Data-rich stock assessments rely on fisheries independent surveys to inform trends in abundance and stock composition. The northeast waters of the United States are sampled by several multi-species trawl surveys. Changing ocean conditions within the footprint of these surveys have resulted in species distribution shifts. Despite these observations, survey indices typically do not account for spatial shifts or sampling changes when included in stock assessment. However, failure to account for changes in population availability to surveys can violate stock assessment model assumptions. Here, we use the northeast black sea bass ( Centropristis striata) stock as a case study to explore two different methods for accounting for changes in fish availability. The first method uses a spatiotemporal model to combine multiple surveys. The second method uses a state-space stock assessment model to explore different process error assumptions on survey catchability and selectivity. Results show the importance of accounting for changes in spatial distribution and demonstrate the benefits and limitations of using spatiotemporal models and process error to account for changes in survey availability.
Hansell et al. (Thu,) studied this question.