Genomewide Predictions for Backcrossing a Quantitative Trait from an Exotic to an Adapted Line

The backcross procedure has been widely used in plants to transfer one or a few known genes from a poorer (e.g., exotic) donor line to a better (e.g., adapted) recurrent line. But for quantitative traits, backcrossing for 6-8 generations has been traditionally ineffective due to a dilution effect and the inability to efficiently select for the best segregants in each backcross generation. My objective was to determine if genomewide predictions are useful for backcrossing a quantitative trait. I simulated a backcrossing program with an elite recurrent parent and a donor line that had the favorable allele (not present in the recurrent parent) at only 10% or 25% of the quantitative trait loci. Genomewide selection was used to identify the best plants, in each backcross generation, for crossing to the recurrent parent to form the next backcross generation. I found that the use of genomewide predictions allowed the recovery of backcross lines that had more than 95% of the recurrent parent genome and that were homozygous for a sufficient number of favorable exotic alleles, to the extent that the backcross lines exceeded the recurrent parent in performance. I concluded that genomewide predictions are useful for backcrossing of quantitative traits in plants.