| Summary: | Thesis (Master's)--University of Washington, 2017-06 This thesis develops a framework for conducting stock assessments that unifies age- and size-structured assessment models, creating what we term “age-size-structured models.” Several studies have explored the application of age-size-structured models to fisheries data. However, most are specific applications or were limited in scope. There is a dearth of generalized single-stock age-size-structured assessment models, which led to the creation and testing of such a model. An age-size-structured assessment method is developed, and examined first using a simulation study, and then using data for two species in the Eastern Bering Sea: Tanner crab (Chionoecetes bairdi) and Pacific cod (Gadus macrocephalus). Simulations are used to test under which conditions the assessment method is biased or imprecise. The simulations are summarized using a measure of the relative error between true and estimated spawning stock biomass. The simulation study suggested little bias and relatively high precision, such that the estimation method was considered acceptable for use in management. The results from the example application were compared to those from the official NOAA assessments. The estimates of mature male biomass for Tanner crab were found to be similar to those from the most recent stock assessment, but the estimates of spawning stock biomass for Pacific cod differed between the age-size-structured model and the latest official assessment, particularly after the mid-2000s. Some of the dissimilarity may be because maturity and growth are modelled based on size-dynamics in the age-size-structured model, while the official assessment considers only age-dynamics. Nevertheless, the age-size-structured model produces relatively robust estimates, and could be used as an alternative model to which current stock assessments could be compared. This generalized age-size-structured fishery stock assessment method could be applied to any data-rich stock globally, particularly ...
|
|---|