Environmental predictability and the cost of imperfect information: Influences on offspring size variability

We examined the potential influence of mean egg size and environmental predictability on the variation of egg size within and among female brook trout, Salvelinus fontinalis. First, we examined egg size variability as a means of balancing the costs of imperfect information and unequal allocation among eggs, predicting that egg size variability will increase as optimal egg size decreases. In the first empirical test of the imperfect information hypothesis, we found support for this prediction, with the variability of egg sizes within females decreasing as mean egg size increased in every population (n = 10) and year (n = 13) examined. Second, we tested the hypothesis that inter-clutch variability in egg size increases as environments become less predictable. When we compared the variability in egg sizes among females across populations and years, we found greater variability in the mean egg size among females in less predictable environments. Finally, we tested the hypothesis that intra-clutch variability increases as environments become less predictable. When we compared the mean variability in egg size within females across populations and years, we found greater variability in egg sizes within females in less predictable environments. Support for these three hypotheses suggests that there is less egg size variability, both within and among females, when environments are more predictable, and that females use variability in egg size to offset the cost of imperfect information when producing smaller eggs.