva-
Jimenez 1998).
The use of selective pesticides is a major consideration in developing an integrated
control program. Utilizing pesticides that are relatively harmless to spiders and other
predatory arthropods could increase the effectiveness of natural predation and
thereby reduce the overall population of injurious insects in lime orchards.
In this laboratory study, we tested the susceptibility of the hunting spider, Hibana
velox (Araneae: Anyphaenidae), to some pesticides commonly used in lime orchards.


Toxicity tests of 14 pesticides recommended for citrus (Knapp 1996) were con-
ducted in the laboratory on spiderlings of H. velox. Table 1 shows the target insect
pests and diseases of the selected 14 pesticides. Test organisms were obtained by col-
lecting egg sacs of H. velox in the field. The egg sacs were transported to the laboratory
for spiderling emergence. After emergence, the spiders were then reared in the labo-
ratory using artificial diet (Amalin et al. 1999). Two-week-old spiderlings were uti-
lized in the tests. Two methods were developed for the bioassay: coated glass vial and
leaf-dip methods. The selective efficiency of these two methods in assessing pesticide
effect on H. velox was compared.

Coated glass vial Method

Surface coating (Fig. 1) was done by exposing the spiders to surface coated glass vi-
als (15 mm diameter x 60 mm long) with pesticide solution. Each vial was coated with
pesticides by dispensing 80 ul of the pesticide solution. The vials were rolled manually
until the whole surface was coated with the solution and air-dried at room tempera-
ture for an hour or until they were completely dry. The spiders were placed individu-
ally in the surface coated vials. A cotton swab saturated with artificial diet was
inserted in each vial and sealed with cotton. The artificial diet consisted of a mixture
of 230 ml soybean beverage, 230 ml homogenized whole milk, 2 fresh chicken egg
yolks, and 5 ml honey (Amalin et al. 1999).

Florida Entomologist 83(3) September, 2000


Trade names Common name Arthropod pests Diseases

Agrimek Abamectin citrus rust mites, broad

(technical grade)
Agrimek + Abamectin +
Petroleum oil Petroleum oil
Petroleum oil FC435

Admire 2F,
Provado 1.6F
Copper WP

Micromite 25W

Guthion 2L

Kelthane 50WP

Tri Basic




mites, citrus leafminer
citrus rust mites, broad
mites, citrus leafminer

citrus rust mites, broad
mites, citrus leafminer,
scale, whitefly, spider
citrus leafminer
citrus leafminer

greasy spot, sooty

phytophthora, foot
rot, root rot, brown
rot, greasy spot,
melanose, citrus
scab, alternaria
brown spot

citrus rust mites, citrus
root weevil, citrus leaf-
orange dog

citrus rust mites, citrus
snow scale
scales, whitefly, mealy-
bug, adult citrus root
weevil, cricket
spider mites, citrus rust

Lorsban 4EC Chlorpyrifos scale, mealybugs, orange
dog, katydid, grasshop-
per, termites, fireants,
aphids, crickets

Neemix 4.5F
Sevin 4L

Azadirachtin citrus leafminer


citrus root weevil,
orange dog, katydids,
grasshoppers, crickets

Three concentrations of each pesticide (for both technical and formulated grade)
were tested, i.e. simulated field rate (SFR), twice SFR, and half SFR. Simulated field
rate is based on 50 gal of spray per acre, in amounts adequate to cover but not to the
point of run off. The SFR for each insecticide used is shown in Table 2. The different
concentrations were prepared using deionized water for all the pesticides except for
the technical grade of abamectin, and Admire for which acetone was used as a sol-
vent. A bioassay using the coated glass vial method was conducted for nine pesticides,

Dipel DF Bacillus
Ethion 4EC Ethion

Amalin et al.: Toxicity of Pesticides to H. velox

Fig. 1. Coated glass vial method.

i.e. abamectin (a compound produced by soil actinomycetes), azinphos-methyl, car-
baryl, chlorpyrifos, dicofol, ethion, Admire@ (a.i., imidacloprid, drench formulation),
Provado@ (a.i., imidacloprid, spray formulation), and Tri-Basic (a copper fungicide).

Leaf-dip Method

Dipping was done by cutting lime leaves into 25 mm diameter circles. The leaf cir-
cles were dipped separately for approximately 30 seconds into different concentra-
tions of five pesticides, i.e., azadirachtin (neem extracts), Bacillus thuringiensis
(insecticidal bacterium), diflubenzuron (insect growth regulator), abamectin + petro-
leum oil (FC-435), and petroleum oil alone. After dipping, the leaves were air-dried for


Pesticides Simulated field rate (SFR)

Abamectin 10.0 ug/ml
Azinphos-methyl 2.5 ul/ml
Carbaryl 2.5 ul/ml
Chlorpyrifos 2.5 ul/ml
Dicofol 3.0 ul/ml
Ethion 3.0 ul/ml
Imidacloprid-Provado 0.27 ul/ml
Imidacloprid-Admire 30.0 ug/ml
Tri-Basic 4.5 mg/ml

Florida Entomologist 83(3) September, 2000

approximately 2 h or until dry and placed singly on the bottom of a 30-hole (25 x 30
mm) cup tray (Fig. 2). A single spiderling was added to each arena and fed with the
artificial combination diet provided on a cotton swab. The cup openings were covered
with a plastic lid.

Experimental Protocol
For both methods, ten individual spiderlings were tested for each concentration.
The control was deionized water for all pesticides except abamectin and Admire for
which acetone was used as a control. The arenas were placed in an incubator condi-
tioned at 27C and 80% RH. The tests were repeated 30 times for each concentration
for all the pesticides used. The spiders were held in the treated substrates for 72 h and
the spider mortality was compared with the control from day 1 to day 3 after treat-
ment. The percent spider mortality was calculated for each concentration for all pes-
ticides. The mean percent spider mortality in treatment was adjusted for that of
control using Abbott's formula (1925).


A wide range of toxicity was exhibited by the different pesticides included in the
bioassay tests. Among the nine pesticides tested using the coated glass vial method,
Admire, and Tri-Basic resulted in a low percentage of spider mortality (10-30%),
producing the lowest mortality even at the highest concentration (twice of SFR) (Fig.
3). This suggests that both pesticides have low or no acute impact on H. velox. Abam-
ectin and Provado had moderate toxicity to H. velox. Abamectin had more than 50%
spider mortality at the highest concentration, whereas Provado had almost 40%.
Both showed less than 35% spider mortality using the SFR (Fig. 3). The organophos-

Fig. 2. Leaf-dip method.

Amalin et al.: Toxicity of Pesticides to H. velox

100 100
abamectin Provado
80 80

60 60

40 40
vi 20 20
ca 0

100 100
80 80
a t Admire Tri Basic 6
". 60 60

40 40

20 20

0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0

Concentration (x SFR)

Fig. 3. Percent mortality of spiders exposed to three different concentrations of ab-
amectin, Provado, Admire, and Tri-Basic using a coated glass vial method.

phates (azinphos-methyl, chlorpyrifos, and ethion), the carbamate (carbaryl), and the
organochlorine (dicofol), were all highly toxic, causing 100% spider mortality even at
the lowest concentration (half of SFR). In the leaf-dip method, oil alone exhibited a
low toxicity to H. velox. Only 15% mortality was recorded from the highest concentra-
tion, 5% from the SFR, and 0% from the lowest concentration (Fig. 4). However, petro-
leum oil + abamectin caused moderate toxicity. This demonstrates that abamectin has
a moderate effect, which was similar using abamectin alone (Fig. 3). The naturally de-
rived products azadirachtinn, Bacillus thuringiensis, and diflubenzuron) showed low
toxicity to H. velox. Less than 20% mortality was recorded at the highest concentra-
tion for all of these products (Fig. 4).
Other laboratory studies have also shown that broad-spectrum insecticides such
as organophosphates, carbamates and organochlorines have significant lethal effects
on spiders in general. For instance, in Israel, laboratory residue studies using grape-
fruit leaves as the substrate showed that the organophosphate chlorpyrifos was
highly toxic to Chiracanthium mildei, a hunting spider known to occur abundantly in
citrus orchards, whereas natural products (i.e. Bacillus thuringiensis and neem ex-
tracts) were virtually non-toxic to spiders (Mansour 1987). Saxena et al. (1984) found

Florida Entomologist 83(3) September, 2000



Oil + abamectin

Bacillus thuringiensis

I I i ii

0.0 0.5 1.0






* 100






1.5 2.0


20 -
0 -

0.0 0.5 1.0 1.5 2.0
Concentration (x SFR)

Fig. 4. Percent mortality of spiders exposed to three different concentrations of
naturally derived pesticides using a leaf-dip method.

that topical application of 50 g of a neem seed kernel extract did not affect the spider,
Lycosa pseudoannulata, a major predator of brown planthopper in Southeast Asia.
Fungicides have been shown to have little or no toxicity for spiders (Stark et al.
1995). Thus, results of our toxicity tests with these different groups of pesticides on
H. velox are comparable to findings of previous laboratory bioassays conducted on
other species of spiders.
Villanueva-Jimenez (1998) evaluated the nontarget effects of some of the pesticides
used by citrus nursery growers in Florida on adults of Ageniaspis citricola Logvi-
novskaya, an introduced parasitoid of CLM. He found that naturally derived products


Amalin et al.: Toxicity of Pesticides to H. velox

(neem, azadirachtin) and the insect growth regulator (diflubenzuron) were not toxic to
A. citricola. Similarly, we found that these types of products had low toxic effect on H.
velox. The impact of imidacloprid on H. velox and A. citricola also showed similar
trends in our study to that of Villanueva-Jimenez (1998). Provado applied to foliage
was highly toxic, while Admire used as drench was less toxic to bothA. citricola and
H. velox. Ethion, a broad-spectrum organophosphate, was highly toxic to both H. velox
andA. citricola. Abamectin was moderately toxic toA. citricola and H. velox. Petroleum
oil was the safest pesticide for spiders. It was also the safest for A. citricola (Villan-
ueva-Jimenez 1998). Previous studies showed that petroleum oil has a short residual
activity on natural enemies (Beattie & Smith 1993, Erkilic & Uygun 1997, Beattie et
al. 1995), and is cost-effective (Beattie 1992, Beattie et al. 1995). Thus, petroleum oil
may be ideal as a component for a pest management program for CLM on limes.
The similarity of our results with the results of other studies on predacious arthro-
pods and parasitoids indicate that the impact of pesticides on the existing pest/natu-
ral enemy complex must be taken into consideration when controlling CLM. The
broad-spectrum pesticides should be used cautiously. The use of naturally-derived
products and petroleum oil should be recommended as an adjunct control measure
with biological control to effectively manage the population of CLM.


We thank Drs. Joseph Knapp, Bruce Schaffer, and Josep Jacas for the review of the
manuscript. We also thank Bayer Corporation for supplying us with imidacloprid
technical grade sample. Florida Agricultural Experiment Station Journal Series No.


ABBOTT, W. S. 1925. A method of computing the effectiveness of an insecticide. J. Econ.
Entomol. 18: 265-267.
AMALIN, D. M., J. E. PENA, AND R. MCSORLEY. 1996. Abundance of spiders in lime
groves and their potential role in suppressing the citrus leafminer population.
P. 72 in M. A. Hoy (ed.). Proceedings, International Meeting: managing the
citrus leafminer, 22-25 April 1996, Orlando, Florida, University of Florida,
Gainesville, Florida.
AMALIN, D. M., J. REISKIND, R. MCSORLEY, AND J. PENA. 1999. Survival of the hunting
spider, Hibana velox (Becker), raised on different artificial diets. J. Arachnol.
27(2): 692-696.
BEATTIE, G. A. C. 1992. The use of petroleum spray oils in citrus and other horticultural
crops. Proc. 1st Nat. Conf. Aust. Soc. Hortic. Sci., Sydney, 1991. Pp. 351-362.
BEATTIE, G. A. C., AND D. SMITH. 1993. Citrus leafminer. Agfact H2.AE.4, second Ed.
NSW Agriculture, Orange, Australia 6 pp.
BEATTIE, G. A. C., Z. M. LIU, D. M. WATSON, A. D. CLIFT, AND L. JIANG. 1995. Evalua-
tion of petroleum spray oils and polysaccharides for control of Phyllocnistis cit-
rella Stainton (Lepidoptera: Gracillariidae). J. Aust. Entomol. Soc. 34: 349-353.
BROWNING, H. W. 1994. Early classical biological control on citrus. Pp. 27-46 in D.
Rosen, F. D. Bennet, J. L. Capinera (eds.), Pest Management in the Subtropics,
Biological Control-A Florida Perspective. Intercept, Andover, UK.
ERKILIC, L. B., AND N. UYGUN. 1997. Studies on the effects of some pesticides on white
peach scale, Pseudaulacaspis pentagon (Targ. Tozz.) (Homoptera: Diaspid-
idae) and its side-effects on two common scale insect predators. Crop Protection
16: 69-72.
HEPPNER, J. B. 1993. Citrus leafminer, Phyllocnistis citrella, in Florida (Lepidoptera:
Gracillariidae: Phyllocnistinae). Trop. Lepidoptera 4: 49-64.

262 Florida Entomologist 83(3) September, 2000

HUANG, M. AND S. LI. 1989. The damage and economic threshold of citrus leafminer
(Stainton) in citrus. Pp. 84-89 in Studies on the Integrated Management of Citrus
Insect Pests. Academic Book and Periodical Press. Beijing (in Chin., Eng. abs.).
KNAPP, J. L. 1996. 1996 Florida Citrus Pest Management Guide. Fla. Coop. Ext. Serv., In-
stitute of Food and Agricultural Sciences. University of Florida, Gainesville, FL.
MANSOUR, F. 1987. Effect of pesticides on spiders occurring on apple and citrus in Is-
rael. Phytoparasitica. 15: 43-50.
PENA, J. E., AND R. DUNCAN. 1993. Control of the citrus leafminer in south Florida.
Proc. Fla. State. Hort. Soc. 106: 47-51.
SAXENA, R. C., H. D. JUSTO JR., AND P. B. EPINO. 1984. Evaluation and utilization of
neem cake against the rice brown planthopper, Nilaparvata lugens, J. Econ.
Entomol. 77: 502-507.
STARK, J. D., P. C. JEPSON, AND C. F. GEORGE THOMAS. 1995. The effects of pesticides
on spiders from the lab to the landscape. Rev. Pestic. Toxic. 3: 83-110.
TAN, B. AND M. HUANG. 1996. Managing the citrus leafminer in China. Pp. 49-52 in M.
A. Hoy (ed.). Proceedings, International Meeting: Managing the Citrus Leaf-
miner, 22-25 April 1996. Orlando, Florida. University of Florida, Gainesville, FL.
VILLANUEVA-JIMENEZ, J. A. 1998. Development of an integrated pest management
program for the citrus leafminer, (Lepidoptera: Gracillariidae) in Florida Nurs-
eries. Ph.D. Dissertation. University of Florida, Gainesville, FL. 133 pp.
VILLANUEVA-JIMENEZ, J. A., AND M. A. HOY. 1997. Current practices in Florida nurs-
eries for managing the citrus leafminer. Citrus Industry 78: 80-85.

Florida Entomologist 83(3)

September, 2000


1Department of Plant Pathology, University of Nebraska-Lincoln
Lincoln, NE 68583

2USDA-ARS, Red River Valley Agricultural Research Center
Biosciences Research Laboratory, Fargo, ND 58105

3USDA-ARS, Midwest Livestock Insects Research Laboratory, Lincoln, NE 68583


Corn rootworms of the genus Diabrotica (Coleoptera: Chrysomelidae) are the most
serious pest of corn in midwestern United States. Despite their economic importance,
phylogenetic relationships within the genus remain unclear. Phylogenetic analysis of
five Diabrotica species and subspecies was undertaken using DNA sequences of the
nuclear rDNA first internal transcribed spacer region (ITS1) and a portion of the
mtDNA cytochrome oxidase I and II genes (COI/COII). Parsimony and maximum
likelihood analysis indicated that southern corn rootworm is sister to banded cucum-
ber beetle, whereas, northern corn rootworm forms a distinct clade with western and
Mexican corn rootworm. ITS1 and COI/COII were found to be useful markers for de-
termining phylogenetic relationships among diabroticites.

Key Words: Diabrotica, rootworm, phylogenetics, mitochondrial DNA, ribosomal DNA

Szalanski et al.: Genetic Relationship Among Diabrotica 263


Gusanos de raiz de maiz del genero Diabrotica (Coleoptera: Chrysomelidae) son la
plaga de mayor seriedad para el maiz en el medio oeste de los Estados Unidos. A pesar
de su importancia econ6mica, relaciones filogeneticas dentro del genero permanecen
confusas. Analisis filogenetico de cinco species de Diabrotica y subespecies fueron lle-
vadas a cabo usando secuencias de ADN del [rDNA] nuclear primer orden, region
[spacer] (ITS1) y una porci6n del [mtDNA cytochrome oxidase] I y II, genes (COI/
COII). Parsimonia y [cladograms] junta-vecinos indicaron que el gusano de raiz de
maiz sureno es hermano del escarabajo bandeado del pepino, mientras que los gusa-
nos de raiz de maiz nortefios forman un [clade] diferente de con los gusanos de raiz de
maiz occidentales y Mexicanos. ITS1 y COI/COII resultaron ser marcadores tiles
para determinar las relaciones filogeneticas entire diabroticitas.

