Content uploaded by Lauren Diepenbrock
Author content
All content in this area was uploaded by Lauren Diepenbrock on Mar 06, 2014
Content may be subject to copyright.
RESEARCH ARTICLE
A seasonal natural history of the ant, Odontomachus brunneus
L. M. Hart
•
W. R. Tschinkel
Received: 28 January 2010 / Revised: 29 April 2011 / Accepted: 6 July 2011 / Published online: 23 July 2011
Ó International Union for the Study of Social Insects (IUSSI) 2011
Abstract A north Florida population of Odontomachus
brunneus, a species of ponerine ants, was studied for a one-
year period to determine the annual cycle of reproduction
and colony growth, including the foraging biology and
seasonal changes in nest architecture. The life cycle of
O. brunneus is strongly seasonal. Colonies produce brood
for 6 months and are broodless for 6 months. Alates are
produced in mixed broods at the beginning of each season,
consuming much of the colony’s energy reserves. These
reserves recover slowly through foraging during the sum-
mer’s worker production, and rapidly after brood production
ceases in October. The foraging population was estimated to
average 77% (SD 22) of the workforce. This proportion was
not related to colony size and female alates were also found
to forage. Nest architecture was found to change seasonally,
with winter nests being more than twice as deep as the
average summer nest.
Keywords Odontomachus brunneus Annual cycle
Seasonal nest architecture Foraging
Introduction
Deciphering life history strategies is a key element in under-
standing a species and its interactions with the environment.
For social insects such as ants, it is the life history of the
superorganism that is of interest—the individual colony
members are the parts that make up the superorganism.
Natural history studies may include, but are not limited to the
seasonal cycle, colony size, queen number, number of nests
per colony, worker size, alate size and number, nest location,
and nest architecture (Ho
¨
lldobler and Wilson, 1990; Gadgil
and Bossert, 1970;Tschinkel,1991, 1993;Laskisand
Tschinkel, 2009).
A large component of life history analysis is ‘‘under-
standing the diversity of reproductive allocation strategies’’
(Bourke and Franks, 1995, p. 301); that is, the patterns of
how colonies allocate resources among the essential tasks of
seasonal and size-related colony growth, sexual reproduc-
tion and colony maintenance (Gadgil and Bossert, 1970;
Kipyatkov, 1993, 1995, 2001; Oster and Wilson, 1978;
Tschinkel, 1993, 1998). The seasonal and life-cycle allo-
cation patterns are best seen in the rates at which the
different types of production (i.e., workers, sexuals) and
maintenance proceed in relation to colony size and season
(Tschinkel, 1993). Each species of ant has a characteristic
annual cycle organized so as to capitalize on the warmest
period of the year for larval development (Kipyatkov,
1993). In considering this, a temperate annual cycle should
have brood production beginning as early in the spring as
possible and continuing until the fall weather becomes too
cold for the successful development, with only the stages
capable of successful over-wintering present at the begin-
ning of the winter season (Kipyatkov, 1993, 2001).
One important component of the annual cycle is the tim-
ing of sexual (alate) versus worker production. In many of
L. M. Hart W. R. Tschinkel
Department of Biological Science, Florida State University,
Tallahassee, FL 32306-4370, USA
e-mail: tschinkel@bio.fsu.edu
Present Address:
L. M. Hart (&)
Division of Plant Sciences, University of Missouri—Columbia,
1-31 Agriculture Building, Columbia, MO 65203, USA
e-mail: lmhd74@mizzou.edu
Insect. Soc. (2012) 59:45–54
DOI 10.1007/s00040-011-0186-6
Insectes Sociaux
123
the temperate ant species that have been studied so far, the
rearing of sexual brood generally occurs prior to worker
production (Brian and Elmes, 1974; Elmes, 1987;Kipyatkov,
1996, 2001; Smith and Tschinkel, 2005; Tschinkel, 1993,
1998). By this account, most of the colony’s initial invest-
ment in brood should be toward the creation of alates for the
sexual reproduction of the colonies followed by a secondary
investment in worker brood to increase the size of the colony
as well as create a worker force that will survive the winter
and begin the cycle all over again. This cyclical colony
growth and reproduction has also been shown to create a
seasonal age structure with young workers predominant in
the fall and older workers in the spring (Rissing, 1987;
Tschinkel, 1998). Because ants have an age-based division of
labor, it is possible that such seasonal variation in worker
age-distribution has consequences for the allocation of labor.
In addition to the reproductive cycle, cyclical variations
in physical structure as a colony moves, grows or prepares
for the environmental change can be an important element
of its yearly phenotype. Nest architecture is species-specific
and presumably serves particular biologically important
functions (Cerquera and Tschinkel, 2009; Tschinkel, 2004).
Descriptive studies of the general nest architecture are in
their infancy, with little if any information on seasonal
adaptations. Odontomachus brunneus nests have only
recently been described as simple, consisting of a single
vertical shaft connecting a varying number of chambers
(Cerquera and Tschinkel, 2009). Despite having only
4 months of nest excavation data, Cerquera and Tschinkel’s
data show a trend toward increasing nest depth into the
cooler winter months (unpublished data). Also previously
described is the seasonal vertical movement of fungus
gardens in the fungus gardening ant Trachymyrmex sep-
tentrionalis (Seal and Tschinkel, 2006). Such data suggest
that ants may have a physiological response to temperature
that induces the workers to alter the structure of their nest
and its components to accommodate these climatic changes.
