Forecasting pacific salmon stocks with directional variability of reproduction conditions

Abstract

The number of salmon offspring depends both on the number of parents and on the conditions of reproduction. In case of random deviations of reproduction conditions from the average long-term values, classical «stock–recruitment» models are usually used. Based on simulation modeling, it is shown that with directed climate changes, the use of such models leads to significant errors in the forecast of the recruitment. For such cases, the necessity of using models that take into account the trend component of the variability of reproduction conditions is justified.