ArticlePDF Available

Clonal ant societies exhibit fertility-dependent shifts in caste ratios

Authors:

Abstract and Figures

Caste differentiation leading to reproductive division of labor is the hallmark of insect societies. Insect colonies typically contain mated queens that reproduce and workers with reduced fertility that undertake the tasks required for colony maintenance and development. Despite the prediction that the proportion of morphological castes should vary to enhance the fitness of colony members in response to environmental conditions, shifts in caste ratios have so far only been reported in a competitive situation. Societies of the ant Cerapachys biroi have evolved in an extraordinary way, in that queens and all workers reproduce through obligatory thelytokous parthenogenesis. Because workers of C. biroi represent the main reproductive force of the colony, the presence of such unmated queens seems puzzling. Here, we show that societies of C. biroi alter caste ratios by considerably increasing the production of queens when larvae are reared by sterile individuals in 2 situations: when senescent colonies are faced with food shortage or when well-fed larvae are reared by callow workers due to persisting plentiful resources. In the absence of these opposite conditions, larvae mostly develop into workers. Additional experiments suggest that these results are consistent with a contact pheromone to which larvae could be exposed when cared for by fertile individuals. In this species in which reproduction mainly relies on young workers with finite fertility, a self-regulated mechanism of caste differentiation could allow the enhancement of colony growth through worker production in fertile conditions or the restoration of colonial fertility through queen production in senescent societies. Copyright 2011, Oxford University Press.
Content may be subject to copyright.
Behavioral Ecology
doi:10.1093/beheco/arq182
Advance Access publication 15 November 2010
Original Article
Clonal ant societies exhibit fertility-dependent
shifts in caste ratios
Emmanuel Lecoutey, Nicolas Cha
ˆ
line, and Pierre Jaisson
Laboratoire d’Ethologie Expe´rimentale et Compare´e, EA4443, Universite´ Paris 13, 99 avenue J.B.
Cle´ment, 93430 Villetaneuse, France
Caste differentiation leading to reproductive division of labor is the hallmark of insect societies. Insect colonies typically contain
mated queens that reproduce and workers with reduced fertility that undertake the tasks required for colony maintenance and
development. Despite the prediction that the proportion of morphological castes should vary to enhance the fitness of colony
members in response to environmental conditions, shifts in caste ratios have so far only been reported in a competitive situation.
Societies of the ant Cerapachys biroi have evolved in an extraordinary way, in that queens and all workers reproduce through
obligatory thelytokous parthenogenesis. Because workers of C. biroi represent the main reproductive force of the colony, the
presence of such unmated queens seems puzzling. Here, we show that societies of C. biroi alter caste ratios by considerably
increasing the production of queens when larvae are reared by sterile individuals in 2 situations: when senescent colonies are
faced with food shortage or when well-fed larvae are reared by callow workers due to persisting plentiful resources. In the absence
of these opposite conditions, larvae mostly develop into workers. Additional experiments suggest that these results are consistent
with a contact pheromone to which larvae could be exposed when cared for by fertile individuals. In this species in which
reproduction mainly relies on young workers with finite fertility, a self-regulated mechanism of caste differentiation could allow
the enhancement of colony growth through worker production in fertile conditions or the restoration of colonial fertility
through queen production in senescent societies. Key words: cannibalism, caste differentiation, Cerapachys biroi, fertility signal,
polyphenism, thelytokous parthenogenesis. [Behav Ecol 22:108–113 (2011)]
A
n important transition in the evolution of insects was the
shift from a solitary life to societies of organisms exhibiting
a division of labor based on specialized castes. This adaptive
division of labor is at the basis of the ecological success of insect
societies and is achieved either by the nestmates assuming
temporary behavioral functions and/or by the evolution of
permanently differentiated morphological castes (Oster and
Wilson 1978; Ho
¨
lldobler and Wilson 1990). Even if genetic
influences on female caste determination or task specializa-
tion have been demonstrated (reviewed in Smith et al. 2008),
female eggs are usually totipotent, and a developmental
switch during the larval stage controlled by nutritional, social,
and other environmental factors results in pronounced
physical and physiological specializations among female
castes (Wilson 1971; Oster and Wilson 1978; Wheeler 1986;
Ho
¨
lldobler and Wilson 1990; Vargo and Passera 1991;
Wheeler 1991; O’Donnell 1998; Karsai and Hunt 2002). In
ants, particularly, queens are typically adapted for dispersal,
mating, colony founding, and are specialized in egg laying. In
contrast, wingless and generally smaller than queens, workers
exhibit reduced ovaries and lack a spermatheca in most spe-
cies. They rarely reproduce and mostly care for the brood,
forage, and defend the colony (Wilson 1971; Ho
¨
lldobler
and Wilson 1990). Moreover, in approximately 15% of the
ant genera, the worker force can also be differentiated into
multiple castes (Oster and Wilson 1978; Ho
¨
lldobler and
Wilson 1990).
Because natural selection acts both at the individual and at
the colony level, it has been suggested that the various physical
castes exhibited by social insects have evolved to enhance the
inclusive fitness of colony members with the prediction that
caste ratios reflect the colony’s needs and should vary with
environmental factors, such as predation, competition, or food
availability (Wilson 1971; Oster and Wilson 1978; Herbers 1980;
Lumsden 1982; Wilson 1985; Schmid-Hempel 1992, Hasegawa
1997). However, this prediction for adaptive shifts in caste ra-
tios has received little evidence and has so far only been re-
ported in a competitive situation (Passera et al. 1996; Harvey
et al. 2000; McGlynn and Owen 2002; Yang et al. 2004).
Far from the social organization, commonly observed in
insect societies, colonies of the ant Cerapachys biroi have
evolved in an extraordinary way, in that there is neither a sex-
ual reproduction (mating) nor a sterile caste. Egg laying is
distributed among all nestmates through obligatory thelyto-
kous parthenogenesis (Tsuji and Yamauchi 1995) and without
any social hierarchy (Ravary and Jaisson 2004). Reproduction
is then divided between 2 discrete morphological castes. On
the one hand, workers that lay eggs (1 or 2 eggs/month) and
care for the brood in their youth, later cease to reproduce as
they become foragers (after 3–4 months on average; Ravary
and Jaisson 2004). On the other hand, atypical queens display
higher (5 eggs/month on average; Ravary and Jaisson 2004)
and lasting laying capacities (at least 1 year; Lecoutey E, per-
sonal observation). Apterous (Brown 1975, Morisita et al.
1989) and devoid of spermathecae (Tsuji and Yamauchi
1995), these ergatoid queens represent less than 6% per col-
ony in field colonies (Ravary F, personal communication) and
are only involved in reproduction and brood care (Ravary and
Jaisson 2004).
Because workers of C. biroi reproduce through thelytoky and
represent the main reproductive force of the colony (94% of
workers laying 1–2 eggs during 3–4 months on average vs. 6%
Address correspondence to E. Lecoutey. E-mail: lecouteymanu
@yahoo.fr.
Received 10 September 2009; revised 6 September 2010; accepted
24 September 2010.
The Author 2010. Published by Oxford University Press on behalf of
the International Society for Behavioral Ecology. All rights reserved.