Along with the species-specific cycle of colony compo-
sition, the flow of energy within a colony provides insight
into the seasonal functions of ant colonies. Annual cycles of
energy allocation have been determined for several other ant
species found in the Apalachicola National Forest including
the Florida harvester ant, Pogonomymex badius (Tschinkel,
1998), and the fire ant Solenopsis invicta (Tschinkel, 1993),
both of which can co-occur with Odontomachus brunneus.
Energy allocation patterns are determined by measuring fat,
the primary energy stores, as well as the weight of new lean
and fat biomass in the various life stages throughout the
annual cycle, revealing colony energy investment patterns.
Additionally, in several ant species, the proportion of fat in a
worker’s body has been shown to decline with age and can
be used to estimate relative age (Wilson, 1985; Tschinkel,
1993, 1998).
In addition to allocating energy, colonies must also allocate
labor. At the most basic level, colonies show an age-based
division of labor in which young workers perform mostly
brood care, and old ones forage. Workers of intermediate age
carry out a variety of other tasks. In species with subterranean
nests, this results in upward movement of older workers (who
are also becoming leaner), creating a vertical pattern of
decreasing average fatness and increasing average age. Per-
haps because they are a primitive species of ant, O. brunneus
colonies generate a division of labor through competitive
duels referred to as ‘‘interaction-based task allocation’’ by
Powell and Tschinkel (1999). Dominant individuals locate
themselves closer to the brood within a nest, participating in
brood care while older, subordinate individuals are forced into
the risky task of foraging. This dominance interaction, cou-
pled with the well-documented phenomenon of age-based
polyethism in most other ant species was used to decipher the
internal social structure in the laboratory O. brunneus
colo-
nies. What is unclear is how faithful workers are to their
designated roles; in laboratory observations, marked workers
were capable of moving among the foraging arena, broodless
and brood zones frequently (personal observation). However,
it is unclear if this process also operates in natural nests to
produce vertical age stratification of O. brunneus workers
as it does in other ant species (Beshers and Fewell, 2001;
Tschinkel, 1998). It is also unclear whether worker fatness is a
useful proxy for age in O. brunneus.
Foraging biology, while a rather active area of myrme-
cological research, has largely neglected the determination
of the proportion of a colony that is actively involved in
foraging. Previous studies of forager populations using
Formica polyctena (Kruk-De Bruin et al., 1977) and Pogo-
nomymex badius (Porter and Jorgensen, 1981) suggest that
(1) forager populations function almost as a separate entity
from the rest of the nest, and (2) that the foraging population
should be comprised of primarily older workers (Golley and
Gentry, 1964; Kruk-De Bruin et al., 1977; Porter and Jor-
gensen, 1981). Despite such studies, it is unknown if the
proportion of foragers in all ant colonies is related to the size
of the colony as it is in Solenopsis invicta (Tschinkel, pers.
comm.), if it is a species-specific trait or if this proportion is
an evolutionarily derived trait such that primitive ants (i.e.,
Odontomachus species) and more derived ants (i.e., Pogo-
nomyrmex species) would have differing proportions of
their workforce participate in foraging.
Methods
Description of the site
This study was completed in management compartment 219
of the Apalachicola National Forest in Leon County, Florida
46 L. M. Hart, W. R. Tschinkel
123
(30°22
0
11N, 84°19
0
32W). The sand hills ecotype is dom-
inated by longleaf pine with an understory of turkey oak and
a ground cover of wiregrass, saw palmetto, assorted shrubs
and Smilax vines. Most of the study was carried out in a
low-lying area near several depression ponds with a
fluctuating water table, resulting in moist, poorly drained
soils.Thesitewasdominatedbyliveoaksandother
broadleaf trees along with occasional longleaf pines, and
the ground was largely covered with a dense layer of
decaying oak leaves (for seasonal temperature and rainfall
information, see Appendix).
Excavation and collection
Six to eight nests were excavated monthly for a total of 76
nests. Excavations were done by digging a pit adjacent to a
focal colony, with the edge of this pit not less than 15 cm
from the nest entrance(s). This distance accommodates
varying chamber sizes as well as the slight deviation of nest
shafts from being entirely vertical. Chambers were exposed
sequentially from the top and their contents accumulated in
20 cm increments down to 60 cm in the total depth (initial
nest depths did not exceed this). Inhabitants residing deeper
than 60 cm were placed in a separate container and the
maximum nest depth recorded (procedure modified from
Tschinkel, 1998).
Nest inhabitants were killedbyrapidfreezing,then
dried at 50°, sorted and counted to determine the number
and vertical locations of the workers, brood, alates and the
queen. For analysis, nests were categorized by size class,
with Class 1 nests having B50 works; Class 2, 51–100
workers; Class 3, 101–150 workers; Class 4, 151–200
workers.
Pupal cocoon dissections
To determine the type of brood early cocoons contained,
dried cocoons from May to July were dissected under a
dissection microscope, and the brood within identified as
sexual, worker or unknown pupae. Sexual pupae were rec-
ognized by the presence of developing wings, and males by
the lack of large mandibles. Unknown pupae were too early
in the developmental process for their caste to be deter-
mined, or were still last instar larvae.
Determination of seasonal energy allocation
throughout 1 year
Dried ants were weighed (mg), assigned an identification
number and placed in a labeled, perforated gelatin capsule.
