Content uploaded by Jelle S. van Zweden
Author content
All content in this area was uploaded by Jelle S. van Zweden
Content may be subject to copyright.
Do Cuticular Hydrocarbons Provide Sufficient Information
for Optimal Sex Allocation in the Ant Formica exsecta?
Jelle S. van Zweden &Emma Vitikainen &
Patrizia d’Ettorre &Liselotte Sundström
Received: 31 July 2011 /Revised: 9 October 2011 /Accepted: 7 November 2011 /Published online: 23 November 2011
#Springer Science+Business Media, LLC 2011
Abstract Split sex ratio theory predicts that when kin
structure varies among colonies of social insects, in order to
maximize the inclusive fitness, colonies with relatively high
sister-sister relatedness should specialize in producing
reproductive females, whereas in those with relatively low
sister-sister relatedness workers should bias their sex ratio
towards males. However, in order to achieve this, workers
need to be able to reliably assess the type of colony in
which they live. The information on colony kin structure
may be encoded in cuticular hydrocarbons (CHCs),
assuming that genetic variability translates accurately into
chemical variability. However, in genetically heterogeneous
colonies, too accurate information may encourage the
pursuit of individual interests through nepotistic behavior
and reduce colony efficiency or cause social disruption. In
this study, we estimated how well variability of CHC
recognition cues reflects colony kin structure in the ant
Formica exsecta. Our results show that CHC variability
does not covary with kin structure or the overall genetic
diversity of the colony, and that patrilines and matrilines
can have distinct CHC profiles in some but not all colonies.
However, within-colony relatedness remains the key deter-
minant of colony sex ratios. Based on our results, CHC
variability cannot serve as accurate information on within-
colony relatedness, kin structure, or full-sib affiliation, nor
do workers seem to use colony CHC variability as a proxy
for sex-ratio adjustment. The use of this type of information
thus could lead workers to make mistakes, and it remains
unclear how colonies of Formica exsecta adjust offspring
sex ratio to their optimal value.
Key Words Insects .Social evolution .Nepotism .Kin
selection .Levels of selection .Genetic variance .Alkanes .
Alkenes
Introduction
Reproductive division of labor is the key contributor to the
ecological success of social insects (Wilson, 1971). How-
ever, despite an impressive degree of cooperation, social
insect colonies are rife with potential conflict, which arises
because colony members, not being clonal, each have an
“incentive”to pursue their individual fitness interests
(reviewed in Ratnieks et al., 2006). The worker-queen
conflict over sex ratio is contingent on the fact that the
haplodiploid system of sex determination of Hymenoptera
Electronic supplementary material The online version of this article
(doi:10.1007/s10886-011-0038-x) contains supplementary material,
which is available to authorized users.
J. S. van Zweden (*):P. d’Ettorre
Centre for Social Evolution, Department of Biology,
University of Copenhagen,
Universitetsparken 15,
2100 Copenhagen, Denmark
e-mail: jvzweden@bio.ku.dk
E. Vitikainen :L. Sundström
Department of Biosciences, University of Helsinki,
PO Box 65, Helsinki 00014, Finland
Present Address:
E. Vitikainen
Centre for Ecology and Conservation, University of Exeter,
Tremough Campus,
TR10 9EZ(Penryn, UK
Present Address:
P. d’Ettorre
Laboratoire d’Ethologie Expérimentale et Comparée (LEEC),
University of Paris 13,
99 av. J.B. Clément,
93430 Villetaneuse, France
J Chem Ecol (2011) 37:1365–1373
DOI 10.1007/s10886-011-0038-x
creates asymmetric genealogies, such that full sisters are
more closely related to each other, than they are to their
brothers (Trivers and Hare, 1976). This relatedness asym-
metry, however, decreases when the queen has mated more
than once (polyandry) or if colonies contain several related
queens (polygyny) (Boomsma, 1993). Consequently, when
kin structure varies among colonies in the same population,
theory predicts that workers can enhance their inclusive
fitness by specializing on the sex to which they are most
related relative to the population average, i.e., females in
monogynous/monandrous colonies (with high relatedness
asymmetry), and males in polyandrous or polygynous
colonies (with low relatedness asymmetry) (split sex ratio
theory; Boomsma and Grafen, 1990,1991). Indeed,
empirical studies, mainly of ant species, have found sex
allocation patterns consistent with worker control in
response to variation in relatedness asymmetry (Chan and
Bourke, 1994; Sundström, 1994; Sundström et al., 1996),
although other studies have found sex allocation patterns
inconsistent with worker control in response to variation in
relatedness asymmetry (Brown and Keller, 2000; Liautard
et al., 2003; Bonckaert et al., 2011).
Precise adjustment of sex ratios by colony workers in
response to colony kin structure requires a mechanism by
which they can assess the genetic diversity within colonies,
and manipulate brood composition accordingly, e.g., by
removing male larvae (Chapuisat et al., 1997). A parsimo-
nious hypothesis is that recognition cues are at least partly
genetically determined, so that within-colony recognition
cue variability to a certain extent reflects genetic diversity
and, thus, relatedness asymmetry. Ants and other social
insects use cuticular hydrocarbons (CHCs) for discriminat-
ing colony members from strangers to maintain colony
integrity. Typically, each colony is characterized by specific
proportions of a set of CHCs (reviewed in van Zweden and
d’Ettorre, 2010), the exact relative compositions of which
can have strong genetic underpinnings (e.g., Stuart, 1988;
van Zweden et al., 2009,2010). On the other hand, if the
individuals’blends of CHCs accurately reflect their genetic
lineage within a colony, this could allow workers in
colonies that contain different patrilines or matrilines to
favor close kin to the detriment of more distant kin, and so
promote their own evolutionary interests. Such nepotistic
discrimination is expected to carry costs with reduced
colony efficiency and social disruption in its wake (Keller,
1997), and may lead to selection against highly diverse
recognition cues (Ratnieks et al., 2007). Therefore, the level
of information encoded in social insect recognition cues is
likely to be the result of these two opposing evolutionary
forces, being good enough to allow efficient discrimination
against non-nestmates but noisy enough to prevent precise
discrimination within colonies (van Zweden et al., 2010).
