STOCK ASSESSMENT AND BIOLOGICAL KNOWLEDGE - CAN PREDICTION UNCERTAINTY BE REDUCED

The paper addresses whether we fully utilise our increasing biological knowledge in fish stock assessments. For illustrative purposes, repea ted references are made to the assessment of the north-east Arctic cod stock. In many cases, existing knowledge may allow us to establish re lationships for predicting short or medium term changes in vital popul ation parameters determining mortality, growth, and recruitment. Ancil lary variables in such relationships, with a priori justification for being chosen, could be parameters describing the state of the fish sto ck in question or the state of its biological or physical environment. Long time-series of biological data exist for several stocks, and the se data series should be used for testing proposed relationships. Expl anatory theories should be used for making predictions not dependent o n the assumption that historical patterns will be repeated, increasing the empirical or informative content of our assessments. (C) 1996 Int ernational Council for the Exploration of the Sea