Content uploaded by Chantal Poteaux
Author content
All content in this area was uploaded by Chantal Poteaux on Jun 02, 2015
Content may be subject to copyright.
Fertility Signaling and Partitioning of Reproduction in the Ant
Neoponera apicalis
Boris Yagound
1
& Rémi Gouttefarde
1
& Chloé Leroy
1
&
Rima Belibel
2
& Christel Barbaud
2
&
Dominique Fresneau
1
& Stéphane Chameron
1
&
Chantal Poteaux
1
& Nicolas Châline
1,3
Received: 14 January 2015 /Revised: 30 March 2015 /Accepted: 19 May 2015
#
Springer Science+Business Media New York 2015
Abstract All individuals in social insect colonies benefit
from being informed about the presence and fertility state of
reproducers. This allows the established reproductive individ-
uals to maintain their reproductive monopoly without the need
for physical control, and the non-reproductive individuals to
make appropriate reproductive choices. Here, we studied
whether fertility signaling is responsible for the partitioning
of reproduction in the ant Neoponera apicalis.Thisspecies
forms small colonies from one single-mated queen, with
workers establishing reproductive hierarchies when hopeless-
ly queenless. Previous studies identified putative fertility sig-
nals, particularly the hydrocarbon 13-methylpentacosane (13-
MeC
25
), and have shown that precise status discrimination
based on these signals could be involved in the regulation of
reproductive activities. Here, we extend these findings and
reveal that all individuals, be they queens or workers, differ
in their cuticular hydrocarbon profile according to fertility
state. Proportions of 13-MeC
25
were a strong predictor of an
individual’s ovarian activity, and could, thus, advertise the
established reproducer(s) in both queenright and queenless
conditions. Furthermore, this compound might play a key role
in the establishment of the reproductive hierar chy, since
workers with low fertility at the onset of hierarchy formation
already have relatively high amounts of 13-MeC
25
. Dyadic
encounters showed that individuals with experimentally in-
creased amounts of 13-MeC
25
triggered less agonistic interac-
tions from top rankers, in accord with them Badvertising^
higher status. Thus, these bioassays supported the use of 13-
MeC
25
by competing ants. This simple recognition system
potentially allows permanent regulation of partitioning of re-
production in this species.
Keywords Cuticular hydrocarbons
.
Dominance
.
Honest
signaling
.
Neoponera (formerly Pachycondyla) apicalis
.
Recognition system
.
Reproductive hierarchy
.
Index
hypothesis
Introduction
The ecological dominance of social insects relies mostly on
their reproductive division of labor, in which only one or a few
individuals (usually queens) invest in egg-laying activities,
with the vast majority of a colony foregoing any direct repro-
duction (Wilson 1971). However, in most eusocial Hymenop-
tera (ants, some bees, and wasps), non-reproductive individ-
uals (workers) maintain functional ovaries, and have the po-
tential to lay their own eggs (Bourke 1988). The non-clonal
kin structure of insect societies creates potential reproductive
conflicts in which, under some circumstances, workers are
predicted to favor their own over the colony’sreproduceroff-
spring (Ratnieks et al. 2006). For example, under single
Electronic supplementary material The online version of this article
(doi:10.1007/s10886-015-0591-9) contains supplementary material,
which is available to authorized users.
* Boris Yagound
boris.yagound@leec.univ-paris13.fr
1
Laboratoire d’Ethologie Expérimentale et Comparée, E.A. 4443,
Université Paris 13, Sorbonne Paris Cité, 93430 Villetaneuse, France
2
Laboratoire de Recherche Vasculaire Translationnelle, INSERM U
1148, Université Paris 13, Sorbonne Paris Cité,
93430 Villetaneuse, France
3
Department of Experimental Psychology, Institute of Psychology,
University of São Paulo, São Paulo, Brazil
JChemEcol
DOI 10.1007/s10886-015-0591-9
mating of the queen, colonies are subject to a queen–wo rker
conflict over male parentage (Bourke 1988; Ratnieks et al.
2006). Levels of worker reproduction then depend on the kin
structure of the colony and on the constraints and costs to
colony-level productivity that prevent or limit the benefits of
direct reproduction (Hammond and Keller 2004; Wenseleers
et al. 2004), with self-restraint and coercion being the two main
mechanisms that regulate the extent of worker reproduction
(Bourke 1988; Ratnieks et al. 2006; Wenseleers and Ratnieks
2006). In this context, the presence and fertility state of a re-
producer in the colony has a crucial influence on the propensity
of worker reproduction. Indeed, theoretical models and empir-
ical observations show that levels of worker reproduction com-
monly are much higher in queenless than in queenright colonies
(Bourke 1988; Ratnieks et al. 2006; Wenseleers et al. 2004).
The often extreme reproductive skew characteristic of insect
societies necessitates that workers are correctly informed about
the presence of a fertile reproducer inside the colony.
Fertility-associated chemical signals are widespread in in-
sect societies (Liebig 2010;Monnin2006; Peeters and Liebig
2009). They consist mostly of long-chain hydrocarbons, a
major part of the lipid layer covering the cuticle of insects,
which primarily act as a barrier to prevent desiccation, but also
are involved in commun ication (Howard and B lomquist
2005). Specific (saturated, methyl-branched, or unsaturated)
hydrocarbons in the chemical profile are typically over-
expressed in fertile individuals, thereby informing nestmates
about their presence and fertility state (de Biseau et al. 2004;
Dietemann et al. 2003; Hartmann et al. 2005; Liebig et al.
2000; Monnin et al. 1998). Numerous studies indicate that
social insects not only detect the presence of a reproducer in
a colony and react accordingly through reproductive self-
restraint and/or coercion toward wou ld-be egg layers
(Dapporto et al. 2007; Endler et al. 2004;Holmanetal.
2010; Smith et al. 2009, 2012;VanOystaeyenetal.2014),
but that they also can perceive various levels of ovarian activ-
ity among fertile individuals (Cuvillier-Hot et al. 2004; Heinze
et al. 2002;Liebigetal.1999; Yagound et al. 2014).
Therefore, the traditional view of queen-produced fertility
chemicals as inhibiting pheromones that actively suppress
worker reproduction (Wilson 1971) has now been largely re-
placed by considering fertility-associated cuticular hydrocar-
bons as honest signals (Heinze and d’Ettorre 2009
;Kellerand
Nonacs 1993;Monnin2006; Peeters and Liebig 2009). The
honest signaling hypothesis predicts that all individuals
should respond to a strong signal (advertising a fully fertile
reproducer), because the benefits of helping to raise a repro-
ducer’s highly related offspring exceed the costs of refraining
or preventing others from reproducing. Thus, a worker’sre-
productive choices follow its own interest, since they ultimate-
ly result in an increase in the worker’s inclusive fitness as long
as the fertility state of the established reproducer is high
enough (Keller and Nonacs 1993).