Evidence to date indicates that this is indeed the case in the
ant Formica truncorum (Boomsma et al., 2003) and the
wasp Vespa crabro (Danietal.,2004), where CHC
variation explained by patriline affiliation is variable among
colonies, although a recent study on Acromyrmex octospi-
nosus shows higher kin specificity (Nehring et al., 2010).
Not all components of the CHC profile are of equal
importance in inter- and intra-colony recognition. Methyl
branched alkanes and alkenes have been especially impli-
cated in nestmate recognition, whereas the role of linear
alkanes in nestmate recognition is less clear (Dani et al.,
2001,2005; Akino et al., 2004; Greene and Gordon, 2007;
Martin et al., 2008; Guerrieri et al., 2009). Conversely, the
relative abundance of linear alkanes appears, for example,
to be involved in within-colony recognition of (worker)
castes in harvester ants (Wagner et al., 2001; Greene and
Gordon, 2003; Martin and Drijfhout, 2009). Therefore, we
hypothesize that CHCs associated with nestmate recogni-
tion may be uniform within colonies, whereas those
involved in the assessment of colony kin structure may be
more variable (c.f., van Zweden et al., 2010).
In the ant Formica exsecta, both the number of queens
that head the colony (monogyny vs. polygyny) and the
number of males that mate with a queen (monandry vs.
polyandry) vary. A previous study showed sex ratio
specialization consistent with the predictions from split
sex ratio theory (Sundström et al., 1996). This implies that
workers can assess colony kin structure also when the
relatedness asymmetry varies solely due to queen mating
frequency. The CHC profile of F. exsecta typically consists
of five to six linear alkanes (n-C
21
:n-C
31
) and five to six
alkenes ((Z)-9-C
21:1
:(Z)-9-C
31:1
), of which the latter have
been implicated as nestmate recognition cues (Martin et al.,
2008). In this study, we compared the CHC variability
among workers in monogynous/monandrous, monogynous/
polyandrous, and polygynous colonies of F. exsecta,to
assess whether workers have enough information to act in
their best inclusive fitness interests. We predicted that CHC
variability would be greater in polyandrous and polygynous
than in monogynous/monandrous colonies, thus reflecting
lower relatedness asymmetry. The second prediction was
that greater CHC variability corresponds to colonies with
male-biased sex ratios. We then combined the information
on CHC and genetic diversity with sex ratio data for the
same colonies to evaluate the degree to which sex
allocation corresponds to colony relatedness on the one
hand and CHC variability on the other.
Methods and Materials
Study Organisms and General Procedures Workers of F.
exsecta were collected from 9 colonies in June 2007 and
from 6 colonies in June 2009 (Table 1), on islands off the
1366 J Chem Ecol (2011) 37:1365–1373
Tvärminne Zoological Station, southwestern Finland. The
population comprises largely monogynous colonies headed
by queens mated with 1 to 3 males, with the exception of a
few polygynous colonies (Haag-Liautard et al., 2009). This
is the same population where Sundström et al. (1996) found
that colonies headed by a singly mated queen specialize in
female brood, and those headed by a multiply mated queen
specialize in male production. Altogether we chose 5
colonies with a single queen that had mated once
(monogynous-monandrous, MG/MA), 5 colonies where
the queen had mated multiply (monogynous-polyandrous,
MG/PA) and 5 colonies with multiple queens (polygynous,
PG) (Table 1). From each colony, 18–24 workers were
collected and killed by freezing at −20°C for later genetic
and chemical analysis.
Sex ratios of the study colonies were assessed by
sampling 50–60 sexual pupae from each colony, in June-
July 2007 and 2009 after all the larvae had pupated and
before any sexuals had emerged. The sex of the pupae was
determined based on morphology, and the sex ratio was
calculated as the proportion of queens of all sexual brood
(c.f., Vitikainen et al., 2011).
Genetic Analysis To confirm the kin structure of the
sampled colonies and to determine patriline or matriline
affiliation, the 18–24 workers of each colony were
genotyped at ten loci: Fe11, Fe13, Fe17, Fe37, Fe38,
Fe42, Fe49 (Gyllenstrand et al., 2002), Fl21 (Chapuisat,
1996), P22 (Trontti et al., 2003), and Fy3 (Hasegawa and
Imai, 2004). Further details of PCR conditions are equal to
Haag-Liautard et al. (2009). The PCR-products were
separated by using automated capillary sequencer (Mega-
BACE 1000) and sized against ET400-R standard (GE
Healthcare). The genotypes were scored with the program
Fragment Profiler v1.2 (GE Healthcare), and allele calling
was confirmed manually.
We used maximum likelihood methods to confirm the
genetic composition of the colonies, and to assign workers
into groups of full- and half-sisters, as implemented in the
program COLONY 1.3. (Wang, 2004). Average within-
colony relatedness was calculated as the pedigree related-
ness, based on the equations in Bourke and Franks (1995),
assuming r=0.75 for MG/MA colonies, and using the
observed paternity shares in MG/PA colonies. For the PG
colonies, we inferred the number of matrilines from the
genotype data assuming the queens were unrelated (r =0).
The true value may be higher, so we ran each test also with
higher values without finding any differences in the
outcome. Pairwise relatedness estimates between individu-
als and colonies were calculated based on individual
genotypes at each of the ten loci using Relatedness 5.0.8
(Goodnight and Queller, 1999), and colony inbreeding
(HL) following the procedure by Aparicio et al. (2006). We
used the background population genetic data from 102
colonies (c.f. Haag-Liautard et al., 2009) to enhance the
accuracy of allele frequency and relatedness estimates.