For permissions, please e-mail: journals.permissions@oup.com
by guest on February 10, 2013http://beheco.oxfordjournals.org/Downloaded from
of ergatoid queens laying 5 eggs/month on average; Ravary
and Jaisson 2004), the presence of such unmated queens
seems puzzling. Because the organization of every social insect
species has evolved in response to the complex environmental
pressures and opportunities unique to its evolutionary history,
we hypothesized that queen production in C. biroi colonies
could be a response to colonial sterility. To determine the role
of the queens and how their production is regulated, we
therefore monitored colonies’ responses to varying internal
(fertility) and external (food resources) colonial parameters.
MATERIALS AND METHODS
Study species
As for most Cerapachyinae ants (Ho
¨
lldobler and Wilson 1990),
colonies of C. biroi are specialized predators of the brood of
myrmicine ants on which they prey through massive raids. In
Okinawa (Japan) and Taiwan, colonies of C. biroi are character-
ized by a phasic reproductive cycle composed of 2 alternating
phases of activity: During the foraging phase (16.07 6 2.27
days), old workers explore for prey, whereas a single cohort of
larvae develop synchronously, nursed by young workers and
queens. Then, during the following statary phase (17.92 6
0.97days), larvae pupate, all workers stay in the nest, and new
eggs are laid by young workers and queens. After the eggs hatch
and the young workers emerge, also synchronously, a new for-
aging phase begins (Ravary and Jaisson 2002, Ravary et al. 2006).
Workers of C. biroi therefore separate into 2 behavioral sub-
castes according to age and thus to reproductive capacities:
Old extranidal workers, with no or weak ovarian activity, are
specialized in foraging and rarely care for larvae, whereas
young intranidal workers remain inside the nest and perform
most of the egg laying as well as most of the nursing activities
(Ravary and Jaisson 2004). Finally, ergatoid queens differ mor-
phologically from workers by a higher ovariole number (4–6
instead of 2 or 3 for workers), the presence of ocelli, more or
less developed vestigial eyes, and an alitrunk with variable
degree of fusion of the thoracic tergites (Ravary and Jaisson
2004)(Figure 1). These ergatoid queens are not all identical
and present a well coordinated and graded development of
morphological characteristics. The deeper the thoracic sutures
are, the more developed the vestigial eyes and the ovaries
(Ravary and Jaisson 2004, Lecoutey E, personal observation).
Rearing conditions of larvae
Larvae of C. biroi were subjected to 8 rearing conditions: They
were reared by 4 homogeneous groups (fertile queens, young
workers, old workers, and callow workers) in 2 feeding con-
ditions (either well fed or undernourished). Among workers,
callow workers are not yet fertile, young workers (1–2 months
old) are fully fertile, whereas old workers (more than 4
months old) present no or weak laying capacities. This allows
us to test for the potential effects of different parameters known
to influence caste differentiation in insect societies (Wilson
1971; Oster and Wilson 1978; Wheeler 1986; Ho
¨
lldobler and
Wilson 1990; Vargo and Passera 1991; Wheeler 1991; O’Donnell
1998; Karsai and Hunt 2002).
Indeed, in many social insects, the queens produce a fertility
signal that commits larvae to the worker caste (Wheeler 1986;
Ho
¨
lldobler and Wilson 1990; Vargo and Passera 1991; Keller
and Nonacs 1993). However, because young workers of C. biroi
are also fertile, we predicted that, whatever the food resour-
ces, queens and young workers will induce worker develop-
ment in larvae. We used old workers as a control. However,
because we cannot rule out potential behavioral manipulation
of larval fate by adults (Bourke and Ratnieks 1999, Hammond
et al. 2002) and because old workers do not care for the brood
as much as young workers and queens (Ravary and Jaisson
2004), we also used callow workers as a second control (indi-
viduals which exhibit nursing activities but are considered un-
fertile because they have not yet developed ovaries) to avoid
any possible misleading conclusions about the respective in-
fluence of feeding and nursing. This second control also
mimics a highly fertile colony where larvae are almost exclu-
sively reared by callow workers after a previous foraging cycle
with plentiful resources.
In addition, it is well known that the amount of the food
given by workers to developing larvae profoundly influences
the future adult morphology, including the features critically
correlated with fecundity. Well-fed larvae usually differentiate
into gyne or soldier, whereas in the absence of sufficient nu-
trition larvae will develop as workers (Wilson 1971; Oster and
Wilson 1978; Wheeler 1986; Ho
¨
lldobler and Wilson 1990;
Wheeler 1991; O’Donnell 1998; Karsai and Hunt 2002).
Therefore, we predicted that, in absence of a fertility signal
(i.e., in callow and old workers colonies), only well-fed larvae
will develop into queens. According to our hypotheses, we
expected queen production in only 2 of the 8 experimental
conditions: in well-fed colonies of callow workers as well as in
well-fed colonies of old workers.
Experimental procedure
Colonies of C. biroi were set up in nests consisting of a single
rectangular (length: 2 cm 3 width: 1 cm 3 height: 0.6 cm) cham-
ber dug in the center of a circular plastic box (8 cm in diameter
and 5 cm in height) half filled with a layer of plaster. Nests were
covered and closed with a glass plate and a red-colored Plexiglas
to allow observations. The plaster was humidified 3 times a week
to keep the nests damp. Colonies were kept at 28C, 70% humid-
ity, and a 12/12 light-dark (LD) photoperiod.
Food and (sub)caste influences
Each of the 32 experimental colonies (8 different conditions
performed with experimental colonies originating from 4 wild
Figure 1
(a) Ergatoid queen and (b) worker of the parthenogenetic ant
Cerapachys biroi. Scale bar, 1 mm.
Lecoutey et al.
Shifts in caste ratios 109
by guest on February 10, 2013http://beheco.oxfordjournals.org/Downloaded from
colonies collected in Taiwan in 1997, 2000, and 2001) was
composed of 28.6 6 7.2 (mean 6 standard deviation [SD])
first instar larvae of the same age and 50 individuals of 1 of the
4 following types: queens (unknown ages and fertile), callow
workers (1 day old and sterile), young workers (1–2 months
old and fertile), or old workers (more than 4 months old and
no or weak laying capacities). Callow workers were isolated at
emergence thanks to their light cuticle. Young fertile workers
were initially picked up at emergence, but experimental col-
onies were kept 1 reproductive cycle before starting the ex-
periments. Old workers used for the experiments were only
foragers with the darkest cuticle. For each of these 4 types of
individuals, 2 colonies were set up with alternate diets: either
well fed or undernourished. During the foraging phase of the
cycle, colonies were fed 3 times a week with white pupae of
Tetramorium bicarinatum. Before supplying food, larvae were
counted. For every 10 larvae, well-fed colonies received 4 pu-
pae of T. bicarinatum, whereas undernourished colonies only
got 1. For instance, if a well-fed colony contained 25 larvae, it
was supplied with 10 prey items. Because cannibalism regu-
larly occurs in this myrmecophaegous species (Lecoutey E, un-
published data), this could be a mean for larvae to selfishly
influence their caste development toward queens or for work-
ers to suppress queen-destined larvae (Bourke and Ratnieks
1999; Faustino et al. 2002; Wenseleers and Ratnieks 2004;
Wenseleers et al. 2005; Ribeiro et al. 2006; Ru
¨
ger et al.