A variety of mechanisms potentially explain the wide-
spread occurrence of reliability in animal communication
(Searcy and Nowicki 2005; Számadó 2011). Honesty has long
been supposed to depend on costs preventing or reducing the
benefits of cheating for low-quality individuals (the handicap
principle; Grafen 1990; Zahavi 1975). As production costs are
likely to be low in the case of fertility signals (Wyatt 2014),
costs guaranteeing honesty may be linked to the maintenance
of such signals, particularly if cheaters suffer social punish-
ment (Smith et al. 2009, 2012). The fundamental assumption
that honest signals need to be costly has, however, been refut-
ed recently (Getty 2006;Számadó2011), showing that hand-
icaps are neither necessary nor sufficient for signals to be
honest. Alternatively, the index hypothesis assumes that phys-
iological constraints cause a non-fakeable connection between
a signal and the trait it advertises (Maynard Smith and Harper
1995). Such causal relationship would, therefore, guarantee
honesty simply because cheating is impossible.
Here, we studied whether honest fertility signaling is re-
sponsible for the regulation o f reproduction in the ant
Neoponera (formerly Pachycondyla) apicalis (Hymenoptera:
Formicidae: Ponerinae). This monogynous and monandrous
species (Yagound et al. unpublished data) forms small colo-
nies typically of less than one hundred individuals, character-
ized by a small queen–worker dimorphism, and workers hav-
ing a high reproductive potential (Fresneau 1994). As in many
social hymenopterans, the partitioning of reproduction is char-
acterized by two distinct phases, namely a queen reproductive
monopoly during most of the colony’s life, and then a final
worker reproductive competition over male parentage,
resulting in the establishment of a reproductive hierarchy
when queen fecundity declines (Blacher et al. 2010;Oliveira
and Hölldobler 1990; Yagound et al. 2014).
Previous studies in hopelessly queenless colonies (colonies
which have lost their queen and where no replacement can
occur) have shown that status discrimination based on puta-
tive fertility signals is involved in the establishment and main-
tenance of the hierarchical structure (Blacher et al. 2010;
Yago und e t a l. 2014). A cuticular hydrocarbon in particular,
13-methylp en taco sane (13-M eC
25
), di fferentiates workers
according to their level of ovarian activity (Yagound et al.
2014). Given that worker reproductive restraint in queenright
nests is supposed to depend on non-volatile queen phero-
mones (Dietemann and Peeters 2000), and because of the
general similarity between queen and worker fertility signals
in other species (Dietemann et al. 2003;Heinzeetal.2002;
Liebig 2010; Smith et al. 2008), we might expect 13-MeC
25
also to be present in queens, and to vary according to fertility
state. It is, therefore, possible that this signal mediates the
regulation of the reproductive division of labor in both
queenright and queenless situations.
Furthermore, the production of fertility signals in hopeless-
ly queenless colonies is usually thought to begin once the
JChemEcol
hierarchical ranks have been determined and the individuals
have started oogenesis (Cuvillier-Hot et al. 2004;Hartmann
et al. 2005; Liebig et al. 2000; Monnin 2006; Peeters and
Liebig 2009). However, workers have been reported to lay
trophic eggs occa sionally in the presenc e of a queen
(Dietemann and Peeters 2000;Fresneau1994;Oliveiraand
Hölldobler 1990). Therefore, the possibility remains that
workers with moderate fertility in queenright nests already
possess relatively high amounts of 13-MeC
25
, and that this
signal could be involved during the first steps of hierarchy
formation to settle dominance/subordinate interactions. Thus,
fertility signals could play a crucial role in the regulation of
reproduction in this species (Yagound et al. 2014).
In this study, we extended previous findings by investigat-
ing the chemical profiles of queens and workers of various
fertility levels in both queenright and queenless conditions to
verify the above predictions and provide additional evidence
supporting the honest fertility signaling hypothesis. To corrob-
orate the correlative chemical evidence and confirm the role of
13-MeC
25
as a badge of status within hierarchies, we conduct-
ed behavioral bioassays in which we monitored the behavioral
responses of high- and low-ranking individuals toward
nestmates of various ranks whose fertility-associated com-
pounds had been manipulated.
Methods and Materials
Ants and Rearing Co nditions We u sed 10 colonies of
Neoponera apicalis morph 4 (Delabie et al. 2008)collected
in Petit Saut (5°04′15.8″N, 53°02′36.3″W), French Guiana, in
2007 and 2011. Each colony comprised a queen, 64.1±8.7
(mean ± SE) workers, and brood. They were reared in plaster
nests (18×14 cm) connected to a foraging area. Ants were
provided twice a week with crickets and honey/apple mixture,
and water ad libitum. Housing conditions were as follows: 60
±5 % RH, 27±2 °C, 12:12 hr L:D cycle. Ant collection, hus-
bandry, and experimental procedures fulfilled all the legal re-
quirements concerning insect experimentation of France.
Experimental Procedures Two consecutive experiments
were carried out to study fertility signals in N. apicalis.In
the first experiment, we compared the chemical profiles of
fertile and unmated queens and workers of varying fertility,
both at the onset of hierarchy formation (orphaning point) and
in well-established reproductive hier archies (hopelessly
queenless situation). This was aimed at: (i) confirming the
honest fertilit y signaling hypothesis, according to which
amounts of 13-MeC
25
should be higher in fertile queens than
in fertile workers, whereas those of unmated queens should be
similar to those of unfertile workers, and (ii) studying the link
between signal expression and ovarian activity both before
and after formation of the dominance hierarchy. In the second
experiment, we manipulated the amount of cuticular 13-
MeC
25
of top and low rankers, and we monitored their subse-
quent behavioral interactions with high- and low-ranking
nestmates during dyadic encounters to confirm the use of this
cuticular hydrocarbon as a badge of status.
Experiment 1
Between-Caste Fertility Signal Comparison Fertile queens
(N=8) and gynes (young unmated queens, N=12) were col-
lected from our stock colonies for sampling cuticular hydro-
carbon profiles, using the non-destructive method of solid-
phase microextraction (SPME). A 100 μm polydimethylsilox-
ane fiber (Supelco, Bellef onte, PA, USA) was ca refully
rubbed against the third and fourth abdominal segments of
live ants for 2 min. The fiber was desorbed in the injection
port of a 7890A gas chromatograph (Agilent Technologies,
Santa Clara, CA, USA), equipped with an HP-5MS capillary
column (30 m×25 μm×0.25 μm) and a split–splitless injec-
tor, coupled to a 5975c mass spectrometer (Agilent Technol-
ogies) with 70 eV electron impact ionization. The carrier gas
was helium at 1 ml.min
−1
. The temperature program was: an
initial hold at 70 °C for 5 min, 70–250 °C at 30 °C.min
−1
,
250–260 °C at 1 °C.min
−1
,260–320 °C at 20 °C.min
−1
,then
held for 5 min. Peak areas were integrated with the MSD
ChemStation software E.02.01.1177 (Agilent Technologies).
