Crown rust (CR), caused by Puccinia coronata f. sp. avenae, is a major constraint to oat (Avena sativa L. ) production in the southern United States. We dissected the genetic architecture of CR resistance and assessed genomic prediction in 234 winter and facultative oat lines adapted to the southern United States. Using multi-environment phenotyping across five CR-prone sites and 8, 234 high-quality single-nucleotide polymorphisms, we conducted genome-wide association study (GWAS) and genomic prediction analyses. GWAS detected 13 significant loci, nine of which co-localized with reported quantitative trait loci or contained plausible resistance candidates such as leucine-rich repeat (LRR), nucleotide-binding site-LRR, and serine/threonine kinases. A stable locus on chromosome 3D was repeatedly identified in Baton Rouge, LA (2016) ; Castroville, TX (2016) ; Winnsboro, LA (2017) ; and in the combined analysis, indicating cross-environment consistency. Additional loci were supported on chromosome 4C (Citra and Quincy, FL, in 2017, and the combined analysis) and 2A (Winnsboro, LA, in 2017 and the combined analysis). Allele stacking was associated with large phenotypic gains. Lines with five or more favorable alleles showed >56% lower crust and 63% lower CrustPer than lines with none, supporting marker-assisted pyramiding. Genomic prediction models achieved high prediction accuracy. Parametric models, particularly BayesA and RRBLUP, performed strongly across cross-validation schemes; random forest was broadly comparable; and gradient boosting was lower. These results underscore the effectiveness of using GWAS and genomic prediction to enhance breeding for durable CR resistance in oat and provide a genomic framework for developing resilient cultivars suited to southern United States growing conditions.
Acharya et al. (Mon,) studied this question.