Corn rootworms are a complex of species in the genus Diabrotica and are the most
serious pest of corn in midwestern United States (Levine and Oloumi-Sadeghi 1991).
Economically important species include western corn rootworm, Diabrotica virgifera
virgifera LeConte (WCR); Mexican corn rootworm, D. v. zeae Krysan & Smith (MCR);
northern corn rootworm, D. barberi Smith and Lawrence (NCR); banded cucumber
beetle, D. balteata LeConte (BCB); and southern corn rootworm, D. undecimpunctata
howardi Barber (SCR). In the United States, 20 to 25 million acres of corn are treated
annually with soil insecticides to protect crops from corn rootworm larval feeding dam-
age (Fuller et al. 1997). In addition, SCR is an economically important pest of cucurbits
and peanuts, and BCB is a pest of sweet potatoes in southeastern Unites States.
Despite their importance as pest species, the phylogenetic relationships within Di-
abrotica are poorly understood, primarily because of the morphological homogeneity of
the genus (Wilcox 1965, Krysan 1986). A phylogenetic study based on UPGMA cluster-
ing of allozymes of 11 Diabroticites by Krysan et al. (1989) supported two distinct
groups, virgifera and fucata. Two molecular DNA markers, the nuclear ribosomal inter-
genic transcribed spacer (ITS1) and a 254 bp DNA sequence of the mtDNA NADH 4
gene have proved useful for differentiating three Diabrotica spp. using DNA sequences
and polymerase chain reaction-restriction fragment length polymorphism (PCR-
RFLP) (Szalanski and Powers 1996, Roehrdanz et al. 1998, Szalanski et al. 1999).
To date, no molecular genetic studies have resolved the phylogenetic relationships
within Diabrotica. The goal of this study was to infer phylogenetic relationships
among five economically important Diabrotica species and subspecies using DNA se-
quences of the nuclear ribosomal DNA ITS1 region and a portion of the mtDNA cyto-
chrome oxidase I and II genes.


Origin of specimens used in this study are listed in Table 1. Corn rootworm beetles
were preserved in 70% ethanol or frozen at -20C. Frozen voucher specimens are
maintained at the USDA-ARS, Red River Valley Agricultural Research Center, Bio-
sciences Research Laboratory, Fargo, ND.
DNA was extracted from individual legs or thoraces using Puregene DNA isolation
kit D-5000A (Gentra, Minneapolis, MN) or using the high salt procedure of Cheung
et al. (1993). The 3' portion of the mtDNA cytochrome oxidase (CO) I gene, tRNA
leucine, and a 5' portion of the CO II gene was amplified with the primers C1-J-2797
(5'-CCTCGACGTTATTCAGATTACC-3') (Simon et al. 1994) and C2-N-3400 (5'-

Florida Entomologist 83(3) September, 2000


Sample Origin Date collected

Diabrotica virgifera virgifera (WCR) Brookings Co., SD 1997
D. v. zeae (MCR) Uvalde Co., TX 1997
D. barberi (NCR) Howard Co., NE 1998
D. undecimpunctata howardi (SCR) Lancaster Co., NE 1998
D. balteata (BCB) Warton Co., TX 1999
Cerotoma trifurcata (BLB) Lancaster Co., NE 1998
Colaspis brunnea Lancaster Co., NE 1999

TCAATATCATTGATGACCAAT-3') (Taylor et al. 1997). The 5' ends of these primers
were located at bp 2797 and 3400 of the Drosophila yakuba mtDNA map (Clary and
Wolstenholme 1985), respectively. The 3' portion of the 18S nuclear rDNA gene, the
entire ITS1 region, and the 3' region of the 5.8S gene was amplified with the primers
rDNA2 (5'-TTGATTACGTCCCTGCCCTTT-3', Vrain et al. 1992) and rDNA5,,, (5'-
ACGAGCCGAGTGATCCACCG-3', Cherry et al. 1997) per Taylor and Szalanski (1999).
The mtDNA PCR protocol was 35 cycles of 94C for 45 s, 42C for 45 s, and 72C for
90 s. The nuclear DNA PCR protocol was 40 cycles of 94C for 45 s, 55C for 45 s, and
72C for 120 s. Amplified DNA from individual beetles was purified, concentrated, and
sequenced per Szalanski et al. (1999). A previous study on the population genetic
structure of WCR and MCR revealed a lack of genetic variation within and among
populations (Szalanski et al. 1999). In NCR, ITS1 DNA sequence variation does occur
(Roehrdanz et al. 1998), but at a level insufficient to influence its relationship relative
to the other Diabrotica taxa. For the phylogenetic analysis, only one representative of
WCR, MCR and NCR were obtained from Szalanski et al. (1999). GenBank accession
numbers for the taxa sequenced in this study are AF195193 to AF195202.
The DNADIST program of PHYLIP v3.57C (Felsenstein 1993) was used to calcu-
late genetic distances according to the Kimura 2-parameter (Kimura 1980) model of
sequence evolution (Table 2). Diabrotica DNA sequences were aligned using two chry-
somelids, grape colaspis Colaspis brunnea (Fabricius), and bean leaf beetle (BLB),
Cerotoma trifurcata (Forster) as the outgroup taxa. Maximum likelihood and un-
weighted parsimony analysis on the alignments was conducted with PAUP* 4.0b2
(Swofford 1999). Gaps were treated as missing characters for all analysis. The reli-
ability of trees was tested with a bootstrap test (Felsenstein 1985). Parsimony boot-
strap analysis included 1000 resamplings using the Branch and Bound algorithm of
PAUP*. For maximum likelihood analysis, the heuristic search and Hasagawa-Kish-
ino-Yano (HKY) model of sequence evolution were used (Hasagawa et al. 1985). Par-



WCR D. virgifera virgifera .0016 .0190 .0767 .0746
MCR D. v. zeae .0000 .0206 .0803 .0763
NCR D. barberi .0542 .0542 .0745 .0741
BCB D. balteata .1147 .1147 .1326 .0368
SCR D. undecimpunctata howardi .1240 .1240 .1322 .0886 -

Szalanski et al.: Genetic Relationship Among Diabrotica 265

meters for the maximum likelihood test were Ti/Tv = 2, and (parameter of gamma
distribution = 4.28 for the rDNA ITS1 sequences and 5.12 for the mtDNA sequences).
For the bootstrap analysis of the maximum likelihood trees the heuristic setting was
used with 100 resamplings.


The ITS1 amplicon for the five Diabrotica taxa ranged from 642 to 758 bp long. The
mtDNA amplicon was 581 bp long for all five Diabrotica. The average base frequen-
cies were A = 0.29, C = 0.17, G = 0. 21, and T = 0.33 for the entire rDNA amplicon, and
A = 0.36, C = 0.15, G = 0.11, and T = 0.38 for the mtDNA amplicon. The aligned DNA
data matrix, including the outgroup taxa, (available upon request, and at the web site
http://ianrwww.unl.edu/ianr/plntpath/nematode/aszalans.htm) resulted in a total of

Colaspis brunnea

Cerotoma trifurcata

i Diabrolica undectmpuncata howardi

Diabroltca balteata

DiabroQtca barber

* Diabrotica virgifera virgifera

Diabrotica virgifera zeae (MCR)

Fig. 1. A phylogenetic tree for Diabrotica based on DNA sequence analysis of the
nuclear rDNA ITS1 region, derived from parsimony analysis and rooted by the out-
group taxa Colaspis brunnea and Cerotoma trifurcata. Bootstrap values are provided.

Florida Entomologist 83(3) September, 2000

879 characters for ITS1 and 583 characters for COI/COII, including gaps. Of the 879
ITS1 characters, 194 characters (22%) were variable and 175 (20%) were parsimony
informative characters. For the 583 COI/COII characters, 255 characters (44%) were
variable and 76 (13%) characters were parsimony informative.
The rDNA dataset had only one most parsimonious tree (Fig. 1), (length = 335,
CI = 0.96, CI excluding uninformative sites = 0.84). Diabrotica undecimpunctata
howardi was depicted as the sister to D. balteata, with D. barberi, D. virgifera vir-
gifera, and D. v. zeae representing a sister clade (Fig. 1). All of the inferred relation-
ships were supported in >90% of the 1000 bootstrap replications. The maximum
likelihood tree (-Ln likelihood = 2608.11827) yielded a phylogenetic relationship iden-
tical to the parsimony tree (Fig. 1). Parsimony (length = 360, CI = 0.89, CI excluding
uninformative sites = 0.74) and maximum likelihood (-Ln likelihood = 2255.76826)
analysis of the mtDNA dataset was identical to that of the ITS1 dataset (Fig. 1).
Results of the present study were congruent with those derived from allozyme and
morphological data (Krysan et al. 1989, Krysan and Smith 1987). Our study supports
the allozyme (Krysan et al. 1989) UPGMA phylogeny with southern corn rootworm
and banded cucumber beetle forming a distinct clade (fucata group) relative to north-
ern, western, and Mexican corn rootworm (virgifera group). The close relationship be-
tween NCR and WCR is supported by field observations of attempted interspecific
mating (Krysan and Guss 1978).
This study provides a baseline for the phylogenetic relationships of this economi-
cally important genus. The ITS1 and COI/COII markers contain adequate informa-
tion for phylogenetic assessment of the five Diabrotica studied and should prove
useful for understanding the relationship of other diabroticites, and could provide the
basis for species specific molecular diagnostic markers.


R. D. Peterson II, and T. 0. Powers provided helpful suggestions and critical re-
views of the manuscript. This is Journal Series No. 12808 Agricultural Research Di-
vision, University of Nebraska-Lincoln.


CHERRY, T., A. L. SZALANSKI, T. C. TODD, AND T. 0. POWERS. 1997. The internal tran-
scribed spacer region of Belonolaimus (Nemata: Belonolaimidae). J. Nematol.
29: 23-29.
CLARY, D. 0., AND D. R. WOLSTENHOLME. 1985. The mitochondrial DNA molecule of
Drosophila yakuba: nucleotide sequence, gene organization, and genetic code.
J. Mol. Evol. 22: 252-271.
CHEUNG, W. Y., N. HUBERT, AND B. LANDRY. 1993. A simple and rapid DNA microex-
traction method for plant, animal, and insect suitable for RAPD and other PCR
analysis. PCR Meth. Appl. 3: 69-70.
FELSTENSTEIN, J. 1985. Confidence limits on phylogenies: An approach using the boot-
strap. Evolut. 39: 783-791.
FELSENSTEIN, J. 1993. PHYLIP: Phylogeny inference package, version 3.5. The Uni-
versity of Washington, Seattle.
mization of soil insecticide rates for managing corn rootworm (Coleoptera:
Chrysomelidae) larvae in the north central United States. J. Econ. Entomol.
90: 1332-1340.

Szalanski et al.: Genetic Relationship Among Diabrotica 267

HASAGAWA, M., H. KISHINO, AND T. YANO. 1985. Dating the human-ape split by a mo-
lecular clock of mitochondrial DNA. J. Molec. Evolut. 22: 160-174.
KIMURA, M. 1980. A simple method for estimating evolutionary rate of base substitu-
tion through comparative study of nucleotide sequences. J. Molec. Evol. 16:
KRYSAN, J. L. 1986. Introduction: biology, distribution, and identification of pest Di-
abrotica. Methods for the study of pest Diabrotica. J. L. Krysan and T. A. Miller
(eds.). Pp. 1-23. Springer-Verlag, New York.
KRYSAN, J. L., AND P. L. GUSS. 1978. Barriers to hybridization between Diabrotica vir-
gifera and D. longicornis barberi (Coleoptera: Chrysomelidae). Ann. Entomol.
Soc. Am. 71: 931-934.
KRYSAN, J. L. AND R. F. SMITH. 1987. Systematics of the virgifera species group of Di-
abrotica (Coleoptera: Chrysomelidae: Galerucinae). Entomography 5: 375-484.
KRYSAN, J. L., I. C. MCDONALD, AND J. H. TUMLINSON. 1989. Phenogram based on al-
lozymes and its relationship to classical biosystematics and pheromone struc-
ture among eleven Diabroticites (Coleoptera: Chrysomelidae). Ann. Entomol.
Soc. America 82: 574-581.
LEVINE, E., AND H. OLOUMI-SADEGHI. 1991. Management of diabroticite rootworms in
corn. Ann. Rev. Entomol. 36: 229-255.
variation in western, Mexican and northern corn rootworms, Diabrotica spp.
(Coleoptera: Chrysomelidae). Keystone Symposia: Toward the genetic manip-
ulation of insects. p. 35.
tion, weighting, and phylogenetic utility of mitochondrial gene sequences and
a compilation of conserved polymerase chain reaction primers. Ann. Entomol.
Soc. America 87: 651-701.
SWOFFORD, D. L. 1999. PAUP*. Phylogenetic Analysis Using Parsimony (*and Other
Methods). Version 4. MA: Sunderland.
SZALANSKI, A. L., AND T. 0. POWERS. 1996. Molecular diagnostics of three Diabrotica
(Coleoptera: Chrysomelidae) pest species. J. Kansas Entomol. Soc. 69: 260-266.
variation in geographical populations of western and Mexican corn rootworm
(Coleoptera: Chrysomelidae). Insect Molec. Biol. 8: 519-526.
chondrial DNA variation among Muscidifurax spp. (Hymenoptera: Pteromal-
idae), pupal parasitoid of filth flies (Diptera). Ann. Entomol. Soc. America 90:
TAYLOR, D. B., AND A. L. SZALANSKI, A. L. 1999. Identification of Muscidifurax spp. by
polymerase chain reaction-restriction fragment length polymorphism. Biol.
Cont. 14: 270-273.
cific rDNA restriction fragment length polymorphism in the Xiphinema amer-
icanum group. Fund. Appl. Nematol. 15: 563-573.
WILCOX, J. A. 1965. A synopsis of the North American Galerucinae (Coleoptera: Chry-
somelidae). New York State Museum Science Services Bulletin 400.

Florida Entomologist 83(3) September, 2000


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


The winged imago of Neotermes mona (Banks) is described for the first time and
the soldier caste is redescribed as two size morphs. The distribution of N. mona in-
cludes Hispaniola, Turks & Caicos Islands, Puerto Rico, and Virgin Islands. It is the
largest kalotermitid in this region.

Key Words: taxonomy, Caribbean, Neotropics, distribution, imago


Se described por primara vez los adults alados de Neotermes mona (Banks) y se
redescribe la casta de soldados como formas de dos tamanos. La distribuci6n de N.
mona incluye Espanola, las Islas Turcas y Caicos, Puerto Rico y las Islas Virgenes.
Este es el mayor kalotermitido de esta region.

For his original description of Neotermes (=Kalotermes) mona from Mona Is.,
Puerto Rico, Banks (1919) offered little more than a brief soldier comparison with In-
cisitermes (=Kalotermes) schwarzi (Banks) and N. (=K.)jouteli (Banks) (Banks & Sny-
der 1920). The description of N. mona lacked measurements and was accompanied
only by a simple line drawing of a soldier's head and pronotum that resembles
Krishna's (1961) definition of the genus Neotermes. In light of recent collections of N.
mona from the Turks and Caicos Is. (Scheffrahn et al. 1990), Mona Is. (Ramos 1946,
Jones 1993), Dominican Republic, Puerto Rico, and Guana Is., B.V.I. (Scheffrahn et al.
1994, Collins et al. 1997), and now Vieques Is. (Puerto Rico) and St. John, U.S.V.I., we
herein redescribe the soldier as a dimorphic caste and describe the winged reproduc-
tive for the first time.


Morphometrics of specimens preserved in 85:15 ethanol:water were made with a
stereomicroscope fitted with a calibrated ocular micrometer. Specimens for measure-
ment were selected from 81 colony series collected during 1988-1999 from 50 localities
on 10 islands in the West Indies (Fig. 1). Measurements of the large and small soldier
morphs are presented separately, but other characters do not differ sufficiently to
warrant separate descriptions.
Scanning electron micrograph prints were scanned at 600 dpi, and the digital im-
age outline traced using photograph-enhancing software (Photo Magic, Micrografx,
Inc., Richardson, TX). The background was converted to black, and the scale bar was

Krecek et al.: Neotermes mona Redescription 269

digitally redrawn. Latitude and longitude coordinates of collection sites were con-
verted to decimal degrees and mapped (Fig. 1) usingArcView GIS version 3.0a software
and relevant map data from Digital Map of the World version 1.0 (Environmental Sys-
tems Research Institute, Inc. Redlands, CA).

Neotermes mona (Banks)

Kalotermes mona Banks 1919: 478 [soldier; Fig. 6].
Kalotermes (Neotermes) mona; Snyder in Wolcott 1948: 62.
Kalotermes mona; Snyder 1949: 18.
Neotermes mona; Krishna 1961: 322.

Imago (Fig. 2 A-B, Table 1).

In dorsal view, general color almost uniformly ferruginous, except for darker, chest-
nut brown frons and anterior vertex in majority of specimens, and dark chestnut
brown posterior halves of three posterior abdominal tergites. Mandibles dark chestnut
brown. Anteclypeus yellowish. Antennae ferruginous orange except for chestnut brown
third article. Compound eyes almost black. Chevron pattern on pterothorax faint and
wide. Femora yellowish, tibiae ferruginous. Sclerotized wing venation ferruginous, re-
mainder of wings, arolia, and abdominal sternites pale ferruginous orange.
In dorsal view, head capsule suboval with sides rectate and faintly converging to
anterior especially in ventral aspect; posterior of head capsule broadly rounded. In ob-
lique view, frons broadly concave, with raised lateral margins, and with delicate stri-
ations. In lateral view, frons plane continuous with plane of vertex. Compound eyes
large and protruding, subcircular, with long subrectate or slightly concave margins
along antennal sockets. Ocelli slightly protruding; comparatively large, oval; broadly
contacting eyes. Mandibular bases with striations. Shallow, small, and circular de-
pression centered at intersection of epicranial suture. Head, pronotum, wing scales,
abdominal tergites, and sternites with numerous and long setae. Antennae with 19 to
22 articles, usually 21 or 22, relative length formula 2 > 3 > 4 = 5. Pronotum about

S00 W
0 Neotermes mono
Provienciales (<7 termite collection sites
Grand Turk
100D km
f Great Inagua


Dominican Republic rg5DN --

_, .- :-,t, -. Vie Gue s n
r t - M, !* 5sst John
Sa -. 1 Viequest John

St. Croix

Fig. 1. Neotermes mona localities and termite collection sites from 1988-1999.

270 Florida Entomologist 83(3) September, 2000

Fig. 2. Scanning electron micrographs ofN. mona. Oblique (A) and dorsal (B) views
of an imago head from Guana Island. Dorsal (C) and lateral (D) views of a large sol-
dier from Guana Island. Dorsal (E) and lateral (F) views of a small soldier from Mona
Island. Scale bar equals 1 mm for A-B and 2 mm for C-F.

twice as wide as median length; anterior and lateral margins with raised and rounded
rim. Anterior margin of pronotum concave; posterior margin slightly concave. Ante-
rior wings with very long subcosta and radius; subcosta terminating at costal margin
near V2 of wing length from suture; radius reaching costal margin at 2/3 of wing length.
Radial sector with 4-5 branches along distal half of wing. Sclerotized media with sev-
eral fine transverse branches to radial sector; posteriorly, few short diagonal and scle-
rotized branches fade into membrane except for most distal branch that terminates at
wing margin. Texture of wing membrane with very faint nodulations. Arolia large.


