Modeling the reproductive potential of rockfishes (Sebastes spp.)

Understanding the reproductive potential of exploited fish populations is critical in the development of sustainable harvest practices. Models that characterize female fecundity often assume that the number of eggs produced per unit body weight (relative fecundity) is independent of size or age. Estimates of average lifetime egg production per female are used to determine target harvest rates for several fish species off the west coast of the United States, and are clearly sensitive to changes in relative fecundity with size. Relative fecundity in rockfishes (genus Sebastes) commonly increases with size, but to a varying degree among species. The extent to which this pattern alters our perception of target harvest rates for rockfish depends on the rate at which relative fecundity changes with size, and rates of somatic growth and natural mortality. The reproductive biology of several rockfish species remains poorly understood, and advice is needed regarding proper characterization of fecundity to aid fisheries managers in the development of harvest guidelines. I develop Bayesian hierarchical models to predict fecundity of data-poor rockfish species. The models use information from closely related (congeneric) species to inform predictions of fecundity at size, quantify uncertainty about those predictions, and provide predictive distributions of model parameters for unobserved species. Trends in size-specific relative fecundity are part of an organism's life history strategy. I use state dependent life history models for optimal resource allocation to evaluate potential mechanisms driving these trends. Patterns of growth, maturation and reproduction observed in rockfishes are consistent with the hypothesis of a trade-off between reproduction and natural mortality.