Table 1 Basic data for the sampled colonies
Colony Year NSex ratio Inbreeding
coefficient
Gene
diversity
Allelic
richness
Nr. of matri/
patrilines
Pedigree
relatedness
All compounds (Z)-9-Alkenes n-Alkanes
MG/MA 1 2007 20 0.85 0.29 0.513 2.275 1 m/1p 0.750 1.664 1.021 1.525
MG/MA 2 2007 18 0.17 0.04 0.548 2.279 1 m/1p 0.750 1.676 0.838 1.469
MG/MA 3 2007 19 1.00 0.04 0.513 2.214 1 m/1p 0.750 1.860 0.843 1.214
MG/MA 4 2009 24 0.96 0.26 0.479 2.130 1 m/1p 0.750 1.871 0.423 1.154
MG/MA 5 2009 23 0.00 0.00 0.585 2.462 1 m/1p 0.750 1.365 0.569 1.017
MG/PA 1 2007 20 0.13 0.26 0.516 2.194 1 m/2p 0.503 2.817 1.337 1.562
MG/PA 2 2007 19 0.55 0.39 0.408 1.983 1 m/2p 0.540 1.946 1.191 1.259
MG/PA 3 2007 19 1.00 0.16 0.647 2.805 1 m/2p 0.590 2.701 1.308 1.970
MG/PA 4 2009 24 0.00 0.12 0.577 2.488 1 m/2p 0.674 1.499 0.534 1.074
MG/PA 5 2009 23 0.00 0.11 0.583 2.415 1 m/2p 0.516 1.531 0.277 1.262
PG 1 2007 18 0.15 0.25 0.718 3.373 8 m 0.094 1.940 0.759 1.063
PG 2 2007 19 0.00 0.28 0.622 2.930 7 m 0.107 1.914 0.856 1.368
PG 3 2007 20 0.00 0.28 0.714 3.305 9 m 0.083 2.774 0.848 1.890
PG 4 2009 23 0.29 0.33 0.446 2.180 4 m 0.083 1.475 0.579 1.343
PG 5 2009 23 0.04 0.40 0.604 2.933 7 m 0.048 1.321 0.683 1.129
Colony name also reflects colony type (MG/MA monogynous-monandrous, MG/PA monogynous-polyandrous, PG polygynous). Nequals the
number of individuals screened per colony. Sex ratio is expressed as the number of females as a proportion of the total number of sexual offspring.
Pedigree relatedness is calculated based on the assumption that queen relatedness is zero. All compounds, (Z)-9-Alkenes, and n-Alkanes refer to
cuticular hydrocarbon (CHC) variability measures based on the respective sets of compounds.
J Chem Ecol (2011) 37:1365–1373 1367
Colony-specific gene diversity and allelic richness were
calculated with colony defined as population using Fstat
2.9.3 (Goudet, 2001). As the two measures correlated
strongly (r=0.94, df=15, P< 0.001), we used the factor
scores from the first axis in a PCA encompassing the two
measures. Comparisons between the different categories of
colonies were done with a one-way ANOVA.
Chemical Analysis Cuticular lipids were obtained by
individually immersing the 18–24 freshly killed workers
(before genotyping them) in 200 μl HPLC-grade pentane
for 10 min, with gentle vortexing for 15 sec at the start and
end of the 10 min. The pentane was left to evaporate at
room temperature under a laminar air-flow hood. Extracts
were resuspended in 50 μl pentane, 2 μl of which were
injected into an Agilent Technologies 6890 N gas chro-
matograph (GC), connected to an Agilent Technologies
5975 mass selective detector (MS), using 70 eV electron
impact ionization. The GC was equipped with an HP-
5MS capillary column (30 m×0.25 mm ID, 0.25 μm
film thickness) and a split-splitless injector. The carrier
gas was helium at 1 ml/min. After an initial hold of
1 min at 70°C, the temperature rose to 180°C at a rate
of 30°C/min, and then to 320°C at 4°C/min with a final
hold of 5 min.
The areas of five (Z)-9-alkenes and five linear alkanes
found on the cuticle of all workers (c.f. Martin et al., 2008)
were integrated for further analysis with Agilent ChemSta-
tion v. D.02.00.237. Peak areas were normalized for each
individual by using a Z-transformation (Aitchison, 1986).
Statistics We estimated within-colony CHC variability both
for single compounds and for sets of multiple compounds
[all compounds (10 variables), (Z)-9-alkenes only (5
variables), and linear alkanes only (5 variables)]. For single
compounds, we calculated the within-colony standard
deviation from the Z-transformed values. For sets of
multiple compounds, we first scaled the Z-transformed
variables to zero mean and unit variance, and then
calculated the absolute Euclidean distance to the mean of
the colony. CHC variability then was expressed per colony
as the mean distance of its individuals.
To assess which factors may influence within-colony
CHC variability, we used ANCOVAs with within-colony
CHC variability as the dependent variable, and colony kin
structure (MG/MA, MG/PA, PG) or within-colony related-
ness, year of sampling (2007 or 2009), within-colony
genetic diversity (as captured in the factor scores for gene
diversity and allelic richness combined), and colony
inbreeding (HL) as the predictor variables. We included
HL in the analysis, given that two earlier studies have
indicated a significant effect of inbreeding on sex ratios
(Haag-Liautard et al., 2009; Vitikainen et al., 2011). To test
the effects of within-colony CHC variability, relatedness,
genetic diversity, and inbreeding on colony sex ratio we
used a multiple regression with stepwise backward elimi-
nation. These analyses were done in Statistix 9 (Analytical
Software, USA).
