Adapting to climate‐driven distribution shifts using model‐based indices and age composition from multiple surveys in the walleye pollock (Gadus chalcogrammus) stock assessment

Abstract The northern Bering Sea is transitioning from an Arctic to subarctic fish community as climate warms. Scientists and managers aim to understand how these changing conditions are influencing fish biomass and spatial distribution in this region, as both are used to inform stock assessments and fisheries management advice. Here, we use a spatio‐temporal model for walleye pollock (Gadus chalcogrammus) to provide two inputs to its stock assessment model: (a) an alternative model‐based biomass index and (b) alternative model‐based age compositions. Both inputs were derived from multiple fishery‐independent data that span different regions of space and time. We developed an assessment model that utilizes both the standard and model‐based inputs from multiple surveys despite inconsistencies in spatial and temporal coverage, and we found that using these data provide an improved spatial and temporal scope of total pollock biomass. Age composition information indicated that pollock density is increasing and moving farther north, particularly for older pollock. We found that including an index of cold pool extent could be used to extrapolate pollock densities in the northern Bering Sea in unsampled years. Stock assessment parameter estimates were similar for standard and model‐based input. This study demonstrates that spatio‐temporal model‐based estimates of a biomass index and age composition can facilitate rapid changes in stock assessment structure in response to climate‐driven shifts in spatial distribution. We conclude that assimilating data from regions neighboring standard survey areas, such as the Chukchi Sea and western Bering Sea, would improve understanding and management efforts as fish distributions change under a warming climate.