The alate of N. mona resembles that of N. jouteli, but the former is larger. Head
width at eyes in N. mona is >2.00 mm, while in N. jouteli it is <1.81 mm; maximum
pronotum width in N. mona is >2.22 mm versus <2.05 mm in N. jouteli, and pronotum
maximum length is >1.49 mm and <1.32 mm for N. mona and N. jouteli, respectively.
Total length with wings of the N. mona alate is >17.89 mm compared to <16.05 mm in
N.jouteli; and wing length from suture is >13.35 mm and <11.79 mm, respectively. The
imago of N. mona differs from all other West Indian congeners by its darkened three

Krecek et al.: Neotermes mona Redescription


Measurement in mm
(n = 5 males, 5 females from 8 colonies) Range Mean + S.D.

Head length with labrum 2.14-2.52 2.38 + 0.12
Head length to postclypeus 1.63-1.97 1.83 + 0.10
Head width, maximum at eyes 2.00-2.17 2.10 + 0.051
Head height without postmentum 1.11-1.21 1.15 + 0.032
Labrum width, maximum 0.73-0.89 0.82 + 0.049
Eye diameter with sclerite, maximum 0.57-0.67 0.62 + 0.033
Eye to head base, minimum from sclerite 0.29-0.38 0.34 + 0.027
Ocellus diameter, maximum 0.22-0.27 0.24 + 0.014
Ocellus diameter, minimum 0.16-0.20 0.17 + 0.011
Eye sclerite to ocellus, minimum 0 0
Pronotum, maximum length 1.49-1.68 1.58 + 0.068
Pronotum, maximum width 2.22-2.47 2.35 + 0.095
Total length with wings 17.89-22.01 19.26 + 1.17
Total length without wings 9.09-11.08 10.08 + 0.65
Fore wing length from suture 13.35-16.47 14.81 + 0.80
Fore wing, maximum width 3.56-4.15 3.87 + 0.17
Hind tibia length 1.77-1.91 1.85 + 0.041

posterior abdominal tergites. The imago of N. mona has a generally darker, ferrugi-
nous coloration compared with a lighter ferruginous orange color in N. jouteli. The N.
mona imago has dense pilosity composed of long setae (<0.3 mm) on the head, prono-
tum, wing scales, and abdominal sternites and tergites, while the N. jouteli imago is
adorned with sparse short setae (-0.03 mm long). The setal follicles of N. mona are
strikingly lighter than the surrounding cuticle, while in N. jouteli the follicles are un-
apparent. The frons of N. mona is on an even plane with the anterior vertex, while in
N. jouteli there is about a 150 slope between the planes of the frons and vertex.
Although alate pilosity characters of N. mona are similar to those ofN. castaneus
Burmeister, N. castaneus is smaller overall, has much smaller compound eyes than N
mona, and its head color is brownish compared to the ferruginous N. mona head. The
frons of the N. mona alate is faintly concave and delicately striate; in N.castaneus the
frons is faintly convex and striations are absent.

Soldier (Figs. 2 C-F, Tables 2-3).

The soldier caste consists of two distinct morphs, large and small, both usually
present in mature colonies. Other than size, there are few distinguishing characters
that separate small and large soldiers of N. mona compared with some congeners and
species in several other kalotermitid genera.
Head capsule generally ferruginous in dorsal view, in some specimens postclypeus,
frontal carinae and antennal carinae darker, chestnut brown. Thorax and abdominal
dorsum ferruginous orange. Mandibles glossy, almost black, with very dark chestnut
brown areas near articulations. Epicranial suture faint or absent. Eyes almost black.
Femora yellow-white; remaining sternum pale ferruginous orange. Darker ferrugi-
nous postmentum contrasting with ferruginous orange genae.

Florida Entomologist 83(3) September, 2000


Measurement in mm (n = 10 from 9 colonies) Range Mean + S.D.

Head length to tip of mandibles 4.37-5.76 4.85 + 0.52
Head length to postclypeus 2.77-3.42 3.04 + 0.22
Head width, maximum 2.31-3.13 2.66 + 0.24
Antennal carinae, outside span 2.18-2.80 2.40 + 0.19
Head height, excluding postmentum 1.52-1.93 1.83 + 0.13
Labrum, maximum width 0.64-0.87 0.71 + 0.069
Postclypeus width, maximum 0.88-1.18 0.98 + 0.079
Left mandible length, tip to most distant
visible point of ventral condyle 2.24-2.84 2.46 + 0.18
Postmentum, length in middle 1.88-2.34 2.08 + 0.14
Postmentum, maximum width 0.78-1.06 0.89 + 0.077
Postmentum, minimum width 0.44-0.64 0.52 + 0.056
Pronotum, maximum width 2.73-3.28 2.99+ 0.18
Pronotum, maximum length 1.63-2.02 1.82 + 0.10
Hind tibia length 1.60-2.15 1.76+ 0.16
Total length 9.94-13.35 11.39 + 1.01

In dorsal view, head capsule subsquare, slightly longer than wide, with sides sub-
parallel in large soldiers, faintly convex in small ones; posterior corners of both mor-
phs rounded; median posterior of head capsule rectate. In some individuals of both
morphs, sides of head capsule faintly converging anteriorly. Head capsule covered


Measurement in mm (n = 10 from 8 colonies) Range Mean + S.D.

Head length to tip of mandibles 5.69-6.49 6.07 + 0.30
Head length to postclypeus 3.81-4.31 4.04 + 0.17
Head width, maximum 2.87-3.43 3.13 + 0.20
Antennal carinae, outside span 2.57-3.10 2.87 + 0.17
Head height, excluding postmentum 1.88-2.47 2.27 + 0.17
Labrum, maximum width 0.72-0.83 0.78 + 0.044
Postclypeus width, maximum 1.05-1.19 1.13 + 0.048
Left mandible length, tip to most distant
visible point of ventral condyle 2.67-3.07 2.87 + 0.13
Postmentum, length in middle 2.77-3.10 2.90 + 0.12
Postmentum, maximum width 0.96-1.14 1.06 + 0.067
Postmentum, minimum width 0.49-0.54 0.52 + 0.022
Pronotum, maximum width 3.37-3.96 3.60 + 0.21
Pronotum, maximum length 2.00-2.37 2.22 + 0.12
Hind tibia length 1.93-2.20 2.04 + 0.091
Total length 11.64-15.62 13.48 + 1.51

Krecek et al.: Neotermes mona Redescription

with dense mat of setae except on occiput. Body also covered with dense mat of setae.
Frons flattened, usually faintly concave. In small soldiers, frons surface smooth; in
some of large soldiers frons with very faint reticulate rugosity. Frontal carinae lobed
with short pointed tubercle. Labrum linguiform. Mandibles comparatively long and
robust; in large soldiers, slightly more robust, with faint, and slightly pilose basal
hump; dentition distinct. Antennae with 13 to 19 articles, usually 16 or 17 in both
morphs; in small soldiers often only 13 or 14 articles present; third antennal article
subclavate, terminal articles usually markedly elongated; antennal formula 2 < 3 > 4 = 5.
Antennal carinae markedly protruding and rugose. Pronotum about twice as wide as
long. Anterior margin of pronotum deeply concave; sides of pronotum slightly convex;
posterior margin weakly emarginate. All soldiers with short wing pads on meso- and
In lateral view, head capsule slightly dorso-ventrally flattened; frons plane sloping
200 from plane of vertex; mandibles moderately curved upward; eyes moderately elon-
gated or less often subcircular, with peripheral satellite facets. Pilosity of frons and
anterior vertex much more dense than on occiput. Hind femora moderately broadened
in small soldiers and noticeably inflated in large ones. Postmentum narrowed near
middle in large soldiers.


Because of size overlap, biometrical separation of small soldiers of N. mona from
those ofN. jouteli is possible only on examination of a series of specimens. Both small
and large soldiers ofN. mona tend to be larger than those ofN. jouteli. In large soldiers,
the following measurements do not overlap between these two species. The maximum
head width is >2.87 mm and inN. jouteli <2.70 mm, left mandible length is >2.67 mm
and <2.42, maximum width of pronotum is >3.37 mm and <3.03 mm, and maximum
length of pronotum is >2.00 mm and <1.85 mm, in N. mona and N. jouteli, respectively.
In large soldiers, the mandibles of N. mona are more robust but with less developed
basal humps and pilosity than those of N. jouteli. In small soldiers, the mandibles are
more robust in N. mona compared to N. jouteli, while pilosity and hump proportions
are similar. The body of N. mona is generally much more pilose than that of N. jouteli.
The frontal carinae and adjacent frontal area of N. mona soldiers exhibit much denser
pilosity than in N. jouteli in both morphs, especially in large soldiers. The tergum and
sternum of N. mona soldiers are conspicuously more pilose than in N. jouteli.

Material Examined and Measured

USA. Puerto Rico. All samples collected by J. Chase, J. Mangold, J. de la Rosa and
R. Scheffrahn. Bosque de Aguirre; 17.93N, 66.15W; 1-VI-1993; 1 small soldier (PR-
175); 2 large soldiers (PR-176); 1 small, 2 large soldiers, 3 alates (PR-177). Mona Is-
land. All samples collected by S. Jones. Uvero Beach; 18.06N, 67.90W; 11-XI-1992; 1
large soldier (PR-409); S. W. airstrip; 18.06N, 67.91W; 12-XI-1992; 1 small soldier, 1
alate (PR-416); same data; 1 small soldier (PR-417). British Virgin Islands. Guana Is-
land. North slope near resort; 18.49N, 64.44W; 27-X-1992; J. Krecek; 1 small soldier
(no. VI-59); same site; X-1991; L. Hernandez; 1 large soldier, 1 alate (no. VI-60); same
site; 19-X-1992; Krecek, light trap; 1 alate (no. VI-61). British West Indies. Turks &
Caicos Islands. Grand Turk Island. 21.46N, 71.14W; 6-II-1990; Scheffrahn, and B.
Diehl; 1 small soldier, 1 alate (TC-21). Dominican Republic. Barahona Prov., Cabral/
Barahona Hwy; 18.23N, 71.13W; 20-VI-1991; Chase, Mangold, de la Rosa, and
Scheffrahn; 1 small, 1 large soldier (DR-27); La Altagracia Prov., Juanillo; 18.480N,

Florida Entomologist 83(3) September, 2000

68.42W; 11-VI-1992; Chase, Mangold, de la Rosa, and Scheffrahn; 2 small, 1 large
soldier (DR-562); Pedernales Prov., 25 km E. Pedernales; 17.92N, 71.53W; 28-X-
1993; Chase and de la Rosa; 1 alate (DR-864); Puerto Plata Prov., Punta Rucia;
19.88N, 71.20W; 21-VIII-1994; Chase, Krecek, de la Rosa, and Scheffrahn; 1 small,
1 large soldier (DR-946); Peravia Prov., 5 km W. Bani; 18.30N, 70.12W; 4-VIII-1995;
Chase; 1 alate (DR-1209); Pedernales Prov., Pedernales, beach, forest; 18.03N,
71.74W; 3-XI-1996; Chase and Krecek; 1 large soldier (DR-1300); Saona Island (new
record). La Romana Prov., Punta Catuano; 18.20N, 68.78W; 14-III-1995; Chase and
de la Rosa; 1 alate (DR-1130).

Additional Material Examined

Paratype: Puerto Rico. Mona Is. No date or collector given; 1 large soldier, 5 small
soldiers, 1 presoldier, and many nymphs (MCZ Type 10076). Vieques Island (new
record). Between Red Beach and U.S. Navy observation post installation; 21.900N,
65.37W; 24-VII-1999; Scheffrahn, Maharajh, and Chase; 1 nymphoid supplementary,
many soldiers and nymphs, 2 larvae (PR-648, 650). U.S. Virgin Islands. St. John (new
record). Terminus of paved road, Hwy 107; 18.31N, 64.71W; 29-VII-1999; Chase; 1
nymphoid supplementary, many soldiers and nymphs (no. VI-97).


Long considered endemic to Mona Is. (Wolcott 1948), N. mona is now recorded from
a wide geography of the central West Indies (Fig. 1). Within this region, Haiti, the
larger Caicos Is., and many of the Virgin Is. have not yet been satisfactorily surveyed
for termites (Fig.1) and may support populations of N. mona. Surveys of Cuba, Ja-
maica, the Bahamas, and the Lesser Antilles have not yielded collections of N. mona
(Scheffrahn & Krecek, unpubl. data). A record of N. mona from Barbados by Bennett
& Alam (1985) is almost certainly based on misidentification. A Neotermes species
listed as "nr. mona" from Cuba (Scheffrahn et al. 1994) is now recognized to be new
species (Krecek & Scheffrahn, unpubl. data).
Neotermes mona is the largest kalotermitid in the West Indies. Snyder (1959) men-
tions that the alate ofN. araguaensis Snyder, comparable in size with N. mona, is the
largest Neotermes in the New World. At 22, the maximum number of antennal articles
for the N. mona imago exceeds Krishna's (1961) diagnoses of_ 21 antennal articles for
the Neotermes and the Kalotermitidae.
Dispersal flights of N. mona are nocturnal. On several occasions, JK observed
alates flying around lights between 0100 and 0200 hours in the Dominican Republic
and Guana Is. in October and December. Compared with some Kalotermitidae (e.g.,
Cryptotermes and Incisitermes), the alates of N. mona exhibited robust flight behavior
and lacked the tendency to shed wings shortly after alighting.
Neotermes mona is usually a coastal inhabitant where it colonizes substantial
woody growth of dry littoral forests, including arboreal cacti and mangroves. This spe-
cies has also been collected from wood in service (Wolcott 1948, Scheffrahn et al.
1990), however its economic significance appears limited. Galleries of N. mona infes-
tations occasionally extend into the xylem elements of living trees or near the tidal
zone of dead mangrove trunks; possibly as moisture refugia during drought. Collins
et al. (1997) rankN. mona as the termite species from the British Virgin Is. having the
greatest moisture requirements. Their ranking was based on climatological factors of
the 19 islands surveyed and not on experimental data. We contend that, contrary to
the rankings of Collins et al. (1997), N. mona has a substantially lower moisture re-

Krecek et al.: Neotermes mona Redescription

quirement than sympatric termite species in the families Rhinotermitidae and Ter-
mitidae that have soil access.
In a study of the phylogeny of 10 kalotermitid species, Luykx et al. (1990) selected
13 morphological characters (7 imago and 6 soldier) for cladistic analysis. In their
data matrix, the eye pigmentation character for the N. mona soldier was erroneously
scored as being absent, when it is indeed heavily pigmented. As a result, there are no
morphological differences for the selected characters among N. mona and its prima-
rily allopatric congeners, N. jouteli and N. luykxi (Nickle and Collins). Therefore, the
morphological cladogram of Luykx et al. (1990) must be revised to show these three
Neotermes as sister species.


The authors thank James A. Chase, John R. Mangold, and Julian de la Rosa for
their unyielding efforts to collect termites in the West Indies. We also thank Diann
Achor at the University of Florida, Lake Alfred Citrus Research and Education Cen-
ter, for assisting with scanning electron microscopy; P. D. Perkins, Museum of Com-
parative Zoology, Harvard University, Cambridge, MA, for the loan of the N. mona
paratype, and F.W. Howard and T. Weissling for their critical reviewing of this manu-
script. Florida Agricultural Experiment Station Journal Series No. R-07236.


BANKS, N. 1919: Antillean Isoptera. Bull. Mus. Comp. Zool. 26: 475-489, 2 pls.
BANKS, N., AND T. E. SNYDER. 1920. A revision of the Nearctic termites with notes on
biology and geographic distribution. Bull. U.S. Nat. Hist. Mus. Washington 108:
BENNETT, F. D., AND M. M. ALAM. 1985. An annotated check-list of the insects and al-
lied terrestrial arthropods of Barbados. Caribbean Agric. Res. and Develop.
Inst. 81 pp.
COLLINS, M. S., M. I. HAVERTY, AND B. L. THORNE. 1997. The termites (Isoptera: Kal-
otermitidae, Rhinotermitidae, Termitidae) of the British Virgin Islands: Distri-
bution, moisture relations, and cuticular hydrocarbons. Sociobiology 30: 63-76.
JONES, S. C. 1993. Termites (Isoptera: Kalotermitidae) of Mona Island: a preliminary
report. Acta Cientifica 5: 73-75.
KRISHNA, K. 1961. A generic revision and phylogenetic study of the family Kaloter-
mitidae (Isoptera). Bull. American Mus. Nat. Hist. 122: 303-408.
LUYKX, P., D. A. NICKLE, AND B. I. CROTHER. 1990. A morphological, allozymic, and
karyotypic assessment of the phylogeny of some lower termites (Isoptera: Kal-
otermitidae). Proc. Entomol. Soc. Washington 92: 385-399.
RAMOS, J. A. 1946. The insects of Mona Island (West Indies). J. Agric. Univ. Puerto
Rico 30: 1-74.
SCHEFFRAHN, R. H, N.-Y. SU, Y., AND B. DIEHL. 1990. Native, introduced, and struc-
ture-infesting termites of the Turks and Caicos Islands, B.W.I. (Isoptera: Kalo-
termitidae, Rhinotermitidae, Termitidae). Florida Entomol. 73: 622-627.
1994. Termites (Isoptera: Kalotermitidae, Rhinotermitidae, Termitidae) of the
West Indies. Sociobiology 24: 213-238.
SNYDER, T. E. 1949. Catalog of the termites (Isoptera) of the world. Smithsonian Mis-
cellaneous Collections 112. 490 pp.
SNYDER, T. E. 1959. New termites from Venezuela, with keys and a list of the described
Venezuelan species. Am. Midland Naturalist 61: 313-321.
WOLCOTT, G. N., 1948. Isoptera: termites in The insects of Puerto Rico. J. Agric. Univ.
Puerto Rico 32: 62-74.

Florida Entomologist 83(3) September, 2000


1Department of Biology, College of Natural Sciences, Andong National University
Andong, Kyungbuk, 760-749 Korea

2Department of Entomology, National Museum of Natural History
Smithsonian Institution, Washington, D.C. 20560-0165, U.S.A.


