Selection from one of several populations

Abstract

Two methods of choosing foundation stock for an animal-breeding programme are compared: selection within one population (chosen in the hope or belief that it is genetically superior) or selection within each of several populations. It is tentatively concluded that selection from several populations is likely to provide the better foundation stock. This conclusion depends on the assumption that variation between populations in mean breeding value is not large relative to genetic variation within flocks, but evidence on this point is desirable. Doubling the number of populations from which selection is made will increase genetic gain by about one-quarter of the heritability multiplied by the phenotypic standard deviation.