Selection during a selfing programme. II. The effects of two or more rounds of selection

Theory presented in a previous paper has been used to provide a description of the responses expected from selection in two or more generations of a selfing series. Optimum two-stages and multi-stage selection programmes are described, for a limited set of circumstances, and compared with the single-stage designs described previously. Simulations show that biases may be expected if epistasis is present or if the number of effective factors is small. If the number of effective factors is large, then progress from even the most efficient designs will be limited by drift.