2008). For each colony, the proportion of cannibalism was
measured with the following ratio: (initial number of larvae2
number of prepupae)/initial number of larvae. Indeed, using
an indirect measure was necessary because we were unable to
separate cannibalism among larvae (self-determination) from
cannibalism of larvae by workers (caste regulation) because the
same larva can be killed by workers, eaten by larvae, and vice
versa. Because old workers tend to forage more than care for
the brood (Ravary and Jaisson 2004), nests of all colonies were
closed to limit differences in feeding and nursing levels
between experimental conditions.
Signal volatility
The results (see below) showed that, in well-fed conditions, col-
onies of fertile workers strongly prevent queen development in
larvae, whereas colonies of callow workers do not. To investi-
gate whether this inhibitory effect was due to a contact or
a volatile pheromone, we replicated these 2 experimental con-
ditions in a new experiment with colonies this time set up side
by side in a special nest composed of 2 chambers (1 cm 3
2 cm each, 1 for each condition), separated by a fine mesh.
Determination of the caste ratio
The potential influence of the rearing caste, food resources,
and signal volatility on larval fate was assessed in each colony
at the emergence of the new cohort of individuals. Callow
individuals were then isolated in a separate nest and dis-
sected 2 weeks later under a stereomicroscope to assign
them as workers or queens according to the morphological
characteristics of each caste (see above and Figure 1).
Statistical analyses
Food, (sub)caste influences
After arcsine transformation of the proportion of queens pro-
duced, the effects of treatments, food, and (sub)castes were
compared with an Analysis Of Variance (ANOVA) followed by
a post hoc Honestly Significant Difference (HSD) Tukey’s test.
Cannibalism
We used an ANOVA after arcsine transformation to compare
the effect of food and (sub)castes on cannibalism. Correla-
tions between the proportion of queens produced and the
proportion of cannibalism were performed using Pearson’s
correlation test.
Signal volatility
The proportion of each caste produced in well-fed colonies
(n ¼ 4) of callow workers separated by a mesh from a well-
fed colony of young fertile workers were compared with those
obtained in well-fed colonies of callow workers (n ¼ 4) using
an ANOVA after arcsine transformation. Statistical signifi-
cance was accepted at a ¼ 0.05. All statistical analyses were
implemented with StatXact-7 and Statistica 8 softwares.
RESULTS
Of the 8 experimental conditions (Figure 2), numerous
queens were reared in undernourished colonies of old work-
ers and well-fed colonies of callow workers (mean 6 SD:
79.8% 6 14 and 75.8% 6 10.7, respectively, n ¼ 4 colonies
for each condition). A moderate proportion of queens was
produced in well-fed colonies of old workers (21.2% 6 11.2,
n ¼ 4), whereas low proportions of queens were reared in the
5 other conditions (mean 6 maximum SD: 0–8.1% 6 5.7,
n ¼ 4 for each condition) (Figure 2).
Consequently, C. biroi larvae highly differ in their caste fate
according to their rearing conditions (one-way ANOVA,
treatment effect: F
7,24
¼ 31.314; P , 10
25
). Overall, there
was a highly significant influence of the rearing individuals
on caste differentiation (one-way ANOVA, (sub)caste effect:
F
3,28
¼ 8.473; P , 0.001). Indeed, callow workers and old
workers reared globally more queens than young fertile
workers and queens (post hoc HSD Tukey’s test, all P ,
0.05). However, although there was no food effect on caste
differentiation (one-way ANOVA, food effect: F
1,30
¼ 0.427;
P ¼ 0.518), the type of rearing individuals cannot solely
explain the results obtained. Indeed, a highly significant
interaction existed between feeding conditions and the type
of individuals that reared larvae (factorial ANOVA: F
3,24
¼
33.353; P , 10
25
).
In addition, when separated by a single fine mesh from
a well-fed colony of young fertile workers, a well-fed colony
of callow workers still reared a high proportion of queens
(64.7% 6 10, n ¼ 4). Because this rate did not differ from
Figure 2
Caste differentiation in larvae of Cerapachys biroi according to
different rearing and feeding conditions. Data for each experimental
condition are mean proportions (6SD) of larvae that developed into
worker (white) or ergatoid queen (black) in the 4 replicates. Sizes of
each caste produced are given in columns. Different letters represent
the rearing conditions that differ significantly (factorial ANOVA
followed by a post hoc HSD Tukey’s test). Feeding conditions: W ¼
Well-fed and U ¼ Undernourished.
110 Behavioral Ecology
by guest on February 10, 2013http://beheco.oxfordjournals.org/Downloaded from
well-fed and isolated colonies of callow workers (one-way AN-
OVA, volatile pheromone effect: F
1,6
¼ 2.508; P ¼ 0.164), this
result rules out the involvement of a volatile pheromone and
is consistent with the widespread hypothesis of a fertility sig-
naling transmitted by contact (Monnin 2006; Peeters and
Liebig 2009).
Another point is that although cannibalism regularly oc-
curred in colonies and was significantly higher (one-way AN-
OVA, food effect: F
1,30
¼ 13.366; P ¼ 0.00097) in
undernourished conditions (mean 6 SD: 39.25% 6 19.8,
n ¼ 16) compared with well-fed conditions (mean 6 SD:
14.69% 6 17.7, n ¼ 16), it cannot account for the results
obtained: There was no difference in cannibalism according
to the type of individuals that reared the larvae (one-way
ANOVA, (sub)caste effect: F
3,28
¼ 0.758; P ¼ 0.526), as well
as no interaction between feeding conditions and the type of
individuals that reared the larvae (factorial ANOVA: F
3,24
¼
2.417; P ¼ 0.091).
Moreover, queen production was never significantly corre-
lated with cannibalism within colonies of C. biroi (Pearson’s
correlation test, n ¼ 32, R ¼ 0.033, P ¼ 0.853). This turned
out to be true in undernourished conditions (n ¼ 16, R ¼
0.172, P ¼ 0.530), well-fed conditions (n ¼ 16, R ¼ 20.006,
P ¼ 0.980), fertile conditions (n ¼ 20, R ¼ 20.294, P ¼ 0.209),
and in sterile conditions (n ¼ 12, R ¼ 20.017, P ¼ 0.954).
Behavioral observations during the experiments also reve-
aled that adults do not starve larvae and that the latter never
present scars.
DISCUSSION
Our study clearly demonstrates that only a low proportion of
larvae differentiate into queens under the influence of fecund
individuals (fertile workers and queens). These low rates of
queens obtained under fertile conditions were similar to the
ones observed in field colonies ( Ravary F, personal commu-
nication) and in colonies reared in the laboratory (3.7–6.3%,
Ravary and Jaisson 2004). Moreover, the results show that
there was no food effect on larval differentiation and that
queen production was not correlated with cannibalism what-
ever the experimental conditions. This indicates that, despite
their high mobility and active feeding behavior and even in
the absence of a fertility signal, larvae of C. biroi do not seem
to selfishly influence their caste development toward queens
by securing more food.
