Behavioral Ecology Vol. 10 No. 4: 409-421
© 1999 International Society for Behavioral Ecology
Sex allocation in a facultatively polygynous ant: between-population and between-colony variation
a Institute of Zoology, Zoological Society of London, Regent's Park, London NW1 4RY, UK b Department of Biology, University College London, Wolfson House, 4 Stephenson Way, London NW1 2HE, UK
Address correspondence to A.F.G. Bourke. E-mail: andrew.bourke{at}ioz.ac.uk
Received 7 May 1998; revised 27 November 1998; accepted 22 December 1998.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
We investigated sex allocation in three U.K. populations of the facultatively polygynous ant Leptothorax acervorum over 1-3 years. The first main finding was that, across sites, the population sex-investment ratio changed from significantly female biased to significantly male biased with increasing polygyny. This was consistent with workers controlling sex allocation and reacting to changes in their population-level relatedness asymmetry. It was also consistent with local resource competition due to reproduction by colony budding under polygyny. Worker control was supported by the finding that queen number had no effect on sex allocation among polygynous colonies. The second main result was that monogynous colonies consistently produced more female-biased sex-investment ratios than polygynous colonies in one site only (Santon). The results from Santon supported both the relative relatedness asymmetry hypothesis and the idea of sex ratio compensation due to colony budding. The workers' response to their population-level relatedness asymmetry reinforced the case for relatedness asymmetry being influential at the colony level. The other populations could have lacked split sex ratios because polygynous colonies were either comparatively rare or common, making them behave as almost entirely monogynous (Aberfoyle) or polygynous (Roydon) populations. In Roydon, this was consistent with the inference from allozyme data that monogynous and polygynous colonies did not differ in their worker relatedness asymmetries. The final principal finding was that, of hypotheses linking the colony sex-investment ratio with sexual productivity, there was support for the constant female hypothesis but not for the constant male, cost variation, or multifaceted parental investment hypotheses.
Key words: allozyme, ants, Formicidae, Hymenoptera, Leptothorax, polygyny, relatedness, sex ratio, social insect, split sex ratio.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
The study of sex ratio evolution in social Hymenoptera is important because of its contribution to our understanding of the general causes of sex ratio variation in animals (Bourke and Franks, 1995
; Crozier and Pamilo,
1996). In ants, sex ratios vary at both the population and colony
levels (Boomsma, 1989;
Nonacs, 1986a;
Pamilo, 1990;
Trivers and Hare, 1976).
Extending Fisher's (1930) sex
ratio theory, Trivers and Hare
(1976) proposed that the
population-level sex-investment ratio equals the relatedness asymmetry (the
ratio of relatedness to the sexes) of whichever party controls sex allocation.
In species with one queen per colony (monogyny), queens are equally related to
the sexes, but workers are three times more closely related to sisters than to
brothers. Trivers and Hare
(1976) therefore predicted
population sex-investment ratios of 1:1 if queens control sex allocation and
3:1 females:males if there is worker control. Average sex-investment ratios in
monogynous, monandrous ants are significantly female biased, supporting the
Trivers and Hare prediction with worker control
(Bourke, 1997;
Bourke and Franks, 1995;
Crozier and Pamilo, 1996).
Many ant species have colonies that contain multiple queens through the
readoption of daughter queens (secondary polygyny). In these colonies, theory
predicts comparatively unbiased or even male-biased population sex-investment
ratios. One reason is that adding extra reproductive queens reduces the
workers' relatedness asymmetry (Trivers
and Hare, 1976). Another is that the fitness of dispersing queens
is discounted in proportion to their now-fractional contribution to the
overall fitness return via female production
(Pamilo, 1990). Last, while
monogynous ant colonies usually reproduce by emitting dispersing sexuals (new
queens and males), polygynous ant colonies often exhibit colony budding, in
which queens leave colonies with workers and brood to establish new colonies
near the parental one (Keller,
1991). This results in local resource competition (relatives
compete for resources; Clark,
1978) among daughter queens, which predicts male-biased population
sex-investment ratios (Pamilo,
1991). In support of these arguments, polygynous ant species
exhibit average male bias in their population sex-investment ratios
(Bourke and Franks, 1995).
However, in polygynous species in which the degree of polygyny varies across
populations, some previous studies have found population sex-investment ratios
to covary with queen number across populations as expected
(Walin, 1997), but others have
not (Elmes, 1987;
Herbers, 1990). In addition,
although population sex-investment ratios in some ant species have been
reported to remain fairly stable across successive years
(Banschbach and Herbers, 1996;
Brian, 1979;
Herbers and Stuart, 1996b), in
others they varied appreciably (Elmes,
1987; Evans,
1996b; Herbers,
1979, 1990;
Herbers and Grieco, 1994;
Pamilo and Rosengren, 1983;
Pearson et al., 1997).
Therefore, in polygynous ants, a need remains for data on population
sex-investment ratios as a function of queen number in more than one
population and year.
Within ant populations, colony sex-investment ratios are typically highly
variable (Nonacs, 1986a). If
different classes of colonies concentrate on producing one sex only, sex
ratios are said to be split (Grafen,
1986). Several authors have recently proposed extensions of sex
ratio theory to explain split sex ratios and, more generally, any systematic
within-population variation in sex allocation
(Table 1). According to these
authors, two classes of factors are involved. The first consists of factors
stemming from variation in the genetic or social structure of colonies, and
the second involves ecological factors that result in variation in colony
resource levels and hence colony productivity. However, the effects of these
classes of factor are not mutually exclusive, and nor are the predictions of
the different hypotheses (Table
1). Evidence for and against some of the hypotheses remains patchy
(Table 1). Moreover, as with
population sex-investment ratios, patterns of sex ratio splitting conceivably
show yearly variation, although reports to date show that such patterns remain
similar across years (Herbers,
1990;
Sundström,
1994;
Sundström et
al., 1996). Therefore, to test the relevant hypotheses for
colony-level variation in the sex-investment ratio
(Table 1), empirical
investigations over several years are required (cf.