The mature larvae ofAltica bicarinata (Kutschera) andA. marevagans Horn, col-
lected in Israel and North America, respectively, are described and illustrated in de-
tail for the first time. Some remarks on their taxonomy and biology are also given
along with some discussion of the state of knowledge of alticine larvae.

Key Words:Altica, larvae, Florida, Israel, Rubus, Oenothera


Las larvas maduras deAltica bicarinata (Kutschera) yA. marevagans Horn, colec-
tadas en Norte America e Israel, estan descritas e ilustradas en detalle por primera
vez. Algunos comentarios sobre su taxonomia y biologia estan incluidos, asi como tam-
bien una discusi6n sobre el estado de conocimiento de las larvas de Alticinae.

Alticinae larvae, including forest and agricultural pests, were studied by many
workers from morphological and biological perspectives. Ogloblin & Medvedev (1971),
Kimoto & Takizawa (1994), and Steinhausen (1994) studied many genera of alticine
larvae taxonomically using the chaetotaxy of the anal plate. Some workers conducted
a variety of studies: Rupertsberger (1894); Sanderson (1902); Henriksen (1927); Reed
(1927); Boving & Craighead (1931); Grandi (1932, 1938); Newton (1933); Anderson
(1938); Paterson (1931, 1943); Dobson (1960);Yano (1963, 1965); Giljarov & Medvedev
(1964); Steinhausen (1966, 1978, 1994); Welch (1972); Westdal & Romanov (1972);
Zaitev & Medvedev (1977); Medvedev & Zaitev (1978); Vig (1989); Lawson (1991); Lee
(1992); Medvedev (1992); Doguet (1994); Lee et al. (1998); but the majority of the stud-
ies have been done recently (i.e., since 1970). For a review of most of these see Stein-
hausen (1996). The present authors describe and illustrate the mature larvae
belonging to two Alticinae species collected in North America and Israel;A. mareva-
gans (Horn) andAltica bicarinata (Kutschera), respectively.
Since studies of chrysomelid larvae conducted by Henriksen (1927), Boving (1927,
1929), and Boving and Craighead (1931) have indicated that there are very few de-
tectable differences between larvae of Galerucinae and Alticinae (see also Marshall
1980 and Lawrence & Britton 1991, but see Lawson 1991 showing differences), these
two largest chrysomelid subfamilies often have been treated together when discuss-
ing the larvae. Boving (1927), apparently following the classification scheme of Leng
(1920), suggested that if the Diabroticini and Phyllobroticini (whose larvae were eas-

Lee & Furth: Altica Larvae

ily separated from the rest of the Galerucinae) were removed from the subfamily Gal-
erucinae and placed with the Systenini, Crepidoderini, and Psylliodini of the
subfamily Halticinae, then it would be possible to separate the rest of the larvae into
the traditional two subfamilies of Galerucinae and Alticinae as with the adults. Later,
in order to solve this problem, Boving and Craighead (1931) used a classification with
the family Galerucidae containing three subfamilies: Galerucinae; Diabroticinae
(containing Phyllobrotica Chevrolat); and Halticinae. These larval studies by Boving
were done with the material available at the U.S. National Museum, which at that
time contained 6 of the 12 Galerucinae tribes. The studies of the Alticinae larvae were
apparently done using only 14 genera, primarily of species from the U.S.A. but with a
few from Denmark.
Boving (1927) said that the Halticini [containingAltica Geoffroy] were well known
because of the various publications by W. C. Woods and considered typical of all other
Halticinae. However, Boving (1927) said that "in general aspect and structural details
the Halticini larvae are more similar to the main bulk of the Galerucinae larvae than
these latter [Galerucinae] are to the Diabroticini and Phyllobroticini larvae and more
than the Halticini larvae themselves are to the Halticinae tribes Systenini, Crepi-
doderini, and Psylliodini." Unlike most other Alticinae (root feeders), Altica [Halti-
cini-sensu Leng (1920) and Boving (1927)] are external leaf feeders and are generally
quite easy to rear. Therefore, it is interesting that it is only recently that most of the
western Palearctic species of Altica larvae have been described (Bartkowska & War-
chalowski 1978, Steinhausen 1994, 1996). The percentage of described larvae in
North American is much smaller. In the western Palearctic Region, only 19% of the Al-
ticinae larvae are known (Steinhausen, 1996), and many of these are either external
leaf feeders, leaf miners (e.g., Argopus Fisher, Dibolia Latreille, Mantura Stephens,
Sphaeroderma Stephens, etc.-relatively easy to rear) or species of significant agri-
cultural importance (e.g., Phyllotreta Chevrolat, Psylliodes Latreille, etc.).


All specimens used in this study were collected by the second author and pre-
served in 70% ethyl alcohol. The final instar larvae were macerated in KOH solution
for 30 minutes, rinsed in water, and dissected under an Olympus stereoscopic micro-
scope. For morphological studies of the minute structures, the parts were mounted on
slides and observed through the compound microscope (Leitz). The terminology of se-
tae in this study is adopted from Anderson (1947). Biological observations were made
by the second author in the field or in captivity during rearing activities. Voucher
specimens of the larvae of both species have been deposited in the National Museum
of Natural History, Washington, D.C.


Five North American species ofAltica larvae (A. bimarginata Say,A. corni Woods,
A. rosae Woods,A. torquata LeConte andA. ulmi Woods) were studied by Woods (1917,
1918). Urban (1928) only superficially described H. lythri Aub6 and H. brevicollis Fou-
dras, without any illustrations. Altica bimarginata was briefly illustrated by Boving
& Craighead (1931). Paterson (1931) briefly described and illustrated Haltica lythri.
Haltica cuprea Jacoby was also described by Paterson (1943).Altica chalybea Illiger
was briefly described and illustrated by Peterson (1960).Altica chalybea andA. corni
were illustrated by Lawson (1991). Oglobin & Medvedev (1971) made a key to two
Palearctic species: Haltica oleracea (Linnaeus) and H. tamaricis Schrank). Bart-

Florida Entomologist 83(3) September, 2000

kowscka & Warchalowski (1978) made a taxonomic key to 9 European species.
Medvedev & Zaitzev (1978) illustrated only 1 species (Altica oleracea), using dorsal
tubercles of the abdominal segments. Phillips (1977) discussed color changes, phenol-
ogy, and morphology of spined tubercles ofA. lythri. Phillips (1979) only very briefly
described the color of A. ericeti (Allard), A. lythri, A. oleracea, and A. palustris Weise.
Two Japanese species,Altica caerulescens (Baly) andA. cirsicola Ohno, were fully de-
scribed and illustrated by Lee (1992). Only the dorsal part of the tubercles of 9 Euro-
pean species were illustrated by Steinhausen (1994). Kimoto & Takizawa (1994, 1997)
illustrated and provided keys for 12 eastern Palearctic species. Most significant larval
studies of species of Altica were done only recently (1978 and since).


Altica bicarinata (Kutschera)
(Figs. 1-10)

Mature larva (Fig. 1). Body blackish brown, nearly straight, elongate, micro-sculp-
tured and densely covered with setae; head and mandibles dark brown, pronotum, tu-
bercles, spiracles and legs pale brown; head (Fig. 3) hypognathous, rounded, slightly
sclerotized; frontal suture narrowly divergent and straight; hind corner of epicranium
slightly produced; coronal suture short; epicranium with 10 pairs of dorsal setae (4
pairs are minute), 2 pairs of lateral setae and 1 pair of dorsal sensilla; frons with 3
pairs of frontal setae and 1 pair of frontal sensilla; endocarina distinct for full length;
epistomal suture developed; stemmata absent; antenna (Fig. 2) 2-segmented, seg-
ment 1 with a large conical sensory papilla, 2 setae and 2 sensilla, segment 2 with 4
setae; clypeus with 3 pairs of clypeal setae; labrum slightly incised in the middle of
anterior margin, with 2 pairs of labral setae and 1 pair of labral sensilla; epipharynx
with 6 pairs of epipharyngeal setae; hypopharynx with many hypopharyngeal
spinules; mandible (Fig. 4) palmate, well sclerotized, with 5 distal teeth, 2 mandibular
setae, 1 sensillum and 3 sharp penicilli; maxillary palp (Fig. 8) 3-segmented, segment
1 without seta, segment 2 with 2 setae and 1 sensillum, segment 3 with 2 setae and 1
sensillum; palpifer with 3 setae; stipes with 2 setae; cardo with 1 seta; galea with 8
setae; lacinia with tightly bunched group of 8 setae located behind galea; labial palp
2-segmented; prementum and postmentum separated by sclerotized membrane; pre-
mentum with 4 pairs of setae and 1 pair of sensilla, postmentum with 4 pairs of setae
and 1 pair of sensilla; pronotum (Fig. 6) brown, well sclerotized, with 14 pairs of setae
and 5 pairs of sensilla; mesothoracic spiracles (Fig. 9) annuliform, situated on epipleu-
ral anterior part, with peritreme strongly sclerotized; epipleuron with 3 setae; legs
(Fig. 10) rather long and slender; tibia with 7 setae; tarsungulus falciform, slightly
curved anteriorly, enlarged base with 1 seta; pulvillus whitish, bladder-like; typical
abdominal segments with two folds; epipleuron with 2 setae; abdomen with 8 pairs of
spiracles; anal plate (Fig. 7) with 7 pairs of setae and 1 pair of sensilla; pygopod (Fig.
1) well developed.
Body length: 7.8 mm (number examined: n = 5). Head width: 0.9 mm (n = 5).
Materials examined. Israel: Golan Heights, Qusbiye, 28 April 1974, larvae col-
lected on the leaves of the perennial edible raspberryRubus sanctus Schreb. (= R. san-
guineus Frivaldszk), and determined by association with adults by D. Furth.
Remarks. The larva of this species closely resembles that ofAltica lythri treated by
Paterson (1931), but differs by having the mandible with a well-developed penicillum.
The cephalic setae are arranged similarly to those of other alticine larvae, but in this
species there are 10 pairs (4 pairs are minute) of setae on the vertex. Voucher speci-
mens will be deposited in the National Museum of Natural History, Washington, D.C.

Lee & Furth: Altica Larvae

Figs. 1-10.Altica bicarinata: (1) mature larva, lateral view; (2) antenna, dorso-lat-
eral view; (3) head, anterior view; (4) mandible, buccal view; (5) clypeus, labrum and
epipharynx, frontal view; (6) pronotum, dorsal view; (7) anal plate, dorsal view; (8)
lower mouth parts, ventral view; (9) spiracle, lateral view; (10) left hind leg, lateral
view. Scale line-1.0 mm (Fig. 1); 0.5 mm (Figs. 3, 7, 8); 0.1 mm (Figs. 2, 4, 5, 6, 9, 10).

Florida Entomologist 83(3) September, 2000

Biological Notes. Adults were collected during all months of the year in Israel
(Furth 1981). The second author has examined adult specimens from the Nile River
area in Egypt; however, otherwise,A. bicarinata is not known from west of central Is-
rael. This species is the most commonAltica in the eastern Mediterranean area and
is monophagous feeding exclusively on Rubus sanctus. The food plant species is found
around the Mediterranean as well as in Iran and Iraq occurring on the banks of or
near rivers, streams, springs, swamps, etc. It has crossed with many other species of
Rubus (Zohary et al. 1980). In modern times, this plant is extremely rare and relictual
(high mountains) in the southern desert areas of Israel and Egypt (Negev and Sinai);
however, presumably in the more moist times of the late Pleistocene it was more com-
mon and widespread across these areas (for further details see Furth 1981). Interest-
ingly, at the Greek Orthodox monastery (Santa Katarina) on Mt. Sinai (Egypt), R.
sanctus is the plant indicated as the supposed "burning bush" from the biblical story
of Moses receiving the Ten Commandments on Mt. Sinai.
Larvae were collected by the second author in Israel from 6 March through 31 May
and in Cyprus on 31 August. Eggs eclose in 7-12 days, and larvae feed (skeleonize) on
both leaf surfaces. Larvae feed for about 30 days, then pupate in the soil beneath their
host. Adults eclose in about 7 days (i.e., complete life cycle [egg to adult] is approxi-
mately 40-50 days). There are two to three generations per year in Israel. Food plant
preference testing indicated thatA. bicarinata is monophagous on R. sanctus (Furth
1981). Adults and larvae could often be found together in large numbers, especially in
March and April, and to a lesser extent May. Large localized areas of the food plants
were often completely defoliated and appeared "burned", but actually the leaves were
skeletonized and dried out or desiccated.
Another interesting phenomenon concerning A. bicarinata concerns its predator
Zicrona coerulea (Linnaeus) (Hemiptera: Pentatomidae), which is widespread
throughout the Palearctic Region. The nymphs and adults of this true bug are well-
known to attack the larvae of a variety of species of Altica as well as some other prey.
The adult bug is exactly the same metallic blue/green color as the adults of Altica.
This phenomenon may have evolved as some sort of Batesian mimicry (see Furth 1981
and Furth 1983 for a detailed discussion).

Altica marevagans Horn
(Figs. 11-20)

Mature larva (Fig. 11). Body blackish brown, slightly curved, micro-sculptured
and with dorsal parts of body, except head covered with club-shaped setae; head and
mandibles dark brown, pronotum, tubercles, spiracles and legs brown; head (Fig. 13)
hypognathous, rounded, slightly sclerotized; frontal suture somewhat divergent and
straight; hind corner of epicranium slightly produced; coronal suture short; epicra-
nium with 8 pairs of dorsal setae (4 pairs of them minute), 3 pairs of lateral setae and
1 pair of dorsal sensilla; frons with 3 pairs of frontal setae and 1 pair of frontal sen-
silla; endocarina distinct for full length; epistomal suture developed; stemmata ab-
sent; antenna (Fig. 12) 2-segmented, segment 1 with a large conical sensory papilla,
2 setae, segment 2 with 4 setae; clypeus (Fig. 15) with 3 pairs of clypeal setae and 1
pair of clypeal sensilla; labrum slightly incised in the middle of anterior margin, with
2 pairs of labral setae and 1 pair of labral sensilla; epipharynx with 6 pairs of epipha-
ryngeal setae; hypopharynx with many hypopharyngeal spinules; mandible (Fig. 14)
palmate, well sclerotized, with 5 distal teeth, 2 mandibular setae and 2 short penicilli;
maxillary palp (Fig. 18) 3-segmented, segment 1 without seta, segment 2 with 2 setae
and 1 sensillum, segment 3 with 2 setae and 1 sensillum; palpifer with 3 setae; stipes

Lee & Furth: Altica Larvae



Figs. 11-20. Altica marevagans: (11) mature larva, lateral view; (12) antenna,
dorso-lateral view; (13) head, anterior view; (14) mandible, buccal view; (15) clypeus,
labrum and epipharynx, frontal view; (16) pronotum, dorsal view; (17) anal plate, dor-
sal view; (18) lower mouth parts, ventral view; (19) spiracle, lateral view; (20) left hind
leg, lateral view. Scale line-1.0 mm (Fig. 11); 0.5 mm (Figs. 13, 17, 18); 0.1 mm (Figs.
12, 14, 15, 16, 19, 20).

Florida Entomologist 83(3) September, 2000

with 2 setae and 1 sensillum; cardo with 1 seta; galea with 8 setae; lacinia with tightly
bunched group of 8 setae located behind galea; labial palp 2-segmented; prementum
and postmentum separated by sclerotized membrane; prementum with 4 pairs of se-
tae and 2 pairs of sensilla, postmentum with 3 pairs of setae and 1 pair of sensilla;
pronotum (Fig. 16) brown, well sclerotized, with 12 pairs of setae (8 pairs long and
club-shaped); mesothoracic spiracles (Fig. 19) annuliform, situated on epipleural an-
terior part, peritreme well sclerotized; epipleuron with 3 setae; legs (Fig. 20) rather
long and slender; tibia with 7 setae; tarsungulus falciform, slightly curved anteriorly,
enlarged base with 1 seta; pulvillus whitish, bladder-like; typical abdominal seg-
ments with two folds; epipleuron with 2 setae; abdomen with 8 pairs of spiracles; anal
plate (Fig. 17) with 7 pairs (6 pairs long and club-shaped) of setae and 1 pair of sen-
silla; pygopod well developed.
Body length: 6.5 mm (n = 5). Head width: 0.7 mm (n = 5).
Materials examined. U.S.A.: Florida, Lido Beach, 16 April1975, larvae collected on
the leaves of Oenothera humifusa Nutt. and determined by association with adults by
D. Furth.
Remarks. The larva of this species closely resembles that ofAltica cirsicola as de-
scribed by Lee (1992), but is different in the following characters: mandibles with pen-
icillus, frons with 3 pairs of setae and prementum with 4 pairs of setae.
Biological Notes. Larvae were collected on the following dates in the Sarasota,
Florida area: 16 April 1975; 9 May 1981. Adults were collected on dates given in Flow-
ers et al. (1994). Larvae were found co-occurring with adults in April and May. The
primary food plant (Oenothera humifusa) has succulent leaves, is low-growing and is
commonly found along beaches even in areas disturbed by active public recreation.


The first author was supported in part by a grant from Andong National Univer-
sity, Korea.


ANDERSON, W. H. 1938. Description of the larvae of Chaetocnema denticulata (Illiger)
and Chaetocnema pulicaria Melsheiner (Col., Chrysomelidae). Proc. Ent. Soc.
Washington 40(6): 161-169.
ANDERSON, W. H. 1947. A terminology for the anatomical characters useful in the tax-
onomy of the weevil larvae. Proc. Ent. Soc. Washington 49: 123-132.
BARTKOWSKA, J., AND A. WARCHALOWSKI. 1978. Studien an Blattkafer-Larven. I. Gat-
tung Haltica Koch (Coleoptera, Chrysomelidae). Polskie Pismo Ent. 48: 583-
BOVING, A. G. 1927. Descriptions of larvae of the genera Diabrotica and Phyllobrotica,
with a discussion of the taxonomic validity of the subfamilies Galerucinae and
Halticinae (Coleoptera: Chrysomelidae). Proc. Ent. Soc. Washington 29(9): 193-
BOVING, A. G. 1929. Beetle larvae of the subfamily Galerucinae. Proc. U.S. Nat. Mus.
75(2): 1-48.
BOVING, A. G., AND F. C. CRAIGHEAD. 1931. An illustrated synopsis of the principal
larval forms of the order Coleoptera. Ent. Americana 11(N.S.)(1): 1-351.
DOBSON, R. M. 1960. The immature stages of the flea beetles Psylliodes cuprea (Koch)
and Psylliodes chrysocephala (L.) (Col., Chrysomelidae). Ent. Month. Mag.
XCVI: 1-4.
DOGUET, S. 1994. Faune de France. No. 80. Coleopteres Chrysomelidae 2, Alticinae.
694 pp.