Hydrocarbons were identified on the basis of mass spectrum
and retention time, and compared to known standards. Iso-
mers of methyl-branched alkanes co-eluting in the same peak
were separated as in Martin et al. (2008).
To link an individual’s chemical profile with its level of
fertility, gynes were frozen for dissection and the mean size
of the six basal oocytes taken as an ovarian index. Head width,
as an indicator of body size, also was measured for each gyne
to account for possible effects of this factor on levels of ovar-
ian activity or amounts of fertility-associated cuticular hydro-
carbons. The fertility level of queens was not determined be-
cause it would result in sacrificing stock colonies. However, as
is u sual in social insects, fertility in N. apicalis queens is
higher than that of reproductive workers (Fresneau 1994),
despite their limited dimorphism. Observations of stock colo-
nies before and after the experiment showed that all sampled
queens were fertile (i.e., laying female-destined eggs). The
chemical profiles of queens and gynes, and the ovarian activ-
ity of the latter, then were compared to data obtained for the
worker caste in both queenright and queenless conditions (see
the following section).
Dynamics of Signal Expression Three groups of 20 workers
were isolated from our stock colonies and placed in a new nest
(same dimensions). This dequeening procedure typically in-
duces the formation of a reproductive hierarchy manifested by
JChemEcol
ritualized agonistic behaviors (Oliveira and Hölldobler 1990;
Yago und e t a l. 2014). Age is known to affect the reproductive
and dominance status of workers (Dietemann and Peeters
2000), with mainly relatively young indiv iduals being in-
volved in the reproductive competition, whereas callow and
old workers hardly ever engage in dominance interactions. To
control for the presence of all age classes (and their associated
physiological characteristics) in our groups, we waited for the
first unambiguous dominance/subordinate interactions to be
clearly visible, which occurred within 2 hr in each case. By
that time, all workers were sacrificed for analysis (as described
in the previous section). Since this situation is virtually iden-
tical to a queenright situation (in terms of chemical profile and
ovarian activity of workers), these individuals are hereafter
considered as queenright workers and represent the physio-
logical state of workers at the onset of hierarchy formation. As
dissections showed that some individuals already exhibited
activated ovaries, workers were classified according to their
number of developed oocytes: moderately fertile workers
(MFW, two to four developed oocytes, N=26) or infertile
workers (IW, zero to one developed oocytes, N=34).
Six additional groups of 20 workers also were isolated by
the same procedure, but this time they remained in queenless
groups for 15 d. This duration is sufficient for a clear near-
linear reproductive hierarchy to be established through ritual-
ized agonistic behaviors and fertility signaling (Yagound et al.
2014). At d 16, the same analyses as for queenright workers
were performed on all individuals. Three groups of varying
fertility (adapted from Yagound et al. 2014) were determined:
highly fertile workers (HFW, 5–6 developed oocytes, N=38),
moderately fertile workers (MFW, 2–4 developed oocytes,
N=37), and infertile workers (IW, 0–1 developed oocytes,
N=41). Some workers (one in two groups, two in one group)
died before d 16 and were, therefore, not included in the
analyses.
Experiment 2
Reproductive Hierarchies Six orphaned colonies of 41
workers were created following the same procedure as the first
experiment. All workers were individually labeled with num-
bered tags glued on their thorax and dots of paint on their
abdomen. Housing and feeding conditions were the same as
above, but this time the nest was connected through a 5-cm
tunnel to a circular chamber (diam. = 3.5 cm) freely accessible
to the workers and subsequently used as the test arena in the
bioassays. Chamber and tunnel walls were blackened so that
ants experienced no discontinuity (namely light stimulation)
with the rest of the nest.
All orphaned colonies were observed for 1 hr a day for
15 d, with orphaning as d 1. We recorded all agonistic behav-
iors (antennal boxing a nd biting) typical of dominance/
subordinate interactions (Oliveira and Hölldobler 1990;
Yago und et al. 2014). This allowed the reconstruction of the
matrix of hierarchical ranks for all individuals (see Blacher
et al. 2010; Yagound et al. 2014). Ants with ranks 1–12 were
considered top rankers (following Yagound et al. 2014),
whereas the last 18 ranks (i.e., ranks 20–37 due to some mor-
tality) were considered low rankers.
Bioassays At d 16, behavioral bioassays were conducted to
confirm the use of 13-MeC
25
as a badge of status. The test
arena was disconnected from the nest and subsequently used
during the tests, thereby providing the ants with a Bcolony
context^, at least for the olfactory modality. Bioassays
consisted of dyadic encounters between a focal ant and a treat-
ed nestmate whose fertility-associated hydrocarbons had been
manipulated. Three treatments were used: increasing amounts
of 13-MeC
25
(putative badge of status), increasing amounts of
eicosane (n-C
20
, control compou nd naturally found on
N. apicalis cuticles but whose quantities are not correlated to
fertility or rank; Yagound et al. 2014), and solvent (hexane,
i.e., manipulation control).
13-MeC
25
was synthesized according to published
methods (Guédot et al. 2009;seeElectronic supplementary
material). To manipulate a stimulus odor, we followed the
protocol of Smith et al. (2012). Stock solutions of 8.4 mg of
synthetic 13-MeC
25
or n-C
20
(Sigma Aldrich, St. Louis, MO,
USA) per 14 ml of hexane were prepared. For each treatment,
25 μl of stock solution were added to the surface of a 10 ml
glass beaker filled with deionized water. Following hexane
evaporation, an ant to be used as a stimulus was anesthetized
briefly at freezing temperature for 30 sec and placed on the
surface of the deionized water and gently swirled, thereby
transferring the surface hydrocarbon film onto its cuticle
(Smith et al. 2012). The treated stimulus was allowed to dry
in a box for 10 min.
To verify the effectiveness of the odor manipulation, a sub-
set of individuals was sampled by SPME both before and after
chemical treatment. The protocol was the same as in the first
experiment, except for the temperature program: an initial
hold of 70 °C for 1 min, 70–250 °C at 40 °C.min
−1
,250–
258 °C at 1 °C.min
−1
,258–320 °C at 40 °C.min
−1
,thenheld
at 320 °C for 3 min. There was a 47.9±11.2 % increase in 13-
MeC
25
(N=16), which is within the natural range of variation
separating top and low rankers (33.6–103.9 %, N=206, data not
shown). The increase in n-C
20
was not significantly different
from that of 13-MeC
25
(28.3±5.8 %, N=19; permutation test:
P=0.12), and was necessarily beyond its normal range of
variation (0.3–1.3 %, N=206, data not shown).