Capsules were threaded onto a wire skewer and placed in a
Soxhlet extractor, extracted with diethyl ether for 48 h,
dried and reweighed (Smith and Tschinkel, 2009). The
difference between the pre- and post-extracted dry weights
represents the extracted fat, and together with the dry weight
allowed the determination of the percentage of fat stored in
each individual ant (modified from Tschinkel, 1993; Seal
and Tschinkel, 2006, 2007; Soxhlet, 1879), and by summa-
tion, in the colony as a whole. From these data, we determined
energy (fat) allocation by life stage and throughout the yearly
cycle as well as the relative age(s) of workers at each level
within the colony, with the assumption that O. brunneus
workers decrease in fat content with age similarly to Pogo-
nomymex badius (Tschinkel, 1998).
Forager collection and marking
Ten nests were chosen for forager monitoring during July
2009. A foraging area with a diameter of 60 cm was cleared
around the nest entrance(s). To ensure collection of only
foragers and not maintenance workers, ants were collected
upon returning to their nest or exiting past 10 cm, this dis-
tance was chosen after observing that ants exiting with
debris deposited such items within 4–5 cm of the nest
entrance, those moving past this point left to forage. Each
ant was then individually marked on the posterior of its head
using Testors enamel and replaced near the nest entrance.
Foragers were collected for two 20-min periods separated
by 48 h allowing time for the marked workers to thoroughly
mix within the population (Ryti and Case, 1986). All col-
lection/observation was performed in the morning (8–11
a.m.), while colonies were actively foraging.
Estimates of forager populations based
on mark-recapture data
The number of foragers in a colony was estimated by a mark
and recapture method (Lincoln Index; Chew, 1959
; Kruk-
De Bruin et al., 1977; Southwood, 1978). The total foraging
population of a colony was estimated from the proportion of
marked workers in a recapture sample. The number of
marked workers initially released is the same proportion of
the total forager population as the number of marked
workers is of the recapture sample. This method makes
several assumptions: (1) all individuals in the forager pop-
ulation have an equal chance of being caught and marked,
(2) the marked ants mix thoroughly with the unmarked ants
before resampling, and (3) marking is permanent for the
duration of the sampling period and does not affect the
behaviors or survival of the marked individuals, and (4)
immigration and emigration are negligible on the time scale
of the estimate (Chew, 1959; Southwood, 1978). There is
little doubt that these assumptions are reasonably met by the
forager populations of ants. Variances and standard errors
were calculated as per Southwood (1978).
A seasonal natural history of the ant, Odontomachus brunneus 47
123
Nest excavation for forager distribution
Nests were hand excavated as described previously to deter-
mine how workers were vertically distributed throughout
the nest. Workers were collected as each chamber was
exposed, and the depth recorded. Within-nest locations were
combined at what appeared to be natural breaks; i.e., ants
located in the topmost chamber were observed to frequently
exit nests and thus were combined with the returning surface
foragers. Collection was done in the morning at similar
times to marking events so that workers would be engaged
in similar daily tasks at the time of excavation.
Using the number of marked foragers (m) in the nest
during the excavation and the proportion of recaptured
marked foragers (p) from the mark-recapture data, the
number of foragers (both marked and unmarked) within the
nest during the excavation was determined (F):
F ¼
m
p
ð1Þ
Equation 1, Foraging population in nest
The number of foragers actively foraging was then found
by subtracting the foraging population within the nest
(F) from the forager population estimate (N). The total
colony size was then determined by adding this value to the
number of ants collected during excavation.
Results
Annual cycle
The monthly excavations and census of Odontomachus
brunneus nests revealed the annual brood cycle of this
species (Fig. 1). Brood production began with the arrival of
the warmer temperatures in late spring and continued until
temperatures were too cool for the brood development in the
mid-fall (see Appendix for yearly temperatures). The first
larvae appeared in small numbers in late April and increased
to a peak in June. Pupal cocoons were present beginning in
May and increased through October with only a few pupae
left to eclose in November. The presence of wing buds in
dissected cocoons (Table 1) revealed the majority of rec-
ognizable pupae to be sexuals in the months of May and
June. Production of alate brood was recorded as late as the
second week of June (6/9/2008), after which colonies pro-
duced only worker brood. Initial brood production in early
spring was a mixture of both worker and sexual brood, not
solely sexuals, although sexual brood predominated. Both
sexual and worker pupae were found in the same nests.
There was no correlation between colony size and sexual
production (r
2
= 19%; df = 12; p = 0.14), or the total
number of pupae present and sexual pupae (r
2
= 21%;
df = 12; p = 0.12). Sexual production is slightly related to
the relative proportion of fat reserves available to the total
colony size (r
2
= 1.7%, df = 63; p = 0.06).
Adult alates appeared in mid-June (6/16–6/23), with a
few female alates remaining as late as the third week of July
(7/21/08), suggesting that mating flights are likely to have
occurred in mid-to-late June.
The lack of over-wintering brood shows this species to
have a heterodynamous annual cycle, with all over-winter-
ing individuals having eclosed prior to the winter inactivity.
Of note, in this regard, several colonies moved brood to the
surface of their nests in late October as temperatures were
beginning to decrease.
Queen presence
Of the 76 colonies excavated, only 26 were queen-right
(34%), with one of these colonies actually containing two
queens. It is unknown if both queens were reproductively
Fig. 1 The annual cycle of production in Odontomachus brunneus
shown as a mean percentage of the colony population in each category.
