Shared loci determine trait relationships and selection responses for milling and nutritional quality in oat

Trait correlations are a common phenomenon in plant biology. Breeding efforts rely on information about trait associations to either improve multiple correlated traits simultaneously or to minimize the negative effect that the selection towards one trait causes on the others. In this study, we used data from five populations of advanced oat breeding lines evaluated across five environments to examine the relationships among 31 quality traits relevant to the oat food industry. A genome-wide association study (GWAS) was performed to locate significant marker-trait (MT) associations for oat quality traits throughout the genome. Routine marker-assisted selection (MAS) and genomic selection (GS) strategies were simulated to assess whether multi-trait loci had an impact on traits that were not under selection. Phenotypic correlation analysis revealed undesirable associations between key oat quality traits including the negative association between test weight, an indicator of milling yield, and β-glucan content (BG) (r = − 0.2, p < 0.001), a soluble fiber considered to improve heart health. The GWAS identified 147 quantitative trait loci (QTLs), of which 82 were associated with a single trait and 65 were associated with two or more traits. Multi-trait loci provided the genetic basis for trait correlations and simulated single-trait selection through MAS and GS indicates a correlated response. Despite partial shared genetic control, our results indicate that simultaneous genetic improvement of multiple desirable quality traits is feasible, as shared loci often explain a large proportion of variation in one trait but only a limited proportion in the correlated undesirable trait. These findings provide valuable insights for designing breeding strategies aimed at optimizing milling and nutritional quality of oat varieties.