Herbers, 1984,
1990).
|
In an earlier study, we showed that in the facultatively polygynous ant
Leptothorax acervorum (Hymenoptera: Myrmicinae), the population
sex-investment ratio was unbiased and split. Monogynous and polygynous
colonies concentrated on female and male production, respectively
(Chan and Bourke, 1994). Of
hypotheses accounting for colony-level variation in the sex-investment ratio,
our findings were consistent with the relative relatedness asymmetry
hypothesis, the hypothesis of sex ratio compensation due to colony budding,
and the constant female hypothesis. They were inconsistent with the
queen-worker conflict hypothesis (Table
1). However, this study was based on 1 year of sampling from a
single population. In the present study, we extended our investigation of sex
allocation in L. acervorum over additional years and populations to
test the hypotheses outlined above more robustly. Nonacs
(1986a,
b) proposed that, at the
proximate level, a positive association of the sex-investment ratio with
sexual productivity (Table 1)
could be linked to a fall in the ratio of new workers to new queens (with
colonies investing in workers under resource-poor conditions and investing in
queens under resource-rich conditions). In the present study we measured new
worker production alongside sexual production to investigate this hypothesis.
Our work adds to that of previous authors who have examined sex allocation in
relation to queen number and colony genetic structure in Leptothorax
ants, including L. acervorum
(Heinze et al., 1995b), L.
ambiguus (Herbers and Grieco,
1994), L. curvispinosus
(Herbers and Stuart, 1996b),
L. longispinosus (Backus,
1995; Herbers,
1984, 1990;
Herbers and Stuart, 1996a,
b), and L. tuberum
(Pearson et al., 1995,
1997).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
Study species
Leptothorax acervorum is a facultatively polygynous ant found in North America, Europe, and Asia (Bourke, 1991
; Buschinger,
1968; Heinze and Ortius,
1991; Ito, 1990).
In pine woods, colonies nest inside dead twigs on the ground. Single colonies
occupy single twigs (monodomy). Genetic data from several populations,
including one of the current study populations (Santon), show that coexisting
queens within colonies are related on average and that inbreeding is absent
(Bourke et al., 1997;
Douwes et al., 1987;
Heinze et al., 1995a,
b;
Stille and Stille, 1992;
Stille et al., 1991).
Additional genetic data suggest that queens mate singly and that workers
produce few or no adult males (Heinze et
al., 1995b, 1997;
Seppä et
al., 1995). In two populations, including the Santon one,
relatedness among workers was significantly higher in monogynous colonies than
in polygynous colonies (Heinze et al.,
1995a, b),
suggesting that workers under monogyny had higher relatedness asymmetries than
those under polygyny. Nonetheless, relatedness among workers fell below 0.75
in the monogynous colonies. This was probably because some of these colonies
were former polygynous ones that had lost queens. The occurrence of such a
process was supported by the discovery of high turnover among reproductive
queens in polygynous colonies (Bourke et
al., 1997). These findings suggest a genuine but flexible link
between queen number and workers' relatedness asymmetry within colonies from
at least some populations.
The lack of evidence for inbreeding implies that L. acervorum
males disperse widely enough before mating to make local mate competition
among them either absent or of low intensity. However, polygynous L.
acervorum colonies are strongly suspected to reproduce by colony budding
(Bourke and Heinze, 1994;
Stille and Stille, 1993). This
suggests the occurrence of local resource competition between daughter queens
and colonies.
Collection and census methods
We collected and censused 284 colonies from 3 pine wood populations of
L. acervorum in the United Kingdom, sampling the populations over 3
years, 2 years, and 1 year, respectively
(Table 2). The populations are
separated by 250 km (Santon and Roydon). 530 km (Santon and Aberfoyle), and
620 km (Aberfoyle and Roydon) (these names will be used throughout for the
three populations in Table 2).
The collection and census methods were as described by Chan and Bourke
(1994), who also presented
data from the first of the collections (Santon in 1993). Censusing involved
two counts, the first immediately after collection
(Table 2) and the second
several months later (mean 3.3 months, range 1-5.5 months) to ensure that all
brood destined to mature in the year of collection had become adult.
Comparison of the two censuses gave each colony's total annual production of
new, alate queens (young, winged queens, which we refer to as
"females"), new males ("males"), and new workers. It
also gave each colony's original number of dealate queens and of adult workers
("old workers"). Dealate queens have shed their wings and are
potential colony queens. During the second census, we removed all dealate
queens and confirmed their reproductive status by ovarian dissections
(Bourke, 1991). Colonies were
classified as monogynous or polygynous on the basis of whether they contained
one or more mated dealate queens (Chan and
Bourke, 1994).
|
Dry weight measurements
We measured the thorax length and dry weight of individual ants sampled
from the Santon colonies (Chan and Bourke,
1994). We used the regression equations between log dry weight and
log thorax length to estimate dry weights from thoracic length measurements
for additional, unweighed individuals. Thorax length was measured in 261 new
workers (10% of which were weighed), 196 alate queens (31% of which were
weighed), and 206 males (34% of which were weighed) collected from 5-6
monogynous and 3-6 polygynous colonies per year, respectively (except that no
sexuals were measured from polygynous colonies in 1995). We assumed that the
relative sizes of the sexes and castes in Aberfoyle and Roydon resembled those
found in Santon. This assumption almost certainly involved some approximation
because in other leptothoracines females tend to increase in weight relative
to males as the frequency of monogyny rises in the population
(Herbers and Stuart, 1996b).
