A rebuilding time model for Pacific salmon Fisheries Research

We describe a new model developed for the purpose of projecting rebuilding periods for overfished Pacific salmon stocks as defined by the Pacific Fishery Management Council. The model has relatively low data requirements as it relies on past estimates of abundance to project future abundance, accounting for positive lag-1 autocorrelation if there is evidence of its existence. Replicate applications of the model allow for computation of the probability of achieving rebuilt status in future years. Application to simulated abundance and escapement data suggested that model-projected rebuilding times generally corresponded to simulated rebuilding times as raw errors were median unbiased. Simulations also suggested that results were generally robust to parameter misspecification and that increased levels of lag-1 autocorrelation in abundance were associated with longer rebuilding periods. The application of the model to five overfished stocks in 2018–2019 illustrated the differences in projected rebuilding times under alternative rebuilding management strategies. The model filled a need for relatively rapid assessment of alternative rebuilding strategies for Pacific salmon stocks, a need that will likely remain given current biological reference points and fluctuations in salmon abundance.