The treated stimulus and focal ant then were introduced
into the test arena, temporarily separated by a microscope
slide for 30 sec to allow them to become accustomed to the
setup. Tests began following the removal of the microscope
slide and lasted 10 min. Each test was recorded by video, and
the ant’s behavior was subsequently analyzed with EthoLog
JChemEcol
2.2 software (Ottoni 2000). We recorded the duration of an-
tennal contacts and the number of all agonistic acts (antennal
boxing, mandible opening, biting). Observers were blind to
the treatment.
Three conditions were used for each treatment: a top ranker
confronting a low ranker (mean gap in their respective rank of
21.7±0.7, N=36), a top ranker confronting another top ranker
(mean gap in their respective rank of 2.0±0.0, N=36), and a
low ranker confronting another low ranker (mean gap in their
respective rank of 1.3±0.3, N=36). Two tests per treatment
per colony were run for each condition. Each ant was involved
only in a single test. Focal ants always had a higher actual rank
compared to treated ants, except in the case of low rankers
whose ranks were similar. All tests for each experimental col-
ony were run within a few hours to avoid any changes in
hierarchical order due to the consecutive removal of tested
individuals. After the completion of the tests, all workers were
frozen for measurement of ovarian activity.
Statistical Analyses We compared the relative quantities of 27
peaks common to all individuals. To avoid problems arising
from multi-colinearity, compounds highly correlated (r
2
>0.8)
were treated as a single variable (Martin and Drijfhout 2009).
This resulted in a final high ratio (7.8) of observations to inde-
pendent variables. A discriminant analysis was performed on
geomean-log transformed proportions according to Reyment
(1989 ): Z
i,j
=ln[Y
i,j
/g(Y
j
)], where Y
i,j
is the area of peak i for
the individu al j,andg(Y
j
) is the geometric mean of all peak
areas for individual j. We studied the general relationship be-
tween fertility, proportion of compounds, and size using Spear-
man rank correlation tests. One-way ANOVAs with the Monte
Carlo procedure also were performed on relative amounts of
cuticular hydrocarbons between all groups, and on ovarian and
size indexes between all groups except queens.
Following hierarchy reconstruction, we calculated the K
index of linearity varying from 0 (no linearity) to 1 (linear
hierarchy) and tested the statistical significance of linearity
with the Appleby (1983) method. We further verified if the
hierarchical rank was correlated with both dominance (propor-
tion of agonistic acts performed) and fertility with the Spear-
man rank correlation test.
We used generalized linear mixed-effects models (GLMM)
to test the effect of treatment (13-MeC
25
, n-C
20
,solvent)on
behavioral responses of ants for each condition (top ranker vs.
low ranker, top ranker vs. top ranker, low ranker vs. low rank-
er) with R-3.2.0 (R Core Team 2012), using the package lme4
(Bates et al. 2013). The total number of agonistic behaviors
(response variable) was compared using GLMMs with a
Poisson error distribution and a log-link function. We com-
pared the duration of antennal contacts (response variable)
using GLMMs with a Gaussian error distribution and an iden-
tity link function. Treatment as a fixed factor and colony as a
random factor were both included in the models.
Statistical analyses were performed using R-3.2.0 and
Statistica 8.0 (StatSoft, Tulsa, OK, USA). Post-hoc corrected
P-values following the Bonferroni–Holm method are denoted
P’. Statistical significance was set at P<0.05.
Results
Experiment 1 All groups diverged in level of fertility (one-
way ANOVA: F
5,188
=189.70, P=0.001; Fig. 1a). As expect-
ed, ovarian activity of gynes was very low and similar to that
of infertile workers (Fig. 1a).
Each group of queens and workers bore a distinct chem-
ical profile, as revealed by discriminant function analysis
(Wilks’s λ =0.003, F
138,981
=11.87, P <0.001; between-
groups comparisons: all P’ < 0.037). Compared with
workers, queens had higher relative amounts of 10-
MeC
19
, x,y-C
22:2
, n-C
22
,11-MeC
23
,13-MeC
25
,andthere
was a strong trend for x,y-C
29:2
, whereas workers had
higher proportions of n-C
29
than queens (Table 1). The
trend was similar for gynes compared with workers for
10-MeC
19
, x,y-C
22:2
, x,y-C
29:2
,andn-C
29
, thus makin g t he-
se compounds, but not n-C
22
,11-MeC
23
, and 13-MeC
25
,
potential caste signals. R elative amounts of x,y-C
22:2
, n-
C
22
, x,y-C
29:2
,andn -C
29
were not different between queens
and gynes, whereas 10-MeC
19
was much higher in gynes
than in queens, and may be involved in courtship or mating.
By contrast, queens had higher relative quantities of 11-
MeC
23
and 13-MeC
25
than gynes (Table 1).
The correlation between relative amount of cuticular hy-
drocarbons and level of fertility proved to be significant for
10-MeC
19
, 11-MeC
23
,13-MeC
25
,andx,y-C
29:2
(Table 1). As
expected, this correlation was much higher for 13-MeC
25
(Spearman rank correlation: r
s
=0.67, N=188, P’ <0.001).
The proportion of 13-MeC
25
and fertility level diverged be-
tween groups in a close relationship (Table 1 and Fig. 1a and
b), in accordance with the fertility signal hypothesis. Queens
indeed had the highest relative amounts of 13-MeC
25
(Fig. 1b), and arguably the highest fertility levels (Fresneau
1994). By contrast, gynes whose ovarian activity was similar
to infertile workers had much reduced relative amounts of 13-
MeC
25
, lying between moderately fertile and infertile workers
(Fig. 1a and b).
Highly fertile, moderately fertile, and infertile workers in
the hopelessly queenless situation showed marked differences
in amounts of 13-MeC
25
and levels of fertility, with a close
relationship between these two variables (Fig. 1a and b), as
already described (Yagound et al. 2014). Interestingly, mod-
erately fertile workers at the onset of orphaning had, by that
time, higher relative amounts of 13-MeC
25
compared with
queenright infertile workers (P’ =0.022; Fig. 1b).
Although individuals exhibited actual differences in their
index of body size (range 1.75–2.24 mm, N=188), this index
JChemEcol
was not correlated with either fertility (r
s
=−0.04, N=188, P’ =
0.60), or with relative amounts of 13-MeC
25
(r
s
=0.06, N=
188, P’ =0.42). No difference in size was found among all
groups of individuals, even when gynes were included (F
5,
188
=1.06, P=0.36).