Nests were broodless throughout half of the year, followed by a period
of production of both sexual and worker brood. Sexual brood were
produced for only 2 months during the spring, followed by 4 months
of worker population
Table 1 Results of cocoon dissections
Month Total
pupae
Worker
pupae (%)
Sexual
pupae (%)
Unknown
pupae (%)
Sexual-to-worker
pupae ratio
May 25 12 20 68 1.67
June 81 8.64 16.05 75.31 1.86
48 L. M. Hart, W. R. Tschinkel
123
active as they had been processed for fat extraction and
therefore could not be dissected. There was no relationship
between the presence of a queen and colony size class (Chi-
square test 11.26; df = 2; p = 0.42).
Seasonal variation on nest depth/architecture
This study revealed changes in the nest depth and structure
through the annual cycle, with nests as deep as 60 cm in the
summer and as deep at 170 cm in the winter. Fall and spring
depths were intermediate between these extremes. In prep-
aration for the winter, the ants extended the final shaft down
60–100 cm from the maximum summer depth. Except on
unseasonably warm days, all winter inhabitants of the nest
were found in the final, nearly circular chamber at the end of
this long shaft.
Regressing the maximum depth of individual nests
against the number of ants showed that, while colony size
did not change much across seasons, the maximum depth
increased with the number of workers in all seasons, except
summer. During the summer, no nest was deeper than 60 cm,
and nest depth was unrelated to the number of workers
(Fig. 2). Winter nests were much deeper than summer nests
of comparable size, with spring and fall nests displaying a
transition between these two extremes.
On March 23, 2009 an excavation showed evidence of a
colony migrating upward in its nest, filling the lower shafts/
chambers with loosely packed soil. Around this time, all
colonies decreased their depth as the ants moved from their
period of winter inactivity into the reproductive period of
their annual cycle. In this particular nest, the majority of
ants were located in a chamber immediately above this shaft
with only a few workers closing the shaft behind the colony.
Dry, lean and fat weights of workers throughout
a one-year cycle
Figure 3 displays the monthly means of worker’s dry
weight, lean weight and percent fat. In April, ants emerging
from their winter diapauses ranged in dry weight from
approximately 2.14–5.27 mg, a range of 3.13 mg with a
mean dry weight of 3.75 mg (COV = 0.15). While the
range of dry weights remained similar throughout most of
the year, the mean dry and lean weights and percent fat
changed with season. However, these seasonal patterns of
weight and fat content were similar across colony size
classes, and are thus shown as monthly means without
regard to colony size in (Fig. 3). After their emergence from
winter inactivity in April and early May, colonies produced
mostly sexual brood (Fig. 1). During this period, fatness of
workers declined, so that the annual minimum for fatness
occurred in June with the completion of sexual production
(Fig. 3). Female alates had a mean of 26% body fat, higher
even than overwintering workers, suggesting that they
sequestered a large amount of available resources on an
individual basis. This spring decline in worker fatness was
largely responsible for the initial decline of their dry weight,
but their lean weight continued to decline until August,
causing dry weight to reach its annual minimum then. Once
the colony switched to worker production in June–July,
fatness slowly increased as new, young workers replaced
old ones that died. When worker production ceased in
October and there was no more brood to feed, worker
Fig. 2 Correlation of maximum nest depth to number of workers by
season. Mean summer nest depth was 43 cm (SD = 16.5) with a little
variation, resulting in a horizontal line. Winter nests were much greater
in depth than summer, with spring and fall occurring in transitional
locations between these extremes. Additionally, nests with a larger
workforce were capable of digging to greater depths in the cooler
seasons
Fig. 3 Lean weight, dry weights and percent fat of workers by month.
Error bars denote 95% confidence intervals. By decomposing the dry
weight into its components, it is apparent that the majority of the
fluctuation in mean dry weight of workers is due to a change in the fat
content
A seasonal natural history of the ant, Odontomachus brunneus 49
123
fatness, dry weight and lean weight increased dramatically
in the preparation for winter inactivity and spring alate
production. The late winter decline is probably associated
with the cost of maintaining workers. Florida winter soil
temperatures are not low enough to reduce metabolism to
near zero.
This change in weight and fatness can also be expressed
in energetic terms. Because fat has about twice the energy
content per mg as lean weight, it has a large effect on energy
content. Energy content of individuals was computed using
39.33 J/mg for fat and 18.87 J/mg for lean matter. The
annual patterns of individual and colony changes in energy
content are shown in Fig. 4a, b. In Fig. 4a, the early spring
peak in the energy content of larvae and pupae was the result
of an initially undetected mixed brood of sexuals and
workers. Later dissection of cocoons showed about 64% of
them to contain sexuals of both sexes. Because females
contain more than triple the energy of males (whose energy
content was somewhat less than workers), the brood energy
content peaked in June. During brood production, worker
energy content declined until about August (Fig. 4).
Thereafter, worker energy content slowly built up, peaking
sharply in the preparation for overwintering and production
of the next spring’s brood. Queen energy content varied
greatly, but was not related to season. Female alates were
present only in June, and were very expensive on an indi-
vidual basis, as was the queen.
Multiplying the energy per individual times the number
of individuals gave the total energy in each type of indi-
vidual, and summing these gave the total energy contained
in the colony. The percent of this total colony energy in each
type of ant is shown in Fig. 4b. Figure 4 shows that an
increasing amount of energy was found in brood during the
breeding season, peaking in October when pupae contain
about half of the colonies’ energy content. This is probably
the result of the replacement of old, lean workers with new,
fatter ones that will overwinter. Larvae are not present later
than September and pupae than October, so that by
November, all energy is found in the workers. Whereas the
seasonal pattern of energy in brood is strong and obvious
(ANOVA: p \ 0.001), these patterns do not differ for col-
onies of different sizes (ANOVA; p = 0.64). It is also
apparent from Fig. 4 how little energy the ants invest in
alates. However, only two colonies contained female alates,
so a little can be said about energetic patterns with respect to
colony size.