However, it is worth noting that, if such a trend is present in L.
acervorum, calculating population sex-investment ratios using
site-specific female and male dry weights would only have exaggerated
differences reported in the Results section (with female bias being even
greater in the relatively monogynous Aberfoyle population and male bias being
even greater in the relatively polygynous Roydon population).
Statistical analyses of sex-investment ratio
We express all sex-investment ratios as the amount of biomass invested in
females as a fraction of the total biomass invested in sexuals. We calculated
standard errors for population and class-specific sex-investment ratios using
Boomsma's method described in Bourke and Franks
(1995: 160), with colony
sex-investment ratios being weighted by the colony's sexual biomass divided by
the average per-colony sexual biomass for the sample. In the regression and
correlation analyses of sex allocation, the variables used were colony
sex-investment ratio (angular transformed), dealate queen and old worker
number (both log10-transformed), female, male, and total sexual
production (all measured as mg of biomass and log10-transformed),
and angular transformed new worker to new queen ratio, calculated as the
biomass of new workers divided by the biomass of total diploid production. We
transformed data sets to ensure (or improve) their fit to a normal
distribution (Kolmogorov-Smirnov tests; Minitab Reference Manual: Release 11
for Windows, Minitab Inc., State College, Pennsylvania, 1996). We excluded
from the regression and correlation analyses colonies producing five or fewer
sexuals and five or fewer new diploids because these could have produced
extreme ratios by chance. We also excluded outlying colonies with
exceptionally high queen numbers (Santon colony SD 93 32, Aberfoyle colonies
AB 94 15 and AB 95 01; see Appendix).
Genetic methods
We genetically examined 20 colonies (5 monogynous and 15 polygynous)
arbitrarily selected from the Roydon 1995 sample. Ten to 12 old workers from
all colonies and 2-10 dealate queens from the polygynous colonies were typed
at 4 allozyme loci. We analyzed 235 workers and 70 queens. The loci examined
were aconitase hydratase-2 (ACOH-2), glucose-6-phosphate
dehydrogenase (G6PDH), glucose-6-phosphate isomerase (GPI),
and phosphoglucomutase-1 (PGM-1). These loci were found to be
polymorphic with, respectively four, two, four, and four alleles. Individual
ants were ground in 8 µl of buffer (0.01% NADP, 0.1%
ß-mercaptoethanol, 0.1% Triton X-100 detergent). After centrifuging, 7
µl of supernatant was applied to presoaked cellulose acetate plates. The
buffers used were a Tris-Gly buffer (0.3% Trizma, 1.4% glycine, pH 8.6) and,
for PGM-1 only, a phosphate buffer (0.36%
NaH2PO4, 0.22% Na2HPO4, pH 6.3).
Plates were run for 20 min at 200 V/2 mA at room temperature and were stained
following protocols in Richardson et al.
(1986). We tested inbreeding
coefficients calculated from the genetic data for significant departures from
zero using the permutation procedure (with 10,000 permutations) in the program
FSTAT Version 2.3 (Goudet,
1995). We calculated relatedness values using RELATEDNESS version
4.2c (K. F. Goodnight and D. C. Queller, Rice University), which estimates
regression relatednesses based on formulae in Queller and Goodnight
(1989).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
Dry weights
The average dry weights of new workers, alate queens, males from monogynous colonies, and males from polygynous colonies were estimated to be 0.353 mg, 0.506 mg, 0.510 mg, and 0.444 mg, respectively. Only males differed significantly in body size across monogynous and polygynous colonies (oneway ANOVA: F1,67 = 7.99, p <.01). We calculated the cost ratio for converting numerical sex ratios to investment sex ratios as the untransformed female:male ratio of dry weights (0.506/0.510 = 0.99 in monogynous colonies and 0.506/0.444 = 1.14 in polygynous colonies). We did not apply Boomsma's 0.7 power conversion (Boomsma, 1989
;
Boomsma et al., 1995) because
this would have made a negligible difference to the cost ratio based on dry
weights alone.
Degree of polygyny
In the Santon population, the frequency of queenless, monogynous, and
polygynous colonies changed significantly from the first to the second and
third years of sampling (Table
3). At the Aberfoyle site, however, there was no significant
change in the frequency of the queen-number classes across years
(Table 3). The Roydon site was
significantly more polygynous than the Aberfoyle and Santon sites (in the
Santon case, for two of the three year-samples), and the Santon site was
significantly more polygynous than the Aberfoyle site for one of the three
year-samples (Table 3).
Therefore, the overall ranking of the sites with respect to the degree of
polygyny was (in order of decreasing polygyny): Roydon (69% colonies
polygynous), Santon (21-53% colonies polygynous), and Aberfoyle (23-26%
colonies polygynous; Table
3).
|
Size and productivity of monogynous and polygynous colonies
We investigated differences in colony size and productivity between years
and queen-number classes using general linear models designed as two-way
ANOVAs (implemented in MINITAB release 11). In the Santon site, the results
showed that, compared to polygynous colonies, monogynous colonies had
significantly smaller colonies, were significantly more sexually productive,
and allocated a significantly greater fraction of resources to sexual
production (Table 4). In the
Aberfoyle site, these differences were repeated only in the case of colony
size. In addition, monogynous Aberfoyle ants showed a non-significant trend
toward lower sexual productivity than polygynous ones and exhibited
significantly lower total productivity and productivity per worker
(Table 4). In Roydon,
monogynous and polygynous colonies showed no significant differences in any
parameter (Table 4).