Experiment 2 The linearity or near-linearity of the hierarchy
was significant in all colonies (K index of linearity ranging
from 0.60 to 1, N=6, all P<0.003 ). The hierarchical rank
further was highly correlated with both dominance (r
s
ranging
from −0.64 to −0.92, N=6, all P<0.001) and fertility (r
s
rang-
ing from −0.70 to −0.84, N=6, all P<0.001). All colonies at
the time of the bioassays, thus, exhibited a near-linear repro-
ductive hierarchy.
The agonistic response of focal high rankers toward low-
ranking ants varied among treatments (GLMM: χ
2
2
=19.19,
P<0.001, N=36; Fig. 2a). The number of agonistic acts per-
formed by focal high rankers toward low rankers with in-
creased amounts of 13-MeC
25
and n-C
20
was smaller than that
of focal high rankers encountering hexane-treated low-rank-
ing ants (post-hoc tests: both P’ <0.005), but there was no
difference between focal high rankers confronting 13-
MeC
25
-andn-C
20
-treated low-ranking ants (P’ =0.29).
Focal high-ranking ants decreased their agonistic response
toward 13-MeC
25
-treated high-ranking ants compared with
both n-C
20
- and hexane-treated high-ranking ants (GLMM:
χ
2
2
=26.69, P<0.001, N=36; both P’ <0.016; Fig. 2b). Focal
individuals confronting n-C
20
-treated high rankers also de-
creased their agonistic response compared to individuals en-
countering hexane-treated high ranking ants (P’ =0.006).
There was no difference among treatments in the number of
agonistic acts of focal low rankers confronting low-ranking
stimuli (GLMM: χ
2
2
=0.98, P=0.61, N=36; Fig. 2c).
The duration of antennal contacts from focal ants toward
treated stimuli was not affected by the treatments i n top
rankers confronting low-ranking treated ants (13-Me C
25
:
86.20±13.20 sec, n-C
20
: 76.20±14.66 sec, hexane: 79.48±
11.56 sec; GLMM: χ
2
2
=1.77, P =0.41, N=36) or in top
rankers confronting top-ranking treated ants (13-MeC
25
:
85.86±15.08 sec, n-C
20
: 71.88±17.91 sec, hexane: 95.21±
18.85 sec; GLMM: χ
2
2
=0.65, P=0.73,N=36). Antennal con-
tacts performed by focal low rankers confronting low-ranking
stimuli were, however, longer to 13-MeC
25
-treated ants than
to n-C
20
- treated ants, but not to hexane-treated ants (13-
MeC
25
: 83.43±15.27 sec, n-C
20
: 50.14±10.43 sec, hexane:
55.49±8.03 sec; GLMM: χ
2
2
=7.47, P=0.024, N=36, P’ =
0.036 and P’ =0.12, respectively).
Discussion
Our study provides new evidence supporting the honest fer-
tility signaling hypothesis. No clear pattern emerges when
looking at which specific compounds or group of compounds
may signal fertility in social Hymenoptera (Liebig 2010;but
see Van Oystaeyen et al. 2014). Indeed, saturated (Smith et al.
2008), methyl-branched (Heinze et al. 2002;Holmanetal.
2010), and unsat urated (Monnin et a l. 1998; Smith et al.
2012) hydrocarbons all correlate with ovarian activity, de-
pending on species. Nevertheless, it appears that alkenes and
particularly methyl-alkanes are the predominant compound
classes that advertise fertility, possibly because of their higher
communicative potential (Monnin 2006). We found the same
pattern in our analysis, with only one alkene (x,y-C
29:2
)and
three monomethyl-branched alkanes (10-MeC
19
, 11-MeC
23
,
and, to a higher extent, 13-MeC
25
) being linked with fertility
both in queen and worker castes. The compound 13-MeC
25
Relative quantity of 13-MeC
25
(%)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
a
bc
e
f
d
b
cd
Ovarian development (mm)
0.0
0.5
1.0
1.5
ab
d
c
a
c
b
a
b
Queenright
Queenless
Queenright
Queenless
HFW
N = 38
MFW
N = 37
MFW
N = 26
IW
N = 41
IW
N = 34
Gynes
N = 12
Queens
HFW
N = 38
MFW
N = 37
MFW
N = 26
IW
N = 41
IW
N = 34
Gynes
N = 12
Queens
N = 8
Fig. 1 Between-group differences in (a) ovarian development (mm) and
(b) relative quantity (%) of 13-MeC
25
. Queens are indicated in (a)evenif
not dissected for reading convenience. Box plots represent 10th, 25th,
50th (median), 75th and 90th percentiles. Sample sizes of individuals
are indicated below each box plot. D ifferent letters denote statistical
differences. HFW, highly fertile workers; MFW, moderately fertile workers;
IW, infertile workers
JChemEcol
was closely correlated with fertility in workers in the hope-
lessly queenless condition, thus confirming previous results
(Yagound et al. 2014). Furthermore, the same relationship
was found in reproductive and non-reproductive queens,
adding to the already reported similarity in the chemical pro-
files of queens and reproductive workers (Dietemann et al.
2003;Heinzeetal.2002; Liebig 2010; Smith et al. 2008).
Another important result in favor of the honest fertility
signaling hypothesis comes from the study of the dynamics
of signal expression before and after the formation of the
reproductive hierarchy. Workers in queenright colonies al-
ready differed in their ovarian development, with moderately
fertile individuals exhibiting higher amounts of 13-MeC
25
compared to their infertile nestmates. It often has been shown
that a behavioral hierarchy precedes, sometimes with a long
time interval, reproductive hierarchy and consequently the
expression of fertility signals (Cuvillier-Hot et al. 2004;
Hartmann et al. 2005; Liebig et al. 2000). In contrast, we
demonstrated that the expression of fertility signals is not de-
layed in N. apicalis.
The formation of social hierarchies depends on a variety of
factors including intrinsic differences between competing in-
dividuals and self-organizing processes (Hsu et al. 2006). Our
results suggest that the signaled fertility state could have a
non-negligible influence on the outcome of agonistic interac-
tions, while reducing injury- and time-related costs of overt
aggression (Hsu et al. 2006). Queenright workers with low
fertility already exhibiting higher amounts of fertility signal
couldberecognizedassuchthroughtheworkers’ fine-scale
status-dis cr imination abilities (Yagound et al. 2014). The
workers’ physiological divergence at the onset of hierarchy
formation, probably resulting from their heterogeneous age-
dependent hormonal states, could be amplified during the sub-
sequent agonistic interactions. We might expect, therefore,
that these moderately fertile individuals could eventually gain
top rank of the repr oductive hierar chy through this self-
sustaining process. Physiological, aggression-mediated
changes, potentially involving juvenile hormone (Hartfelder
2000), also could participate in activation of the ovaries
(Lamba et al. 2007), thus reinforcing reproductive status. The-
se results suggest that fertility signaling is not restricted to
maintenance of the hierarchy, and could also be involved in
establishment of the hierarchical structure.