To test for an association between genetic and CHC
similarity of colonies and individuals within colonies, we
created distance matrices, one with the pairwise genetic
relatedness based on the ten genotyped loci, and one with
the pairwise Euclidean CHC distance. Pairwise Euclidean
CHC distance was calculated for sets comprising either all
compounds (10 variables), (Z)-9-alkenes only (5 variables),
or linear alkanes only (5 variables). The correlation
between the genetic and CHC matrices then was tested
using Mantel tests in the program Genodive (v. 2.0b20)
(Meiermans, 2010). Correlations between relatedness and
CHC distances were performed both between individuals
within colonies (i.e., for each colony separately) and
between colonies.
Finally, to estimate variation among hydrocarbon pro-
files explained by patriline or matriline affiliation, we
performed a PCA using the MGPA and PG colonies, based
on the three sets of multiple compounds [all compounds,
linear alkanes only, and (Z)-9-alkenes only]. This was
followed by a MANOVA in the program R (v. 2.12.0) with
those PCs explaining >95% of the variation as dependent
variables, and colony and patri- or matriline nested within
colony as the explanatory variables. We repeated this
analysis for each colony separately, except that colony
was not used as an explanatory variable in the MANOVA.
We corrected the above analyses for multiple testing by
using the false discovery rate as appropriate (Benjamini and
Hochberg, 1995).
Results
All putative MG/MA colonies were confirmed to be
composed of full sisters only. All five MG/PA colonies
had two patrilines, and the samples from PG colonies
contained between 4 and 9 matrilines (Table 1). The
diversity of (Z)-9-alkenes was in both sampling years
significantly lower than that of n-alkanes, and the diversity
of both was lower in 2009 than in 2007 (full factorial
ANOVA, compound type: F
1,26
=46.3, P<0.001, year: F
1,26
=
23.5, P<0.001, compound type x year: F
1,26
=1.08, P=0.31).
We found no significant differences in CHC variability
between PG, MG/PA, and MG/MA colonies, neither when
all compounds were considered together, nor when the
subsets of (Z)-9-alkenes or n-alkanes were considered
separately [Fig. 1; one-way ANOVA with colony type as
explanatory variable: (Z)-9-alkenes: F
2,12
=0.57, P=0.58,
1368 J Chem Ecol (2011) 37:1365–1373
n-alkanes: F
2,12
=0.31, P=0.74, all compounds: F
2,12
=0.84,
P=0.46]. Similarly, in the more comprehensive model,
neither within-colony relatedness, genetic diversity (factor
scores for the combined effects of allelic diversity and gene
diversity), nor colony inbreeding were significantly associ-
ated with CHC variability in any of the three multi-
compound sets [all compounds, (Z)-9-alkenes, n-alkanes;
Table 2], or any of the individual compounds (Appendix 1).
The outcome was qualitatively the same when colony type
(PG, MG/PA, and MG/MA) was used instead of within-
colony relatedness (Appendix 1). Indeed, within-colony
CHC variability was not significantly correlated to any
of the other measures, except for sampling year
(Table 2). However, genetic diversity did not significantly
differ among the three colony types either (PG, MG/PA,
MG/MA) (F
2,12
=3.25, P=0.07), suggesting that the
increased number of genetic lineages in MG/PA and PG
colonies did not necessarily increase genetic diversity.
We found a portion of CHC variation was explained by
patri- and matrilines when all compounds or (Z)-9-alkenes
only were considered (all compounds: Wilks’λ=0.333,
approx. F
148,632.1
=1.36, P<0.01; (Z)-9-alkenes: Wilks’λ=
0.420, approx. F
111,477.1
=1.44, P<0.01; n-alkanes: Wilks’
λ=0.521, approx. F
111,477.1
=1.05, P=0.367), although
these differences were only pronounced in some of the
colonies (Table 3). Matrilines differed significantly in their
CHCs in one of the five PG colonies, regardless whether all
compounds or only (Z)-9-alkenes were used. Similarly, in
one of the five MG/PA colonies patrilines differed
significantly with respect to the (Z)-9-alkenes (all corrected
for false discovery rate; Table 3). We also found a negative
correlation between pairs of colonies for pairwise related-
ness and (Z)-9-alkene distance after correction for false
discovery rate (r=−0.342, P=0.001), but not for n-alkanes
or when all compounds were considered (r=−0.021, P=
0.401, and r=−0.213, P=0.016, respectively) (Fig. 2). This
shows that the relative abundance of (Z)-9-alkenes has a
strong genetic component with respect to between-colony
variation. However, within colonies, we found no signifi-
cant associations between genetic and CHC distances for
any of the considered compound sets (Appendix 2).
The sex ratios of all PG colonies were highly male
biased, as expected based on their low relatedness (Fig. 3).
The monogynous colonies were more variable. Two of the
five MG/PA colonies produced a female biased sex ratio,
and two of the five MG/MA colonies produced mainly
males (Fig. 3). When the effects of colony-specific CHC
variability on sex ratio were considered in conjunction with
within-colony relatedness and colony inbreeding (HL), the
only significant determinant of colony sex ratios was
within-colony relatedness (full model including relatedness
and HL as explanatory factors: R
2
=0.40, F
2,12
=4.04, P=
0.045; HL: F
2,12
=3.13, P=0.102; relatedness F
2,12
=8.07,
P=0.015; all other variables, year, all compounds, (Z)-9-
alkenes, n-alkanes: F
2,12
<0.35, P>0.50).
Discussion
In this study, we showed that in Formica exsecta the
within-colony CHC variability is not contingent on colony
kin structure, yet genetically related colonies have more
similar chemistry, showing a genetic component to the
cuticular chemistry without a direct link with the genetic
composition of the colony. Nonetheless, genetic lineages
were chemically distinct in some but not all colonies. This
suggests that chemical cues by which workers can assess
colony kin structure are occasionally available, but that
these cues generally are unreliable, so that workers would
make too many mistakes when using them as a proxy for
the sex ratio to be produced by the colony.
