Divergent genomic selection for herbage accumulation and days-to-heading in perennial ryegrass

Perennial ryegrass (Lolium perenne L.) is the principal source of nutrition for ruminant livestock in temperate environments. Increasing the rate of genetic gain for dry matter (DM) yield in this species is an important goal for breeders. Genomic selection (GS) is a strategy that aims to improve genetic gain by using molecular marker information to predict breeding values in selection candidates. A rapid assessment of GS for herbage accumulation (HA; proxy for DM yield) and days-to-heading (DTH), was conducted by completing one cycle of divergent GS from four selection populations (Pop I G1 and G3; Pop III G1 and G3), using previously-developed genomic prediction models. The G1 selection populations were offspring of the training set while G3 populations had advanced through additional selection cycles from the training generation. In field trials, mean HA of High GEBV selection group (SG) progenies, averaged across all selection populations, was 16.5% higher (P < 0.05) than Low GEBV SG’s in the target environment but did not differ significantly in a second environment. Mean divergence was greater in the Pop I lineage (41%; P<0 .05) than Pop III (12%) and response in G1 was higher on average (30%) than in G3 (23%) . A GS cycle for DTH also resulted in a significant difference (P<0.05) between High and Low GEBV SG’s of up to 8.2 days and was more effective in G1 populations (mean difference 6.5 days) than G3 populations (4.3 days). This study shows that GS can be used in perennial ryegrass to improve traits with contrasting genetic architectures and highlights the importance of target environment selection for training prediction models, as well as the influence of relatedness between training set and selection populations. Indicative GS responses in populations generationally-distant to the training set suggests that genomic selection may be effective over multiple cycles of recurrent selection.