Our results can therefore be explained according to our
main hypothesis that the fertility of the nursing individuals
modulates the queen production. Because fertile individuals
reared low proportion of queens in opposite feeding condi-
tions, we suggest that, whatever the food availability, fecund
individuals (mainly young workers and queens) could produce
a fertility signal that induces a worker differentiation in the de-
veloping larvae they care for.
However, because there is an interaction between the type of
rearing individuals and the food resources, we also suggest that
when this signal is weak (callow and old workers), queen pro-
duction could be affected by food availability because it may
lead to differences in the pheromone regime experienced
by larvae. Indeed, well-fed conditions allow larvae to pupate
in a few days (mean 6 SD: 12.6 days 6 1.7, n ¼ 16). There-
fore, most well-fed larvae reared by callow workers developed
into queens because they pupate before workers are able to
lay eggs (i.e., before workers produce the fertility signal). On
the contrary, undernourished conditions lead to an increase
(24.5 days 6 6.8, n ¼ 16) of the typical duration of the larval
stage. As 20 days old workers may produce the fertility signal
because they usually lay eggs during their first statary phase
(Lecoutey E, personal observation), we hypothesize that un-
dernourished larvae initially reared by callow workers were
submitted to the fertility signal during their last larval instar
and thus developed into workers. Consequently, we predict
that if callow workers were continually replaced by younger
ones every few days, a high proportion of larvae would de-
velop into queens even if undernourished.
Furthermore, whatever the amount of food received, the
production of the fertility signal may decrease in old workers
due to their ovarian/physiological senescence, which could ex-
plain the increased production of queens. This is reinforced by
the fact that only few eggs were laid during the statary phase in
each experimental colony of old workers. It has also been
reported that, with age, a behavioral shift occurs in old workers,
which then hardly ever exhibit brood care in the presence of
younger workers (Ravary and Jaisson 2004). Therefore, as for-
aging activity is regulated by brood satiety in insect societies
(Pankiw 2004; Ravary et al. 2006; Pankiw 2007; Mas and
Ko
¨
lliker 2008) and mainly performed by old workers in
C. biroi (Ravary and Jaisson 2004), the caste ratio difference
observed between the 2 feeding conditions in colonies of old
workers could be interpreted according to the amount of
brood stimulation: In starving conditions, high brood stimu-
lation may have led most of the old workers to neglect nursing
activities for foraging (i.e., nest exploration in our experiment
because the nests were closed to limit differences in feeding
and nursing levels between experimental conditions),
whereas in well-fed conditions, old workers may have provided
more care to satiated larvae and hence more fertility signal.
This latter hypothesis leads to the testable prediction of a pos-
itive correlation between ‘brood contact by workers’ and
food availability in colonies of old workers.
In the light of all these results, we therefore suggest that
worker differentiation is probably induced by fecund individ-
uals via a nonvolatile signal of fertility to which larvae are
exposed during brood care. The action of this possible con-
tact pheromone is likely to be dose-dependent and not an all-
or-nothing mechanism: The less fertility signal a larva receives,
the more pronounced the queen differentiation will be. This
hypothesis is supported by 2 observations: First, queens ex-
hibit a well coordinated and graded development of morpho-
logical characteristics ranging from different ovariole
numbers to variable degrees of fusion of the thoracic tergites.
Second, the most developed queens (deep thoracic sutures,
vestigial eyes, and 6 ovarioles) were mostly produced in sterile
conditions (Lecoutey E, personal observation).
In most social insect species, allocating egg laying to 1 or
a few reproductives has some obvious advantages (Oster and
Wilson 1978; Ho
¨
lldobler and Wilson 1990). However, every
rose has its thorn: The queens’ death usually leads the colony
to dwindle to its death (but see Peeters and Molet 2010). In
contrast, as thelytokous reproduction is distributed among all
nestmates (Ravary and Jaisson 2004), colonies of C. biroi are
virtually everlasting. In natural colonies of C. biroi composed
of queens and workers of different ages, brood care is per-
formed by young individuals (callow and fertile individuals)
and foraging activities by old workers (Ravary and Jaisson
2004). Consequently, whatever the food resources, fertile
colonies may produce only a few queens thus promoting
colonial productivity by compelling most larvae to develop
into workers. However, workers of C. biroi represent the main
reproductive force of the colony (94% of workers laying 1–2
eggs during 3–4 months on average vs. 6% of ergatoid
queens laying 5 eggs/month on average; Ravary and Jaisson
2004) and only reproduce in their youth (1 or 2 eggs each
month during 3–4 months on average, Ravary and Jaisson
2004). Specialized in myrmicine brood predation, colonies
of C. biroi may sometimes have to face long periods of food
shortage resulting in a weak generation turnover. In that
Lecoutey et al.
Shifts in caste ratios 111
by guest on February 10, 2013http://beheco.oxfordjournals.org/Downloaded from
case, colonies are then mainly composed of old workers with
decreasing fertility. This may allow many larvae to differen-
tiate into queens especially when food resources are low
because a high brood stimulation increases foraging, thus
reducing nursing activities. Because queens lay between 4
and 8 eggs during each statary phase (Ravary and Jaisson
2004), producing them instead of workers in a undernour-
ished and senescent society of C. biroi could boost the
reproductive capacity of the colony by a fourth fold in a sole
reproductive cycle.
In senescent societies, such a mechanism also equips the
colony with lastingly fecund individuals (laying activities for at
least 1 year; Lecoutey E, personal observation) who keep the
society fertile even through long periods of starvation. In this
obligatory-dependent colony founding species, the produc-
tion of many queens may also promote budding toward more
favorable ecological niches.
In addition, a high proportion of queens is also obtained
when larvae are reared by callow workers. Although rarer, this
situation can occur when highly fertile colonies experience 2
consecutive cycles of foraging bonanza. In that case and due
to the centrifugal polyethism of social insects, well-fed larvae
are mainly reared by many callow workers (originating from
the previous reproductive cycle) that do not become fertile
before they pupate. In these flourishing conditions, the pro-
duction of numerous queens in C. biroi could be a prerequisite
to colonial dispersion through budding events during the
following cycles.
We therefore propose that a self-regulated mechanism of
caste differentiation emerges from the uncommon reproduc-
tive division of labor of C. biroi in which young workers with
finite fertility care for the brood and senescent ones forage.
This could be an adaptive mean for C. biroi colonies to mod-
ulate their caste ratio in response to their level of fecundity.
The genus Cerapachys is composed of more than 240 essen-
tially monogynous species (Brown 1975, www.antweb.org),
among whom some possess permanently wingless queens
(i.e., ergatoid queens). It seems likely that C. biroi evolved
from an ancestor that had ergatoid queens that mated. As
a result of thelytokous parthenogenesis evolving in queens
and workers, the queens lost their reproductive monopoly.
It is likely that some ecological constraints (i.e., finding
a mate) have then favored thelytoky instead of mating. In such
a species where all individuals reproduce, the production of
thelytokous queens thus became optional and these latter are,
at present, mostly produced under the extreme conditions
emphasized in our study.