Polygynous and monogynous/polyandrous colonies
showed no higher amounts of CHC variability than
monogynous/monandrous ones. This is in agreement with
the result obtained by S.J. Martin et al. (unpublished data)
for the same species, but stands in stark contrast with the
general expectation of greater CHC variability in geneti-
cally more heterogeneous, i.e., polygynous and/or mono-
Fig. 1 CHC variability, colony type and CHC group (mean ±95%CI;
MG/MA monogynous-monandrous, MG/PA monogynous-polyandrous,
PG polygynous)
Table 2 Ancova results for effects on CHC variability
Effect All compounds (Z)-9-Alkenes n-Alkanes
F P FPFP
Year 5.63 0.04 20.09 0.001 3.71 0.08
Relatedness 0.12 0.73 0.33 0.58 0.01 0.92
Genetic diversity 0.39 0.55 0.12 0.73 0.07 0.80
Inbreeding 0.25 0.63 1.64 0.23 0.06 0.81
Within-colony relatedness was calculated assuming zero relatedness
among queens, and genetic diversity as the combined effects of
allelic richness and gene diversity captured in their factor scores.
df1=1, df2=10 in all cases.
J Chem Ecol (2011) 37:1365–1373 1369
gynous/polyandrous societies, which has formed the foun-
dation for assuming lower discrimination abilities in
polygynous societies (e.g., Fletcher and Michener, 1987;
Starks et al., 1998). However, we also found that the
within-colony genetic diversity was not higher in the
polygynous and the monogynous/polyandrous colonies
than in the monogynous/monandrous ones. Indeed, al-
though within-colony relatedness varied 15-fold, the
corresponding ranges for both gene diversity and allelic
richness were only 1.7 fold in both cases (Table 1). Thus,
the genetic basis for a greater diversity is not present, which
may explain the lack of correlation between colony kin
structure and CHC variability.
The (Z)-9-alkenes overall were less variable within
colonies than n-alkanes. This supports the existing evidence
that mainly alkenes mediate nestmate recognition in F.
exsecta, whereas alkanes are less informative in this respect
(Martin et al., 2008). Moreover, the between-colony CHC
distance for (Z)-9-alkenes, but not the n-alkanes, decreased
significantly with increasing between-colony pairwise
relatedness. This, in conjunction with the result that
within-colony CHC variability showed no clear association
with genetic diversity, suggests a strong genetic foundation
of the (Z)-9-alkene profile that is to a large extent blurred
within colonies by odor cue transfer between nestmates (c.f.
van Zweden et al., 2010). It also suggests unequal mixing
Fig. 3 Sex ratios (numerical proportion of females in the brood) as a
function of within-colony relatedness, grouped according to colony
type (MG/MA monogynous-monandrous, MG/PA monogynous-
polyandrous, PG polygynous)
Z
Fig. 2 Correlations between colony pairwise relatedness and colony
CHC distances. Association is significant, after correction for false
discovery rate, for (Z)-9-alkenes (Mantel’s test; (Z)-9-alkenes, r=−0.342,
P<0.001), but not when n-alkanes or all compounds were considered
(n-alkanes, r=−0.021, P=0.401; all compounds, r=−0.213, P=0.016)
Table 3 Discriminant analyses between genetic lineages within MGPA and PG colonies
Colony All compounds (3 PCs) (Z)-9-Alkenes (2 PCs) n-Alkanes (2 PCs)
Wilks’λFPWilks’λFPWilks’λFP
All colonies 0.333 F
148,632
=1.36 0.007 0.420 F
111,477
=1.44 0.005 0.521 F
111,477
=1.05 0.367
MG/PA 1 0.795 F
3,16
=1.38 0.285 0.891 F
2,17
=1.05 0.373 0.858 F
2,17
=1.41 0.272
MG/PA 2 0.502 F
3,14
=4.63 0.018 0.810 F
2,15
=1.76 0.206 0.577 F
2,15
=5.49 0.016
MG/PA 3 0.730 F
3,15
=1.85 0.182 0.711 F
2,16
=3.25 0.065 0.851 F
2,16
=1.40 0.276
MG/PA 4 0.732 F
3,20
=2.44 0.094 0.539 F
2,21
=8.99 0.002 0.829 F
2,21
=2.16 0.140
MG/PA 5 0.718 F
3,19
=2.49 0.092 0.821 F
2,20
=2.18 0.140 0.902 F
2,20
=1.09 0.355
PG 1 0.098 F
20,20.7
=1.23 0.323 0.482 F
14,16
=0.50 0.899 0.221 F
14,16
=1.29 0.310
PG 2 0.290 F
18,28.8
=0.88 0.606 0.255 F
12,22
=1.80 0.113 0.325 F
12,22
=1.38 0.246
PG 3 0.220 F
24,26.7
=0.76 0.746 0.413 F
16,20
=0.70 0.767 0.379 F
16,20
=0.78 0.690
PG 4 0.088 F
9,41.5
=7.92 <0.001 0.023 F
6,36
=33.52 <0.001 0.468 F
6,36
=2.77 0.026
PG 5 0.323 F
18,40.1
=1.09 0.394 0.351 F
12,30
=1.72 0.111 0.436 F
12,30
=1.29 0.277
A PCA was performed on the Z-transformed CHC data of either all colonies together or single colonies; PCs explaining >95% of the variation
were then included as the dependent variable in a MANOVA with genetic lineage (nested within colony, when all colonies were considered
together) as the explanatory variable.
1370 J Chem Ecol (2011) 37:1365–1373
of CHCs, which could be achieved by accumulating
alkenes to a greater extent than alkanes in the postpha-
ryngeal gland (PPG; Soroker et al., 1994). The PPG does
not synthesize hydrocarbons, but acts as a reservoir for
these compounds to be homogenized and exchanged via
grooming and trophallaxis (Soroker et al., 1994). Thus,
inequalities in diversity could arise if alkenes are released
more readily, or workers metabolize alkanes more rapidly,
with alkenes becoming better mixed than alkanes among
colony members. Altogether, these results suggest a pattern
similar to that observed in F. rufibarbis, where cues used in
nestmate recognition are genetically determined but prefer-
entially mixed among nestmates (van Zweden et al., 2010).
