Defining a biomass limit reference point under time-varying productivity: a case study of Atlantic cod in the southern Gulf of St. Lawrence, Canada

Many commonly used fisheries reference points assume stable population processes, yet time-varying productivity is increasingly observed among fish stocks. Atlantic cod (Gadus morhua) in the southern Gulf of St. Lawrence (sGSL) collapsed in the 1990s and has not recovered, with non-stationarity in recruitment, growth, maturity, and natural mortality preventing equilibrium conditions. In Canada, the biomass limit reference point (LRP) defines the threshold below which a stock is considered to experience serious harm to its productivity and requires rebuilding. We used life-history data, stock assessments, stock–recruit models, and surplus production models to evaluate candidate LRPs while accounting for non-stationarity in sGSL cod productivity. Two indicators of serious harm were identified: a persistent low production–low biomass state and a breakpoint in production suggesting the stock crossed an Allee threshold in 1992. Three candidate LRPs were supported and produced similar stock status, whereas other commonly used static or dynamic methods generated values below the level of serious harm. These findings highlight the importance of testing multiple approaches when estimating LRPs for stocks experiencing shifting productivity.