It seems plausible that selection at the colony level may favor
the production of queens under extreme conditions (food
shortage and senescent colonies or ad libitum food resources
and highly fertile colonies), whereas promoting colony growth
through worker production when these extreme situations are
absent. Interestingly, queen differentiation could also be
driven by an individual selection for selfish reproduction. This
occurs in Pristomyrmex punctatus, another parthenogenetic ant
in which it has been recently shown that queens can form
a cheater lineage isolated from workers (Dobata et al.
2009). Such parasitic traits are expected to evolve in geneti-
cally heterogeneous colonies. However, these possibilities
remain to be investigated in C. biroi.
FUNDING
French Ministry of Research; Ecole Doctorale Galile´e of the
University Paris 13 (to E.L.).
We thank Fabien Ravary and 2 anonymous referees for comments on
the manuscript, as well as Jean-Francxois Hamel for statistical advices.
REFERENCES
Bourke AFG, Ratnieks FLW. 1999. Kin conflict over caste determina-
tion in social Hymenoptera. Behav Ecol Sociobiol. 46:287–297.
Brown WLJ. 1975. Contribution toward a reclassification of the
formicidae. V. Ponerinae, tribes Platythyreini, Cerapachyini, Cylin-
dromyrmecini, Acanthostichini, and Aenictoginiti. Search Agricul-
ture, 5. Entomol (Ithaca). 15:1–115.
Dobata S, Sasaki T, Mori H, Hasegawa E, Shimada M, Tsuji K. 2009.
Cheater genotypes in the parthenogenetic ant Pristomyrmex puncta-
tus. Proc R Soc B. 276:567–574.
Faustino CD, Silva-Matos EV, Mateus S, Zucchi R. 2002. First record of
emergency queen rearing in stingless bees (Hymenoptera, Apinae,
Meliponini). Insectes Soc. 49:111–113.
Hammond RL, Bruford MW, Bourke AFG. 2002. Ant workers selfishly
bias sex ratios by manipulating female development. Proc R Soc B.
269:173–178.
Harvey JA, Corley LS, Strand MR. 2000. Competition induces adaptive
shifts in caste ratios of a polyembryonic wasp. Nature. 406:183–186.
Hasegawa E. 1997. The optimal caste ratio in polymorphic ants: esti-
mation and empirical evidence. Am Nat. 149:706–722.
Herbers JM. 1980. On caste ratios in ant colonies: population re-
sponses to changing environments. Evolution. 34:575–585.
Ho
¨
lldobler B, Wilson EO. 1990. The ants. Cambridge (MA): Belknap
Press.
Karsai I, Hunt JH. 2002. Food quantity affects traits of offspring in the
paper wasp Polistes metricus (Hymenoptera: vespidae). Environ
Entomol. 31:99–106.
Keller L, Nonacs P. 1993. The role of queen pheromones in social
insects: queen control or queen signal? Anim Behav. 45:787–794.
Lumsden CJ. 1982. The social regulation of physical castes: the super-
organism revived. J Theor Biol. 95:749–781.
Mas F, Ko
¨
lliker M. 2008. Maternal care and offspring begging in social
insects: chemical signalling, hormonal regulation and evolution.
Anim Behav. 76:1121–1131.
McGlynn TP, Owen JP. 2002. Food supplementation alters caste allo-
cation in a natural population of Pheidole flavens, a dimorphic leaf-
litter dwelling ant. Insectes Soc. 49:8–14.
Monnin T. 2006. Chemical recognition of reproductive status in social
insects. Ann Zool Fennici. 43:515–530.
Morisita M, Kubota M, Onoyama K, Ogata K, Terayama M, Kondoh M,
Imai HT. 1989. A guide for the identification of Japanese ants. I.
Ponerinae, Cerapachyinae, Pseudomyrmecinae, Dorylinae, and
Leptanillinae (Hymenoptera: formicidae). [in Japanese]. Tokyo
(Japan): Myrmecological Society of Japan.
O’Donnell S. 1998. Reproductive caste determination in eusocial
wasps (Hymenoptera: vespidae). Annu Rev Entomol. 43:
323–346.
Oster GF, Wilson EO. 1978. Caste and ecology in the social insects.
Princeton (NJ): Princeton University Press.
Pankiw T. 2004. Brood pheromone regulates foraging activity of honey
bees (Hymenoptera: apidae). J Econ Entomol. 97:748–751.
Pankiw T. 2007. Brood pheromone modulation of pollen forager turn-
around time in the honey bee (Apis mellifera L.). J Insect Behav. 20:
173–180.
Passera L, Roncin E, Kaufmann B, Keller L. 1996. Increased soldier
production in ant colonies exposed to intraspecific competition.
Nature. 397:630–631.
Peeters C, Liebig J. 2009. Fertility signalling as a general mechanism
of regulating reproductive division of labor in ants. In: Gadau J,
Fewell J, editors. Organization of insect societies : from genome
to socio-complexity. Cambridge (MA): Harvard University Press.
p. 220–242.
Peeters C, Molet M. 2010. Colonial reproduction and life histories. In:
Lach L, Parr C, Abbott K, editors. Ant Ecology. Oxford (UK):
Oxford University Press. p. 159–176.
Ravary F, Jahyny B, Jaisson P. 2006. Brood stimulation controls the
phasic reproductive cycle of the parthenogenetic ant Cerapachys
biroi. Insectes Soc. 53:20–26.
Ravary F, Jaisson P. 2002. The reproductive cycle of thelytokous colo-
nies of Cerapachys biroi Forel (Formicidae, Cerapachyinae). Insectes
Soc. 49:114–119.
Ravary F, Jaisson P. 2004. Absence of individual sterility in thelytokous
colonies of the ant Cerapachys biroi Forel (Formicidae, Cerapachyi-
nae). Insectes Soc. 51:67–73.
112 Behavioral Ecology
by guest on February 10, 2013http://beheco.oxfordjournals.org/Downloaded from
Ribeiro Mde F, Wenseleers T, Santos-Filho PS, Alves DA. 2006. Mina-
ture queens in stingless bees: basic facts and evolutionary hypothe-
ses. Apidologie. 37:191–206.
Ru
¨
ger MH, Fro
¨
ba J, Foitzik S. 2008. Larval cannibalism and worker-
induced separation of larvae in Hypoponera ants: a case of conflict
over caste determination? Insectes Soc. 55:12–21.
Schmid-Hempel P. 1992. Worker castes and adaptive demography.
J Evol Biol. 5:1–12.
Smith CR, Toth AL, Suarez AV, Robinson GE. 2008. Genetic and
genomic analyses of the division of labour in insect societies. Nat
Rev Genet. 9:735–748.
Tsuji K, Yamauchi K. 1995. Production of females by parthenogenesis
in the ant Cerapachys biroi. Insectes Soc. 42:333–336.
Vargo EL, Passera L. 1991. Pheromonal and behavioral queen control
over the production of gynes in the Argentine ant Iridomyrmex
humilis (Mayr). Behav Ecol Sociobiol. 28:161–169.
Wenseleers T, Ratnieks FLW. 2004. Tragedy of the commons in Meli-
pona bees. Proc R Soc B. 271:S310–S312.
Wenseleers T, Ratnieks FLW, Ribeiro Mde F, Alves Dde A, Imperatriz-
Fonseca VL. 2005. Working-class royalty: bees beat the caste system.
Biol Lett. 1:125–128.