Lee & Furth: Altica Larvae

FLOWERS, R. W., D. G. FURTH, AND M. C. THOMAS. 1994. Notes on the distribution and
biology of some Florida Leaf Beetles (Coleoptera: Chrysomelidae). Coleopt.
Bull. 48(1): 79-89.
FURTH, D. G. 1981. Altica of Israel (Coleoptera: Chrysomelidae: Alticinae). Israel J.
Ent. 14 (1980): 55-66.
FURTH, D. G. 1983. Flea beetle mimicry. Chrysomela No. 3: 3.
GILJAROV, M. C., AND L. N. MEDVEDEV. 1964. Familie Chrysomelidae-Blattfresser. Pp.
507-530 In Bestimmung der im Boden lebenden Larven der Insekten. Nauka
Moskau (in Russian).
GRANDI, G. 1932. Morfologia ed etologia comparata dei Insetti a regime specializzato
II. La morfologia delle larve minatrici di due Coleotteri delle sottofamiglia degli
Alticini. Mem. R. Accad. Sci. Ist. Bologna (VIII) IX: 3-10.
GRANDI, G. 1938. Morfologia ed etologia comparata dei Insetti a regime specializzato
XIV. La morfologia e l'etologia delle larve minatrici di due Coleotteri Crisome-
lidi delle tribu degli Alticini. Mem. R. Accad. Sci. Ist. Bologna (IX) V: 165-170.
HENRIKSEN, K. 1927. VII. Bladbiller, Laverne. In Hansen, V. Danmarks Fauna 31:
290-376. Kobenhaven, G.E.C.
KIMOTO, S., AND H. TAKIZAWA. 1994. Leaf Beetles (Chrysomelidae) of Japan. Tokai
Univ. Press, 539 pp. (in Japanese).
KIMOTO, S., AND H. TAKIZAWA. 1997. Leaf Beetles (Chrysomelidae) of Taiwan. Tokai
Univ. Press, 581 pp. (in Japanese).
LAWRENCE, J. L., AND E. B. BRITTON. 1991. Coleoptera. Pp. 543-683 In Insects of Aus-
tralia. Volume 2. Cornell University Press, Ithaca.
LAWSON, F. A. 1991. Chrysomelidae. In F. W. Stehr (ed.). Immature Insects. Kendall-
Hunt Publishing Co., Dubuque. Volume 2: 568-585.
LEE, J. E. 1992. Larval description of four alticine species of generaAltica andArgo-
pistes from Japan (Coleoptera; Chrysomelidae). Korean J. Ent. 22(4): 287-295.
LEE, J. E., S. W. LINGAFELTER, AND A. S. KONSTANTINOV. 1998. Larval morphology of
Systena blanda Melsheimer (Coleoptera: Chrysomelidae: Altcinae). Proc. Ent.
Soc. Wash. 100(3): 484-488.
LENG, C. W. 1920. Catalogue of the Coleoptera of America, North of Mexico. J. D. Sher-
man, Mt. Vernon, NY. 479 pp.
MARSHALL, J. E. 1980. A key to some larvae of the British Galerucinae and Halticinae
(Coleoptera: Chrysomelidae). Ent. Gaz. 31: 275-283.
MEDVEDEV, L. 1992. Systematics and Ecology of Insects from Vietnam. Moscow, Sci-
ence. 185 pp. (in Russian).
MEDVEDEV, L. N., AND Y. M. ZAITZEV. 1978. Leaf beetle larvae of Siberia and Far East.
183 pp. Nauka, Moscow (in Russian).
NEWTON, H. C. F. 1933. On the biology of some species of Longitarsus (Col., Chry-
somelidae) living on ragwort. Bull. Ent. Res. 24: 511-520.
OGLOBLIN, D. A., AND L. N. MEDVEDEV. 1971. The larvae of the leaf beetles (Coleop-
tera, Chrysomelidae) of the European part of the USSR. Isdat. Nauka Opred.
Faune SSSR, Leningrad. 124 pp. (in Russian).
PATERSON, N. F. 1931. Studies on the Chrysomelidae II. The bionomics and morphol-
ogy of certain Chrysomelidae. Proc. Zool. Soc. London, 42: 879-949.
PATERSON, N. F. 1943. Early stages of two species of Halticinae (Coleoptera; Chry-
somelidae). Ent. Soc. S. Africa. 6: 29-36.
PETERSON, A. 1960. Larvae of Insects. Part II. 461 pp. Columbus, OH.
PHILLIPS, W. M. 1977. Observations on the biology and ecology of the chrysomelid ge-
nus Haltica Geoff. in Britain. Ecol. Ent. 2: 205-216.
PHILLIPS, W. M. 1979. A contribution to the study of species relations within the chry-
somelid genus Altica Muller in Britain. Zool. J. Linnean Soc. 66: 289-308.
REED, H. 1927. Some observations on the leaf-mining flea-beetle Dibolia borealis
Chevrolat. Ann. Ent. Soc. Amer. 20: 540-549.
RUPERTSBERGER, M. 1894. Die biologische Literatur aber die Kafer Europas von 1880
an. Mit Nachtragen aus fruherer Zeit. Linz: 246-261.

Florida Entomologist 83(3) September, 2000

SANDERSON, E. D. 1902. Notes upon the structure and classification of Chrysomelid-
larvae. Proc. Ent. Soc. Wash. 5: 21-30.
STEINHAUSEN, W. R. 1966. Vergleichende Morphologie des Labrum von Blattkaferlar-
ven. Dtsch. Ent. Z. (N.F.) 13 (IV/V): 313-322.
STEINHAUSEN, W. R. 1978. Bestimmungstabelle fur die Larven der Chrysomelidae
(partim). In B. Klausnitzer. Ordnung Coleoptera (Larven). Akad. Verlag Berlin
10: 336-343.
STEINHAUSEN, W. R. 1994. 116. Family: Chrysomelidae. In B. Klausnitzer (ed.). Die
Larven der Kafer Mitteleuropas. Goecke & Evers, Bd. 2: 231-314 (in German).
STEINHAUSEN, W. R. 1996. Status of west Palearctic Leaf Beetle larvae research. Pp.
65-91 in P. Jolivet and M. Cox (eds.). Chrysomelidae Biology, Vol. 3. SPB Aca-
demic Publishing, Amsterdam.
URBAN, C. 1928. Aus dem Leben einiger einheimischer Haltica-Arten. Koleopterol.
Rundsch. 14(4): 151-157.
VIG, K. 1989. The morphology of Phyllotreta larvae (Col., Chrysomelidae) and deter-
mination keys to larvae of more common species. N6v6ny-v6delen 25(9): 412-
419 (in Hungarian).
WELCH, C. 1972. The biology of Hermaeophaga mercurialis F. (Coleoptera, Chry-
somelidae). Ent Gaz. 23(3): 153-166.
WESTDAL, P. H., AND W. ROMANOV. 1972. Observations on the flea-beetle Phyllotreta
cruciferae (Coleoptera: Chrysomelidae). Manitoba Ent. 6: 35-45.
WOODS, W. C. 1917. The biology of the Alder flea-beetle. Maine Agric. Exp. Stat. Bull.
265: 249-284.
WOODS, W. C. 1918. The biology of the Maine species ofAltica. Maine Agric. Exp. Stat.
Bull. 273: 149-204.
YANO, T. 1963. Coleopterous leaf-miners of Japan. VI. The larvae of Mantura clavar-
eui Heikertinger (Chrysomelidae). Trans. Shikoku Ent. Soc. 8(1): 19-22.
YANO, T. 1965. Larval stages of the leaf-miners found in Shikoku (Coleopterous leaf-
miners of Japan). Trans. Shikoku Ent. Soc. 8(4): 115-132.
ZAITZEV, Y. M., AND L. N. MEDVEDEV. 1977. Some chrysomelid-beetle larvae from
Mongolia (Coleoptera: Chrysomelidae). Ins. Mongol. 5: 353-371 (in Russian).
ZOHARY, M., C. C. HEYN, AND D. HELLER. 1980. Conspectus Florae Orientalis. An an-
notated catalogue of the flora of the Middle East. Fascicle 1 (Papaverales and
Rosales). Israel Academy of Sciences and Humanities, Jerusalem. 107 pp.

Florida Entomologist 83(3) September, 2000


University Of Kentucky, Department Of Entomology


A new species of braconid (Hymenoptera: Braconidae: Agathidinae), Bassus ag-
athoides, is described from Florida. The phylogenetic position of this species within
Bassus is discussed. It is a parasitoid of Samea multiplicalis (Guenee) (Lepidoptera:
Pyralidae), a natural control agent of waterlettuce, Pistia stratiotes L. (Araceae), an
important aquatic weed. Characters to distinguish B. agathoides from other species of
Nearctic Bassus are provided.

Newton & Sharkey:A New Species of Bassus

Key Words: Weed biological control, systematics, tritrophic interactions,Agathis


Una nueva especie de braconido (Himen6ptera: Braconidae: Agathidinae), Bassus
agathoides, es descrita de la Florida. La posici6n filogenetica de esta especie dentro de
Bassus es discutida. Es un parasito de Samea multiplicalis (Guenee) (Lepid6ptera:
Pyralidae), un agent de control natural de la lechuga de agua, Pistia stratiotes L.
(Araceae), una maleza acuatica important. Caracteristicas para distinguir a B. aga-
thoides de otras species de Bassus cercanos al artico son proveidos.

In Florida, Samea multiplicalis (Guenee) (Lepidoptera: Pyralidae) is the most
common natural control agent associated with waterlettuce, Pistia stratiotes L.
(Araceae), an aquatic weed that can affect navigation and flood control. S. multiplica-
lis is native to South America and the southeastern United States, and its larval
stages can cause severe damage to P. stratiotes (Wheeler and Halpern 1999).
Knopf and Habeck (1976) reared several parasitoids from specimens of S. multi-
plicalis from Florida, including three ichneumonoids and a tachinid fly. Further re-
search by G.S. Wheeler (USDA/ARS Aquatic Weed Research Unit, Ft. Lauderdale, FL,
personal communication) suggests that the most common parasitoid attacking Flor-
ida populations of S. multiplicalis is a braconid wasp in the subfamily Agathidinae. In
their hearings, 13.4% of 732 S. multiplicalis larvae were consumed by this agathidine
wasp. This was also one of the parasitoids reared from S. multiplicalis by Knopf and
Habeck (1976). Although Knopf and Habeck placed the parasitoid in the Agathidinae,
they referred to it as an undetermined species of Agathis. Our study describes and
names the parasitoid, and briefly discusses the phylogenetic relationship of this spe-
cies with other Agathidinae species. It is hoped that this taxonomic study will facili-
tate future research into this economically important tritrophic system.


The description is of the holotype female with variations in parentheses. Morpho-
logical terminology follows Sharkey (1996) and Sharkey and Wharton (1997).


Bassus agathoides, NEW SPECIES

HOLOTYPE FEMALE: Length. 4.73 mm (females 3.63-4.73 mm, males 3.80-4.53
Color. Flagellomeres (with antennae directed anteriorly) dark brown dorsally, fad-
ing to dark orange ventrally (ventrally ranging from entirely black to yellow); anterior
orbit of eye black, the posterior orbit orange (ranging to entirely black); mouthparts
pale yellow with black highlights, remainder of head black dorsally with orange
patches laterally (ranging from entirely black to mostly orange with dark highlights);
fore leg orange with tarsus darkened distally; middle leg orange with tibia darkened
distally, tarsomeres mostly dark; hind coxae dark orange (ranging to nearly black, es-
pecially in males); hind femur dark orange (ranging to black with some orange, espe-
cially in males); basal black band present on hind tibia; hind tibia black in distal half,
otherwise orange; wings clear; mesosoma black with orange tegula (ranging from

Florida Entomologist 83(3) September, 2000

black with black tegula to black with orange highlights, often with an orange spot on
the mesopleuron); metasoma pale yellow ventrally (ranging to dark orange); with ter-
gum 1 entirely black, tergum 2 black in the posterior half and orange anteriorly (or
black with only the anterior margin orange), tergum 3 black with orange posterior
margin, remaining terga orange with dark highlights.
Head. Number of flagellomeres = 29 (27-30); ratio, distance between ocellus and
compound eye to distance between lateral ocelli = 1.2 (1.1-1.3); temple not bulging as
viewed dorsally (tmin, Fig. 1); ratio, malar space to eye height = 0.57 (0.50-0.69); gena
rounded posteroventrally (ge, Fig. 2); median ridge connecting face and median ocel-
lus present between antennae (mr, Fig. 1); interantennal space raised to converge on
single point anteromedially (Fig. 1); antennal depressions shallow (ad, Fig. 1); median
line between antennae without leather-like coriarus sculpturing.
Mesosoma. Bump of propleuron absent; notaulus crenulate to punctate (ranging to
weakly punctate) along entire length (na, Fig. 3); posterior semicircular depression of
scutellum absent; posterior transverse ridge of scutellum weak (pr, Fig. 3), or absent;
posterior surface of scutellum rugose (ps, Fig. 3); metapleuron granulate (mp, Fig. 4);
propodeum evenly rugose (Fig. 5); propodeal psuedosternite without strong trans-
verse carina; hind coxal cavity closed, with a complete sclerite separating it from
metasomal cavity (cc and mf, Fig. 6); ratio, distance between the hind coxal cavity and
metasomal foramen to the diameter of the hind coxal cavity = 0.22-0.26 (hind leg not
removed from holotype).
Legs. Ratio, hind femur length to hind femur width = 4.20 (3.82-4.49); spines of
foretibia absent; hind tibia with eight spines (6-10) (sp, Fig. 7); basal lobe of tarsal
claws absent (tc, Fig. 8).
Wings. Last abscissa of RS vein of forewing weakly curved (RS, Fig. 12); basal por-
tion of free distal abscissa of CU of hind wing tubular to nebulus (Cu, Fig. 12).
Metasoma. Pair of longitudinal carinae on median tergum 1 absent; ratio, length
of median tergum 1 to apical width of median tergum 1 = 1.04 (1.03-1.18); median
terga 1, 2, and 3 granulate (Figs. 9, 10, and 11); ratio, length of ovipositor to length of
metasoma = 1.64 (1.49-1.88); ovipositor sheath narrower than apex of tibia.
Hosts and Biology. First instar caterpillars of Samea multiplicalis (Guenee)
(Pyralidae) are attacked and the parasitoid pre-pupa emerges from the last larval in-
star of the host (Wheeler, unpublished data).
Etymology. The specific epithet refers to the fact that this species has several fea-
tures that are convergently present in many species ofAgathis (see Discussion).
Material Examined. Holotype Female, USA: Florida: Palm Beach Co., 27-IX-96
(United States National Museum). Paratypes, USA: Florida: 16 females, 12 males,
Palm Beach Co., 27-IX-96. 8 females, 3 males, Marion Co., 18-IX-96 (paratypes are de-
posited in the United States National Museum, the Florida State Collection of Arthro-
pods, Gainesville, FL, and The Insect Collection of the Department of Entomology at
the University of Kentucky).


Bassus agathoides has several features that are sometimes associated with spe-
cies of Agathis and other Agathidini. These include a somewhat narrow, rostriform
face (Fig.2), and the lack of a basal lobe on the tarsal claws (tc, Fig. 8) (Sharkey 1991).
The wide sclerite between the hind coxal cavities and the metasomal foramen (cc and
mf, Fig. 6) as well as the reduced size of the third labial palpomere (pa, Fig. 2) pre-
clude placement in the genusAgathis (Sharkey 1991). The complex rugose propodeal
sculpturing exhibited by B. agathoides (Fig. 5) is not found in species of Agathis, but
is common in Bassus species (Sharkey 1991). Similarly, the distinct granulate sculp-

Newton & Sharkey:A New Species of Bassus

Fig. 1. Dorsal aspect of head; tm = temple; ad = antennal depression; mr = median
ridge. Fig. 2. Lateral aspect of head; ge = gena; pa = third palpomere. Fig. 3. Dorsal
aspect of mesonotum; na = notaulus; pr = posterior transverse ridge of scutellum; ps
= posterior surface of scutellum. Fig. 4. Lateral aspect of mesosoma; mp = metapleu-
ron. Fig. 5. Dorsal aspect of propodeum. Fig. 6. Posterior aspect of mesosoma with legs
and metasoma removed; mf = metasomal foramen; cc = coxal cavity.

Florida Entomologist 83(3)

Fig. 7. Apex of third tibia; sp = spines. Fig. 8. Apex of Tarsus; tc = tarsal claw. Fig.
9. Dorsal aspect of median terga 1, 2, and 3. T1 = median tergum 1; T2 = median ter-
gum 2; T3 = median tergum 3. Fig. 10. Dorsal aspect of median tergite 1. Fig. 11. Sur-
face detail of median tergite 1.

September, 2000

Newton & Sharkey:A New Species of Bassus 289

,-_V I" -_^

CU -----------

Fig. 12. Front and hind wings.

turning on metasomal terga 1, 2, and 3 (T1, T2, and T3, Fig. 9) and especially on the
first metasomal tergum (Figs. 10 and 11) is present on many other species of Bassus,
including B. cintus (Cresson), B. discolor (Cresson), and B. agilis (Cresson), but is un-
known inAgathis. Based on outgroup analysis, this granulate sculpturing is a derived
character state within Bassus and may define a monophyletic group.
B. agathoides runs to couplet 36 in the key to the Nearctic species of Bassus
(Muesebeck 1927). The simple tarsal claws, lacking basal lobes (tc, Fig. 8), distinguish
B. agathoides from all Bassus spp. in couplet 36 and following in the Muesebeck's key.


