Statistical Aspects of Design of Animal Breeding Programs: A Comparison Among Various Selection Strategies

Abstract

The goal of animal breeding is to change populations in such a way that overall economic merit is increased as fast as possible over many generations. Available genetic variation and any possible genetic variation arising from mutations need to be effectively utilized. Because of the conflict between maximizing short- and long-term responses compromises need to be found. To compare different strategies, a full-sib structure was simulated and five strategies (family selection, restricted and non-restricted within family selection, mass selection and index selection) were investigated. Comparisons were made for the first generation, short-term and long-term responses. In the first generation, the correlation between estimated and true breeding value and the expected value of the standardized selection differential have to be considered. In the short term the reduction in variance due to linkage disequilibrium also has to be taken into account and, in the long term, the reduction in genetic variance due to drift must be considered. This last factor depends on both the systematic selection pressure and effective population size.