The shifts in the total energy content in Fig. 4b were the
result of the changing fat and lean mass in the workers
shown in Figs. 3 and 4a, and not of colony growth—colony
size did not change significantly across the monthly samples
(ANOVA: number of workers by collection month; p =
0.9) or even when these were lumped into four seasons (p =
0.2). However, it should be remembered that the samples
were terminal—no colony was followed across seasons, and
the statement of ‘‘no change’’ in colony size applied to the
population of colonies through the year. It is possible that
individual colonies increase in size during the year.
Vertical distribution of worker fatness within the nest
The fatness of workers was analyzed by nest level within
each month (Fig. 5). In 9 out of 12 months, workers were
the leanest in the uppermost stratum and the fattest in the
lowest (with a few exceptions). This pattern was absent
December–February for two reasons. First, the entire
worker population becomes fatter, and second the great
majority of the colony assembles in the lowest chambers,
with only occasional workers occurring in other levels,
especially the uppermost (note the large error bars in Jan-
uary and February). This was caused by a small sample size
of fatty workers to emerge on warmer days to forage.
Fig. 4 Seasonal distribution of energy content per individual (a) and
colony totals (b) for 5 types of ants through the annual cycle. The
investment in worker pupae gradually increased, peaking in October
with the last brood of the year. Alates represent a small proportion of
the total energy of the colony when they are present. These shifting
investments represent both energy gained from forage, and energy
from metabolic reserves
50 L. M. Hart, W. R. Tschinkel
123
Assuming that the proportion of fat decreases with
worker age (as it does in other ants; Porter and Jorgensen,
1981; Tschinkel, 1993, 1998), there appeared to be strati-
fication by age within the nests of O. brunneus for the
warmer months of the year (Fig. 3).
Estimation of the forager population
Through mark-recapture and excavation, colony sizes and
the proportion of foragers per colony were determined.
Female alates were included in the population census of
workers as they were observed repeatedly foraging in the
majority of the nests. Four colonies with large standard
errors of the estimate were not used, and the estimates were
based on the remaining six. Recapture samples ranged from
8 to 28 workers, of which 27–67% were marked (mean
42%). Colonies ranged in total size from 41 to 107 workers
with 51–88% of the workforce participating in foraging,
with no relationship to colony size. There is a degree of
uncertainty in these estimates that stems from a generous
definition of foragers. All workers returning to the nest from
a distance and those departing beyond 10 cm were desig-
nated foragers. It would include, for example, workers
leaving the nest for the first time on exploratory forays, or
midden workers venturing farther than 10 cm. These esti-
mates therefore need to be verified by the future studies.
Excavation: location of marked and unmarked workers
Figure 6 shows that ants marked as foragers were distrib-
uted throughout the nest upon excavation, but unequally
among the levels. A Chi-square test with an expected equal
distribution of marked foragers within each nest showed that
the marked foragers were not equally distributed among
nest levels (Chi-square = 63.78831, p = 0.0002), but were
more abundant in the upper levels with a smaller, but
varying number in the middle and bottom levels. In all nests,
brood was found in the bottom and middle chambers toge-
ther with both marked and unmarked ants. Brood was also
found in the top chambers of nests F10 and F12. It should be
noted that the total number of ants collected was not the total
individuals per colony as excavations were performed while
colonies were actively foraging. An estimate of colony sizes
is provided in Table 2.
Discussion
The life cycle of O. brunneus is strongly seasonal. Colonies
produce brood for 6 months and are broodless for 6 months.
Because northern Florida represents the northern range limit
of the genus Odontomachus, it is possible that this long
broodless period is induced by less favorable environmental
conditions. For example, in Myrmica rubra, brood pro-
duction is the greatest in mid-range and declines at both
range extremes (Elmes et al., 1999). O. brunneus sexuals are
produced in mixed broods at the beginning of the season,
concurrent with a marked decrease in the colony’s fat
reserves, suggesting a high energetic investment toward the
production of these individuals. Reserves recover slowly
through foraging during the summer’s worker production,
and rapidly after brood production ceases in October. This
seasonal pattern is generally similar to those of S. invicta
and P. badius in which early spring sexual production is
associated with a large decline in worker fat stores
(Tschinkel, 1993, 1998). Presumably, this pattern evolved
in all three species because it is advantageous to produce
sexuals early in the season before much forage is available,
but it is probably also driven by the inability to forage
Fig. 5 Monthly within-nest allocation of worker fat by depth. Error
bars denote 95% confidence intervals. Throughout most of the year,
leaner (older) workers were found in the upper chambers of all nests
with fatter (younger) workers in the lower regions
Fig. 6 Location(s) of marked and unmarked ants by nest. Marked
foragers were distributed throughout nests, but this distribution was
unequal, with the majority located in the upper region of the nests
A seasonal natural history of the ant, Odontomachus brunneus 51
123
during the cold of winter and early spring. With respect to
brood rearing, stored fat and forage are probably fungible to
a large degree, for that when forage is less available, stored
fat can substitute. The extreme of this life cycle is found in
Prenolepis imparis in which both sexuals and workers are
produced from huge fat stores, entirely without feeding
(Tschinkel, 1987), thus completely separating the foraging
phase of the seasonal cycle during which reserves are stored
from the brood production phase when these reserves are
used. Considering that these four species are not closely
related, it seems likely that the association of declining fat
stores with early sexual and worker production is wide-
spread in seasonal ants. How workers convert fat stores into
larval food is unknown, but two routes seem possible—the
laying of trophic eggs, or the production of nutritional
secretions (or both).