|
Population sex-investment ratio
At the Santon site, the population sex-investment ratios across the 3 years
of sampling did not differ significantly (one-way ANOVA:
F2,127 = 2.17, p >.05). However, there was a
trend for female bias to increase over successive years
(Table 5), corresponding with a
decrease in the frequency of polygynous colonies in the population
(Table 3). At Aberfoyle, there
was also no significant change in population sex-investment ratio
(Table 5) over the 2 years of
sampling (t = 0.26, df = 45, p >.05). Across the three
sites, the overall population sex-investment ratios were, in order of
increasing female bias, 0.13 (Roydon), 0.60 (Santon), and 0.71 (Aberfoyle)
(Table 5). These means were
significantly different (F2,247 = 5.92, p
<.01). Overall, the population sex-investment ratio was significantly male
biased at Roydon and significantly female biased at Santon and Aberfoyle
(Table 5).
|
Colony-level variation in sex-investment ratios as a function of
queen number
Comparison of sex-investment ratio in monogynous and polygynous
colonies
We examined sex-investment ratio as a function of year and queen-number
class within sites using general linear models designed as two-way ANOVAs. In
both the Santon and Aberfoyle sites, neither year alone, nor the interaction
of year and queen-number class, had significant effects. Queen-number class
was a significant factor affecting sex allocation in Santon
(F1,136 = 17.1, p <.001), but not in Aberfoyle
(F1,29 = 0.02, p >.9) or Roydon
(F1,23 = 0.39, p >.5). Specifically, at the
Santon site, monogynous colonies produced female-biased broods, and polygynous
colonies produced male-biased broods consistently over the 3 years of
sampling, whereas at the other two sites monogynous and polygynous colonies
produced similar sex-investment ratios
(Figure 1).
|
In the Roydon population, we divided colonies on the basis of their inferred relatedness asymmetry independently of the observed queen number. Twelve colonies contained workers that could have come from monogynous, monandrous groups (two or fewer genotypes per locus; see "Genetic results") and were inferred to have high relatedness asymmetries. Six colonies contained workers that could not have come from such groups (greater than two genotypes per locus; see "Genetic results") and were inferred to have low relatedness asymmetries. These two groups did not differ significantly in their sex-investment ratios (t = 0.11, df = 15, p >.9).
Effect of queen number on colony sex-investment ratio in polygynous
colonies
We investigated the effects of queen number on colony-level sex allocation
within the polygynous colony class by regression analysis, controlling for the
effects of colony size and sexual productivity
(Nonacs, 1986a,
b). In five data sets (the six
year-samples with the Aberfoyle 1994 and Aberfoyle 1995 samples pooled due to
small sample sizes), there was no significant regression of colony
sex-investment ratio on queen number, controlling separately for colony size
and total sexual production. There was also no such regression in the pooled
data sets (Figure 2).
|
Colony-level variation in sex-investment ratios as a function of
colony sexual productivity
We performed preliminary multiple regressions with colony sex-investment
ratio as the dependent variable and, as independent variables, total colony
sexual production and the two categorical variables queen-number class (colony
monogynous or polygynous) and year (if applicable). We dropped from the
initial model terms and interactions whose exclusion showed that they had no
significant effect. There were no significant interaction terms in any
population. In the Santon population, the final model involved only total
sexual production and queen-number class as significant factors
(F2,131 = 13.7, p <.001). We therefore pooled
samples across years and regressed colony sex-investment ratio on total sexual
production for monogynous and polygynous colonies separately. Monogynous
colonies showed no significant regression of sex-investment ratio on total
sexual production, whereas polygynous colonies showed a significant negative
regression (Figure 3A,
3B).
|
In the Aberfoyle population, neither queen-number class nor year proved to be significant factors in the preliminary multiple regressions. Consequently, we pooled colonies from both queen-number classes and years. There was no significant regression of sex-investment ratio on total sexual production, although the trend was for a positive association (Figure 3C). In the Roydon population, queen-number class was again not a significant factor. In the pooled data set, there was no significant regression of sex-investment ratio on total sexual production, although the trend was for a negative association (Figure 3D).
Relative resource allocation to new worker production and total
sexual production
To investigate the association predicted by Nonacs
(1986a,
b), we calculated the partial
correlation coefficient between the ratio of new workers to new queens and
total colony sexual production (with old worker number held constant). This
partial correlation was significantly negative in monogynous Santon colonies
(r = -.59, df = 95, p <.001), not significant in
polygynous Santon colonies (r = -.19, df = 33, p >.1),
significantly negative in all Aberfoyle colonies (r = -.52, df = 29,
p <.01), and not significant in all Roydon colonies (r
=.24, df = 20, p >.1).
Genetic results
In the Roydon population, the allele frequencies at the four allozyme loci
were as follows: ACOH-2: 0.92, 0.03, 0.04, 0.01; G6PDH:
0.60, 0.40; GPI: 0.70, 0.11, 0.10, 0.09; PGM-1: 0.63, 0.35,
0.01, 0.01. The population-wide inbreeding coefficient (F) measured
from the dealate queen data was -0.40. There was significant homozygote excess
at the locus ACOH-2 (F = 0.25, p =.01) and
significant heterozygote excess at the remaining three loci (G6PDH,
F = -0.79; GPI, F = -0.17; PGM-1,
F = -0.44; all p <.001), the reasons for which were
unknown. Similar results were found when the F values were calculated
from the worker data. Because relatedness calculations assume Hardy-Weinberg
equilibrium, we only examined relatedness values at all four loci pooled.
Relatedness among dealate queens over all loci was 0.17 ± 0.08 SE
(n = 70 queens from 15 polygynous colonies) and was not significantly
greater from zero (t = 2.13, df = 14, p >.05).
