Using multi-species surplus production models to estimate ecosystem-level maximum sustainable yields

We used time series of biomass and catches for the major commercial groundfish stocks in the Eastern Bering Sea/Aleutian Islands region and in he Gulf of Alaska to compute annual surplus production for 1977–2004. Annual surplus production and average annual biomass were aggregated cross stocks within each region to examine the relationship between total surplus production and total aggregated biomass. We fit two surplus roduction models (Graham-Schaefer and Pella-Tomlinson) to the observed relationships to estimate maximum multi-species surplus production equivalent to maximum sustainable yield) of the groundfish complexes in these ecosystems. Maximum multi-species surplus production was stimated to be approximately 2.5× 106 t in the Bering Sea/Aleutian Islands and 330× 103 t in the Gulf of Alaska. These point estimates were maller than the sum of single-species MSY proxies from recent stock assessments, and estimates for the Gulf of Alaska were much smaller than arlier estimates that were used to specify optimum yield ranges for the Gulf of Alaska groundfish complex. Therefore, optimum yield ranges for he Gulf of Alaska may need to be re-assessed to reflect more recent conditions. We further estimated the effects of environmental variability on annual surplus production using correlations and generalized surplus production odels with environmental covariates. Results suggest that surplus production in the Bering Sea may be lower during conditions associated with he positive phase of the Pacific Decadal Oscillation, which is characterized by warmer temperatures, reduced ice, and reduced wind mixing over he shelf. Similarly, surplus production on the Gulf of Alaska shelf was lower during years with warm bottom temperatures coupled with low alinities in late winter. If maximum multi-species surplus production varies as a result of environmental variability, optimum yield ranges for the roundfish complexes in these ecosystems should take such fluctuations into account. 2006 Elsevier B.V. All rights reserved.