Behavioral Ecology Vol. 13 No. 5: 653-656
© 2002 International Society for Behavioral Ecology
Adaptive biases in offspring sex ratios established before birth in a marsupial, the common brushtail possum Trichosurus vulpecula
School of Tropical Biology, James Cook University, Townsville, Queensland 4811, Australia
Address correspondence to C.N. Johnson. E-mail: christopher.johnson{at}jcu.edu.au.
Received 2 July 2001; revised 14 December 2001; accepted 2 January 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Offspring sex ratios in the common brushtail possum are male biased in many populations, and there is evidence that inter-population differences in sex ratios represent adaptive responses to local conditions. However, how these biases are produced is not known. Using comparisons between populations with and without biased offspring sex ratios, we show that biases in this species are not produced by sex-differential mortality between birth and weaning or sex-selective termination of pregnancy. Rather, adjustment in the sex ratio of offspring are evidently due to shifts in the probability of conceiving male and female offspring.
Key words: brushtail possums, local resource competition, Marsupialia, primary sex ratio, sex ratio mechanisms, Trichosurus vulpecula.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Anumber of theories identify circumstances in which adaptive variation in the sex ratio of offspring should occur, and there is often an excellent match between predicted and observed variation in sex ratios among invertebrates (Godfray and Werren, 1996
).
For vertebrates, however, the evidence on adaptive variation in the sex ratio
is less clear, and it has been argued that vertebrates may have limited
capacity for adaptive control of primary sex ratios
(Packer et al., 2000;
Palmer, 2000;
Williams, 1979). Sex ratios of
offspring have been reported for many species of marsupials, and significant
departures from parity are common
(Cockburn, 1990). Not only are
biased sex ratios found at the level of species, but in some species there is
evidence of significant variation among populations
(Cockburn et al., 1985;
Dickman, 1988;
Johnson et al., 2001;
Johnson and Jarman, 1983). In
Antechinus and the common brushtail possum Trichosurus
vulpecula, such interpopulation variation is consistent with the
hypothesis of local resource competition
(Clark, 1978). Male-biased sex
ratios occur where resource competition between mothers and daughters is
likely to be high (Cockburn et al.,
1985; Johnson et al.,
2001); in T. vulpecula the limiting resource has been
identified as den sites in tree hollows.
These observations suggest that females in these species have the capacity
to adjust the sex ratio of their offspring in response to local ecological
conditions. Biases in the sex ratio observed for dependent young in marsupials
could originate in a number of ways: (1) sex-biased mortality between birth
and weaning; (2) a longer duration of dependence in one sex; (3)
discrimination by mothers against one sex of offspring at birth; (4)
sex-selective loss between conception and birth; or (5) a biased probability
of conception of males and females. Here we discriminate among these
mechanisms using comparisons of populations of T. vulpecula with and
without biased offspring sex ratios. Before doing this, we review data on sex
ratios of offspring in T. vulpecula to determine the prevalence of
biased sex ratios in populations of this species. Because marsupial young are
born at an early stage of development and then spend a long period attached to
a teat (often in a pouch), data on offspring sex ratios are easy to collect.
This accounts for the fact that many estimates of the sex ratio of offspring
in marsupials have been reported, but there is a correspondingly high risk of
selective reporting in favor of data sets that show significant departures
from parity or that support one of the theories of adaptive sex ratio
adjustment. Selective reporting can be revealed by plotting estimates of sex
ratio against sample size, as described by Palmer
(2000). We used this approach
to test whether our understanding of sex ratio variation in T.
vulpecula has been distorted by selective reporting and to determine
whether the populations analyzed in this study are representative of the range
of variation in sex ratios for the species.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
In T. vulpecula, a single young is born after a gestation of approximately 18 days, attaches to a teat, and remains in the pouch for about 5 months, after which it rides on its mother's back for another 1-2 months. Females are capable of breeding at any time of the year and can conceive again to replace a young that dies before being weaned, but in many populations there is a concentration of births between April and July, and most females produce only one offspring per year (Fletcher and Selwood, 2000
).
We compiled all available data on sex ratios of offspring between birth and
weaning in T. vulpecula. We restricted the survey to data sets from
single geographic localities, representing local populations, and with sample
sizes equal to or above the minimum required to demonstrate a significant
departure from parity by a binomial test (i.e., n = 6 offspring).
Twenty-eight estimates of the sex ratio met these criteria; 23 represented
populations from throughout the species' natural range in Australia, and a
further 5 were from populations in New Zealand, where T. vulpecula
has been introduced. We conducted a more detailed analysis of data from six
North Queensland populations of T. vulpecula (described in
Johnson et al., 2001); in this
series of populations, male biases in the sex ratio were significantly
associated with relative shortages of den sites. All populations occupied open
eucalypt woodland with a grassy understory and were distributed over an area
of about 40-km radius. Populations were studied by live-trapping on 40-ha
grids for periods up to 4 years, and sex was determined for as many
pouch-young and young-on-back as possible. In all, 311 offspring from 142
mothers were sexed. Head length was measured for all offspring, and
pouch-young were aged from head length by a standard curve established for
these populations (Johnson, unpublished data). Because growth of the head was
linear during pouch life, a standard measure of growth rate could be
calculated for any animal measured twice by dividing the change in head length
by the time interval between measurements. We calculated growth rate only when
the two measurements were separated by more than 30 days because estimates
over shorter intervals were subject to a relatively large measurement error.