The seasonal cycle can thus be understood as a single
entity driven by both season and nutritional status. Accu-
mulating data on ant seasonal cycles (Kipyatkov, 1996,
2001; Passera and Keller, 1987; Ricks and Vinson, 1972;
Rissing, 1987; Tschinkel, 1993, 1998) is revealing that even
for ants in warm temperate zones, the accumulation of
metabolic reserves late in the year not only is necessary for
overwintering, but also is an integral part of early spring
sexual production in the next year. Although sexuals may be
produced from overwintered brood in some boreal ant
species (Gamanilov and Kipyatkov, 2000;Ho
¨
lldobler and
Wilson, 1990), early production from metabolic reserves is
probably widespread among temperate ants for the simple
reason that sexual production as early as possible in the
spring makes the success of colony founding more likely.
Earlier incipient colonies have a competitive advantage
over later ones, and can accumulate more workers and
reserves before the stress of overwintering sets in. In a
number of ant species, as in O. brunneus, the early brood is
actually composed of both workers and sexuals, with sex-
uals predominating (MacKay, 1981; Tschinkel, 2006). The
reasons for mixed broods are not clear. Candidate reasons
include ecological conditions that promote high survival of
founding queens, primitive social organization and worker
life span synchronized to the seasons so that there is a high
worker turnover at one season. For O. brunneus, while
worker life span is unknown, it is likely that workers live for
at least a year—worker life span generally increases with
body size, and O. brunneus is a large-bodied ant (Tschinkel,
personal observation). The high rate of worker production
late in the season also suggests that there is high turnover in
workers in late summer. Because we did not track individual
colonies through the year, we cannot say with certainty that
colonies grew during the warm season. However, mean
colony size showed no significant variation during the year
suggesting that the colony growth during late summer is
modest if not absent.
O. brunneus founds new colonies independently, without
the company of workers, and thus female sexuals contain a
lot of metabolic reserves, especially fat. Nevertheless,
founding queens of this species are semi-claustral, that is,
they forage during the founding period (Ho
¨
lldobler and
Wilson, 1990, 2005, 2009). Thus, their 26% fat is lower than
the threshold value of about 50% for independent founding
reported by Keller and Passera (1989). However, female
alate ants gain their metabolic reserves during early adult
life, and the females we measured may not all have been
flight-ready.
These colony-level attributes are essential to under-
standing colonies as a unit. It is also important to consider
the life cycle of the individual ants that make up the colony.
It has been shown in more derived species such as Pogo-
nomymex badius (Tschinkel, 1998
) that individual ants
follow a particular sequence throughout their lives: they
eclose on the brood pile where they remain as brood care
workers, as they age and new ants eclose, the older ants
move away from the brood to perform nest maintenance
tasks, and ultimately end their lives as foragers (Bourke and
Franks, 1995; Oster and Wilson, 1978; Tschinkel, 1998).
This process, termed adaptive demography (Wilson, 1985),
is commonly accepted as the core of division of labor
of most species. This clear age-related task distribution
Table 2 Population estimates resulting from mark-recapture and nest excavation
Nest
ID
Number
marked
and
released
(a)
Recapture
sample (n)
Proportion
of
recaptured
ants marked
(p)
Forager
estimate
(N)
Standard
error of
the mean
Total nest
population
(excavation)
Total
marked ants
(excavation)
Total
foraging
population
in nest (F)
Foragers
out of
nest
(N - F)
Estimated
total
colony
size
Percent of
colony
involved in
foraging
F1 31 21 0.67 47 3.33 71 18 27 20 91 51
F4 10 11 0.27 37 6.33 35 8 29 7 42 87
F5 21 20 0.30 70 9.06 76 20 67 3 79 88
F6 18 14 0.36 50 7.95 41 7 20 31 72 70
F8 27 28 0.43 63 5.12 95 22 51 12 107 59
F10 11 8 0.50 22 3.67 25 3 6 16 41 54
52 L. M. Hart, W. R. Tschinkel
123
appears to be more flexible in O. brunneus (Figs. 5, 6):
while, there are distinct separations throughout most of the
annual cycle, the locations of active foragers (Fig. 6) show
that workers can and do migrate within their nests. It is
possible that an age-related division of labor is less apparent
for this species in natural settings, or that foragers are not as
restricted to the upper regions of the nest. Perhaps they
deliver food directly to the larvae, rather than to interme-
diate transport workers. Stratification also appears to be
relaxed in the month of January (Fig. 5), during the coolest
time of the year. This lack of apparent stratification can
probably be attributed to the fact that all workers have
gained similar amounts of fat for overwintering. Thus, their
ages can no longer be divined through their fat content. It is
also likely that the proportion of the workers that are young
is higher during this period because most of the overwin-
tering workers were probably born during the summer and
fall.