A previous draft of this manuscript was critically reviewed by Dr. Lee Townsend
and Jason Leathers (University of Kentucky). Research was facilitated by Martha
Potts (University of Kentucky) and Dr. Gary Wheeler (USDA/ARS, Ft. Lauderdale,
FL). Support for this investigation was provided by Kentucky Agriculture Experiment
Station project no. 99-08-44.


KNOPF, K. W., AND D. H. HABECK. 1976. Life history and biology of Samea multipli-
calis. Environmental Entomology 5: 539-542.
MUESEBECK, C. F. W. 1927. A revision of the parasitic wasps of the subfamily Bracon-
inae occurring in America north of Mexico. Proceedings of the United States
National Museum. 69: 1-73.
SHARKEY, M. J. 1996. The Agathidinae (Hymenoptera: Braconidae) of Japan. The Bul-
letin of the National Institute of Agro-Environmental Sciences, No. 13. 100 pp.
SHARKEY, M. J., AND R. A. WHARTON. 1997. Morphology and terminology. In R. A.
Wharton, P. M. Marsh, and M. J. Sharkey (eds). Manual of the New World gen-
era of the family Braconidae (Hymenoptera). Special Publication of the Inter-
national Society of Hymenopterists. 1: 1-439.
WHEELER, G. S., AND M. D. HALPERN. 1999. Compensatory responses of Samea mul-
tiplicalis larvae in response to different fertilization levels of the aquatic weed
Pistia stratiotes. Entomologia Experimentalis et Applicata. 92: 205-216.

Florida Entomologist 83(3) September, 2000


1Tropical Research and Education Center, University of Florida, IFAS
18905 SW 280 Street, Homestead, FL 33031

2Fort Lauderdale Research and Education Center, University of Florida, IFAS
3205 College Avenue, Fort Lauderdale, FL 33314


We documented the decline of a 2-hectare Canary Island date palm (Phoenix ca-
nariensis) nursery caused by the palmetto weevil (Rhynchophorus cruentatus) in
Dade County, FL. External palm symptoms were defined, divided into nine categories,
and representative palms were destructively harvested to assess internal weevil as-
sociations. Apparently healthy palms declined and died in a mean of 49 days. At the
beginning of the study, 42% of 950 palms appeared healthy but within seven months
only 3% were alive. Economic losses were estimated at $285,000-$380,000 for the
nursery studied. Palm decline was patchily distributed in the field. The mean palm
weevil counts ranged from 0.3 to 223.3 weevils per palm, for healthy to collapsing
palms, respectively. Twenty-four weevil grubs were sufficient to kill one mature palm.
External symptoms did not allow preventative diagnosis and treatment of internal R.
cruentatus infestations. By the time that external symptoms were unambiguous, the
mean total weevil counts per palm were over 100 with more than 65% as larvae and
more than one quarter of these were >2.5 cm in length. Palms in these categories were
dying because of irreparable damage to their apical meristems and attempts to save
them would have been ineffectual. Thus, phytosanitation (palm removal and destruc-
tion) for management of R. cruentatus in Canary Island date palms should be imple-
mented as soon as host leaves droop and weevil frass is observed. Growers and buyers
of P. canariensis in regions where R. cruentatus exists should be aware of the potential
lethal risk that it poses for this non-native palm. The costs of aggressive phytosanita-
tion at the first symptoms of R. cruentatus infestation and prophylactic pesticide
treatment at times of pruning, stress, or transplanting should be factored into the pre-
dicted cost of production and maintenance of Canary Island date palms in Florida.

Key Words: integrated pest management, palm decline, palmetto weevil, Phoenix


Documentamos el deterioro de un vivero de 2-hectareas de palmas datileras Isla
Canaria (Phoenix canariensis) causado por el gorgojo "palmetto" (Rhynchophorus
cruentatus) en el condado de Miami-Dade, FL. Sintomas externos de la palma fueron
definidos, divididos en nueve categories, y palmas representatives fueron cosechadas
destructivamente para evaluar asociaciones internal de los gorgojos. Palmas aparen-
temente saludables se deterioraron y murieron en un promedio de 49 dias. Al co-
mienzo del studio, 42% de 950 palmas parecian saludables pero dentro de siete
meses solo el 3% estaban vivas. P6rdidas econ6micas fueron estimadas entire
$285.000-$380.000 para el vivero del studio. El deterioro de las palmas en el campo
fue distribuido irregularmente. Cuentas de promedios de gorgojos variaron entire 0,3
y 223,3 gorgojos por palma, de palmas saludables a enfermas, respectivamente. Vein-
ticuatro orugas de gorgojo fueron suficiente para matar una palma madura. Sintomas

Hunsberger et al.: Rhynchophorus cruentatus in Date Palms 291

externos no permitieron diagnosis preventive y tratamientos de infestaciones inter-
nas de R. cruentatus. Para el moment en que los sintomas externos eran inequivocos,
la cuenta promedio total de gorgojos por palma era sobre 100 con mas de 65% en forma
de larva y mas de un cuarto de estos tenfan un largo de > 2,5 cm. Palmas en estas ca-
tegorias morian por dano irreparable a su meristemo apical e intentos para salvarlos
hubieran resultado inutiles. Por lo tanto, fito-saneamiento (la practice de remover y
destruir las palmas) para administraci6n de R. cruentatus en palmas datileras Isla
Canaria deberia ser implementado en cuanto las hojas del huesped se inclinen y se ob-
serven residues de dano de gorgojo. Vendedores y compradores de P. canariensis en re-
giones donde R. cruentatus existe deben estar al tanto de este riesgo potencialmente
letal para esta palma no-nativa. El costo de fito-saneamiento agresivo en los primeros
sintomas de infestaci6n de R. cruentatus y tratamiento de pesticide profilactico en mo-
mentos de podar, estres, o transplant deben de ser facturados en los costs predichos
de producci6n y mantenimiento de palmas datileras Isla Canaria en la Florida.

The palmetto weevil, Rhynchophorus cruentatus (F.) is distributed from the Flor-
ida Keys through the coastal regions of the southeastern U.S. including South Caro-
lina and Texas (Wattanapongsiri 1966) and is co-distributed with the sabal palm,
Sabal palmetto [Walker] Loddiges ex J. A. et Schultes (Woodruff 1967). Although not
considered a major pest of palms, R. cruentatus has recently reached pest status in
several Canary Island date palm (Phoenix canariensis Hortorum ex Chabaud) nurs-
eries in southern Florida. Canary Island date palms are a valuable ornamental crop
in Florida with retail values of $1,000-4,000 per palm (Anonymous 1997) for large
specimen trees.
Adult R. cruentatus were first collected from Canary Island date palms from Flor-
ida in 1907 and recorded from Dade County, FL in 1909 (Wattanapongsiri 1966). The
Canary Island date palm and sabal palm are the most commonly reported hosts for
R. cruentatus (Giblin-Davis & Howard 1989). Other hosts include saw palmetto (Ser-
renoa repens [Bartram] Small), date palm (P. dactylifera [L.]), Pritchardia sp., Wash-
ingtonia sp., royal palms (Roystonea sp.), coconut palm (Cocos nucifera L.), Latania
sp., Caryota sp. (Wattanapongsiri 1966, Giblin-Davis & Howard 1988) and the Florida
thatch palm (Thrinax radiata Loddiges ex J. A. & J. H. Schultes) (A. G. B. H., unpub-
lished data).
Rhynchophorus cruentatus usually attacks transplanted or otherwise stressed
palms but can attack healthy palms. Adults are often attracted to stressed, damaged
or dying palms (Wattanapongsiri 1966) and rely on semiochemicals for aggregation
(Giblin-Davis et al. 1996a). Mating takes place on the host and eggs are laid deep in
the petiole bases, or in wounds. Females lay 207 + 19 eggs during a 42-day oviposition
period (Weissling & Giblin-Davis 1994). Eggs hatch within 64 h and larvae bore into
the palm stem. In southern Florida, R. cruentatus is mutivoltine and development can
take less than 84 days from egg to adult emergence (Giblin-Davis & Howard 1989).
Late instar larvae can kill palms by destroying the palm heart (bud) (Giblin-Davis &
Howard 1988). Palms are usually asymptomatic until the apical meristem has been
damaged. Therefore, weevil presence is usually not detectable until fatal damage and
associated rot has occurred, making control efforts ineffective. Males can survive for
87 days and females for 74 days (Giblin-Davis & Howard 1989).
We observed a Canary Island date palm nursery for natural palm mortality caused
by R. cruentatus. The objectives of this study were to monitor the rate of decline of
infested palms and rate of infestation of healthy palms under natural field conditions,
and to determine the weevil population dynamics in palms in various stages of decline.

Florida Entomologist 83(3) September, 2000


Study Site

Sampling and monitoring was conducted from March October 1997 at a 2-hectare
Canary Island date palm field nursery in Florida City, Miami-Dade County, FL. Ap-
proximately 90% of the palms at this nursery were planted 10 years ago, and the re-
maining were planted 5 to 6 years ago. Ten randomly selected, healthy 10-year-old
Canary Island date palms were measured at the beginning of the study to give an in-
dication of the size and health of the trees in this nursery for future comparisons.
Mean bare stem height was 3.9 cm (range 0-11.4 cm) (measured from the ground to
the 'booted' [with attached petioles] portion of the stem) (Fig. 1). Mean bare stem di-
ameter was 42.6 cm (33.0-53.3 cm). Mean booted stem height was 58.4 cm (30.5-81.3
cm) (Fig. 1). Mean maximum booted stem diameter was 56.4 cm (45.7-68.6 cm) (Fig.
1). Mean overall palm height measured from the ground to the top of the spear leaves
unexpandedd frond) was 217.7 cm (182.9-251.5 cm) (Fig. 1).
The soil type was 'Krome very gravelly loam' with a plant spacing of 7.9 m x 2.7 m
in 0.9 m hilled planting beds. No insecticides were used in the past 6 to 7 years, and
no fertilizer or herbicides were applied or palms pruned in the year preceding this
study. The surrounding area included avocado (Persea americana Mil.) and vegetable
plantings with a natural area located just west of the palm field. Native saw palmet-
tos were present in the area but no sabal palms were found nearby. There was a 2-
hectare planting of approximately 12-year-old Chinese fan palms (Livistonia chinen-
sis [Jacq.] R. Br. ex Mart.) near the study site.

Distribution and Rate of Decline

The site was mapped by assigning each of the 950 living or dead palms a row and
tree number. Each palm was surveyed and visually monitored every 1 to 4 weeks for
7 months. We designated nine categories for external palm symptoms. (1) old-dead:
crown and stem completely collapsed, often dry and falling apart, fronds completely
necrotic (brown) and pithy, palm dead for several months or more; (2) dead: crown and
stem collapsed, fronds completely necrotic; (3) recently dead: crown and stem col-
lapsed, spear leaf and 1 to 2 adjacent leaves green, other fronds completely necrotic;
(4) crown and stem collapsed but most of the fronds still green; (5) crown collapsing
(spear leaf leaning >450); (6) crown beginning to collapse (spear leaf leaning 5-45),
oldest fronds reclining; (7) 3 to 4 oldest fronds ... .11. -apparently healthy, weevil
frass detectable only after close inspection; (9) apparently healthy, no detectable wee-
vil frass (Fig. 2). Weevil frass is easily detected at or near petiole bases in categories
1 through 7.

Weevil Population Dynamics within Palms

A total of 45 randomly selected palms from differing categories of decline were cut
at the soil line with a chainsaw, placed into separately labeled 150-liter plastic con-
tainers for transportation to the laboratory and dissected to determine weevil popu-
lation dynamics. Booted and bare palm stems were quartered with a chainsaw and
the remaining tissue was finely sectioned with a sharp, heavy-bladed knife. Com-
pletely collapsed and dead palms (categories 3 through 1) were harvested by hand by
pulling the crowns and trunks apart in the field. A chainsaw and knives were used as
needed. All larvae and adults, as well as empty and occupied cocoons, were removed
from dissected palms.

Hunsberger et al.: Rhynchophorus cruentatus in Date Palms 293

Maximum Booted
Stem Diameter

Overall Palm

14- Spear Leaf

Fig. 1. Schematic drawing of Phoenix canariensis showing how various palm size
measurements were made.

Larvae were separated by length into two categories, small (<2.5 cm) and large
(>2.5 cm). All cocoons were opened to determine the stadium of the weevil (last instar
larva, prepupa, pupa, or adult). Adults ready to emerge from the cocoons were sexed.
Empty cocoons were counted to estimate the numbers of adults that had successfully
emerged. Free-living adults found on dissected palms were also sexed and counted.

Florida Entomologist 83(3) September, 2000

N 6

4-3 A 2-1

Fig. 2. Photographs and illustrations depicting the various categories of Canary Is-
land date palm decline caused by Rhynchophorus cruentatus larval feeding damage.
Some categories are pooled together because differentiating symptoms are not visible
at this magnification or without color. Categories 9 and 8; (9) Palms healthy, asymp-
tomatic, (8) weevil frass detected in petioles. Category 7; dying, 3 to 4 oldest fronds
drooping, weevil frass detected in petioles. Category 6; dying, spear leaf and stem be-
ginning to lean (5-450), weevil frass detected in petioles. Category 5; dying, spear leaf
and stem collapsing (>450), weevil frass detected in petioles. Categories 4 and 3; (4)
dying, spear leaf and stem completely collapsed but crown leaves still green, weevil
frass detected in petioles (3); recently dead, spear leaf and 1 to 2 adjacent leaves
green, other fronds completely necrotic, weevil frass detected in petioles. Categories
2 and 1; (2) dead, crown and stem collapsed, fronds completely necrotic; weevil frass
detected (1); old-dead: crown and stem completely collapsed, dry and falling apart,
fronds completely necrotic (brown) and pithy, weevil frass detected, palm dead for sev-
eral months or more.


Hunsberger et al.: Rhynchophorus cruentatus in Date Palms 295

These represented weevils recently emerged from cocoons or weevils that flew into
palms before they were harvested.

Temporal Dynamics

We compared the number of weevils and empty cocoons from palms that died be-
fore the beginning of this study (categories 1 and 2) with palms that died or were de-
clining during the study (categories 2, 5 and 6) for differences in adult weevil
production per palm.


Temporal dynamics data were subjected to an analysis of variance (ANOVA) using
PROC GLM (SAS Institute 1985) and means were separated by a Duncan's Multiple
Range test when the ANOVA was significant at P < 0.01.


Distribution and Rate of Decline

Palms declined from category 9 to 3 in a mean of 49 days (_-16 SD; range 11-100
days) and from category 7 to 3 in a mean of 31 days (_-15 SD; range 8-80 days). Palm
decline was patchily distributed throughout the study site (Table 1). When initially
surveyed on March 28, 1997 (Julian Date 87), the three middle rows of the site (rows
4, 5 and 6) had the greatest proportion of dead palms (mean mortality 58, 77, 64%,
rows 4 6 respectively) (Table 1). More dead palms were found near the ends of rows,
except for row 5 where dead palms were found throughout the row. In the beginning of
this study, 54.7% of all palms were dead, 3.3% were dying, and 41.8% were apparently
healthy (Fig. 3). However, seven months later (October 1997; Julian Date 286), the pro-
portion of healthy palms decreased to 3.1% (Fig. 3). Over the same time period, the pro-
portion of dead palms (categories 1 to 3) increased from 54.7% to 88.5%. The proportion
of dying palms (categories 4 through 7) was fairlyconstant, with a mean of 9.4% (range
3.3-13.1%) per month. Also, by October 13 (Julian Date 286), almost all of the palms in
rows 3, 4, 5, and 6 were dead. No healthy palms were found in row 6 (Table 1).



March 28, 1997 1 2 3 4 5 6 7 8

% Palms Dead 57.1 43.4 41.6 57.8 73.5 63.8 47.5 56.0
% Palms Dying 4.8 2.8 4.4 2.2 3.8 3.8 2.9 5.0
% Palms Healthy 38.1 53.8 54.0 40.0 22.0 32.3 49.6 39.0

October 13, 1997
% Palms Dead 85.7 89.5 87.6 94.0 96.2 93.8 82.7 90.0
% Palms Dying 9.5 6.3 6.6 3.0 2.3 6.2 13.7 6.0
% Palms Healthy 4.8 4.2 5.8 3.0 1.5 0 3.6 4.0

Florida Entomologist 83(3) September, 2000

80 100 120 140 160 180 200 220 240 260 280 300

80 100 120 140 160 180 200 220 240 260 280 300

Julian Date (1997)

Fig. 3. Plots of the percentage of Rhynchophorus cruentatus-infested field-grown
Canary Island date palms in each symptom category for each Julian date in 1997 from
a nursery in southern Florida. See Figure 2 for descriptions of the symptom catego-
ries. A. Palm symptom categories 9 through 1. B. Palm symptom categories consoli-

Hunsberger et al.: Rhynchophorus cruentatus in Date Palms 297

Weevil Population Dynamics within Palms

We found that as few as twenty-four R. cruentatus larvae were capable of killing a
10-year-old Canary Island date palm. Only two early instar larvae were found in a
healthy (category 9) dissected palm with no associated internal damage (Table 2).
Adults [1:1 (M:F)] were found within debris collected at the petiole bases of all of the
healthy palms surveyed. Palms in category 8, where the palm appears healthy except
for larval frass, were rare (Fig. 3) and none were dissected (Table 2). This category was
difficult to detect because externally visible frass was normally associated with larger
larvae preparing to pupate.
Eighty to 92% of the total live weevils from palms beginning to decline (categories
7 and 6) were larvae and most of these were small (Table 2). As symptoms progressed,
the percentage of total live weevils that were pupae and prepupae increased from
11.5% to 38.7% (categories 7 to 3, respectively). Palms categorized as 7, when symp-
toms were first noticed, had large numbers of weevils (>100 live weevils per palm)
(Table 2). The greatest number of weevils recovered was 308 (mostly immatures) in a
palm whose crown was just starting to collapse (category 6). On the average, palms
with collapsing crowns (category 5) had the most weevils (>200 per palm) (Table 2).
Palms with completely collapsed crowns (category 4) also contained large populations
of larvae (mean = 113.4 per palm) (Table 2). Recently dead palms (category 3) still had
larvae (mean = 29.0 per palm) compared with palms that had been dead for awhile
which had less than 1 larva per palm (Table 2). Dead palms (categories 2 and 1) con-
tained mostly empty cocoons (89-99% of the total weevils per palm) (Table 2). Up to
two generations of weevils were found within many palms indicating the continued
recruitment of adults to infested palms or recidivism. Dead palms had the fewest total
weevils per palm (mean = 40.0, 88.3, 75.5; categories 1 through 3, respectively). For
all palms, the sex ratio was 1:1 (M:F) for adults inside cocoons.