Relatedness among workers over all loci was 0.47 ± 0.15 SE in
monogynous colonies (n = 58 workers from 5 colonies), 0.44 ±
0.04 SE in polygynous colonies (n = 177 workers from 15 colonies),
and 0.45 ± 0.05 SE across all colonies (n = 235 workers from
20 colonies). Relatedness among workers did not differ significantly between
monogynous and polygynous colonies (t = 0.38, df = 18, p
>.1).
Because the lack of Hardy-Weinberg equilibrium at the marker loci may have affected the relatedness calculations, we examined the genotypes of workers in colonies classified as monogynous or polygynous on the basis of queen number. Assuming monandry, the possession by workers from any one colony of more than two genotypes per locus indicates a departure from strict monogyny. Inspection of the data showed that two of five "monogynous" colonies contained workers with more than 2 genotypes at 1 or more loci, whereas the figure for "polygynous" colonies was 5 of 15 colonies. This finding again suggested that the two colony classes had similar kin structures, as indicated by the relatedness calculations. We summarize key genetic and other features of all three study populations in Table 6.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
Population sex-investment ratio
In Leptothorax acervorum, population sex-investment ratios were variable across sites but, across years within sites, remained stable or changed only gradually. Overall, population sex-investment ratios covaried with the degree of polygyny (Table 3 and 5) as predicted by theory (Pamilo, 1990
, 1991;
Trivers and Hare, 1976). The
Roydon population was the most polygynous and had a significantly male-biased
population sex-investment ratio. The Santon population, with a variable but
intermediate level of polygyny, had a population sex-investment ratio that was
significantly female biased but significantly less than 0.75. Furthermore, in
this population, there was a trend for relative female bias to rise as the
level of polygyny fell across the 3 years of sampling. Finally, the Aberfoyle
population had a comparatively low level of polygyny, and its population
sex-investment ratio did not differ significantly from 0.75 (Tables
3 and
5).
In the Santon and Roydon populations, greater relative male bias in the
population sex-investment ratio could have been due to either the workers'
reduced relatedness asymmetry, or local resource competition stemming from
colony budding, or both. If queens control sex allocation, female bias in the
population sex-investment ratio is not predicted by any model for sex ratio
evolution under polygyny (Pamilo,
1990,
1991). Therefore, the
existence of significant overall female bias in Santon and Aberfoyle suggested
that sex allocation was worker controlled.
The reasons that the three sites differed in their overall level of polygyny are unknown. There were no obvious causative ecological differences among the sites, but our study did not address this issue. For example, we did not measure the density of nests. However, our findings show that, whatever the cause of variation in the degree of polygyny, such variation was associated with variation in population sex allocation in the expected direction.
Colony-level variation in sex-investment ratios as a function of
queen number
The Santon population showed a consistent pattern of sex ratio splitting
over 3 successive years, with monogynous colonies concentrating on female
production and polygynous colonies concentrating on male production
(Figure 1). Furthermore, in
this population workers in monogynous colonies were significantly more highly
related than those in polygynous colonies
(Table 6). Together, these
findings were consistent with the predictions and assumptions of the relative
relatedness asymmetry hypothesis (Table
1). They were also consistent with sex ratio compensation due to
colony budding (Table 1).
However, the female-biased overall sex-investment ratio in Santon indicated
that workers responded to their relatedness asymmetry at the population level.
This strengthened the case that sex ratio splitting in Santon stemmed, at
least in part, from workers responding to their colony's relatedness
asymmetry. Within the polygynous colony class, sex-investment ratios did not
vary with queen number in the Santon population (or in any of the other
populations: Figure 2).
Therefore, if workers do indeed respond to their colony relatedness asymmetry,
it appears that they can assess the type of colony they belong to (monogynous
or polygynous), but not the precise degree of polygyny. The lack of
association of colony sex-investment ratio with queen number at all sites also
showed that there was no support for the queen-worker conflict hypothesis
(Table 1).
In both Aberfoyle and Roydon, monogynous and polygynous colonies coexisted
within populations but did not differ significantly in their class-specific
sex-investment ratios. The absence of sex ratio splitting was consistent
across the 2 years of sampling in Aberfoyle
(Figure 1). We cannot offer a
certain explanation for the absence of split sex ratios at Aberfoyle and
Roydon, but suggest the explanations discussed below based on the current
findings (Table 6). General
reasons populations may unexpectedly lack sex ratio splitting are discussed by
Ratnieks and Boomsma (1997).
If none of these suggestions apply, the findings from Aberfoyle and Roydon
must count as contradictions of both the relative relatedness asymmetry and
sex ratio compensation due to colony budding hypotheses.
In the Aberfoyle population, although the polygynous colonies had high queen numbers and were productive (Table 4; Appendix), their frequency of occurrence was relatively low and remained low across years (Tables 3 and 6). We suggest, therefore, that polygyny was insufficiently frequent in this population for workers to have evolved a sex ratio response to its presence. This would have led to the population behaving as if it were completely monogynous. This was consistent with sex ratios not being split by queen number (Figure 1) and with the population sex-investment ratio not being significantly different from 0.75 (Table 5). Conceivably, in facultatively polygynous populations, workers can only successfully assess their colony's specific relatedness asymmetry if the frequency of the minority queen-number class exceeds a critical threshold, with the threshold being set by nonlinearities in the workers' ability to detect systematic variation in brood odor above the prevailing background level. Another possibility is that polygynous colonies in the Aberfoyle population retained comparatively high worker relatedness asymmetries through being functionally monogynous (having several mated queens, only one of which lays eggs at any one time). However, dissection data (not shown) revealing active ovaries (with corpora lutea and yolky eggs) among all mated queens within the polygynous Aberfoyle colonies contradict this possibility.