Finally, this apparent inseparable association between pro-
duction of 13-MeC
25
and activation of ovaries adds to the
evidence in favor of the index hypothesis (Maynard Smith
and Harper 1995) as the mechanism maintaining honesty of
fertility signals. Numerous studies in insects point toward the
implication of common endocrinological mechanisms involv-
ing gonadotropic hormones (juvenile hormone and
ecdysteroids) that may be responsible for such a connection
(Blomquist 2010; Cuvillier-Hot et al. 2004; Holman 2012;
Tabl e 1 Caste- and fertility-
related differences in relative
amounts of cuticular
hydrocarbons
Compound One-way
ANOVA
all groups
Between-caste comparison Correlation with
fertility level
Queens (Q) vs
workers
a
(W)
Gynes (G) vs
workers (W)
Queens (Q) vs
gynes (G)
10-MeC
19
F
6,196
=27.08
P<0.001
Q>W
all P’ <0.011
G>W
all P’ <0.003
Q<G
P’ =0.003
r
s
=−0.23, N=188
P’ =0.002
x,y-C
22:2
F
6,196
=5.39
P<0.001
Q>W
all P’ <0.016
G>W
b
all P’ <0.065
P’ =0.70 r
s
=−0.07, N=188
P’ =0.37
n-C
22
F
6,196
=4.96
P<0.001
Q>W
all P’ <0.037
all P’ >0.56 P’ =0.61 r
s
=0.01, N=188
P’ =0.87
11-MeC
23
F
6,196
=9.74
P<0.001
Q>W
all P’ <0.026
all P’ >0.26 Q > G
P’ =0.030
r
s
=0.41, N=188
P’ <0.001
13-MeC
25
F
6,196
=24.02
P<0.001
Q>W
all P’ <0.002
MFW > G >
IW
c
Q>G
P’ =0.002
r
s
=0.67, N=188
P’ <0.001
x,y-C
29:2
F
6,196
=5.47
P<0.001
Q>W
b
all P’ <0.052
G>W
b
all P’ <0.052
P’ =0.74 r
s
=0.23, N=188
P’ =0.002
n-C
29
F
6,196
=31.40
P<0.001
Q<W
all P’ <0.002
G<W
all P’ <0.002
P’ =0.17 r
s
=0.06, N=188
P’ =0.43
Only cuticular hydrocarbons with clear between-caste differences are presented. See Yagound et al. (2014) for a
comprehensive description of the cuticular hydrocarbon profile of Neoponera apicalis. Significant values are
highlighted in bold
a
Comparisons are made for highly fertile, moderately fertile and infertile workers
b
Some comparisons are not different, but there is a strong trend
c
MFW, moderately fertile workers; IW, infertile workers. See Fig. 1b
JChemEcol
Monnin 2006; Peeters and Liebig 2009). These constraints
can guarantee the honesty of the signaling system without
the necessity for additional production or maintenance costs.
We note, however, that this does not necessarily imply a cost-
free system, as costs and constraints should not be considered
mutually exclusive (Holman 2012). As has been suggested
previously (Heinze and d’Ettorre 2009; Keller and Nonacs
1993; Smith et al. 2009), the index hypothesis of honest fer-
tility signaling is likely to have a general significance across
various taxa of social insects.
The short time window available for successful male pro-
duction in hopelessly queenless nests and the unpredictability
of the onset of worker reproductive competition in this tropical
species exert strong ecological pressures on the reproductive
strategies of workers (Dietemann and Peeters 2000;Fresneau
1994). Indeed, the few eggs having an actual chance of devel-
oping into adult males are most likely among the very first
eggs to be laid because of policing from the top rankers and
severe food supply limitations due to the death of the foraging
workers. These strong selective pressures could have favored
the use of cuticular hydrocarbons whose biosynthesis is close-
ly linked to ovary activation to signal fertility. An index-based,
honest signaling system would, thus, allow a quick resolution
of reproductive competition and the possibility for all
nestmates to gain inclusive fitness benefits by raising a few
males before the colony eventually collapses.
Corroborating chemical co rrelations through behavioral
bioassays is important but rarely performed due to its inherent
difficulties (Howard 1993;Liebig2010; Martin and Drijfhout
2009;Monnin2006; Peeters and Liebig 2009). Here, manip-
ulation of the amounts of 13-MeC
25
induced a decrease in the
agonistic response of high rankers toward high- and low-
ranking treated nestmates, in accord with these treated stimuli
advertising a higher status compared to sham-treated controls.
Low rankers are seldom involved in agonistic interactions
(Yagound et al. 2014). Their similar behavioral response in
all conditions is probably the consequence of a bottom effect,
although ants increased their duration of antennal contacts
toward 13-MeC
25
- compared with n-C
20
-treated stimuli, pos-
sibly indicating an increased interest toward a high-ranking
signal. No effect in the level of response was expected toward
n-C
20
-treated high- and low-ranking individuals, since this
compound is not linked to fertility. This contradictory result
is puzzling and is possibly due to a perturbation of the recog-
nition system (Howard and Blomquist 2005), although there
was no effect on the duration of antennation to treated stimuli.
Bioassays depend heavily on the motivation of tested ants and
are context-dependent (Buczkowski and Silverman 2005;
Howard 1993). Although we tried to provide ants with a real-
istic context and, despite the relatively long duration of the
tests, the few agonistic acts observed also could have influ-
enced our results. This methodological limitation is due main-
ly to the lateness of when the tests were performed relative to
13-MeC
25
n-C
20
Hexane
No. of agonistic behaviors
0
5
10
15
20
25
a
b
c
13-MeC
25
n-C
20
Hexane
0
5
10
15
20
25
a
a
b
13-MeC
25
n-C
20
Hexane
0
5
10
15
20
25
a
a
a
a
b
c
No. of agonistic behaviorsNo. of agonistic behaviors
JChemEcol
orphaning, which was necessary to assess an individuals’ rank
correctly. Additional studies, such as testing multiple com-
pounds or introducing stimuli in groups of individuals (e.g.,
Smith et al. 2012) are necessary to confirm the results of our
bioassays.
