Hypotheses for Facultative Adjustment
Trivers-Willard hypothesis
Trivers and Willard (1973) provided theoretical justifica-
tion for facultative manipulation of sex ratios by mothers in
polygynously breeding mammals under certain conditions.
These authors argued that mothers in poor condition or
with limited resources should bias their investment toward
offspring of the sex with the lowest reproductive variance.
Their reasoning was that offspring of mothers in poor con-
dition will be smaller than those of mothers in better con-
dition and, in species in which males compete for access
to females, sons from these poor mothers are less likely to
breed than daughters. The Trivers-Willard hypothesis was
provocative and stimulated much empirical research, as well
as additional theoretical treatments on manipulating sex ra-
tios. Follow-up work has extended incentives for parental
manipulation of sex ratios to the effects of local competi-
tion for resources (Myers 1978; Silk 1983; Johnson 1988),
and differential advantage of the sexes to temporal changes
in resources over the course of a breeding season (Werren
and Charnov 1978; Lambin 1994b,c). Each of these hy-
potheses focuses on a different aspect of mammalian biol-
ogy that should be open to selective pressures, though not
necessarily in all species.
Let us consider tests for each of the hypotheses for fac-
ultative manipulation of sex ratios in rodents in more detail,
with an eye toward identifying deficiencies that limit our
ability to reject explanations. As the Trivers and Willard
(1973) hypothesis set the stage for this area of research,
it provides a suitable starting point. The Trivers-Willard
(T-W) hypothesis predicts biased investment based upon
reproductive potential of offspring as influenced by mater-
nal condition during parental investment. The hypothesis
is based on three simplifying assumptions: (1) that the
young’s condition at the end of parental investment is cor-
related with the mother’s condition during her investment;
(2) that differences in body condition of young at the end
of parental investment are maintained into adulthood; and
(3) that differences in body condition affect males more
than females. More specifically, Trivers and Willard predict
that “natural selection must favor one or more genes that
adjust the sex ratio produced by an adult female to her own
condition at the time of parental investment. In taxa such as
mammals where males determine sex of offspring, female
control of the sex ratio must involve differential mortality
by sex, either of sperm cells or of growing young during pa-
rental investment” (Trivers and Willard, 1973, p. 91). From
this statement and as developed elsewhere in their argu-
ment, there are two conditions necessary for a given data
set to be consistent with the T-W hypothesis: first is the
existence of differential investment by a mother between
sexes, and second is that differential investment is faculta-
tive (conditional) and under maternal control. Failure to
satisfy both of these conditions undermines support for the
hypothesis.
Trivers and Willard’s hypothesis is often misinterpreted
as applying only to the secondary sex ratio (the sex ratio of
a litter at birth) rather than biased investment in general,
despite the fact that the authors explicitly include the entire
period of postnatal investment in their model (Trivers and
Willard 1973, p. 91). This interpretation perhaps is under-
standable given the title of their paper and the fact that
most of their discussion revolves around sex ratios. Their
nearly exclusive focus on sex ratios almost certainly stems
from the fact that, to evaluate available data relative to their
model, they made the assumption “that sex ratio at birth
in mammals is a measure of the tendency to invest in one
sex more than in the other” (p. 91). Nevertheless, the fun-
damental requirement is for differential investment between
the sexes, which may or may not result in biased sex ratios.
A very important consequence of this clarification is that a
study need not document skewed sex ratios to support the
T-W model.
The second condition of maternal control of investment
is equally important, but dependent on the first. Differential
allocation of resources by the mother based on maternal
condition constitutes facultative adjustment in that alloca-
tion “decisions” are optional and conditional. This is a crit-
ical requirement because differential mortality or growth
caused by sex differences in energy or nutrient requirements
of the young would not constitute facultative adjustment
and hence provide no support for the T-W hypothesis.
The Trivers-Willard hypothesis has drawn much atten-
tion from researchers, but results of studies posed as tests
of the model are equivocal. One of the major difficulties in
evaluating existing data is that, with few exceptions, they
seldom were designed as explicit tests of the hypothesis.
Instead, the T-W hypothesis, or some other explanation,
is used to interpret skewed sex ratios detected in other
work (recall, however, that the Trivers-Willard model is not
about sex ratios per se, but differential investment in the
sexes, with one potential outcome being biased sex ratios).
Further, although Trivers and Willard included a consider-
ation of polytocous species in their discussion, the potential
for concurrent manipulations of litter size in such species
could complicate interpretations of sex ratio adjustment
(see Williams 1979).
Key studies using rodents, which may be used to evalu-
ate the Trivers-Willard hypothesis or its assumptions, are
presented in table 11.1. Among these studies, maternal con-130 Chapter Eleven