When more than two measurements were available for a pouch young, we used the
earliest and latest to calculate a single estimate of growth rate.
Reproductive rates of adult females were measured as the number of offspring
produced per year for females monitored for 2 years or more.
The six North Queensland populations were divided into two groups: those with sex ratios above the mean for the whole sample of populations (three populations with male proportions of 0.683, n = 72; 0.620, n = 159; 0.618, n = 13) and those with sex ratios below the mean (three populations with male proportions of 0.500, n = 37; 0.500, n = 14; 0.438, n = 16). We pooled data in each set of three populations to increase the power of the analysis to detect differences that would indicate the stage of development at which sex biases were generated.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Male-biased offspring sex ratios in T. vulpecula
Data on the sex ratio of unweaned offspring in 28 populations of T. vulpecula are shown in Figure 1. There were male biases in 19 cases and female biases in only 7 (binomial test, p = .05; in 2 cases sex ratios were exactly even); all 8 cases of significant departure from parity favored males (binomial test, p = .01).
|
In Figure 1, sex ratios are
plotted against sample sizes to produce a funnel graph. Extreme values of the
sex ratio are more likely to arise by chance when sample sizes are small,
hence the increase in variance of estimates at low sample sizes, producing the
funnel shape of the scatter of points. If there is a tendency to selectively
report ratios that depart strongly from parity, this will be revealed by an
absence of estimates close to 0.5, particularly when sample sizes are small;
that is, the funnel would appear to be hollow. A tendency to selectively
report values that deviate in one direction would produce a relationship
between sample size and estimated sex ratio (e.g., a preference for reporting
male biases would produce a negative regression of sex ratio on sample size;
Palmer, 2000). The scatter of
values for T. vulpecula falls into a solid funnel shape, and there is
no relationship of sex ratio to (log10) sample size (F1,26
= 0.26, p = .61), suggesting that selective reporting of sex ratios
has not distorted our understanding of variation in sex ratios in this
species. The six populations on which the present analysis is based are
typical of the variation in sex ratios reported among other populations of
T. vulpecula.
Origins of male biases
Generation of biased sex ratios by either sex-biased mortality between
birth and weaning or a longer duration of dependence in one sex would result
in a trend for sex ratios to become more male biased with offspring age in
populations with an overall male bias, but not in populations lacking such a
bias. For neither set of populations was there a significant relationship
between the age of young when sexed and their probability of being male
(Figure 2, by logistic
regression, 
21 = 2.48, p = .12,
21 = 1.34, p = .25 in male biased and
unbiased populations, respectively). There was no tendency for the rate of
growth of the head to differ between the sexes (F1,59 =
1.68, p = .20) and no effect of an interaction of sex and population
type (with or without a sex ratio bias) on growth rates (F = 0.04,
p = .85; see Table
1).
|
|
Male biases could also be produced if some females who conceive female
offspring prevent them from entering the pouch immediately after birth or
abort them. Because T. vulpecula produces single young, both
mechanisms would require females to terminate breeding attempts. Given that
breeding is seasonal, females could either resume breeding in the same
breeding season or defer reproduction to a subsequent season. A tendency for
mothers to discriminate against female offspring early in the season and
produce replacement young in the same season (thereby biasing the population
sex ratio toward males) would be revealed by a later mean birth date of female
than of male offspring. Figure
3 shows that, in populations with male-biased sex ratios, the
seasonal distributions of births of male and female offspring were similar.
The birth dates (in number of days after January 1) of male and female
offspring did not differ significantly (t175 = 0.65,
p = .51); the mean birth days for males and females (with 95%
confidence limits) were 130.33 (125.29-135.36) and 127.76 (120.96-134.58),
respectively (considering only the main birth period; i.e., up to the 155th
day of the year). There was no relationship between birth day and offspring
sex ratio through the main birth period (by logistic regression,
21 = 0.67, p = .41). If females who
terminated breeding attempts did not breed again until the following year,
there would be no sex difference in mean birth date within a season, but
fecundity would be reduced in populations with male-biased sex ratios. We
measured fecundity as the mean number of pouch-young produced per year for
females who were captured regularly over periods of 2 years or more; this did
not differ between the two sets of populations
(Table 1;
t154 = 1.55, p = .13).
