Data needs and spatial structure considerations in stock assessments with regional differences in recruitment and exploitation

Abstract

This study uses a simulation experiment to demonstrate that bias in estimates of spawning biomass is influenced by the spatial configuration of a stock assessment model, whether survey data are used or not and whether an environmental index is available to inform the spatial distribution of recruitment. Stocks with limited movement of postsettlement fish may be spatially structured due to environmental forces that affect larval dispersal and recruitment distribution or from nonuniform spatial exploitation. Data are frequently aggregated across space in stock assessments, thus disregarding this complex spatial structure and possibly introducing bias into estimates of stock status. An operating model (OM) is created that simulates data that are used in a set of estimation models to assess bias. The following experimental factors are considered: (i) using survey data and environmental indices in the assessment; (ii) using disaggregated data (two regions, as generated by the OM) or aggregated data (one region); and (iii) incorporating different patterns in the OM’s regional exploitation and environmentally driven recruitment distribution.