Although the within-colony CHC variability showed no
correlation with colony kin structure, relatedness, or any of
the measures of genetic diversity, we found significant
CHC divergence between genetic lineages within colonies.
This was, however, only the case in two of the ten colonies
that comprised more than one lineage. Earlier analyses in
the wasps Polistes dominulus,Vespa crabro (Dani et al.,
2004), and Dolichovespula saxonica (Bonckaert et al.,
2011), and the ant Acromyrmex octospinosus (Nehring et al.,
2010), found more consistent CHC differences between
matrilines and patrilines. However, the CHC profile of F.
exsecta exhibits far fewer hydrocarbons with fewer structural
groups than these species (P. dominulus:44compounds;V.
crabro:25compounds;D. saxonica:56compounds;A.
octospinosus: 38 compounds), and so offers less scope for
both divergence and statistical discrimination between
genetic lineages. However, in F. truncorum,Boomsmaet
al. (2003) nonetheless were able to discriminate a greater
fraction of patrilines although the analysis entailed only nine
compounds.
At least two mutually non-exclusive explanations for the
relative lack of CHC discrimination among genetic lineages
are possible. First, active cue scrambling may render the
colony odor more uniform or blur the recognition code. This
may be an evolved response that either mitigates the
opportunities for workers to selfishly favor close kin and
disfavor more distant kin with social disruption in its wake, or
increases discrimination between members of different colo-
nies (Keller, 1997; van Zweden et al., 2010). Theory predicts
depauperate CHC profiles and cue mixing when potential
conflicts are rife in the colony (e.g., Sundström and
Boomsma, 2001). This is indeed the case in F. exsecta,but
many more compounds have been found in F. truncorum
(Akino, 2006), which goes against this interpretation.
Second, given that inbreeding is rife in the population of F.
exsecta studied here (Sundström et al., 2003;Haag-Liautard
et al., 2009; Vitikainen et al., 2011), genetic variation for cue
diversity may have been purged. We found no significant
effects of inbreeding on CHC variability, but the measure we
used as a covariate in the analyses does not reflect the
absolute level of inbreeding, only the relative level. Thus, a
significant effect of inbreeding may not be detectable.
Within-colony CHC variability had no significant pre-
dictive value on sex ratio, but colonies with higher
relatedness produced more females, when also accounting
for colony inbreeding. This suggests that workers do not
use the CHC variability of adult individuals as a proxy for
sex ratio decisions, in contrast to the results obtained by
Boomsma et al. (2003) for F. truncorum. It begs the
question by what mechanism this is achieved. All poly-
gynous colonies produced a male-biased brood, but against
expectations based on the kin value of brood, one
polyandrous colony produced only female brood and one
monandrous colony only male brood. In the case of
polyandrous colonies, errors in the assessment of colony
kin structure may occur if the two fathers are closely
related, and so mediate similar CHC profiles to their
offspring. However, the polyandrous colony that produced
an all-female brood had the lowest relatedness between
fathers, and not an exceptionally unequal contribution by
the two fathers (Table 1). It remains to be shown what
information the workers use to assess colony kin structure,
and why such apparent errors occur. A possibility is that
CHC information on eggs or larvae more accurately reflects
the genetic make-up of the colony, owing to no or less
transfer of cues between genetic lineages, and that nurse
workers bias the sex ratio accordingly. Another possibility
is that nurse workers can assess the number of queens
directly or by colony egg production, and use this as a
proxy to adjust sex ratio in polygynous colonies. This is
consistent with split-sex ratio theory (Boomsma, 1993),
because polygynous colonies will tend to have higher egg-
production and a lower than population-level-average
relatedness asymmetry, and also with the idea that
specialization in the cheaper sex (males weigh less in F.
exsecta; Vitikainen et al., 2011) requires greater initial egg-
production for equal total biomass production. On the other
hand, this does not explain the tendency of monogynous/
polyandrous colonies to specialize in male production.
Environmental factors also are likely to influence the
optimal sex ratio for a colony, and hence may explain the
discrepancies between the observed pattern and the out-
come expected based on purely inclusive fitness arguments
(Nielsen et al., 1999; Liang and Silverman, 2000).
In summary, we found no support for the suggested
association between colony kin structure and variability of
recognition cues (c.f. Fletcher and Michener, 1987), but this
lack of relationship may be explained by the uniform levels
of genetic variation we found among the different colony
types. The breeding patterns and the genetic structure of the
population, which indicate non-trivial levels of inbreeding
(Sundström et al., 2003; Haag-Liautard et al., 2009;
Vitikainen et al., 2011), may have contributed to the
J Chem Ecol (2011) 37:1365–1373 1371
depauperate chemical profile through genetic purging, a
question which will have to be addressed in future studies.
We also found that the within-colony CHC variability does
not allow accurate assessment of colony kin structure, and
that workers do not seem to use any single compound or set
of compound as a proxy for sex ratio adjustments. Rather,
genetically informative cues appear to be mixed among
workers to create a colony odor that is used in between-
colony discrimination. The cuticular hydrocarbon variabil-
ity of Formica exsecta may thus be subject to balancing
selection on the accuracy of genetic variability, allowing
colony members to serve their collective but not selfish
fitness interests, possibly blended with loss of genetic
diversity owing to inbreeding in this population.
