Application of Historical Data from Commercial Ryegrass Breeding to Enable Rapid Implementation of Genomic Selection

The application of genome-wide molecular markers and quantitative genetics to predict plant performance will change the way pasture plants are bred. Genomic selection offers the ability to predict the performance of selection candidates with only genotypes and marker effects estimated in a reference population comprised of genotyped and phenotyped individuals or populations.  The main factors affecting the accuracy of genomic selection are reference population size, trait heritability, and breeding population genetic diversity.  The global perennial ryegrass population is diverse and would need a large reference population to achieve moderate accuracy of genomic selection. However, the diversity within a breeding program may be less, which allowed us to apply genomic selection across cultivars/varieties with moderate reference size.  We combined historical phenotype records for seasonal biomass yield and heading date from 15 years of commercial perennial ryegrass breeding with high-density population based transcriptome genotyping-by-sequencing. The ability to genomically predict the observed phenotypic performance in each progressive year was explored by using all varieties from previous years as a reference population and applying a range a genomic selection equations. Accuracies achieved for biomass and heading date were moderate to high, respectively, reflecting the relative heritabilities of these two agronomic traits. These results highlight the ability to predict sward-based phenotypic performance early in the breeding program of individual plants, rapidly reducing the breeding cycle and consequently at least doubling the rate of genetic gain. This study along with the commercially relevant reference population now enables rapid adoption and implementation of genomic selection in ryegrass.