Temporal Dynamics

Palms that died in the year before this study was started had been infested with a
lower (F = 5.33; df = 2,18; P = 0.0152) number of weevils than palms that died during
the study (Table 3). A mean of 40.0 weevils per palm was recovered in palms that died
before the spring 1997, 88.3 weevils in palms that died during the spring of 1997, and
115.0 weevils in palms that died during the summer of 1997 (Table 3).
The number of empty cocoons within trees served as an estimator of the adult weevil
production potential of Canary Island date palms (killed by lethal infestations of wee-
vils) because each empty cocoon represented an adult that emerged. The adult weevil
production per palm was related to when the palms were infested. It was a maximum
of 148 weevils (mean = 78.5) for those palms that died during the spring-early summer
of 1997 (pre-spring 1997 infestation) compared with a maximum of 142 weevils (mean
= 105.6) for palms that died during the summer 1997 (spring 1997 infestation) and a
maximum of 59 (mean = 39.5) for palms that were infested and died prior to the spring
of 1997 (Table 3). Because these observations involved palms that died over the course
of 18 months, host size could have been a factor. However, gross comparisons of stem
size did not confirm that the trees had grown very much during this time.
Palms (category 6) (Table 3, season of infestation = spring 1997) were selected and
some were harvested while others of this cohort were allowed to decline and were sub-
sequently harvested as category 5 and 2 palms at different times in 1997. There was
a decrease (F = 9.35; df = 4, 22; P = 0.0001) in total weevil counts as palms declined
(Table 3). The mean total weevils per palm in categories 5 and 6 were 223.3 and 183.0,
respectively. In contrast, category 2 palms that had been infested at the same time


Mean (--SE) no. weevils per palm

Larvae Pupae Adults in cocoons
Palm health Total live Empty Total
category' (n)2 Small Large Prepupae Pupae Females Males weevils cocoons3 weevils4

1 4 0 0.5 (0.5) 0 0 0 0 0.5 (0.5) 39.5 (7.3) 40.0 (7.1)
2 12 0 0.6 (0.3) 0 2.3 (1.0) 3.3 (0.7) 3.6 (0.9) 9.8 (1.0) 78.5 (9.3) 88.3 (10.6)
3 2 0.5 (0.5) 28.5 (26.5) 7.5 (7.5) 14.0 (7.0) 2.0 (1.0) 3.0 (2.0) 55.5 (26.5) 20.0 (15.0) 75.5 (15.5)
4 5 44.0 (21.5) 69.4 (21.7) 11.4(4.6) 19.2 (2.4) 6.4 (0.8) 4.8 (1.6) 155.2 (23.5) 17.2 (5.3) 172.4 (23.4)
5 3 89.7 (15.4) 85.3 (8.5) 10.0(4.4) 14.7 (5.8) 7.3 (2.3) 10.7 (3.7) 217.6 (20.0) 5.7 (1.5) 223.3 (19.4)
6 3 111.0 (43.8) 51.7 (15.7) 3.3 (1.5) 6.0 (4.0) 3.0 (1.0) 2.3 (0.7) 177.3 (62.5) 6.0 (3.1) 183.3 (62.3)
7 3 62.3 (13.3) 23.7 (10.7) 7.7 (2.3) 4.7 (3.7) 4.7 (4.7) 4.3 (3.3) 107.4 (34.2) 1.3 (1.3) 108.7 (34.7)
8 0
9 7 0.3 (0.2) 0 0 0 0 0 0.3 (0.2) 0 0.3 (0.2)

See Figure 2 for legend.
'n = number of palms dissected.
Empty cocoon= emerged weevil.
Total weevils = total live weevils + empty cocoons.


Mean number (Range) Mean number (Range)
Palm Palms of larvae per palm Percentage of pupae per palm Percentage
Season of health harvested of total of total
infestation category' (n) Small Large Total2 weevils3 Prepupae Pupae Total weevils

Pre-spring 1 4 0 (0) 0.5 (0-2) 0.5b 1.3 0 (0) 0 (0) Ob 0
1997 2 12 0 (0) 0.6 (0-3) 0.6b 0.7 0 (0) 2.2 (0-12) 2.3b 2.6

Spring 1997 2 5 0 (0) 0.8 (0-4) 0.8b 0.7 1.0 (0-2) 3.0 (0-7) 4.0b 3.5
5 3 89.7 (65-118) 85.3 (72-101) 175a 78.4 10.0 (2-17) 14.7 (5-25) 24.7a 11.1
6 3 111.0 (50-196) 51.7 (26-80) 162.7a 88.8 3.3 (1-6) 6.0 (2-14) 9.3b 5.1

Palm Mean number (Range) adults in cocoons Percentage Mean number Percentage
Season of health of total (Range) of of total Total
infestation category Females Males Total weevils empty cocoons weevils weevils

Pre-spring 1997 1 0 (0) 0 (0) Ob 0 39.5 (25-59)bc 98.8 40.0c
2 3.1(0-7) 3.6 (0-9) 6.9b 7.8 78.5 (21-148)ab 89.0 88.3bc

Spring 1997 2 1.0 (0-4) 1.0 (0-4) 4.6b 4.0 105.6 (51-142)a 91.8 115.0b
5 7.3 (3-11) 10.7 (6-18) 18.0a 8.1 5.7 (3-8)c 2.6 223.3a
6 3.0 (2-5) 2.3 (1-3) 5.3b 2.9 6.0 (0-10)c 3.3 183.3a

See Figure 2 for legend.
'Means in a column followed by a different lower case letter are significantly different according to a Duncan's Multiple Range test.
Total weevils = total live weevils + empty cocoons.

Florida Entomologist 83(3) September, 2000

had a significantly lower mean total number of weevils (115.0 per palm) (Table 3) sug-
gesting unknown mortality or other factors. Most of the weevils in category 5 and 6
palms were larvae (78.4 and 88.8%, respectively); whereas, category 2 palms con-
tained 0.7% larvae (Table 3).


This study has confirmed that apparently healthy Canary Island date palms are
suitable and susceptible hosts for R. cruentatus (Giblin-Davis & Howard 1989, Giblin-
Davis et al. 1996a,b). This contrasts with most other species of palms that appear to
be suitable and susceptible to R. cruentatus infestation only after some major stress
(Giblin-Davis & Howard 1989). Adult R. cruentatus seek harborage between palm
sheaths or in rotten parts of the palm, presumably in search of moisture and oviposi-
tion sites (Weissling & Giblin-Davis 1993, 1994). Rhynchophorus cruentatus may at-
tack healthy Canary Island date palms because of their large and tender leaf bases
which may be more semiochemically apparent and easily infested by neonates than
the split and hardened petioles of palms such as S. palmetto (R. M. G.-D., unpublished
data). None of the palms (Chinese fan palm or saw palmetto) in areas next to the
study site declined. Rhynchophorus cruentatus appeared to be the sole cause of death
for all dead or dying Canary Island date palms (922 of 950) observed during this study.
In a similar study of 197 six to eight-year-old Canary Island date palms in Indian
River County, FL during 1994-1995,61% died with heavy infestations ofR. cruentatus
(R. M. G.-D. & J. T., unpublished data). At that site, none of the > 100 S. palmetto (aged
5 to 10 years old), > 300 Syagrus romanzoffiana (Chamisso) Glassman (Queen palm)
(aged 5 to 8 years old), > 500 Aceolorrhaphe wrightii (Grisebrach & H. A. Wendland)
(Paurotis palm) (> 5 years old), > 200 P. reclinata N. J. Jacquin (Reclinata palm) (> 5
years old), 214 Livistonia australis (R. Brown) Martius (Australian fan palm) (> 5
years old), and > 30 Washingtonia robusta H. A. Wendland (Washington palm) (> 5
years old) declined due to R. cruentatus infestations during or within 3 years of the
study (R. M. G.-D. & J. T., unpublished data).
A few adult R. cruentatus may initially be found on a palm, but recruitment occurs
when male-produced aggregation pheromones and kairomones from stressed, dam-
aged, or dying trees are released (Giblin-Davis et al. 1996a). When healthy and wee-
vil-infested Canary Island date palms in this study were dissected, flying R.
cruentatus were attracted to the area within minutes of cutting. In most palms, R.
cruentatus infestations are secondary to some major stress factor such as pruning,
transplanting, drought, and/or earlier damage from the West Indian sugarcane borer,
Metamasius hemipterus sericeus (Olivier) (Giblin-Davis et al. 1996b). However, we
found R. cruentatus attacking apparently healthy Canary Island date palms that had
not been pruned within the past year. The population increase of R. cruentatus may
have started after hurricane Andrew occurred in August, 1992 causing severe damage
to vegetation. Damaged palms at the study site may have provided resources for an
increase in the population of this weevil that went unnoticed and became an epizootic.
At high densities, R. cruentatus may be more of a threat to healthy Canary Island
date palms than when present at low densities.
Canary Island date palms are also a suitable host for M. h. sericeus, but it does not
cause lethal damage (Giblin-Davis et al. 1996b). There was some old M. h. sericeus
damage and empty cocoons (mean = 3.6 per palm) in the dried petioles of R. cruenta-
tus-infested palms in this study. Because of similar chemical ecology, infestation and
damage by M. h. sericeus can result in the recruitment of the more damaging R. cruen-
tatus to palms (Giblin-Davis et al. 1996a).

Hunsberger et al.: Rhynchophorus cruentatus in Date Palms 301

The symptom categories proposed in this study were ineffective for designating
Canary Island date palms that had sub-lethal or potentially lethal internal infesta-
tions of R. cruentatus. Palms did not react to weevil infestations until there was an ac-
cumulation of late instar larvae (>20) (Table 2). As a result, substantial damage
occurred before there was any observable manifestation of injury by the palm (Wolfen-
barger 1958). The earliest infestation category (8) proved rare and unreliable because
of the cryptic nature of R. cruentatus. Frass production, the most reliable external
symptom of R. cruentatus infestation, was not easily observed until palms were clas-
sified as category 7. The attached petiole bases of P. canariensis obscure visibility
making the discovery of frass difficult until large numbers of late instar larvae begin
to pupate. Removal of petiole bases before they have completely dried and are ready
for abscission to discover R. cruentatus is laborious and creates wounds that could at-
tract more weevils.
There are several possible scenarios for category 8 palms where frass was detected
before the palm declined. First, an apparently healthy host palm could have been near
category 7, but the large number of larvae had not done sufficient damage to critical
zones of the palm to induce symptoms. Palms are monocots with large numbers of vas-
cular bundles that connect leaves to roots. Borers can do a lot of random mechanical
damage to these vascular elements before enough are destroyed to induce a systemic
decline. This would help explain the short time difference between apparently healthy
plants (category 9) through death (category 3) (49 days) versus onset of frond reclining
(category 7) through death (category 3) (31 days). Second, there might have been a
smaller weevil infestation where the weevils had metamorphosized and frass was for-
tuitously deposited where it could easily be observed. This would explain the long
time to death from category 9 through 3 in some palms (range up to 100 days). A palm
in this situation, would be the most likely candidate for preventative systemic insec-
ticide treatment, if external symptoms were more reliable and there were proven sys-
temic insecticides available. When a palm in this situation is left untreated, the next
generation of adult weevils recruit new weevils or mate among themselves and pro-
duce progeny that may kill the palm.
Category 7 was a reliable indicator of a dying palm but some variability existed in
the time to death from category 7 through 3 palms (range 8-80 days) suggesting dif-
ferences in weevil-host dynamics. Fast host death is consistent with a host that was
on the verge of severe decline with large numbers of mature weevils when it was rated
(Table 2). Slow host death is consistent with a palm that had been structurally com-
promised by a few weevils allowing for early display of symptoms. The lack of reliabil-
ity of asymptomatic palms in categories 9 and 8, the similarities of weevil-host
dynamics and time to death for categories 7 through 4 and categories 3 through 1 (Ta-
ble 2), all suggest that the nine categories designated at the start of this study can be
consolidated into the following three categories for diagnostic purposes; APPAR-
ENTLY HEALTHY (categories 9 and 8), DYING (categories 7 through 4), and DEAD
(categories 3 through 1) (Fig. 3).
External symptoms did not allow preventative diagnosis and treatment of internal
R. cruentatus infestations in P. canariensis. By the time that external symptoms were
unambiguous (categories 7 through 4), the mean total weevil counts per palm were
over 100 with more than 65% as larvae and more than one quarter of these were >2.5
cm in length (Table 2). Palms in these categories were dying because of irreparable
damage to their apical meristems and attempts to save them would have been inef-
fectual. Thus, phytosanitation (palm removal and destruction) for management of R.
cruentatus in Canary Island date palms should be implemented as soon as host leaves
droop and weevil frass is observed.

Florida Entomologist 83(3) September, 2000

Weevil populations during the present study were high in the late spring and early
summer of 1997 (a total of 3,016 R. cruentatus and 3,043 M. h. sericeus adults were
captured and killed in semiochemical-baited traps [5 traps per hectare] over the
course of the seven month study [unpublished data]). We hypothesize that the popu-
lations expanded in 1997 because of significant increases in the number of empty co-
coons harvested in spring of 1997 from category 2 palms compared with category 1
palms from pre-spring 1997 (Table 3). The lack of chemical control and/or phytosani-
tation were the most likely factors responsible for the weevil population increase. Ad-
ditionally, the semiochemicals released from dying palms and adult weevils may have
recruited more R. cruentatus from surrounding areas into the site. Even though
>3,000 R. cruentatus were mass-trapped from the site, the within field populations of
R. cruentatus did not decrease enough during the summer to prevent most of the
palms at the site from being destroyed by this weevil. Adult weevil density estimates
for the entire site were 34,390 (417 palms with 78.5 empty cocoons per palm, category
2; 57 palms with 20.0 empty cocoons per palm, category 3; 30 palms with 17.2 empty
cocoons per palm, category 4) between April May 1997 (Fig. 3 and unpublished data).
Therefore, the mass trapping efforts removed about 9% of R. cruentatus adults
present. We speculate that the M. h. sericeus that we trapped were recruited from
other sites because we did not find any immatures in dissected palms. The only phy-
tosanitation done at this site was the removal of palms for dissection and represented
about 5% of the total suggesting that semiochemical-based mass trapping was inef-
fective for small field plot control of R. cruentatus without aggressive concurrent phy-
tosanitation, especially during a weevil epizootic. Replicated studies are needed to
verify the usefulness of mass trapping in small and large stands of Canary Island date
palms. Care should be taken when implementing mass trapping in small plots of
highly susceptible palms, such as Canary Island date palms, because pheromone
could call more weevils into a site.
The theoretical R. cruentatus yield for this site for 1997 was 82,696 adult weevils
(522 palms, mean = 64.1 weevils per palm; 428 palms, mean = 115.0 weevils per
palm). Loss to this grower was estimated at $285,000-$380,000 (based on the number
of palms per hectare and a $300-$400 wholesale price for this size palm). Growers and
buyers of P. canariensis in regions where R. cruentatus exists should be aware of the
potential lethal risk that it poses for this non-native palm. The costs of aggressive
phytosanitation at the first symptoms of infestation by R. cruentatus and prophylactic
pesticide treatment at times of pruning, stress or transplanting (Giblin-Davis &
Howard 1989) should be factored into the predicted cost of production and mainte-
nance of Canary Island date palms in Florida. To prevent the spread ofR. cruentatus
populations, Canary Island date palms shouldn't be transplanted from sites with an
epizootic because early weevil infestations are not easily diagnosed and stressing of
palms can call in colonizing weevils before removal from the nursery.


We thank M. Diaz Farms and Tripson Trail Nursery and Tree Farm for the gener-
ous use of their land. We are grateful to J. Tripson, A. Jimenez, J. Leidi, N. DeCrappeo,
C. Vanderbilt, A. McCall, B. Center, F. Bilz, J. Rodreguez and V. Slanac and for their
technical assistance and F. W. Howard and J. E. Pena for critical review of the manu-
script. This research was supported by the United States Department of Agriculture
(USDA) through a grant in Tropical and Subtropical Agriculture CRSR-95-34135-
1763. This manuscript is Florida Agricultural Experiment Station Journal Series R-

Brambila:A review of Cligenes


ANONYMOUS. 1997. PlantFinder, wholesale guide to foliage and ornamental plants,
Aug. 15. Betrock Information Systems, Hollywood, FL. Pp. 165-166.
GIBLIN-DAVIS, R. M., AND F. W. HOWARD. 1988. Notes on the palmetto weevil, Rhyn-
chophorus cruentatus (Coleoptera: Curculionidae). Florida State Hort. Soc.
101: 101-107.
GIBLIN-DAVIS, R. M., AND F. W. HOWARD. 1989. Vulnerability of stressed palms to at-
tack by Rhynchophorus cruentatus (Coleoptera: Curculionidae) and insecti-
cidal control of the pest. J. Econ. Entomol. 82: 1185-1190.
AND L. M. GONZALEZ. 1996a. Chemical and behavioral ecology of palm weevils
(Curculionidae: Rhynchophorinae). Florida Entomol. 79: 153-167.
GIBLIN-DAVIS, R. M., J. E. PENA, AND R. E. DUNCAN. 1996b. Evaluation of an ento-
mopathogenic nematode and chemical insecticides for control of Metamasius
hemipterus sericeus (Coleoptera: Curculionidae). J. Entomol. Sci. 31: 240-251.
SAS INSTITUTE. 1985. SAS user's guide: statistics, 5th ed. SAS Institute, Cary, NC.
WATTANAPONGSIRI, A. 1966. A revision of the genera Rhynchophorus and Dynamis.
Dept. of Agricultural Science Bulletin, Bangkok 1:1-328.
WEISSLING, T. J., AND R. M. GIBLIN-DAVIS. 1993. Water loss dynamics and humidity
preference of Rhynchophorus cruentatus (Coleoptera: Curculionidae) adults.
Environ. Entomol. 22: 93-98.
WEISSLING, T. J., AND R. M. GIBLIN-DAVIS. 1994. Fecundity and fertility of Rhynchopho-
rus cruentatus (Coleoptera: Curculionidae). Florida Entomol. 77: 373-376.
WOLFENBARGER, D. 0. 1958. Palm insects and their control. Principes 2: 107-112.
WOODRUFF, R. E. 1967. A giant palm weevil, Rhynchophorus cruentatus (Fab.), in
Florida (Coleoptera: Curculionidae). Florida Dept. of Agriculture Division of
Plant Industry, Entomology circular no. 63.