Wheeler DE. 1986. Developmental and physiological determinants of
caste in social Hymenoptera: evolutionary implications. Am Nat.
128:13–34.
Wheeler DE. 1991. The developmental basis of worker caste polymor-
phism in ants. Am Nat. 138:1218–1238.
Wilson EO. 1971. The insect societies. Cambridge (MA): Harvard
University Press.
Wilson EO. 1985. The sociogenesis of insect colonies. Science.
28:1489–1495.
Yang AS, Christopher M, Nijhout HF. 2004. Geographic variation
of caste structure among ant populations. Curr Biol. 14:514–519.
Lecoutey et al.
Shifts in caste ratios 113
by guest on February 10, 2013http://beheco.oxfordjournals.org/Downloaded from
... (Modified after Matsuura 2011). The numbered references are given in the "References of (Lecoutey et al. 2011;Ravary and Jaisson 2004), which could be considered either as a discrete worker caste (Ravary and Jaisson 2004) or atypical, unmated queens (Lecoutey et al. 2011). The production of these unmated individuals with larger fecundity is tuned in response to colony needs. ...
... (Modified after Matsuura 2011). The numbered references are given in the "References of (Lecoutey et al. 2011;Ravary and Jaisson 2004), which could be considered either as a discrete worker caste (Ravary and Jaisson 2004) or atypical, unmated queens (Lecoutey et al. 2011). The production of these unmated individuals with larger fecundity is tuned in response to colony needs. ...
... Aside from these ants where queens reproduce asexually, earlier research also revealed several remarkable cases of obligate or near-obligate worker thelytoky. For example, workers produce female offspring parthenogenetically in the ants Cerapachys biroi (Lecoutey et al. 2011;Ravary and Jaisson 2004;Tsuji 1995) and Pristomyrmex punctatus (Dobata et al. 2009;Dobata et al. 2011;Nishide et al. 2007;Tsuji 1988;Tsuji 1990;Tsuji 1995;Tsuji and Dobata 2011) and in some populations of the Neotropical ant Platythyrea punctata Hartmann et al. 2005;Heinze and Hölldobler 1995;Kellner et al. 2010;Kellner and Heinze 2011a;Kellner and Heinze 2011b;Schilder et al. 1999b) (Table 1). In these species, there is a trend for the queen caste to be lost (Table 1), and reproductive division of labour is divided among the workers relatively equally, with all or most of the workers laying eggs when young and switching to foraging later in life (in C. biroi and P. punctatus, Ravary and Jaisson 2004;Tsuji 1988;Tsuji 1990;Tsuji 1995); P. ...
... The colonies of C. biroi include two worker subcastes that differ in morphology, behavior and fertility levels, referred to as high and low reproductive individuals, or HRIs and LRIs (Teseo et al. 2013(Teseo et al. , 2014. These are also called intercastes and workers by Ravary and Jaisson (2004) and ergatoid queens and workers by Lecoutey et al. (2011). Low reproductive individuals (LRIs) have two ovarioles and lay eggs exclusively during their first four-five months of life, during which they also provide care to the developing brood. ...
... From a functional perspective, colonies therefore comprise two groups: old LRIs that behave as reproductively-inactive foragers, as well as young LRIs and HRIs of all ages that act as nurses and are fertile. In C. biroi, the production of HRIs is regulated via a feedback system based on the actual fertility level of the colony (Lecoutey et al. 2011). The more fertile a colony is, the greater its proportion of HRIs and young LRIs, and the less HRIs it produces, and vice versa. ...
... In social Hymenoptera, CHC signatures often are correlated with fertility, and are used by reproducers to signal their presence and reproductive status (reviewed by Monnin 2006). As a previous study on C. biroi indicated that the HRI regulation system is not based on volatile chemical signals (Lecoutey et al. 2011), we hypothesized that non-volatile, cuticular chemical cues could signal fertility and/or reproductive status. We further hypothesized that these cues act as primer pheromones that inhibit the HRI developmental trajectory. ...
Article
Full-text available
Although cuticular hydrocarbons (CHCs) have received much attention from biologists because of their important role in insect communication, few studies have addressed the chemical ecology of clonal species of eusocial insects. In this study we investigated whether and how differences in CHCs relate to the genetics and reproductive dynamics of the parthenogenetic ant Cerapachys biroi. We collected individuals of different ages and subcastes from several colonies belonging to four clonal lineages, and analyzed their cuticular chemical signature. CHCs varied according to colonies and clonal lineages in two independent data sets, and correlations were found between genetic and chemical distances between colonies. This supports the results of previous research showing that C. biroi workers discriminate between nestmates and non-nestmates, especially when they belong to different clonal lineages. In C. biroi, the production of individuals of a morphological subcaste specialized in reproduction is inversely proportional to colony-level fertility. As chemical signatures usually correlate with fertility and reproductive activity in social Hymenoptera, we asked whether CHCs could function as fertility-signaling primer pheromones determining larval subcaste fate in C. biroi. Interestingly, and contrary to findings for several other ant species, fertility and reproductive activity showed no correlation with chemical signatures, suggesting the absence of fertility related CHCs. This implies that other cues are responsible for subcaste differentiation in this species.
... Reproduction occurs in stereotypical colony cycles, and a new cohort of workers emerges synchronously approximately every 34 days [29][30][31][32] . Colonies collected in the field usually contain tens to a few hundreds of workers, and in the laboratory, colonies composed of as few as five workers are fully viable. ...
... Because workers are totipotent and queens are absent, the number of individuals produced by a given genotype during each colony cycle serves as a straightforward measure of genotype fitness. Finally, colonies of C. biroi consist of two worker subcastes that differ in morphology, behaviour and reproductive physiology [29][30][31] (Fig. 1). High-reproductive individuals (HRIs), which have four to six ovarioles, usually do not forage and lay up to eight eggs per reproductive cycle, and low-reproductive individuals (LRIs), which have two ovarioles, lay at most two eggs per cycle and become sterile foragers at four to five months of age [29][30][31] . ...
... Finally, colonies of C. biroi consist of two worker subcastes that differ in morphology, behaviour and reproductive physiology [29][30][31] (Fig. 1). High-reproductive individuals (HRIs), which have four to six ovarioles, usually do not forage and lay up to eight eggs per reproductive cycle, and low-reproductive individuals (LRIs), which have two ovarioles, lay at most two eggs per cycle and become sterile foragers at four to five months of age [29][30][31] . ...
Article
Full-text available
In social species, the phenotype and fitness of an individual depend in part on the genotype of its social partners. However, how these indirect genetic effects affect genotype fitness in competitive situations is poorly understood in animal societies. We therefore studied phenotypic plasticity and fitness of two clones of the ant Cerapachys biroi in monoclonal and chimeric colonies. Here we show that, while clone B has lower fitness in isolation, surprisingly, it consistently outcompetes clone A in chimeras. The reason is that, in chimeras, clone B produces more individuals specializing in reproduction rather than cooperative tasks, behaving like a facultative social parasite. A cross-fostering experiment shows that the proportion of these individuals depends on intergenomic epistasis between larvae and nursing adults, explaining the flexible allocation strategy of clone B. Our results suggest that intergenomic epistasis can be the proximate mechanism for social parasitism in ants, revealing striking analogies between social insects and social microbes.