Overall, this study reinforces the hypothesis that 13-MeC
25
constitutes, or at least contributes to, an honest signal of fer-
tility in N. apicalis. Interestingly, our results indicate that 13-
MeC
25
could be involved in the partitioning of reproduction in
both queenless and queenright conditions, negating the neces-
sity for specific caste signals advertising a queen’spresence.
Monitoring the amounts of 13-MeC
25
could allow workers to
make appropriate reproductive choices favoring their own in-
clusive fitness (Keller and Nonacs 1993). Indeed, this signal
identifies the best egg layer in the colony, namely the mother
queen, during the vast majority of the colony’s life, as well as
the worker(s) with the highest reproductive potential when the
colony becomes hopelessly queenless. This signal could be
sufficient for establishing the reproductive monopoly of repro-
ducers without any need for physical control because of
shared fitness interests between all nestmates (Reeve and
Jeanne 2003). Indeed, a queen is never aggressive toward
workers (Dietemann and Peeters 2000) and she maintains
her reproductive monopoly through fertility signaling, as ex-
pected on grounds of relatedness. The fitness interests of
workers in hopelessly queenless colonies are, however, not
aligned, and agonistic behaviors are expressed frequently dur-
ing the establishment of the reproductive hierarchy (Oliveira
and Hölldobler 1990; Yagound et al. 2014). However, the
short time-window available for worker reproduction con-
strains a quick resolution of the reproductive conflict
(Dietemann and Peeters 2000) and, therefore, a switch to fer-
tility signaling and status recognition. The fine tuning of
relatedness-mediated benefits, productivity-associated costs
and life-history constraints seems to have resulted in a simple
yet efficient recognition system in N. apicalis, in which a
single cuticular hydrocarbon may be sufficient to regulate
the reproductive division of labor permanently.
Acknowledgments We thank Paul Devienne for help during ant col-
lection and the building of our experimental setup, Philippe Cerdan for
logistical support in the field, Jocelyn Millar and Isabelle Bataille for
advice with the synthesis of 13-MeC
25
, and Margot Perez for statistical
guidance. Two anonymous referees provided helpful comments on the
manuscript. This study was supported by the Fundação de Amparo à
Pesquisa do Estado da Bahia/Conselho Nacional de Desenvolvimento
Científico e Tecnológico (PNX0011/2009 PRONEX).
References
Appleby MC (1983) The probability of linearity in hierarchies. Anim
Behav 31:600–608
Bates D, Maechler M, Bolker B, Walker S (2013) lme4: Linear mixed-
effects models using Eigen and S4. R package version 1.0-4. http://
CRAN.R-project.org/package=lme4. Accessed 2 Oct 2013
Blacher P, Lecoutey E, Fresneau D, Nowbahari E (2010) Reproductive
hierarchies and status discrimination in orphaned colonies of
Pachycondyla apicalis ants. Anim Behav 79:99–105
Blomquist GJ (2 010) Biosynthes is of cuticular hydrocarbons. In:
Blomquist GJ, Bagnières A-G (eds) Insect hydrocarbons: biology,
biochemistry, and chemical ecology. Cambridge University Press,
Cambridge, pp 35–52
Bourke AFG ( 1988) Worker reproduction in the higher eusocial
Hymenoptera. Q Rev Biol 63:291–311
Buczkowski G, Silverman J (2005) Context-dependent nestmate discrim-
ination and the effect of action thresholds on exogenous cue recog-
nition in the Argentine ant. Anim Behav 69:741–749
Cuvillier-Hot V, Lenoir A, Crewe R, Malosse C, Peeters C (2004)
Fertility signalling and reproductive skew in queenless ants. Anim
Behav 68:1209–1219
Dapporto L, Santini A, Dani FR, Turillazzi S (2007) Workers of a Polistes
paper wasp detect the presence of their queen by chemical cues.
Chem Senses 32:795–802
de Biseau J-C, Passera L, Daloze D, Aron S (2004) Ovarian activity
correlates with extreme changes in cuticular hydrocarbon profile in
the highly polygynous ant, Linepithema humile. J Insect Physiol 50:
585–593
Delabie JHC, Mariano CSF, Mendes LF, Pompolo SG, Fresneau D
(2008) Problemas apontados por estudos morfológicos, ecológicos
e citogenéticos no gênero Pachycondyla na região Neotropical: o
caso do complexo apicalis. In: Vilela EF, Santos IA, Schoereder JH,
Neto JL, Serrão JE, Campos LAO (eds) Insetos sociais: da biologia à
aplicação. Editora UVF, Viçosa, pp 197–222
Dietemann V, Peeters C (2000) Queen influence on the shift from trophic
to reproductive eggs laid by workers of the ponerine ant
Pachycondyla apicalis. Insect Soc 47:223–228
Dietemann V, Peeters C, Liebig J, Thivet V, Hölldobler B (2003)
Cuticular hydrocarbons mediate discrimination of reproducti ves
and nonreproductives in the ant Myrmecia gulosa.ProcNatlAcad
Sci U S A 100:10341–10346
Endler A, Liebig J, Schmitt T, Parker JE, Jones GR, Schreier P,
Hölldobler B (2004) Surface hydrocarbons of queen eggs regulate
worker reproduction in a social insect. Proc Natl Acad Sci U S A
101:2945–2950
Fresneau D (1994) Biology and social behaviour of a Neotropical
ponerine ant (Pachycondyla apicalis). Diss ertation, Université
Paris 13
Getty T (2006) Sexually selected signals are not similar to sports handi-
caps. Trends Ecol Evol 21:83–88
Grafen A (1990) Biological signals as handicaps. J Theor Biol 144:517–
546
Guédot C, Millar JG, Horton DR, Landolt PJ (2009) Identification of a
sex attractant pheromone for male winterform pear psylla,
Cacopsylla pyricola. J Chem Ecol 35:1437–1447
Hammond RL, Keller L (2004) Conflict over male parentage in social
insects. PLoS Biol 2:1472–1482
Fig. 2 Total number of agonistic behaviors performed by focal workers
toward nestmates with different treatments (13-MeC
25
, n-C
20
and
solvent) in each condition. (a) High-ranking focal ants confronting low-
ranking treated stimuli (ants), (b) high-ranking focal ants confronting
high-ranking treated stimuli, (c) low-ranking focal ants confronting
low-ranking treated stimuli. Box plots represent 10th, 25th, median
(continuous line), mean (dotted line), 75th and 90th percentiles. Sample
sizes of individuals are N=12 for each box plot. Different letters denote
differences among treatments
JChemEcol
Hartfelder K (2000) Insect juvenile hormone: from Bstatus quo^ to high
society. Braz J Med Biol Res 33:157–177
Hartmann A, d’Ettorre P, Jones GR, Heinze J (2005) Fertility signalling–
the proximate mechanism of worker policing in a clonal ant.
