Genetic Diversity among landraces and commercial oats and association of SSR markers with resistance to biotic and abiotic stresses

Abstract

Genetic diversity studies are an important tool to be exploited by breeding programs. However, one of the factors that have limited their success in applying results to breeding is that most genomic studies have been conducted in populations derived from a bi-parental cross and hence, the identified markers often turn out to be unique to a specific genetic background. Association analysis which exploits the variation in a collection of genetically diverse materials to uncover a significant association between a trait and a gene or molecular marker on the basis of linkage disequilibrium has emerged as a promising and valuable tool to exhaustively identify QTL in plants. In this study, genetic diversity among 176 Avena accessions (A. sativa, A. byzantina, A. strigosa), was studied using 31 simple sequence repeat (SSR) markers. SSRs were chosen following preliminary studies which indicated that they were highly polymorphic and were well distributed throughout the oat genome. Polymorphic Information Content (PIC) ranged from 0.46 to 0.96, with an average of 0.80 indicating the usefulness of many of the SSR for genotype identification. A total of 455 alleles were detected with an average of 14.22 alleles per locus. This was sufficient to discriminate all the accessions analysed. UPGMA clustering grouped the genotypes into four clusters with genetic similarity of 0.43. Cluster 1 corresponded to A. sativa commercial varieties, cluster 2 to A. byzantina landraces and commercial genotypes and cluster 3 and 4 to the different A. sativa landraces. 176 alleles were unique to accessions in clusters 2, 3 and 4 indicating the considerable genetic variation that exists in landraces that is not present in the commercial accessions. Principal component analysis (PCA) and the software STRUCTURE 2.3 based on Bayesian clustering approach were also applied as alternative methods to infer population structure with similar results. Association analysis between the marker alleles and scores assigned for resistance to powdery mildew, rust and drought to the different genotypes was conducted using TASSEL 2.1 software. Four different general and mixed linear models combining structure coefficients, PCA axis or kinship matrix and three different statistical tests, an F test for specific markers, a permutation-based test, and the experiment- wise p value that controls the error-rate for the entire set of hypotheses were carried out to reduced false positive and increase the power of the analysis. According to this, 6 marker alleles were significantly associated with rust resistance, 2 associated to powdery mildew resistance and 2 associated to drought tolerance.