... (65). Lecoutey et al. (65) referred to individuals with more than two ovarioles as ergatoid queens. ...
... A divergence-dating analysis inferred that the origin of thelytoky occurred relatively recently in M. smithii, approximately 0.5-1. 65 106), or L-types (22) (hereafter ergatoid queens). Individuals of a third morph, distinctly larger than the ergatoid queens, also have a spermatheca and four ovarioles, but possess three ocelli. ...
Article
Full-text available
Female parthenogenesis, or thelytoky, is particularly common in solitary Hymenoptera. Only more recently has it become clear that many eusocial species also regularly reproduce thelytokously, and here we provide a comprehensive overview. Especially in ants, thelytoky underlies a variety of idiosyncratic life histories with unique evolutionary and ecological consequences. In all eusocial species studied, thelytoky probably has a nuclear genetic basis and the underlying cytological mechanism retains high levels of heterozygosity. This is in striking contrast to many solitary wasps, in which thelytoky is often induced by cytoplasmic bacteria and results in an immediate loss of heterozygosity. These differences are likely related to differences in haplodiploid sex determination mechanisms, which in eusocial species usually require heterozygosity for female development. At the same time, haplodiploidy might account for important preadaptations that can help explain the apparent ease with which Hymenoptera transition between sexual and asexual reproduction. Expected final online publication date for the Annual Review of Entomology Volume 58 is December 03, 2013. Please see http://www.annualreviews.org/catalog/pubdates.aspx for revised estimates.
... Given limited resources available to a colony, the optimal resource allocation must be achieved through a trade-off between various colony functions. In many insect colonies these include alteration of the number and size of offspring produced (Passera et al. 1996;Lecoutey et al. 2011;Schmidt et al. 2011), or morphological and behavioural plasticity (Wilson 1971;Ishikawa and Miura 2012). ...
Article
Full-text available
Phenotypic diversification among colony members often leads to formation of physical castes which are morphologically specialised for particular tasks within the colony. The relative abundance of these castes and their body sizes represent two key aspects of the demography of a colony that may reflect the colony’s needs and conditions, and ultimately influence its survival and reproductive success. In a recently discovered social trematode, Philophthalmus sp., which exhibits a reproductive division of labour, the role of competition and colony composition in shaping reproductive success and behaviour of colony members has been documented. As body size variation within physical castes often influences colony efficiency, we investigated how the growth of reproductive and non-reproductive morphs of Philophthalmus sp. responds to competitive pressure, and to other attributes of colony demography such as colony size and composition. Our survey of a natural population and in vitro experiments demonstrate that the growth of reproductive colony members reflects the interaction between colony composition and the presence of a competitor, while the non-reproductive members simply grow larger in the presence of the intra-host competitor, Maritrema novaezealandensis. Furthermore, the close association between the volume and reproductive capacity of the reproductive members corroborates an adaptive value of colony member size in determining the fitness of the trematode colony as a whole. The present study is the first to demonstrate a fitness consequence, and identify the determinants, of the growth of colony members in social trematodes.
... The reproductive cycle of Cerapachys biroi is similar to that of some army ants in that C. biroi colonies produce brood in distinct cohorts, synchronized with cycles of alternating statary and nomadic phases (RAVARY & JAISSON 2002, RAVARY & al. 2006). Cerapachys biroi has received particular attention because, unlike most ants, all workers can produce diploid eggs through thelytokous parthenogenesis, although workers differ in their number of ovaries and their potential reproductive output (TSUJI & YAMAUCHI 1995, RAVARY & JAISSON 2004, LECOUTEY & al. 2011). This means that any colony fragment can theoretically found a new population. ...
Article
Full-text available
Cerapachys biroi FOREL, 1907 is a small, inconspicuous ant that has spread around the world through human commerce. To examine the worldwide distribution of C. biroi, we compiled and mapped specimen records from > 100 sites. We documented the earliest known C. biroi records for 24 geographic areas (countries and island groups), including several for which we found no previously published records: Comoro Islands, Guadeloupe, Îles Éparses, Indonesia, Madagascar, the Seychelles, Turks & Caicos Islands, and the US Virgin Islands. All continental records of Cerapachys biroi come from Asia, where populations show notable geographic variation in morphology, suggesting that the species is native to this region. Conversely, the lack of obvious morphological variation among C. biroi specimens from outside mainland Asia suggests that C. biroi is exotic to these regions. Outside Asia, all records of C. biroi come from islands, possibly due to reduced competition with dominant ants in island habitats. Perhaps the many dominant native and exotic ant species in continental regions have prevented widespread establishment of C. biroi.
... The model relies on a strong assumption: clonal reproduction. In ants, this is the exception rather than the rule (but see4647484950), and it raises some questions about our results. In particular, to what extent can conclusions from clonal models of adaptive dynamics be applied to populations of sexual organisms? ...
Article
Full-text available
This study aims to better understand the evolutionary processes allowing species coexistence in eusocial insect communities. We develop a mathematical model that applies adaptive dynamics theory to the evolutionary dynamics of eusocial insects, focusing on the colony as the unit of selection. The model links long-term evolutionary processes to ecological interactions among colonies and seasonal worker production within the colony. Colony population dynamics is defined by both worker production and colony reproduction. Random mutations occur in strategies, and mutant colonies enter the community. The interactions of colonies at the ecological timescale drive the evolution of strategies at the evolutionary timescale by natural selection. This model is used to study two specific traits in ants: worker body size and the degree of collective foraging. For both traits, trade-offs in competitive ability and other fitness components allows to determine conditions in which selection becomes disruptive. Our results illustrate that asymmetric competition underpins diversity in ant communities.
... In the latter, adaptive adjustments to varying environmental conditions are more feasible than in the former, via altered investments in individuals to maintain colony efficiency. This is achieved through alteration of caste ratios, where caste allocation decisions vary according to environmental fluctuations [1,2] or through morphological and behavioural plasticity [3]. ...
Article
Full-text available
Members of some social insects adjust their behaviours depending upon social context. Such plasticity allows colonies to sustain efficiency of the whole without the cost of additional production of individuals or delayed responses to perturbations. Using the recently discovered social clonal stage of trematode parasites, we investigated whether members of the reproductive caste adjust their defensive behaviour according to the local availability of non-reproductive defensive specialists, and if so whether the plasticity affects the short-term reproductive success of reproductive morphs. In vitro experiments demonstrated plasticity in competitive interactions of the reproductive morphs depending on the number of non-reproductive defensive specialists present nearby, which lead to differences in reproductive output at the individual level. This study provides support for the benefit of maintaining non-reproductive morphs in competitive situations, arising through socially mediated behavioural plasticity.
