From research to application: evaluation of literature based and newly identified GWAS and GP-derived loci for anthracnose resistance in white lupin, across validation panels and environments

Summary

Resistance to anthracnose disease, caused by Colletotrichum lupini, in white lupin is polygenic and rare in global germplasm, making breeding for resistance challenging. Several studies have identified genomic regions associated with resistance, but reported QTLs rarely overlap. Overall, anthracnose-associated loci have not yet been appropriately exploited in breeding due to the lack of validation, especially in unrelated material. We assessed three literature-based SNP-markers associated to a previously described QTL, as well as six originating from two published GWAS studies. In addition, by re-analysing two published datasets, we newly (i) identified five SNP-markers through a genome-wide association study (GWAS) and (ii) assembled a genomic prediction (GP) model based on a set of 42 SNP-markers. All SNP-markers were transformed to PCR-based assays to facilitate cost-efficient application in breeding and tested in a diversity panel and two breeding panels, phenotyped in field and/or controlled conditions. We validated the association to anthracnose resistance for the previously reported QTL on chromosome Lalb_Chr10, tagging a source of resistance in the resistant cultivar 'Frieda'. Two markers showed significant associations (p < 0.05) with anthracnose resistance in at least two panels, and an additional seven markers were significant in one. The GP model had a prediction accuracy of 0.84 (± 0.13) in the training panel, but is not applicable as such in the unrelated validation panels. However, re-training the GP model with the respective validation panel data showed clear prediction improvement buffering the limitation of environmental context and narrower genepools applied in a breeding programme.