The finding of foraging female alates in both O. brunneus
and Neoponera apicalis (Fresneau and Dupuy, 1988) lends
support to either a lack of division of labor or a more
primitive version of this distribution. Because nests produce
a very few sexuals per season, it seems maladaptive to allow
these alates, who represent a large colony energy investment
(Fig. 4) to engage in this risky endeavor. Why colonies do
not retain these female alates under safer conditions until
such time that they mate as is seen in derived species
(Tschinkel, 1993) is not yet understood. It is possible that
the colonies require the alates to forage in order to maintain
sufficient resources not only to nourish the brood but also to
maintain the fat stores of the alates until it is time for them to
leave the nest. Also, because Ponerine queens found colo-
nies in a semi-claustral fashion, it may just be in the very
nature of the female alate to forage during the early portion
of her life.
While a few studies of this nature have been performed
on primitive ants, there is a similarity in the foraging of
O. brunneus with Neoponera apicalis, a fellow Ponerine
species that also engages a large proportion of its work-
force, including female alates, in foraging (Fresneau, 1985;
Fresneau and Dupuy, 1988). Similar to more derived species,
O. brunneus displays a spatial partitioning of workers by age,
with the youngest residing primarily in the lower regions of
nests and the older in the upper region, likely performing
primarily as foragers. Due to the diversity of combinations of
both primitive and derived traits, Fresneau and Dupuy (1988)
suggest that the subfamily of Ponerinae is a potential model
for studying the evolution of social organization; such
studies could yield powerful insight into how and when the
various traits of eusociality were derived within the diverse
family of ants.
The high proportion of queenless nests has several pos-
sible origins. It is unlikely that we failed to find queens that
were present, as the excavation procedure rarely missed
ants. True queenlessness is also unlikely, for workers can
only produce males. It is more likely that O. brunneus is
polydomous. If this is the case, then the average colony
occupied three nests and average colony size (as opposed to
nest size) would be approximately triple our estimate. It
would also seem that each nest functions as a largely
independent unit. Similar polydomy was found in Campo-
notus socius in which the average colony occupied 2.3 nests
(Tschinkel, 2005).
Acknowledgments We would like to thank Dr. Josh King for
assistance in project design, as well as Carli Seeba, Jacob Kline, Justin
Diepenbrock, David Hart and Joshua Gold for their hours of field
assistance.
Appendix
References
Beshers S.N. and Fewell J.H. 2001. Models of division of labor in
social insects. Annu. Rev. Entomol. 46: 413-440
Brian M.V. and Elmes G.W. 1974. Production by the ant Tetramorium
caespitum in a Southern English heath. J. Anim. Ecol. 43: 889-903
Bourke A.F.G. and Franks N.R. 1995. Social Evolution in Ants.
Princeton, New Jersey: Princeton University Press.
Cerquera L.M. and Tschinkel W.R. 2009. The nest architecture of the
ant, Odontomachus brunneus. J. Insect Sci. 10:64
Seasonal temperatures and rainfall during April 2008–May 2009
Month Avg.
Temp. (°C)
Avg.
Temp (F)
Rainfall
(mm)
Rainfall
(in)
Apr-08 19 67 3.6 0.14
May-08 24 76 2.8 0.11
Jun-08 27 81 5.08 0.2
Jul-08 28 82 4.1 0.16
Aug-08 28
a
82
a
15 0.59
Sep-08 26 79 1 0.04
Oct-08 20 68 4.3 0.17
Nov-08 13 56 4.8 0.19
Dec-08 14
b
57
b
1.3 0.05
Jan-09 11 52 1 0.04
Feb-09 11 51 2.8 0.11
Mar-09 17 62 4.3 0.17
Apr-09 19 67 9.7 0.38
May-09 24 76 6.6 0.26
a
Tropical Storm Fay influenced a decrease in temperatures in mid and
late August 2008
b
Mid-December had several warm days in the high 70s and low 80s
A seasonal natural history of the ant, Odontomachus brunneus 53
123
Chew R.M. 1959. Estimation of ant colony size by the Lincoln Index
Method. J. N.Y. Entomol. Soc. 67: 157-161
Elmes G.W. 1987. Temporal variation in colony populations of the ant
Myrmica sulcinodis. II: Sexual production and sex ratios. J. Anim.
Ecol. 56: 573-583
Elmes G.W., Wardlaw J.C., Nielsen M.G., Kipyatkov V.E., Lopatina
E.B., Radchenko A.G. and Barr B. 1999. Site latitude influences
on respiration rate, fat content and the ability of worker ants to
rear larvae: A comparison of Myrmica rubra (Hymenoptera:
Formicidae) populations over their European range. Eur. J. Ento-
mol. 96: 117-124
Fresneau D. 1985. Individual foraging and path fidelity in a ponerine
ant. Insect. Soc. 32: 109-116
Fresneau D. and Dupuy P. 1988. A study of polyethism in a ponerine
ant: Neoponera apicalis (Hymenoptera, Formicidae). Anim. Behav.
36: 1389-1399
Gadgil M. and Bossert W.H. 1970. Life historical consequences of
natural selection. Am. Nat. 104: 1-24
Gamanilov P.N. and Kipyatkov V.E. 2000. Seasonal cycle of
development and annual brood production in the ant, Myrmica
scabrinodis Nyl. (Hymenoptera, Formicidae), in the environs of
St. Petersburg. Vestn. Zool. 34: 55-64
Golley F.B. and Gentry J.B. 1964. Bioenergetics of the Southern
harvester ant, Pogonomyrmex badius. Ecology 45: 217-25
Ho
¨
lldobler B. and Wilson E.O. 1990. The Ants. Cambridge, Massa-
chusetts: Harvard University Press.
