Time-varying predation as a modifier of constant natural mortality for Gulf of Alaska walleye pollock

Although highly variable in time and space, predation remains the greatest source of mortality for juvenile and lower trophic-level fishes. As such, predation can have substantial and long-term effects on the dynamics of these prey. Gulf of Alaska walleye pollock (Gadus chalcogrammus) has shown considerable variability in biomass over the past four decades. During this same time, the demersal fish community transitioned from being dominated by pollock to a system comprised primarily of upper trophic-level predators. We estimated time-varying predation mortality to better understand its effects on the population dynamics of pollock in this currently “top heavy” system. Our index of predation accounted for spatiotemporal variation in predator biomass, bioenergetics-based rations, and age-specific proportions of pollock consumed (1990–2019). To evaluate population-level impacts of predation, we included an index of removals as part of the stock assessment model. This formulation allowed for non-annual data inputs and included a proportionality constant with which to scale predation. Age-specific natural mortality was allowed to vary according to a penalized random walk. We found that natural mortality ranged from 37% higher to 17% lower than the long-term mean. Resulting estimates of total pollock biomass differed by as much as 37% relative to a model without time-varying natural mortality, though the maximum difference for exploitable biomass was only 14%. Using an empirically-derived predation index to modify constant natural mortality allows stock assessment scientists to evaluate impacts of time-varying predation on assessed species. This approach provides a relatively simple way of incorporating ecological information into single-species stock assessments and may reduce bias compared to conventional models that do not account for changes in predation mortality. Notably, including predation mortality in single-species assessments may help identify inconsistencies in biomass estimates that warrant further consideration.