adjustments to parental investment, skewed sex ratios at
weaning are equally consistent with gender-specific mortal-
ity (Werren and Charnov 1978), which is a form of nonfac-
ultative (although potentially adaptive) litter manipulation
(see the following).
A second important distinction of Gosling’s work with
regard to the T-W model in rodents is that differential in-
vestment was manifested amongrather than withinlit-
ters. This pattern is consistent with studies supporting the
T-W hypothesis outside of rodents, such as in red deer
(Cervus elephas;Clutton-Brock et al. 1986). If maternal ad-
justments are more likely among as opposed to within lit-
ters, then might some of the listed rodent studies provide in-
sight concerning where to look to for additional support,
even without evidence that skewed sex ratios are facultative?
Work by Rivers and Crawford et al. (1974) and Wright
et al. (1988) on domestic mice (Mus musculus), Labov et al.
(1986) on golden hamsters (Mesocricetus auratus), and
Clark et al. (1990) on Mongolian gerbils (Meriones un-
guiculatus) provides good examples. In these studies, a bias
in maternal investment was reported in litters following
a period of nutritional stress induced by food restriction
(Rivers and Crawford 1974; Labov et al. 1986; Wright
et al. 1988), or by rearing a previous litter of all-male or
all-female offspring (Clark et al. 1990). But in each case
the adverse condition for the mother was in place prior to
conception.
Timing of stress may be a critical factor. In fact, one can
quickly glean from table 11.1 that of the 15 studies report-
ing biased sex ratios or other forms of differential mater-
nal investment, adverse conditions potentially eliciting dif-
ferential investment were present before parturition in 12.
In contrast, authors reported biased sex ratios in only 3
of 7 studies where adverse conditions were imposed post-
natally, and two of these studies raise further questions.
One of these is McClure’s (1981), results of which have not
been duplicated despite repeated attempts in several species
(Sikes 1995; 1996a, 1996b; Moses et al. 1998). Another
was the study by Moses et al. (1998), in which skewed ra-
tios in bushy-tailed woodrats clearly did not result from
maternal manipulation. I have argued previously, based on
my inability to validate assumptions of the T-W hypotheses
when stress is imposed postnatally, that application of the
T-W hypothesis probably should be restricted to situations
where adverse conditions are in place prior to parturition
(Sikes, 1996a). The data summarized here make applica-
tion of the Trivers-Willard hypothesis in response to post-
natal conditions in rodents even more questionable. This
conclusion is supported in part by Cameron’s (2004) meta-
analysis of data on T-W-like sex ratios in mammals, where
she found strongest support for sex ratio adjustments when
maternal condition at conception was posed as the deter-
mining factor.
In summary, with the exception of the excellent work
by Gosling (1986), unequivocal support for the Trivers and
Willard hypothesis in rodents is lacking because clear evi-
dence of facultative adjustment of reproductive investment
by the mother is not provided. Given the interest that this
hypothesis has generated, even with extremely limited sup-
port, any carefully designed study that addresses this criti-
cal weakness and also validates the underlying assumptions
is likely to profoundly influence life-history research in
mammals far into the future.Local resource competition hypothesis
The local resource competition hypothesis (LRC) was de-
veloped primarily from work with primates (Clark 1978;
Silk 1983), but has since been extended to other taxa, in-
cluding ungulates (Clutton-Brock et al. 1982; Caley and
Nudds 1987) and marsupials (Cockburn et al. 1985). This
hypothesis holds that in species where one sex is philopat-
ric, selection should favor the dispersing sex (usually males
in mammalian systems) when resources are limited because
mothers will compete with daughters for resources. Types
of limiting resources are varied but include food, suitable
den sights, or mates. Lambin (1994b, 1994c) provides sup-
port for this hypothesis in rodents from an observational
study of Townsend’s voles (Microtus townsendii), in which
he noted female-biased sex ratios of litters in spring seasons
where vole density was low (see also table 11.2). However,
he concluded that sex ratios in his system were altered due
to broad-scale competition caused by overall high popula-
tion densities, rather than local competition as influenced
by neighborhood or family size. Bond et al. (2002) pre-
sented data from a manipulative study of gray-tailed voles
(Microtus canicaudatus) that are partially consistent with
the LRC model. These investigators monitored recruitment
in naturally increasing populations of voles with approxi-
mately even adult sex ratios in the spring and summer, but
then manipulated adult sex ratios by adding or removing
adults in the autumn. Bond et al. (2002) reported results
consistent with the LRC model for early season litters, but
as sex ratios later in their study were biased in the opposite
direction, even though population densities and hence like-
lihood of local competition did not change, they concluded
that their data were best explained by seasonal rather than
local resource competition effects.
Local mate competition may be viewed as a special case
of resource competition and potentially brings in new con-
siderations based upon mechanisms of sex determination
in some rodents. Local mate competition as proposed by134 Chapter Eleven