Evaluating two assesment methods for Gulf of Maine northern shrimp based on simulations

Abstract

Analyses of simulated data with measurement errors, process errors, and model specification errors similar to those for the fishery for Gulf of Maine northern shrimp, Pandalus borealis, demonstrate that a stage-based model (catch-survey analysis) provides accurate estimates of stock abundance, and that a surplus production model (ASPIC: A stock production model incorporating covariates) indicates accurate trends in stock biomass when catch and relative stock size indices are measured with moderate to high precision. However, the accuracy of catch-survey estimates decreases when the precision of input data is low (i.e. greater than 30% coefficient of variation). Results suggest that catch-survey analysis of data with a wide range of observations is more reliable than those with a restricted range. Despite bias in estimates of stock biomass from ASPIC, trends in stock size are generally accurate.