Estimating movement with a spatially-explicit stock assessment model of eastern Bering Sea walleye pollock, Theragra chalcogramma

Abstract

The standard Eastern Bering Sea (EBS) walleye pollock (Theragra chalcogramma) stock assessment model has no spatial dimension. To advance the understanding of EBS walleye pollock movements and spatial structure on finer temporal and spatial scales, a spatially explicit migration model of EBS walleye pollock was developed. However, there are no estimates of movement rates for this population. Using standard sample size formulae in a Petersen-type experiment, we showed that only a moderate mark-recapture program [minimum number of tags for ages-1 + was 9,475 (all sectors included) and 20,924 (only catcher-vessel shoreside sector included)] is needed to estimate abundance. Given these sample size requirements determined for abundance estimation, the Darroch method was used to estimate movement parameters between two regions, the northwest (NW) and southeast (SE) EBS in a simple compartment model. Directed movement could be reasonably estimated with Monte Carlo simulation. To develop the EBS walleye pollock age-structured movement model, the standard stock assessment model was extended into a two-region (NW and SE EBS), two-season, age-specific movement model. Movement could be estimated from disaggregated data without mark-recapture information, but with low precision. However, the uncertainty indicates that a mark-recapture study is needed before such a model could be applied for management applications.