Historical trends in productivity of 120 Pacific pink, chum, and sockeye salmon stocks reconstructed by using a Kalman filter

Abstract

Temporal trends in productivity of Pacific salmon ( Oncorhynchus spp.) stocks are important to detect in a timely and reliable manner to permit appropriate management responses. However, detecting such trends is difficult because observation error and natural variability in survival rates tend to obscure underlying trends. A Kalman filter estimation procedure has previously been shown to be effective in such situations. We used it on a Ricker spawner–recruit model to reconstruct indices of annual productivity (recruits per spawner (R/S) at low spawner abundance) based on historical data for 120 stocks of pink ( Oncorhynchus gorbuscha ), chum ( Oncorhynchus keta ), and sockeye ( Oncorhynchus nerka ) salmon. These stocks were from Washington, British Columbia, and Alaska. The resulting estimated temporal trends in productivity show large changes (on average 60%–70% differences in R/S and average ratios of highest to lowest R/S between 5.4 and 7.9 for the three species). Such changes suggest that salmon stock assessment methods should take into account possible nonstationarity. This step will help provide scientific advice to help managers to meet conservation and management objectives. The Kalman filter results also identified some stocks that did not share temporal trends with other stocks; these exceptions may require special monitoring and management efforts.