Acknowledgements We thank V. Nehring for his help in analyzing
the chemical data, J.J. Boomsma for critical discussions, and all
members of the Copenhagen Centre for Social Evolution for providing
a stimulating working environment. This work was supported by the
EU Marie Curie Excellence Grant CODICES-EXT-CT-2004-014202
(to PdE), Academy of Finland grants 206404,121216, and 135975 (to
LS), LUOVA graduate school in Helsinki (to EV), and the NordForsk
research network “Social evolution in insects”. JSvZ was also
supported by a postdoctoral fellowship from the Danish Council for
Independent Research (09–066595).
References
AITCHISON, J. 1986. The statistical analysis of compositional data. The
Blackburn Press, Caldwell.
AKINO, T. 2006. Cuticular hydrocarbons of Formica truncorum
(Hymenoptera: Formicidae): Description of new very long
chained hydrocarbon components. Appl. Entomol. Zool.
41:667–677.
AKINO, T., YAMAMURA, K., WAKAMURA, S., and YAMAOKA, R. 2004.
Direct behavioral evidence for hydrocarbons as nestmate recog-
nition cues in Formica japonica (Hymenoptera: Formicidae).
Appl. Entomol. Zool. 39:381–387.
APARICIO, J. M., ORTEGO, J., and CORDERO, P. J. 2006. What should
we weigh to estimate heterozygosity, alleles or loci? Mol. Ecol.
15:4659–4665.
BENJAMINI,YandHOCHBERG, Y. 1995. Controlling the false
discovery rate: A practical and powerful approach to multiple
testing. J. Roy. Stat. Soc. B 57:289–300.
BONCKAERT,W.,VA N ZWEDEN, J. S., D’ETTORRE, P., BILLEN, J., and
WENSELEERS, T. 2011. Colony stage and not facultative policing
explains pattern of worker reproduction in the Saxon wasp. Mol.
Ecol. 20:3455–3468.
BOOMSMA, J. J. 1993. Sex ratio variation in polygynous ants, pp. 86–
109, in L. Keller (ed.), Queen number and sociality in insects.
Oxford University Press, New York.
BOOMSMA, J. J. and GRAFEN, A. 1990. Intraspecific variation in ant
sex ratios and the Trivers-Hare hypothesis. Evolution 44:1026–
1034.
BOOMSMA,J.J.andGRAFEN, A. 1991. Colony-level sex ratio
selection in the eusocial Hymenoptera. J. Evol. Biol. 4:383–407.
BOOMSMA, J. J., NIELSEN, J., SUNDSTRÖM, L., OLDHAM, N. J.,
TENTSCHERT, J., PETERSEN, H. C., and MORGAN, E. D. 2003.
Informational constraints on optimal sex allocation in ants. Proc.
Natl. Acad. Sci. USA 100:8799–8804.
BOURKE, A. F. G. and FRANKS, N. R. 1995. Social evolution in ants.
Princeton University Press, Princeton.
BROWN, W. D. and KELLER, L. 2000. Colony sex ratios vary with
queen number but not relatedness asymmetry in the ant Formica
exsecta. Proc. R. Soc. B 267:1751–1757.
CHAN,G.L.andBOURKE, A. F. G. 1994. Split sex ratios in a
multiple-queen ant population. Proc. R. Soc. B 258:261–266.
CHAPUISAT, M. 1996. Characterization of microsatellite loci in
Formica lugubris B and their variability in other ant species.
Mol. Ecol. 5:599–601.
CHAPUISAT, M., LISELOTTE, S., and KELLER, L. 1997. Sex-ratio
regulation: The economics of fratricide in ants. Proc. R. Soc. B
264:1255–1260.
DANI,F.R.,JONES,G.R.,DESTRI,S.,SPENCER, S. H., and TURILLAZZI,S.
2001. Deciphering the recognition signature within the cuticular
chemical profile of paper wasps. Anim. Behav. 62:165–171.
DANI, F. R., FOSTER, K. R., ZACCHI, F., SEPPÄ, P., MASSOLO, A.,
CARELLI, A., ARÉVALO, E., QUELLER, D. C., STRASSMANN, J.,
and TURILLAZZI, S. 2004. Can cuticular lipids provide
sufficient information for within-colony nepotism in wasps?
Proc. R. Soc. B 271:745–753.
DANI, F. R., JONES, G. R., CORSI, S., BEARD, R., PRADELLA, D., and
TURILLAZZI, S. 2005. Nestmate recognition cues in the honey
bee: differential importance of cuticular alkanes and alkenes.
Chem. Senses 30:477–489.
FLETCHER, D. J. C. and MICHENER, C. D. 1987. Kin recognition in
animals. Wiley, New York.
GOODNIGHT, K. F. and QUELLER, D. C. 1999. Computer software for
performing likelihood tests of pedigree relationship using genetic
markers. Mol. Ecol. 8:1231–1234.
GOUDET, J. 2001. FSTAT, a program to estimate and test gene
diversities and fixation indices. Version 2.9.3.
GREENE, M. J. and GORDON, D. M. 2003. Cuticular hydrocarbons
inform task decisions. Nature 423:32–32.
GREENE, M. J. and GORDON, D. M. 2007. Structural complexity of
chemical recognition cues affects the perception of group
membership in the ants Linepithema humile and Aphaenogaster
cockerelli.J. Exp. Biol. 210:897–905.
GUERRIERI, F. J., NEHRING, V., JØRGENSEN, C. G., NIELSEN, J.,
GALIZIA, C. G., and D’ETTORRE, P. 2009. Ants recognize foes
and not friends. Proc. R. Soc. B 276:2461–2468.
GYLLENSTRand, N., GERTSCH,P.J.,AND PAMILO, P. 2002. Poly-
morphic microsatellite DNA markers in the ant Formica exsecta.
Mol. Ecol. Notes 2:67–69.
HAAG-LIAUTARD, C., VITIKAINEN, E., KELLER, L., and SUNDSTRÖM,
L. 2009. Fitness and the level of homozygosity in a social insect.
J. Evol. Biol. 22:134–142.
HASEGAWA, E. and IMAI, S. 2004. Characterization of microsatellite
loci in red wood ants Formica (s. str.) spp. and the related genus
Polyergus.Mol. Ecol. Notes 4:200–203.