Ho
¨
lldobler B. and Wilson E.O. 2005. The rise of the ants: A phylo-
genetic and ecological explanation. Proc. Natl Acad. Sci. USA
102: 7411-7414
Ho
¨
lldobler B. and Wilson E.O. 2009. The Superorganism. New York,
NY: W.W. Norton & Company Ltd.
Keller L. and Passera L. 1989. Size and fat content of gynes in relation
to the mode of colony founding in ants (Hymenoptera; Formici-
dae). Oecologia 80: 236-240
Kipyatkov V.E. 1993. Annual cycles of development in ants: diversity,
evolution and regulation. In: Proc. Coll. Social Insects Russian-
speaking Section IUSSI, Vol. 2 (Kipyatkov V.E., Ed), Socium, St.
Petersburg, pp 25-48
Kipyatkov V.E. 1995. Role of endogenous rhythms in regulation of
annual cycles of development in ants (Hymenoptera, Formici-
dae). Entomol. Rev. 74: 1-15
Kipyatkov V.E. 1996. Godichnyie tsikly razvitiya murav’ev (Annual
Cycles of Development in Ants). Unpublished Doctoral Thesis, St.
Petersburg University, St. Petersburg.
Kipyatkov V.E. 2001. Seasonal life cycles and the forms of dormancy
in ants (Hymenoptera: Formicoidea). Acta Soc. Zool. Bohem. 65:
211-238
Kruk-De Bruin M., Rost L.C.M. and Draisma F.G.A.M. 1977.
Estimates of the number of foraging ants with the Lincoln-Index
Method in relation to the colony size of Formica polyctena. J.
Anim. Ecol. 46: 457-470
Laskis K.O. and Tschinkel W.R. 2009. The seasonal natural history of
the ant, Dolichoderus mariae, in northern Florida. J. Insect Sci. 9:2
MacKay W.P. 1981. A comparison of the nest phenologies of three
species of Pogonomymex harvester ants (Hymenoptera: Formici-
dae). Psyche 88: 25-74
Oster G.F. and Wilson E.O. 1978. Caste and Ecology in the Social
Insects. Princeton, New Jersey: Princeton University Press.
Passera L. and Keller L. 1987. Energy investment during the differen-
tiation of sexual and workers in the Argentine ant, Iridomyrmex
humilis (Mayr). Mitt. Schweiz. Entomol. Ges. 60: 249-260
Porter S.D. and Jorgensen C.D. 1981. Foragers of the harvester ant,
Pogonomyrmex owyheei: A disposable caste? Behav. Ecol.
Sociobiol. 9: 247-256
Powell S. and Tschinkel W.R. 1999. Ritualized conflict in Odonto-
machus brunneus and the generation of interaction-based task
allocation: a new organizational mechanism in ants. Anim. Behav.
58: 965-972
Ricks B.L. and Vinson S.B. 1972. Changes in nutrient content during
one year in workers of the imported fire ant. Ann. Entomol. Soc.
Am. 65: 135-138
Rissing S.W. 1987. Annual cycles in worker size of the seed-harvester
ant Veromessor pergandei. Behav. Ecol. Sociobiol. 10: 117-124
Ryti R.T. and Case T.J. 1986. Overdispersion of ant colonies: a test of
hypotheses. Oecologia 69: 446-453
Seal J.N. and Tschinkel W.R. 2006. Colony productivity of the fungus-
gardening ant Trachymymex septentionalis (Hymenoptera: For-
micidae) in a Florida pine forest. Ann. Entomol. Soc. Am. 99:
673-682
Seal J.N. and Tschinkel W.R. 2007. Energetics of newly-mated queens
and colony founding in the fungus-gardening ants Cyphomyrmex
rimosus and Trachymymex septentrionalis (Hymenoptera: For-
micidae). Physiol. Entomol. 32: 8-15
Smith C.R. and Tschinkel W.R. 2009. Ant Fat Extraction with a
Soxhlet Extractor. Cold Spring Harbor Protocols. 3 pp
Southwood T.R.E. 1978. Ecological Methods. New York, NY:
Chapman and Hall
Soxhlet F. 1879. Die Gewichtsanalytische Bestimmung des Milch-
fettes. Polytechn. J. 232: 461-465
Tschinkel W.R. 1987. Seasonal life history and nest architecture of the
winter-active ant, Prenolepis impairis. Insect. Soc. 34: 143-164
Tschinkel W.R. 1991. Insect sociometry, a field in search of data.
Insect. Soc. 38: 77-82
Tschinkel W.R. 1993. Sociometry and sociogenesis of the fire ant
Solenopsis invicta during one annual cycle. Ecol. Monogr. 63:
425-457
Tschinkel W.R. 1998. Sociometry and sociogenesis of colonies of the
harvester ant, Pogonomyrmex badius: worker characteristics in
relation to colony size and season. Insect. Soc. 45: 385-410
Tschinkel W.R. 2004. The nest architecture of the Florida harvester
ant, Pogonomyrmex badius. J. Insect Sci. 4:19pp
Tschinkel W.R. 2005. The nest architecture of the ant, Camponotus
socius. J. Insect Sci. 5:9
Tschinkel W.R. 2006. The Fire Ants. Belknap Press of Harvard Univ.
Press. Cambridge, MA.
Wilson E.O. 1985. The sociogenesis of insect colonies. Science 228:
1489-1495
54 L. M. Hart, W. R. Tschinkel
123