|
These analyses exclude the first four mechanisms identified above. We
therefore conclude that biases were established at conception. This agrees
with the only other study of a marsupial that determined the stage of
development at which biased sex ratios are generated. Davison and Ward
(1998) found that in a
population of Antechinus agilis a strongly female-biased sex ratio
was established at conception. Krackow
(1995) has identified many
physiological mechanisms that have the potential to bias the probabilities of
conceiving sons and daughters in mammals, but it is not possible to say which
was most likely to have been operating in T. vulpecula. It is clear,
however, that sex ratio adjustment was achieved with little cost, as females
who produced offspring in the direction of the population-wide bias were able
to do so without terminating reproductive attempts. The fact that biased sex
ratios are established before birth in both of the known cases of adaptive
adjustment of population sex ratios in marsupials, representing two orders,
together with the high incidence of biased sex ratios of offspring in
marsupial species (Cockburn,
1990), suggests that the capacity to adjust sex ratios at
conception might be widespread among marsupials.
|
ACKNOWLEDGEMENTS |
|---|
We thank Alison Payne for help with data collection, Murray Efford for comments and provision of unpublished data, the Australian Research Council for support, D. Sheahan for permission to work on his land, and David and Erica Murray for hospitality in the field.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONREFERENCES |
|---|
Caughley G, Kean RI, 1964. Sex ratios in marsupial pouch young. Nature 204: 491.[Medline]
Clark AB, 1978. Sex ratio and local resource
competition in a prosimian primate. Science
201: 163-165.
Clinchy M, 1999. Does immigration "rescue" populations
from extinction? (PhD dissertation). Vancouver: University of British Columbia
(available at
www.zoology.ubc.ca/
clinchy).
Cockburn A, 1990. Sex ratio variation in marsupials. Austral J Zool 37: 467-479.
Cockburn A, Scott MP, Dickman CR, 1985. Sex ratio and intrasexual kin competition in mammals. Oecologia 66: 427-429.[Web of Science]
Coleman JP, Green WQ 1984. Variation in the sex and age distributions of brush-tailed possum populations. N Z J Zool 11: 313-318.
Davison MJ, Ward SJ, 1998. Prenatal bias in sex ratios in a marsupial, Antechinus agilis. Proc R Soc Lond B 265: 2095-2099.[Medline]
Dickman CR, 1988. Sex ratio variation in response to interspecific competition. Am Nat 132: 289-297.[Web of Science]
Dunnett GM, 1956. A live-trapping study of the brush-tailed possum Trichosurus vulpecula (Kerr) (Marsupialia). CSIRO Wildl Res 1: 1-18.
Efford M, 1998. Demographic consequences of sex-biased dispersal in a population of brushtail possums. J Anim Ecol 67: 503-517.
Fletcher T, Selwood L, 2000. Possum reproduction and development. In: The brushtail possum: biology, impact and management of an introduced marsupial (Montague TL, ed). Lincoln, New Zealand: Manaaki Whenua Press; 58-67.
Godfray HCJ, Werren JH, 1996. Recent developments in sex ratio studies. Trends Ecol Evol 11: 59-62.
Hope RM, 1972. Observations on the sex ratio and the position of the lactating mammary gland in the brush-tailed possum, Trichosurus vulpecula (Kerr) (Marsupialia). Austral J Zool 20: 131-137.
Johnson CN, Clinchy M, Taylor AC, Krebs CJ, Jarman PJ, Payne A, Ritchie EG, 2001. Adjustment of offspring sex ratios in relation to availability of resources for philopatric offspring in the common brushtail possum. Proc R Soc Lond B 268: 2001-2005.[Medline]
Johnson CN, Jarman PJ, 1983. Geographical variation in offspring sex ratios in kangaroos. Search 14: 152-153.
Kerle JA, Howe CJ, 1992. The breeding biology of a tropical possum, Trichosurus vulpecula arnhemensis (Phalangeridae: Marsupialia). Austral J Zool 40: 653-665.
Krackow S, 1995. Potential mechanisms for sex ratio adjustment in mammals and birds. Biol Rev 70: 225-241.[Medline]
Packer C, Collins DA, Eberly LE, 2000. Problems with
primate sex ratios. Phil Trans R Soc Lond B
355: 1627-1635.
Palmer AR, 2000. Quasireplication and the contract of error: lessons from sex ratios, heritabilities and fluctuating asymmetry. Annu Rev Ecol Syst 31: 441-480.[Web of Science]
Pieters CW, 1985. The ecology of the common brushtail possum (Kerr, 1972) on Magnetic Island with particular reference to its adaptability to a modified environment (Honours). Townsville, Australia: James Cook University.
Rohlf FJ, Sokal RR, 1995. Statistical tables, 3rd ed. New York: W. H. Freeman.
Williams GC, 1979. The question of adaptive sex ratio in outcrossed vertebrates. Proc R Soc Lond B 205: 567-580.[Medline]
Winter JW, 1976. The behaviour and social organisation of the brushtail possum (Trichosurus vulpecula) (PhD dissertation). Brisbane: University of Queensland.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||