Stock prediction using stochastic recruitment numbers with empirical stock-dependent distribution

Abstract

Probability distributions for the recruitment, conditional on the spawning stock biomass (SSB) were made using a kernel method. Predictions were made with recruitments according to these distributions for 10 years, using a Monte Carlo procedure, assuming constant weight at age, maturity ogive and natural mortality. Examples are given for 3 stocks, Western mackerel, North Sea sandeel and North Sea herring. Three management strategies were studied, a fixed F, a fixed yearly catch and a regime aiming at stabilizing the SSB. Due to the variable recruitment, attempts to stabilize the SSB increases the year to year variations in the yield. Attempting to take a fixed catch every year is hazardous, since the appearance of a few poor year classes may detoriate the stock so that the future recruitment suffers. This simple approach may be a useful tool for evaluating management strategies in terms of risks and possible outcome of the fisheries.