KELLER, L. 1997. Indiscriminate altruism: unduly nice parents and
siblings. Trends Ecol. Evol. 12:99–103.
LIANG, D. and SILVERMAN, J. 2000. “Youarewhatyoueat”:Diet
modifies cuticular hydrocarbons and nestmate recognition in
the Argentine ant, Linepithema humile.Naturwissenschaften
87:412–416.
LIAUTARD, C., BROWN, W. D., HELMS, K. R., and KELLER, L. 2003.
Temporal and spatial variations of gyne production in the ant
Formica exsecta.Oecologia 136:558–564.
MARTIN, S. J. and DRIJFHOUT, F. 2009. Nestmate and task cues are
influenced and encoded differently within ant cuticular hydro-
carbon profiles. J. Chem. Ecol. 35:368–374.
MARTIN, S. J., VITIKAINEN, E., HELANTERÄ, H., and DRIJFHOUT,F.P.
2008. Chemical basis of nest-mate discrimination in the ant
Formica exsecta.Proc. R. Soc. B 275:1271–1278.
MEIERMANS P. G. 2010. GenoDive. Version 2.0b20.
1372 J Chem Ecol (2011) 37:1365–1373
NEHRING, V., EVISON, S. E. F., SANTORELLI, L. A., D’ETTORRE,P.,
and HUGHES, W. O. H. 2010. Kin-informative recognition cues in
ants. Proc. R. Soc. B 278:1942–1948.
NIELSEN,J.,BOOMSMA,J.J.,OLDHAM,N.J.,PETERSEN,H.C.,
and MORGAN, E. D. 1999. Colony-level and season-specific
variation in cuticular hydrocarbon profiles of individual workers
in the ant Formica truncorum.Insectes Soc. 46:58–65.
RATNI EK S ,F.L.W.,FOSTER, K. R., and WENSELEERS, T. 2006. Conflict
resolution in insect societies. Ann. Rev. Entomol. 51:581–608.
RATN I EK S ,F.L.W.,HELANTERÄ, H., and FOSTER, K. R. 2007. Are
mistakes inevitable? Sex allocation specialization by workers
can reduce the genetic information needed to assess queen
mating frequency. J. Theor. Biol. 244:470–477.
SOROKER,V.,VIENNE,C.,HEFETZ, A., and NOWBAHARI, E. 1994. The
postpharyngeal gland as a “gestalt”organ for nestmate recogni-
tion in the ant Cataglyphis niger.Naturwissenschaften 81:510–
513.
STAR KS ,P.T.,WATS O N,R.E.,DIPAOLA,M.J.,andDIPAOLA,C.P.
1998. The effect of queen number on nestmate discrimination
in the Facultatively Polygynous Ant Pseudomyrmex pallidus
(Hymenoptera: Formicidae). Ethology 104:573–584.
STUART, R. J. 1988. Collective cues as a basis for nestmate
recognition in polygynous leptothoracine ants. Proc. Natl. Acad.
Sci. USA 85:4572–4575.
SUNDSTRÖM, L. 1994. Sex ratio bias, relatedness asymmetry and
queen mating frequency in ants. Nature 367:266–268.
SUNDSTRÖM, L. and BOOMSMA, J. J. 2001. Conflicts and alliances in
insect families. Heredity 86:515–521.
SUNDSTRÖM, L., CHAPUISAT, M., and KELLER, L. 1996. Conditional
manipulation of sex ratios by ant workers: A test of kin selection
theory. Science 274:993–995.
SUNDSTRÖM, L., KELLER, L., and CHAPUISAT, M. 2003. Inbreeding and
sex-biased gene flow in the ant Formica exsecta.Evolution
57:1552–1561.
TRIVERS, R. L. and HARE, H. 1976. Haplodiploidy and the evolution
of the social insects. Science 191:249–263.
TRONTTI, K., TAY, W. T., and SUNDSTRÖM, L. 2003. Polymorphic
microsatellite markers for the ant Plagiolepis pygmaea.Mol.
Ecol. Notes 3:575–577.
VAN ZWEDEN, J. S., BRASK, J. B., CHRISTENSEN, J. H., BOOMSMA,J.
J., LINKSVAYER, T. A., and D’ETTORRE, P. 2010. Blending of
heritable recognition cues among ant nestmates creates distinct
colony gestalt odours but prevents within-colony nepotism. J.
Evol. Biol. 23:1498–1508.
VAN ZWEDEN, J. S. and D’ETTORRE, P. 2010. Nestmate recognition in
social insects and the role of hydrocarbons, pp. 222–243, in G. J.
Blomquist and A.-G. Bagnères (eds.), Insect hydrocarbons:
biology, biochemistry, and chemical ecology. Cambridge Uni-
versity Press, Cambridge.
VA N ZWEDEN, J. S., DREIER, S., and D’ETTORRE, P. 2009. Disentangling
environmental and heritable nestmate recognition cues in a
carpenter ant. J. Insect Physiol. 55:158–163.
VITIKAINEN, E., HAAG-LIAUTARD, C., and SUNDSTRÖM, L. 2011.
Inbreeding and reproductive investment in the ant Formica
exsecta.Evolution 65:2026–2037.
WAGNER,D.,TISSOT,M.,andGORDON,D.2001.Task-related
environment alters the cuticular hydrocarbon composition of
harvester ants. J. Chem. Ecol. 27:1805–1819.
WANG, J. 2004. Sibship reconstruction from genetic data with typing
errors. Genetics 166:1963–1979.
WILSON, E. O. 1971. The insect societies. Harvard University Press,
Cambridge.
J Chem Ecol (2011) 37:1365–1373 1373