In the Roydon population, we make a similar suggestion as for Aberfoyle but
with the relative frequencies of the colony classes reversed. We propose that
monogynous colonies were present at such low frequency that the Roydon
population behaved as an almost completely polygynous one. Specifically, sex
ratios were not split by queen number
(Figure 1) or by the workers'
inferred relatedness asymmetry, and the population sex-investment ratio was
significantly male biased (Table
5). The relatedness results indicated that there was no
association between a colony's queen-number class and its workers' relatedness
asymmetry. This would also have contributed to a lack of sex ratio splitting
by queen number in the Roydon population. A decoupling of queen-number class
and worker relatedness asymmetry was reported by Evans
(1995) in the polygynous ant
Myrmica tahoensis. Similarly, in a German L. acervorum
population, relatedness among new queens in polygynous colonies was not
significantly different from the relatedness expected under monogyny
(Heinze et al., 1995b). A
possible proximate cause of these phenomena is a high level of queen turnover
(Bourke et al., 1997;
Evans, 1995,
1996a;
Heinze et al., 1995b). In the
U.K. L. acervorum, decoupling of the queen-number class and worker
relatedness asymmetry might have occurred in Roydon but not in Santon because
of the higher level of polygyny in Roydon. This could have meant that, in
Roydon, many more colonies that were monogynous at the time of sampling had
experienced a developmental history involving polygyny.
Production schedules and colony-level variation in sex-investment
ratios as a function of colony sexual productivity
The Santon population demonstrated fundamental differences in the
production schedules of monogynous and polygynous colonies within a
facultatively polygynous population (Table
6). Monogynous colonies invested significantly more resources in
sexuals overall (Table 4),
their sex-investment ratios did not change systematically as sexual
productivity rose (Figure 3A),
and their ratios of new workers to new queens decreased with increasing sexual
productivity. Polygynous colonies invested fewer resources in sexuals
(Table 4), their sex-investment
ratios became significantly more male biased as sexual productivity rose
(Figure 3B), and their ratios
of new workers to new queens did not change with rising sexual productivity.
The results from monogynous colonies therefore provided no support for the
constant male, cost variation, or multifaceted parental investment hypotheses,
whereas those from the polygynous colonies supported the constant female
hypothesis (Table 1). In
addition, the results from the monogynous colonies showed that, contrary to
results of Nonacs (1986a,
b), a fall in the new worker
to new queen ratio with rising sexual productivity need not accompany an
increase in the female bias of the colony sex-investment ratio with rising
sexual productivity, because such an increase was absent
(Figure 3A). Instead, in the
monogynous Santon colonies, the decrease in the new worker to new queen ratio
appeared to be associated with an increase in the proportionate allocation of
resources to all sexuals as overall colony productivity (of new sexuals plus
new workers) increased (r =.43, df = 97, p <.001).
In Aberfoyle, monogynous and polygynous colonies did not differ significantly in the proportion of resources allocated to sexuals (Table 4). Colonies as a whole did not alter their sex-investment ratios as sexual productivity rose, although the trend was for increasing female bias (Figure 3C). This trend was consistent with, but did not provide firm support for, the constant male, cost variation, and multifaceted parental investment hypotheses (Table 1). Colonies also showed a significant negative association of the new worker to new queen ratio with total sexual production. Given their similarity with the results from the monogynous Santon colonies (Table 6), these findings were consistent with the interpretation of the Aberfoyle population as behaving like a largely monogynous population.
Finally, in Roydon, monogynous and polygynous colonies again did not differ significantly in their production schedules (Table 4). Colonies as a whole did not alter their sex-investment ratios as sexual productivity rose, although the trend was for male bias to increase with increasing productivity (Figure 3D). This trend was consistent with, but not conclusively supportive of, the constant female hypothesis (Table 1). Colonies showed no association of the new worker to new queen ratio with total sexual production. Given their similarity with the results from the polygynous Santon colonies, these findings were consistent with the interpretation of the Roydon population as behaving like an almost entirely polygynous one.
We conclude that, overall, between-population and between-colony variation
in sex-investment ratios in facultatively polygynous L. acervorum
populations is complex (cf. Herbers,
1984, 1990). Some
of the variation is explicable with current hypotheses, but a complete
explanation of all the observed patterns requires further investigation.
A1,
A2,
A3,
A4,
A5
|
|
|
|
|
|
APPENDIX |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
Composition and annual production of Leptothorax acervorum colonies. DQ, number of dealate, colony queens; OW, number of original, old workers; NW, number of new workers produced; F, number of new queens produced; M, number of males produced; dash indicates missing data (in the case of DQ data, due to unsuccessful ovarian dissections). Data for the Santon, 1993 collection were published in Chan and Bourke (1994
). |
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
We thank the members or former members of the Ecology Group of the Institute of Zoology for support throughout this work, especially Steve Albon, Peter Cotgreave, Tim Coulson, and Ian Owens. We also thank the following for help and comments: Jürgen Heinze, Rob Hammond, the late Gordon Mackenzie, Jim Mallet, Andrew Pomiankowski, Lotta Sundström, and the anonymous referees. The Forestry Commission kindly provided collecting permission. This work was supported by a Natural Environment Research Council studentship held by G.L.C.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXACKNOWLEDGEMENTSREFERENCES |
|---|
Backus VL, 1995. Rules for allocation in a temperate forest ant: demography, natural selection, and queen-worker conflict.Am Nat 145:775-796.[Web of Science]
Banschbach VS, Herbers JM, 1996. Complex colony structure in social insects: II. Reproduction, queen-worker conflict, and levels of selection. Evolution 50:298-307.[Web of Science]
Boomsma JJ, 1989. Sex-investment ratios in ants: has female bias been systematically overestimated? Am Nat 133:517-532.[Web of Science]
Boomsma JJ, 1993. Sex ratio variation in polygynous ants. In: Queen number and sociality in insects (Keller L, ed). Oxford: Oxford University Press; 86-109.
