Lack of relationship between simulated fish population responses and their life history traits: inadequate models, incorrect analysis, or site-specific factors?

Abstract

Relationships between fish population responses to changes in their vital rates and commonly available life history traits would be a powerful screening tool to guide management about species vulnerability, to focus future data collection on species and life stages of concern, and to aid in designing effective habitat enhancements. As an extension of previous analyses by others, I analyzed the responses to changes in fecundity and yearling survival of age-structured matrix and individual-based population models of 17 populations comprising 10 species. Simulations of the matrix models showed that the magnitude of population responses, but not the relative order of species sensitivity, depended on the state (sustainable or undergoing excessive removals) of the population. Matrix and individual-based models predicted population responses that appeared to be unrelated to their species-level life history traits when responses were plotted on a three-end-point life history surface. Density-dependent adult growth was added to the lake trout (Salvelinus namaycush) matrix model, and simulations demonstrated the potential importance to predicted responses of density-dependent processes outside the usual spawner–recruit relationship. Four reasons for the lack of relationship between population responses and life history traits related to inadequate population models, incorrect analysis, inappropriate life history model, and important site-specific factors are discussed.