Comparative simulation studies on the effects of selection for quantitative traits in autogamous crops: Early selection versus single seed descent

Abstract

A comparison is made between two contrasting breeding procedures for self-pollinating crops by means of computer simulation studies. The first is an early cross selection method, based on cross prediction by F3 line estimates of the cross mean and between-line variance. Subsequently, line selection is performed. Selection is postponed in the second procedure to a more homozygous, F6 generation, which is obtained by single seed descent. Only then is line selection performed, regardless of the pedigree. The two procedures are compared for the cases of one and five crosses under selection. If only one cross is concerned, the early selection method reduces to early line selection and it was found to supply better inbred lines than SSD. But when more crosses are involved, the SSD procedure performed just as well as the early selection method and, at low heritability, even better because the early cross prediction was often poor. Dominance appears to be of very little influence on either selection procedure. © The Genetical Society of Great Britain.