Boomsma JJ, Grafen A, 1990. Intraspecific variation in ant sex ratios and the Trivers-Hare hypothesis. Evolution 44:1026-1034.[Web of Science]
Boomsma JJ, Grafen A, 1991. Colony-level sex ratio selection in the eusocial Hymenoptera. J Evol Biol 4:383-407.
Boomsma JJ, Keller L, Nielsen MG, 1995. A comparative analysis of sex ratio investment parameters in ants. Funct Ecol 9:743-753.
Bourke AFG, 1991. Queen behaviour, reproduction and egg cannibalism in multiple-queen colonies of the ant Leptothorax acervorum. Anim Behav 42:295-310.
Bourke AFG, 1997. Sociality and kin selection in insects. In: Behavioural ecology: an evolutionary approach, 4th ed (Krebs JR, Davies NB, eds). Oxford: Blackwell;203 -227.
Bourke AFG, Franks NR, 1995. Social evolution in ants. Princeton, New Jersey: Princeton University Press.
Bourke AFG, Green HAA, Bruford MW, 1997. Parentage, reproductive skew and queen turnover in a multiple-queen ant analysed with microsatellites. Proc R Soc Lond B 264:277-283.[Medline]
Bourke AFG, Heinze J, 1994. The ecology of communal breeding: the case of multiple-queen leptothoracine ants. Phil Trans R Soc Lond B 345:359-372.
Brian MV, 1979. Habitat differences in sexual production by two coexistent ants. J Anim Ecol 48:943-953.
Buschinger A, 1968. Mono- und Polygynie bei Arten der Gattung Leptothorax Mayr (Hymenoptera Formicidae). Insectes Soc 15:217-225.
Chan GL, Bourke AFG, 1994. Split sex ratios in a
multiple-queen ant population. Proc R Soc Lond B
258:261-266.
Clark AB, 1978. Sex ratio and local resource
competition in a prosimian primate. Science
201:163-165.
Crozier RH, Pamilo P, 1993. Sex allocation in social insects: problems in prediction and estimation. In: Evolution and diversity of sex ratio in insects and mites (Wrensch DL, Ebbert MA, eds). New York: Chapman and Hall;369 -383.
Crozier RH, Pamilo P, 1996. Evolution of social insect colonies: sex allocation and kin selection. Oxford: Oxford University Press.
Douwes P, Sivusaari L, Niklasson M, Stille B, 1987. Relatedness among queens in polygynous nests of the ant Leptothorax acervorum. Genetica 75:23-29.[Web of Science]
Elmes GW, 1987. Temporal variation in colony populations of the ant Myrmica sulcinodis. II. Sexual production and sex ratios. J Anim Ecol 56:573-583.
Evans JD, 1995. Relatedness threshold for the
production of female sexuals in colonies of a polygynous ant, Myrmica
tahoensis, as revealed by microsatellite DNA analysis. Proc Natl
Acad Sci USA
92:6514-6517.
Evans JD, 1996a. Queen longevity, queen adoption, and posthumous indirect fitness in the facultatively polygynous ant Myrmica tahoensis. Behav Ecol Sociobiol 39:275-284.
Evans JD, 1996b. Temporal and spatial variation in reproduction in the facultatively polygynous ant Myrmica tahoensis (Hymenoptera: Formicidae). Insectes Soc 43:309-317.
Fisher RA, 1930. The genetical theory of natural selection. Oxford: Clarendon Press.
Frank SA, 1987. Variable sex ratio among colonies of ants. Behav Ecol Sociobiol 20:195-201.[Web of Science]
Goudet J, 1995. FSTAT (version 1.2): a computer
program to calculate F-statistics. J Hered
86:485-486.
Grafen A, 1986. Split sex ratios and the evolutionary origins of eusociality. J Theor Biol 122:95-121.
Hamilton WD, 1967. Extraordinary sex ratios.Science
156:477-488.
Hasegawa E, Yamaguchi T, 1995. Population structure, local mate competition, and sex-allocation pattern in the ant Messor aciculatus. Evolution 49:260-265.[Web of Science]
Heinze J, Lipski N, Hölldobler B, Bourke AFG, 1995a. Geographical variation in the social and genetic structure of the ant, Leptothorax acervorum.Zoology 98:127-135.
Heinze J, Lipski N, Schlehmeyer K,
Hölldobler B, 1995b. Colony
structure and reproduction in the ant, Leptothorax acervorum.Behav Ecol
6:359-367.
Heinze J, Ortius D, 1991. Social organization of Leptothorax acervorum from Alaska (Hymenoptera: Formicidae).Psyche 98:227-240.
Heinze J, Puchinger W, Hölldobler B,1997 . Worker reproduction and social hierarchies in Leptothorax ants. Anim Behav 54:849-864.[Web of Science][Medline]
Herbers JM, 1979. The evolution of sex-ratio strategies in Hymenopteran societies. Am Nat 114:818-834.[Web of Science]
Herbers JM, 1984. Queen-worker conflict and eusocial evolution in a polygynous ant species. Evolution 38:631-643.[Web of Science]
Herbers JM, 1990. Reproductive investment and allocation ratios for the ant Leptothorax longispinosus: sorting out the variation. Am Nat 136:178-208.[Web of Science]
Herbers JM, Grieco S, 1994. Population structure of Leptothorax ambiguus, a facultatively polygynous and polydomous ant species. J Evol Biol 7:581-598.[Web of Science]
Herbers JM, Stuart RJ, 1996a. Multiple queens in ant nests: impact on genetic structure and inclusive fitness. Am Nat 147:161-187.[Web of Science]
Herbers JM, Stuart RJ, 1996b. Patterns of reproduction in southern versus northern populations of Leptothorax ants (Hymenoptera: Formicidae). Ann Entomol Soc Am 89:354-360.
