Inbreeding Method Effects on Genetic Mean, Variance, and Structure of Recurrent Selection Populations

Abstract

Devising an optimal recurrent selection program within autogamous crops requires knowledge of which inbreeding method will provide the biggest short‐term and long‐term gains. We compared genetic gains using either doubled haploid (DH) or single seed descent (SS1)) methods by computer simulation of recurrent selection for high phenotypic value. The first simulation assumed an additive model with homogeneous recombination frequencies and allele effects. Genetic means and variances, number of fixed alleles, and linkage disequilibrium were modeled over 15 cycles of selection in a factorial of the two inbreeding methods, recombination frequencies between adjacent loci of 0.05, 0.15, 0.50, and environmental variances corresponding to initial heritabilities of 0.58, 0.26, and 0.13. The second simulation allowed for heterogeneity of recombination frequencies and allele effects. Short‐term gain determined by DH was superior. However, the difference in gain between DH and $SD declined as recombination frequency decreased from 0.50 to 0.05. At the two lower recombination frequencies, more (‐) alleles were fixed, and long‐term gain determined by DH was inferior compared with SSD. The recombination frequency heterogeneity treatment did not affect the response variables. The allele effect heterogeneity treatment increased short‐term and decreased long‐term gain but did not affect the relative performance of the two inbreeding methods. These results suggest that inbreeding method should be chosen on the basis of knowledge or assumptions about recombination frequencies between loci affecting the trait of interest.