Where genetic differences between age groups change during a selection programme, proportional contributions of each age group to replacements should also change in order to maximize genetic responses to selection. Sums of discounted returns were estimated for such a ‘dynamic’ selection policy and these were compared with those from ‘static’ selection policies in which genetic contributions of different age groups remained unchanged. Four selection methods, two sets of genetic differences between age groups at the outset, three age structures and two discount rates were used.

The results showed that the superiority of the dynamic policy was greatest where genetic differences at the outset differed most markedly from their steady-state values, where discount rates are high and where there are large numbers of age groups and the selection method allows culling in all age groups. Under most conditions initial genetic differences are probably the most important factor. The proportional advantage in sums of discounted returns is variable and generally small within the framework of this study ranging from zero to roughly 8% unless breeding objectives before and during the programme were negatively correlated. In the latter case the proportional superiority of the dynamic policy was much greater.