Selection for Adult Size in Female Coho Salmon (Oncorhynchus kisutch)

Publisher Website
Google Scholar
Abstract
Several recent studies have presented evidence that large size confers a selective advantage to female Pacific salmon. Nevertheless, a wide range of female size is normally present in any spawning population. Two possible explanations exist for the observed range in female size. First, average female size might be determined by an optimizing process with variation around the optimum size due to individual differences in success at obtaining food. Second, various sizes of females might coexist as a mixed evolutionary stable strategy. Under the first explanation, females of sizes other than the optimum would display lower fitness whereas, under the second explanation, females of all sizes would be equally fit. We investigated factors affecting survival of eggs, fry, and smolts of coho salmon (Oncorhynchus kisutch) in Carnation Creek on Vancouver Island [British Columbia, Canada] with a view to determining the relative fitness of different sized females. Egg-to-fry mortality was best explained by a model that included only the effects of stream bed scour and gravel quality. Including an effect of female size, expressed through depth of egg burying, worsened the model''s predictive capability. we could find no evidence that the eggs of large females consistently survived better during incubation than those of small females. In fact, we observed three instances in which it appeared that the eggs of smal females survived better. In Carnation Creek, large 1- and 2-yr-old smolts did not consistently survive better in the marine environment than small smolts. Thus, we were unable to demonstrate that the reproductive success of large females was consistently higher than that of small females, contrary to the hypothesis that female size is the result of an optimizing process. In Carnation Creek the observed range of female sizes probably represents an evolutionary stable strategy in which all sizes have equal fitness. We propose a model that predicts female size and variance in size based on the conflicting selective effects of gravel quality, scour, and competition for nest sites.