Article
Division of labor among castes in social insect colonies increases ergonomic efficiency and colony-level fitness, and has played a key role in the ecological success of social insects. Knowledge of the factors that regulate castes is important for understanding adaptive social organization. Our previous study on the termite Reticulitermes speratus demonstrated that the presence of a pair of nymphoid reproductives during development affected offspring caste ratios. In the present study, we investigated further the influence of individual neotenics on offspring caste ratios. Parthenogenetically-produced offspring were reared in worker-tended experimental colonies with the addition of different forms (nymphoid or ergatoid) and numbers of neotenics, and compared the caste ratios of the offspring between the different experimental treatments. We found that all offspring in worker-only tended colonies became nymphs, while a proportion of offspring in colonies with a single neotenic (with the exception of male ergatoids) differentiated into workers. These results show offspring caste ratios are influenced by the presence of single female ergatoids, single female and male nymphoids, while they remain unaffected by the presence of male ergatoids.
Article
Full-text available
We review some aspects of the biology of the ant Pristomyrmex punctatus, in which the winged queen caste is absent and wingless females reproduce by thelytokous parthenogenesis. The majority of females have two ovarioles, whereas up to 50% of colonies contain large-bodied females which have four ovarioles. We call the former workers and the latter ergatoid queens. Males are rare. Some ergatoid queens have a spermatheca, but no inseminated individual has been found so far. Castes are morphologically defined, and workers engage in asexual reproduction in all colonies regardless of the presence of ergatoid queens. In colonies containing only workers, reproductive division of labor is regulated by age-polyethism: All young workers reproduce and fulfill inside-nest roles, and old workers become sterile and fulfill outside-nest roles. Colonies are founded by fission or budding, and consequently neighboring colonies are often re-lated. Nevertheless, populations are multi-colonial, with strong hostility among neighboring colonies. A genetic analysis revealed that colonies often have multiple genotypes (parthenogenetic lineages), and suggested that the majority of those lineages can produce both workers and ergatoid queens. However, a lineage in a population in central Japan pro-duces only ergatoid queens. We define these queens as cheaters, as they fulfill no other task than oviposition and therefore depend on the work force of other non-cheater lineages. Ergatoid queens in cheater lineages have three distinct ocelli, but those in non-cheater lineages usually have zero to two. As cheaters are likely to be horizontally transmitted, we draw an analogy to transmissible cancers. The coexistence of cheaters and non-cheaters is discussed in the frameworks of multilevel selection in the short term, and local extinction-immigration in the long term. However, many things re-main to be studied, such as the developmental origin of the ergatoid queens, the frequency of sexual reproduction, and how colony identity is maintained.
Article
Full-text available
Some stingless bees are known to produce both large queens, reared from larger royal cells, and small "miniature" queens, reared from worker cells. Here we review what is known about miniature queens, and evaluate some major evolutionary hypotheses as to why they are produced. One hypothesis - that miniature queens are females who selfishly evade an intended worker fate - is shown to receive significant support. In particular, there is increasing evidence that the decision to become a miniature queen may be under genetic control of the developing females themselves. In addition, data from several species show that females gain significant fitness benefits from doing so, since miniature queens are frequently observed heading colonies and often are as productive as normal-sized queens. On the other hand, in some species miniature queens have a reduced fecundity or may have lower chances of being chosen as a new queen. This shows that the strategy may also have costs. Queens of the genus Melipona, which are also reared from worker-sized cells, are suggested to have the same evolutionary origins as miniature queens.
Article
Full-text available
textlessptextgreaterAmong social Hymenoptera, the evolution of the worker caste has reached its apex in the ants, in which some taxa have evolved complex physical worker caste systems. Diverse worker caste systems can be generated through regulation of three aspects of larval growth: critical size, growth parameters, and reprogramming of these factors. Even the most complex caste systems could have evolved simply by the addition of revised programs to the end of an ancestral developmental pathway for workers. Worker castes in ants provide a system in which to study the evolution of reaction norms and developmental switches. In ants, the simplest developmental switch, revision of critical size alone, does not lead to discontinuous phenotypes. Only when changes in growth rules are tied to the decision to revise critical size are distinct phenotypes produced from the alternative developmental programs. The addition of ever more physical castes may be limited by both developmental and ecological factors.
Article
Colonies of Formica obscuripes were subjected to stress, either by resource depletion or artificial predation, in order to examine the effects of changing environmental parameters on their population structures. This species is polymorphic, with a division of labor based on morphological differences among individuals, and the frequencies of the morphological castes should be sensitive to colony disturbance. Changes in population structure relative to controls were observed in experimental nests. In undisturbed nests, the frequencies of majors and medias fluctuated from spring through fall significantly more than in stressed nests. Although mean caste proportions were similar in controls and experimentals, there was reduced variability around those means in the latter group. This result is interpreted with caste ergonomic theory, and is consistent with the hypothesis that environmental uncertainty affects the evolution of caste ratios. The data suggest that for mound-building formicines, the fluidity of caste proportions is the feature of evolutionary significance rather than the absolute frequencies.
Article
A simple model is proposed for estimation of the optimal caste composition in polymorphic ants. The model gives the optimal caste composition as the highest point on a hypersurface that represents the total efficiency of the colony as a whole. The hypersurface is the product of all partial efficiency surfaces for all tasks. The model was applied to the dimorphic ant Colobopsis nipponicus to test its utility, The ergonomic efficiency of the colony as a whole was estimated from the quantitative ability of each caste to perform each role. When all partial functions of the major workers were included in the model, the predicted optimal ratio closely reflects the observed caste ratio in the field. In addition, the pattern of changes in brood productivity upon experimental manipulation of the caste ratio was consistent with the pattern predicted from the model. These results suggest that the model is useful for estimations of the optimal caste ratio and that the caste ratio in field colonies is tuned to maximize the ergonomic efficiency of the colony as a whole through colony-level selection.
Article
The eusociality threshold, marked by cooperative brood care by mother and daughters, is crossed by both halictine bees and polistine wasps. Beyond the eusociality threshold lies the continued evolution and elaboration of social systems. Two particularly important thresholds of advanced colony organization are marked by morphological commitment to caste: 1) a commitment to queen-worker dimorphism, and 2) a commitment to morphological diversity within the worker form. Queen-worker dimorphism is a defining character of highly eusocial caste systems. If morphological differences exist between queens and workers, the pattern of queen and worker development must diverge no later than the larval stage. Larvae must be able to translate their own developmental, especially nutritional, history into a development decision. The linking and coordinate expression of gyne characters are advantageous when intermediates are less fit than full queens or full workers. Once such a developmental system evolves, individuals become extremely vulnerable to control by other colony members that have an interest in their developmental fate. In highly eusocial Hymenoptera, queens use 2 principal mechanisms to control offspring development: direct control of nutrition (eg honeybees Apis) and interference with larval response to nutrition (eg Bombus terrestris, many ants). A 2nd major threshold in the evolution of morphologically complex caste systems is the addition of multiple physical worker castes. -from Author
Article
Among the ants, the population of worker phenotypes made by a colony is closely regulated. For the first time, a model of the regulation of a particular size class, that of major workers, is developed within a new theory of the superorganismic activity of social insect colonies. A remarkable linkage is found to exist between caste homeostasis in a model polymorphic society and three well-known but previously unrelated social insect colony properties: the form of the caste distribution, the short-term nature of a colony's “memory”, and the existence of co-operative interactions among nestmates. The linkage consists of mass social processes and takes the form of a colony-level, negative feedback circuit controlling the production of the major workers. The implications of this case study for rigorous, substantive theories of social insect mass phenomena are discussed.