Ito F, 1990. Functional monogyny of Leptothorax acervorum in northern Japan. Psyche 97:203-211.
Keller L, 1991. Queen number, mode of colony founding, and queen reproductive success in ants (Hymenoptera Formicidae). Ethol Ecol Evol 3:307-316.
Keller L, Aron S, Passera L, 1996. Internest sex-ratio
variation and male brood survival in the ant Pheidole pallidula.Behav Ecol
7:292-298.
Nonacs P, 1986a. Ant reproductive strategies and sex allocation theory. Q Rev Biol 61:1-21.
Nonacs P, 1986b. Sex-ratio determination within colonies of ants. Evolution 40:199-204.[Web of Science]
Nonacs P, 1993. The effects of polygyny and colony life history on optimal sex investment. In: Queen number and sociality in insects (Keller L, ed). Oxford: Oxford University Press;110 -131.
Pamilo P, 1990. Sex allocation and queen-worker conflict in polygynous ants. Behav Ecol Sociobiol 27:31-36.
Pamilo P, 1991. Evolution of colony characteristics in social insects. I. Sex allocation. Am Nat 137:83-107.[Web of Science]
Pamilo P. Rosengren R, 1983. Sex ratio strategies in Formica ants. Oikos 40:24-35.[Web of Science]
Pamilo P, Seppä P, 1994. Reproductive competition and conflicts in colonies of the ant Formica sanguinea. Anim Behav 48:1201-1206.
Pearson B, Raybould AF, Clarke RT, 1995. Breeding behaviour, relatedness and sex-investment ratios in Leptothorax tuberum Fabricius. Entomol Exp Appl 75:165-174.
Pearson B, Raybould AF, Clarke RT, 1997. Temporal changes in the relationship between observed and expected sex-investment frequencies, social structure and intraspecific parasitism in Leptothorax tuberum (Formicidae). Biol J Linn Soc 61:515-536.
Queller DC, Goodnight KF, 1989. Estimating relatedness using genetic markers. Evolution 43:258-275.[Web of Science]
Ratnieks FLW, Boomsma JJ, 1997. On the robustness of split sex ratio predictions in social Hymenoptera. J Theor Biol 185:423-439.
Richardson BJ, Baverstock PR, Adams M, 1986.Allozyme electrophoresis: a handbook for animal systematics and population studies . London: Academic Press.
Rosenheim JA, Nonacs P, Mangel M, 1996. Sex ratios and multifaceted parental investment. Am Nat 148:501-535.[Web of Science]
Seppä P, Sundström L, Punttila P, 1995. Facultative polygyny and habitat succession in boreal ants. Biol J Linn Soc 56:533-551.
Stille M, Stille B, 1992. Intra- and inter-nest variation in mitochondrial DNA in the polygynous ant Leptothorax acervorum (Hymenoptera; Formicidae). Insectes Soc 39:335-340.
Stille M, Stille B, 1993. Intrapopulation nestclusters of maternal mtDNA lineages in the polygynous ant Leptothorax acervorum (Hymenoptera: Formicidae). Insect Mol Biol 1:117-121.[Medline]
Stille M, Stille B, Douwes P, 1991. Polygyny, relatedness and nest founding in the polygynous myrmicine ant Leptothorax acervorum (Hymenoptera; Formicidae). Behav Ecol Sociobiol 28:91-96.[Web of Science]
Sundström L, 1994. Sex ratio bias, relatedness asymmetry and queen mating frequency in ants.Nature 367:266-268.[Web of Science]
Sundström L, 1995. Sex allocation and colony maintenance in monogyne and polygyne colonies of Formica truncorum (Hymenoptera: Formicidae): the impact of kinship and mating structure. Am Nat 146:182-201.[Web of Science]
Sundström L, Chapuisat M, Keller L,1996
. Conditional manipulation of sex ratios by ant workers: a
test of kin selection theory. Science
274:993-995.
Trivers RL, Hare H, 1976. Haplodiploidy and the
evolution of the social insects. Science
191:249-263.
Vargo EL, 1996. Sex investment ratios in monogyne and polygyne populations of the fire ant Solenopsis invicta. J Evol Biol 9:783-802.
Walin L, 1997. Colony characteristics and worker-queen conflicts in Myrmica ants (PhD thesis). Helsinki: University of Helsinki.
Ward PS, 1983. Genetic relatedness and colony organization in a species complex of ponerine ants. II. Patterns of sex ratio investment. Behav Ecol Sociobiol 12:301-307.[Web of Science]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. C. Wiernasz and B. J. Cole Dioecy and the evolution of sex ratios in ants Proc R Soc B, June 7, 2009; 276(1664): 2125 - 2132. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Meunier, S. A. West, and M. Chapuisat Split sex ratios in the social Hymenoptera: a meta-analysis Behav. Ecol., March 1, 2008; 19(2): 382 - 390. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Bromham and R. Leys Sociality and the Rate of Molecular Evolution Mol. Biol. Evol., June 1, 2005; 22(6): 1393 - 1402. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Ruppell, J. Heinze, and B. Holldobler Intracolonial patterns of reproduction in the queen-size dimorphic ant Leptothorax rugatulus Behav. Ecol., March 1, 2002; 13(2): 239 - 247. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||