| Summary: | Multispecies fisheries with technical interactions and sparse data present a challenge for assessment scientists, many of whom are now legally required to provide species-specific management advice. A key question is can these data be used to provide species-specific assessments and management advice? Secondly, in light of new assessment approaches, how would new management procedures perform with respect to the status quo? This thesis uses a novel multispecies model to reconstruct historical abundances using fishery-dependent data from the Hawaiian bottomfish fishery. In this handline fishery, hook competition and time spent handling fish limits catch rates, resulting in nonlinear relationships between CPUE and abundance. The model is jointly fit to species-specific catch and is conditioned on historical fishing effort. The model allows for hook competition and partitions time spent fishing and time spent handling fish, h. Simulation experiments showed this approach provided nearly unbiased parameter estimates unless an incorrect assumption was made about h. Empirical h estimates were unavailable, so a range of h values were imputed from information on fishing gears and sensitivity to h was evaluated. Species-specific information was disentangled using h, but no single value best described all species-specific catch equally. Leading and management parameter estimates were relatively insensitive to the assumed value of h, except for catchability at low stock size. Given the uncertainty about h, management procedure evaluation was used to evaluate management performance of the multispecies model to an aggregate model and evaluate alternative management procedures (MPs), defined by the decision rule, assessment model, and data types. Trade-offs among management objectives describing stability in catch, fishery productivity, and over-exploitation risk were compared. Trade-offs between catch and conservation objectives were largely determined by the decision rule. MPs associated with an aggregate assessment represented a balance from the aggregate perspective, but led to over-exploitation of one or more species. Average annual variation in catch was high for all MPs and highest for MPs using the multispecies approach in combination with a tag-recapture program. These results were largely robust to the assumptions about h in the assessment model and demonstrated the difficulty of co-managing multiple species.
|
|---|