Brambila:A review of Cligenes


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

Cligenes grandis (Lygaeoidea: Rhyparochromidae: Antillocorini), a new species
from Mexico and Central America, is described and illustrated. New geographic
records are given for Cligenes distinctus. Valeris, a new genus, is described and illus-
trated for Cligenes subcavicola Scudder, Darlington, and Hill.
Key Words: Hemiptera, Lygaeoidea, Lygaeidae, Antillocorini, Cligenes, Botocudo, Val-
eris, cave, bat, Ficus, seeds

Se describe e ilustra Cligenes grandis (Lygaeoidea: Rhyparochromidae: Antilloco-
rini), una especie nueva de M6xico y Centroamerica. Se reportan nuevas localidades

Florida Entomologist 83(3) September, 2000

de Cligenes distincuts. El g6nero nuevo Valeris es descrito e ilustrado para la especie
Cligenes subcavicola Scudder, Darlington, and Hill.

Slater (1964) included 35 species in the Neotropical lygaeoid genus Cligenes in his
Catalogue of the Lygaeidae of the World. All but the type species, Cligenes distinctus
Distant, and C. subcavicola Scudder, Darlington, and Hill, were subsequently as-
signed to other genera (see Slater and O'Donnell [1995]). Cligenes is distinguished
from other antillocorine genera by a groove on the prosternum where the labium lies
at rest. This character was not included in the original description of Cligenes by Dis-
tant (1893). It was mentioned under the name of "rostral groove" by Scudder (1962) to
distinguish Botocudo from Cligenes. However, when describing Cligenes subcavicola,
which lacks this character, Scudder et al. (1967) concluded that the rostral groove was
not a diagnostic generic character for Cligenes because it was variable. I consider the
prosternal groove as an autapomorphic character for Cligenes that defines this genus
and is of high value since no other antillocorine is known to have it.
A new species of Cligenes is described below.
Cligenes subcavicola was originally placed in Cligenes because of its explanate
pronotal lateral margins; however, in addition to the lack of a prosternal groove in C.
subcavicola these margins are upturned. Cligenes subcavicola also differs from other
Cligenes by having a caudal projection on the genital capsule of the males. C. subcav-
icola is here removed from Cligenes and designated as the type species of a new genus,
All measurements are in millimeters.
The following abbreviations are used in the text: AMNH (American Museum of Nat-
ural History, New York); BMNH (The Museum of Natural History, London); CASC (Cal-
ifornia Academy of Sciences, San Francisco, CA); EMEC (Essig Museum of Entomology,
Berkeley, California); FSCA (Florida State Collection of Arthropods, Gainesville, Flor-
ida); RMB (R. M. Baranowski collection, Homestead, Florida); UNAM (Colecci6n Ento-
mol6gica, Instituto de Biologia, Universidad Nacional Aut6noma de M6xico, D.F.,
Mexico); NMNH (National Museum of Natural History, Washington, D.C.).

Cligenes grandis Brambila, NEW SPECIES
(Figs. 1, 3, 6-15)

Diagnosis. A reddish brown antillocorine with a longitudinal groove on proster-
num; anterior lobe of pronotum carinate, punctate, and strongly convex; and corium
creamy white with apex brown. Measuring 2.30 to 3.20 mm in total body length it is
one of the largest species among the Antillocorini and larger than most Cligenes dis-
Description. Male. Head brownish black. Anterior lobe of pronotum reddish brown
with collar and lateral margins yellowish brown; posterior lobe yellowish brown with
meson and humeral angles reddish brown and creamy white on each side of meson
and anterior to humeral angles. Scutellum dark yellowish brown mesally, becoming
dark brown laterally, apex white (Fig. 3). Clavus creamy white. Corium creamy white
with apex brown and with light brown marking at midpoint next to margin. Mem-
brane translucent white. Body below reddish brown with capsule yellowish brown.
Antennae reddish brown, distal ends of segments II-IV yellowish brown and labium
yellowish brown. Legs pale brownish yellow, coxae yellowish brown.

Brambila:A review of Cligenes

Dorsum of head rugose; tylus reaching middle of first antennal segment. Gula nar-
row reaching anterior margin of prosternum, contiguous with prosternal groove (Figs.
1 & 7). Labium reaching mesocoxae.
Pronotum with surface shining, glabrous, and punctate, with row of punctures
along carina at lateral margins, row at indentation separating lobes, and irregular
row of punctures marking a collar. Anterior lobe of pronotum strongly convex, nearly
2x longer than posterior lobe (Fig. 6). Lateral margins explanate and indented at area
of transverse impression; posterior margin concave. Scutellum evenly punctate, ex-
cept apex impunctate. Thoracic pleura with large punctures. Scent gland auricle on
metapleuron curved posteriorly; evaporative area surrounding auricle covering less
than half of metapleuron (Fig. 10). Meso- and metasternum with a median keel.
Clavus with three rows of punctures. Corium mesally with two rows of punctures
parallel to clavus, followed laterally by smooth area, then by irregular rows of punc-
tures except for uniform outer-most row. Lateral corial margins explanate with ante-
rior half upturned and slightly sinuate; apical margin with mesal half deeply concave
(Figs. 3 & 6). Abdominal sternum with decumbent setae; trichobothria on abdominal
sternite V linear, anterior to spiracle V, and closer to each other than to trichobothria
of segment IV (Fig. 8). Fore femur moderately incrassate (1.5x as wide as middle fe-
mur) with row of 7 ventral spines distally (Fig. 9). Genital capsule, parameres, and
spermatheca as in Figs. 11-15.
Head length 0.44, width across eyes 0.60, interocular distance 0.33. Pronotal
length 0.70, width across humeral angles 1.11. Scutellar length 0.60, width 0.66.
Length of claval commissure 0.18. Wing length from base of corium 1.92. Length of
antennal segments I 0.26, II 0.43, III 0.34, IV 0.42. Length of labial segments I 0.32,
II 0.46, III 0.5, IV 0.22. Total body length 2.72.
All specimens are macropterous. Total body length range: 2.30 to 3.20 mm. Several
specimens have the posterior lobe of pronotum nearly entirely creamy white or yel-
lowish white. Some specimens have a diffuse brown macula on the corium adjacent to
the lateral margin at midpoint; some have the wing membrane clear and colorless. Fe-
males are similar to males, except they have with fewer spines on fore femur.
HOLOTYPE: 6 Mexico, OAXACA, Chacahua, 31-V-1987, L. Cervantes (UNAM).
The holotype is in good condition, glued on its right side to a paper point, with the gen-
ital capsule removed and stored in a vial with glycerin. In UNAM.
PARATYPES: 19 6, 19 9. MEXICO, OAXACA: 1 2 Chacahua, 31-V-1987, coll. E.
Barrera (UNAM). 1 2 Puerto Escondido, 23-VII-1975, Noct., coll. H. Brailovsky
(UNAM); CHIAPAS: 1 6, 1 2 Tapachula, 19-IV-1983, coll. H. Brailovsky (UNAM); 1 2
Tapachula, 19-IV-1983, coll. H. Brailovsky, (UNAM); 5 6, 1 2 Tapachula, 19-IV-1983,
coll. E. Barrera (UNAM); 1 2 5 Mi. N.E. of Chiapa, W9258': N1645', 22-VIII-1966,
colls. J. & W. Ivie (AMNH); 1 2 Rosario, Izapa, 20-IV-1983, coll. H. Brailovsky
(UNAM); 1 2 Frontera [=international border with Guatemala], 6-IV-1979, 400 m.,
coll. E. Barrera (UNAM); and 1 2 El Zapotal. 2 Mi. S. Tux. Gutierrez, 5-VII-1957, coll.
P. D. Hurd (EMEC). EL SALVADOR: 1 6 San Salvador, 31-V-1959, coll. P. A. Berry
(NMNH). COSTA RICA: 1 6 Prov. San Jose, 8 Km. S. Orotina, 16-II-1983, colls. R. M.
Baranowski, F. Gilstrap (RMB); 1 6 Guanacaste, Prov. Sta. Rosa National Park, 22-
V-1985, colls. Doyen & Powell (EMEC). PANAMA: 1 2 Cabima, 22-V-?1 [might be
1911], coll. August Busck (NMNH); 1 6 Coco Solo Hosp., Canal Zone, light trap, 5-V-
1974, coll. D. Engleman (J. A. Slater Collection); 1 2 Coco Solo Hosp., C. Z., light trap,
20-VI-1975, coll. D. Engleman (UNAM); 1 9, 1 6 Coco Solo, 20-VI-1975, Noct., coll.
D. Engleman (UNAM); 1 6, 1 2 Coco Solo Hosp., C. Z., 21-V-1976, light trap, coll.
D. Engleman (UNAM); 7 6, 6 2 Coco Solo Hosp., C. Z., 21-V-1976, light trap, coll.
D. Engleman (J. A. Slater Collection). In AMNH, EMEC, NMNH, UNAM, J. A. Slater
and R. M. Baranowski collections.

Florida Entomologist 83(3) September, 2000

This species was first discovered among specimens from Mexico believed to be Cli-
genes distinctus because of the prosternal groove. Comparison of the capsules,
parameres and spermathecae revealed that these specimens were not conspecific with
C. distinctus (Figs. 11-15, 20-24). Cligenesgrandis n. sp. is larger and lighter in color-
ation than C. distinctus (with the exception of some specimens of the latter from Trin-
idad); the body size range for C. distinctus is 1.80 to 2.85 mm. An excellent illustration
of the latter species is in Slater and Baranowski (1990). Cligenes distinctus is ex-
tremely variable in pronotal color and shape, making comparisons difficult. Although
both species have four pale markings on the posterior lobe of the pronotum, C. distinc-
tus usually has most of the posterior lobe concolorous with the anterior lobe (Fig. 4)
while specimens of C. grandis n. sp. have most of the posterior lobe pale (Fig. 3). An-
tennal segment IV is reddish brown in C. grandis n. sp. but usually white in C. dis-
tinctus. Cligenes distinctus has two distinct brown maculae on each wing while most
C. grandis n. sp. have only an apical macula and at most a diffuse macula at midpoint.
The genital capsules of males differ in part by the absence in C. grandis n. sp. of pos-
terior-pointing decumbent thick setae seen in C. distinctus ventrally, that is, below
the cuplike sclerite (Fig. 22); furthermore, the arms of the cuplike sclerite are closer
to each other and meet at a sharp angle on C. grandis n. sp.(Fig. 15) while in C. dis-
tinctus this angle is rounded and the arms are farther apart (Fig. 22).
ETYMOLOGY. The specific name refers to the body size of this species, larger than
most Cligenes distinctus.
DISTRIBUTION. Mexico (southern states of Chiapas and Oaxaca), Guatemala, El
Salvador, Costa Rica, and Panama, i.e. Mesoamerican distribution.
BIOLOGY. Unknown. Several specimens were captured at night in light traps.

New Localities of Cligenes distinctus Distant
Cligenes distinctus is sympatric with C. grandis n. sp. in Mexico and Central
America (Fig. 35). Cligenes distinctus was described from Panama (Distant 1893) and
has also been reported from Cuba (Barber 1954) and Florida (Blatchley 1926) and as
Tomopelta munda (Uhler 1893), a junior synonym, from St. Vincent Island. In Florida
it is known from Brevard, Dade, and Monroe counties (Slater and Baranowski 1990).
However, it is widespread. The following are new country records (numbers in paren-
theses refer to the number of specimens):
MEXICO: (1) Veracruz, Puente Nacional, 6 Mi. S.E. Rinconada, 30-IX-1975, at light,
colls. J. Powell & J. Chemsak (EMEC). NICARAGUA: (1) Dept. Rivas, 10 Km. N.W. Sa-
poa, Rio Canas Gordos, 9-VI-1964, colls. Blanton, Broce (RMB). BAHAMAS, GRAND
BAHAMA ISL.: (1) Freeport, 20-27-VI-1987, colls. W. E. Steiner, M. J. & R. Molineaux
(NMNH); MAYAGUANA ISL.: (1) 27-VIII-1963, black light trap, coll. C. Murvosh
(FSCA). DOMINICA: (1) 5 mi. E. Dublanc, 1250', 20-VIII-1986, C. W. and L. O'Brien
(O'Brien Collection); (2) 4 mi. E. Salisbury, 19-VIII-1986, C. W. and L. O'Brien (O'Brien
Collection); (1) Morne Trois Pitons N. P., Freshwater Lake Rd., 2600', 13-VIII-1986, C. W.
and L. O'Brien (O'Brien Collection); (1) Trafalgar Falls, ca. 1200', 12-VIII-1986, C. W. and
L. O'Brien (O'Brien Collection). DOMINICAN REPUBLIC: (1) Prov. Barahona, Bara-
hona, 9-VI-1998, black light trap, colls. P. H. Freytag, B. K. Dozier, & R. E. Woodruff
(R. E. Woodruff Collection). GUADELOUPE: (2) Duclos, 25-VI-1971, colls. J. A. Slater,
R. M. Baranowski, J. E. Harrington (J. A. Slater Collection). JAMAICA: (1) Parish of St.
Ann, 3 Mi. W. Ocho Rios, 4-VII-1971, colls. J. A. Slater, R. M. Baranowski, J. E. Har-
rington, A. [adults] and N. [nymphs] under Ficus sp. (UNAM); (1) Parish of St. Cath-
erine, Worthy Park, 10-VI-1975, black light trap, coll. R. E. Woodruff (FSCA).
MARTINIQUE: (1) 12 km. N. Fort de France (N-3), 23-VIII-1986, C. W. and L. O'Brien

Brambila:A review of Cligenes

(O'Brien Collection). PUERTO RICO: (1) Municipio de Lares, near Lares, at cross of
roads 129 and 134, from leaf litter and soil picked at mouth of Cueva Golondrinas (across
from Cueva Catedral), 31-VII-1999, coll. J. Brambila (FSCA). SABA: (1) Mt. Scenary
[Mount Scenery], 800-840 m., 12-14-1-1968 (UNAM). TRINIDAD AND TOBAGO, TRIN-
IDAD: (2) Arima Valley, 800-1200 ft., 10-22-II-1964, colls. Rosen & Wygodzinsky collec-
tors (AMNH); (3) St. George Co., Simla, Arima Valley, 12-VII-1978, black light trap, coll.
M. Ramla (NMNH). BRAZIL: (6) Pernambuco, Caruaru, 900 m., IV-1972, coll. M. Alva-
renga (AMNH); (1) same, V-1972 (AMNH); (4) same, V-1972, J. Lima (AMNH); (1) [State
of Rio de Janeiro], Guanabara, Corcovado, XI-1971, coll. M. Alvarenga (AMNH).
The following are previously unpublished collection records from Cuba, St. Vincent
Isl., and Panama (unpublished records from Florida are numerous and will be pre-
sented in a different manuscript): CUBA: (1) Baragua, T.P.R.F., Ent. No. 379, at light,
coll. C. F. Stahl (AMNH). ST. VINCENT ISL.: (4) Charlotte Parish, 3 Mi. N.W. George-
town, 21-VI-1973, 2000', under Ficus sp., colls. R. Baranowski, F. O'Rourke, V. Picchi,
J. Slater (UNAM). PANAMA: (1) Barro Colorado [Isl.], CZ [Canal Zone], VIII-IX-1949,
Berl. funnel, Zetek 5427 (NMNH); (1) Cerro Campana, 800 m., R. de Pan. [Republica
de Panama], 8040'N, 79056'W, 28-IV-1973, coll. Engleman (J. A. Slater Collection).

Valeris Brambila, NEW GENUS

Type species: Cligenes subcavicola Scudder, Darlington & Hill, 1967. Monobasic.

Gula wide and shallow with bucculae meeting in a round arc (Fig. 25). Pronotum
with lateral margins sinuate, explanate and upturned (Figs. 5 & 26), posterior margin
concave. Transverse division between pronotal lobes moderately impressed. Metatho-
racic scent gland auricle straight, with apex pointed, elevated, and directed posteri-
orly at a diagonal (Fig. 27). Anterior half of lateral corial margin expanded and
upturned. Trichobothria on abdominal tergites IV and V in linear configuration, with
trichobothria on tergite V closer to each other than to trichobothrium on segment IV,
and posterior trichobothrium of segment V directly below spiracle of segment V to
slightly caudad (Fig. 28). Fore femur with two rows of spines (Fig. 29), spines of pos-
terior row larger in males than in females. Male genital capsule, paramere, and sper-
matheca as in Figs. 30-34; capsule with a median caudal projection (Fig. 33).
ETYMOLOGY. The name Valeris is from a Russian dancer. Learning about dance
history is one of the author's interests.
Valeris subcavicola was originally placed in the antillocorine genus Cligenes due to
the explanate pronotal lateral carina. However, Valeris differs from Cligenes by lack-
ing a groove for the labium on the prosternum (Fig. 26) and by having a caudal pro-
jection on the genital capsules of the males (Fig. 33). It differs from Cligenes also by
having only the anterior trichobothrium on abdominal sternal tergite V anterior to
the spiracle of tergite V (Fig. 19) and by having two rows of spines on the fore femur
instead of a single row (Fig. 18). Valeris also differs from Botocudo Kirkaldy because
the type species of Botocudo (B. diluticornis [Stal]) has the two posterior trichobothria
on segment V located dorso-ventrally in a diagonal relative to each other, lacks spines
on the fore femur, and the pronotal lateral carinae are not explanate and upturned.

Valeris subcavicola (Scudder, Darlington & Hill), NEW COMBINATION
(Figs. 2, 5, 25-34)

Cligenes subcavicola Scudder, Darlington and Hill, 1967: 465-469
Male, female, immatures, and eggs were described by Scudder, Darlington and Hill,
but not illustrated. For antillocorines, V subcavicola is large, with the holotype male

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