Naturwissenschaften 92:282–286
Heinze J, d’Ettorre P (2009) Honest and dishonest communication in
social Hymenoptera. J Exp Biol 212:1775–1779
Heinze J, Stengl B, Sledge MF (2002) Worker rank, reproductive status
and cuticular hydrocarbon signature in the ant, Pachycondyla cf.
inversa. Behav Ecol Sociobiol 52:59–65
Holman L (2012) Costs and constraints conspire to produce honest sig-
naling: insights from an ant queen pheromone. Evolution 66:2094–
2105
Holman L, Jorgensen CG, Nielsen J, d’Ettorre P (2010) Identification of
an ant queen pheromone regulating worker sterility. Proc R Soc
Lond B Biol Sci 277:3793–3800
Howard RW (1993) Cuticular hydrocarbons and chemical communica-
tion. In: Stanley-Samuelson DW, Nelson DR (eds) Insect lipids,
chemistry, biochemistry and biology. University of Nebraska
Press, Lincoln, pp 177–226
Howard RW, Blomquist GJ (2005) Ecological, behavioral, and biochem-
ical aspects of insect hydrocarbons. Annu Rev Entomol 50:371–393
Hsu Y, Earley RL, Wolf LL (2006) Modulation of aggressive behaviour
by fighting experience: mechanisms and contest outcomes. Biol Rev
81:33–74
Keller L, Nonacs P (1993) The role of queen pheromones in social in-
sects: queen control or queen signal? Anim Behav 45:787–794
Lamba S, Kazi YC, Deshpande S, Natesh M, Bhadra A, Gadagkar R
(2007) A possible novel function of dominan ce behaviour in
queen-less colo nies of the primitively eusocial wasp Ropalidia
marginata. Behav Process 74:351–356
Liebig J (2010) Hydrocarbon profiles indicate fertility and dominance
status in ant, bee, and wasp colonies. In: Blomquist GJ, Bagnières
A-G (eds) Insect hydrocarbons: biology, biochemistry, and chemical
ecology. University Press Cambridge, Cambridge, pp 254–281
Liebig J, Peeters C, Hölldobler B (1999) Worker policing limits the num-
ber of reproductives in a ponerine ant. Proc R Soc Lond B Biol Sci
266:1865–1870
Liebig J, Peeters C, Oldham NJ, Markstädter C, Hölldobler B (2000) Are
variations in cuticular hydrocarbons of queens and workers a reli-
able signal of fertility in the ant Harpegnathos saltator? Proc Natl
Acad Sci U S A 97:4124–4131
Martin SJ, Drijfhout FP (2009) How reliable is the analysis of complex
cuticular hydrocarbon profiles by multivariate statistical methods? J
Chem Ecol 35:375–382
Martin SJ, Helanterä H, Drijfhout FP (2008) Colony-specific hydrocar-
bons identify nest mates in two species of Formica ant. J Chem Ecol
34:1072–1080
Maynard Smith J, Harper DGC (1995) Animal signals: models and ter-
minology. J Theor Biol 177:305–311
Monnin T (2006) Chemical recognition of reproductive status in social
insects. Ann Zool Fenn 43:515–530
Monnin T, Malosse C, Peeters C (1998) Solid-phase microextraction and
cuticular hydrocarbon differences related to reproductive activity in
queenless ant Dinoponera quadriceps. J Chem Ecol 24:473–490
Oliveira PS, Hölldobler B (1990) Dominance orders in the ponerine ant
Pachycondyla apicalis (Hymenoptera, Formicidae). Behav Ecol
Sociobiol 27:385–393
Ottoni EB (2000) EthoLog 2.2: a tool for the transcription and timing of
behavior observation sessions. Behav Res Methods Instrum Comput
32:446–449
Peeters C, Liebig J (2009) Fertility signaling as a general mechanism of
regulating reproductive division of labor in ants. In: Gadau J, Fewell
J (eds) Organization of insect societies: from genome to socio-com-
plexity). Harvard University Press, Cambridge, pp 220–242
R Core Team (2012) R: a language and environment for statistical com-
puting. R Foundation for Statistical Computing, Vienna, Austria.
http://www.R-project.org/. Accessed 18 June 2012
Ratnieks FLW, Foster KR, Wenseleers T (2006) Conflict resolution in
insect societies. Annu Rev Entomol 51:581–608
Reeve HK, Jeanne RL (2003) From individual control to majority rule:
extending transactional models of reproductive skew in animal so-
cieties. Proc R Soc Lond B Biol Sci 270:1041–1045
Reyment RA (1989) Compositional data analysis. Terra Nov. 1:29–34
Searcy WA, Nowicki S (2005) The evolution of animal communication:
reliability and deception in signaling systems. Princeton University
Press, Princeton
Smith AA, Hölldobler B, Liebig J (2008) Hydrocarbon signals explain
the pattern of worker and egg policing in the ant Aphaenogaster
cockerelli. J Chem Ecol 34:1275–1282
Smith AA, Hölldober B, Liebig J (2009) Cuticular hydrocarbons reliably
identify cheaters and allow enforcement of altruism in a social in-
sect. Curr Biol 19:78–81
Smith AA, Millar JG, Hanks LM, Suarez AV (2012) Experimental evi-
dence that workers recognize reproductives through cuticular hydro-
carbons in the ant Odontomachus brunneus. Behav Ecol Sociobiol
66:1267–1276
Számadó S (2011) The cost of honesty and the fallacy of the handicap
principle. Anim Behav 81:3–10
Van Oystaeyen A, Oliveira RC, Holman L, van Zweden JS, Romero C,
Oi CA, d’Ettorre P, Khalesi M, Billen J, Wackers F, Millar JG,
Wenseleers T (2014) Conserved class of queen pheromones stops
social insect workers from reproducing. Science 343:287–290
Wenseleers T, Ratnieks FLW (2006) Comparative analysis of worker
reproduction and policing in eusocial Hymenoptera supports relat-
edness theory. Am Nat 168:E163–E179
Wenseleers T, Helanterä H, Hart A, Ratnieks FLW (2004) Worker repro-
duction and policing in insect societies: an ESS analysis. J Evol Biol
17:1035–1047
Wilson EO (1971) The insect societies. Harvard University Press,
Cambridge
Wyatt TD (2014) Pheromones and animal behavior: chemical signals and
signatures, 2nd edn. Cambridge University Press, Cambridge
Yagound B, Blacher P, Fresneau D, Poteaux C, Châline N (2014) Status
discrimination through fertility signalling allows ants to regulate
reproductive conflicts. Anim Behav 93:25–35
Zahavi A (1975) Mate selection–a selection for a handicap. J Theor Biol
53:205–214
JChemEcol
