• size-selective fishing;
  • individual growth variation;
  • yield per recruit;
  • IBM

In a release experiment with cod in Norway, the apparent mean growth rates of 3+ cod, calculated by sampling the released cohorts at different ages, were very slow (<0·08 mm day−1). However, when individual growth rates of individual tagged cod of the same size range were measured, the mean growth rates were much faster (0·24 mm day−1). These observations were attributed to size-selective fishing mortality and were illustrated by an individual based simulation model of a cohort of cod with variable individual growth rates. The effects on mean length at age of the surviving cohort of increasing fishing intensity were demonstrated. The model showed that size-selective fishing with the observed individual growth variation, removed the fastest-growing individuals at proportionally higher rates than the slower-growing ones, leading to decreased apparent mean growth rate. The fishing pattern which gave the optimum yield, changed when individual variation was included, and when the apparent growth rate was used in the model the yield per recruit reduced dramatically. This study has shown that individual growth heterogeneity and size-selective mortality are factors which should be considered in future fisheries management models.