Phenotypic and genetic diversity in the foxtail millet (setaria italica (l.) p. beauv.) core collection)

Foxtail millet core collection consisting of 155 accessions was evaluated at three environments for 12 qualitative and 13 quantitative traits to study the phenotypic diversity and to identify trait specific accessions. Foxtail millet core collection was also molecularly profiled using 84 SSR markers to study molecular genetic diversity, population structure and to identify SSR markers associated with the agronomic traits. In REML analysis variance due to genotypes (σ2 g) and genotype × environment (σ2 ge) were significant for all the 13 quantitative traits. On the basis of phenotypic dissimilarity between pair of accessions, ten pairs of most diverse accessions were identified for breeding program for the developing high yielding cultivars with a broad genetic base and for the development of mapping populations. On the basis of pooled BLUPs (Best Linear Unbiased Predictors) of three environments, we have identified trait specific accessions for economically important traits such as yield and its traits contributing to yield (15 accessions for each trait). These accessions could be used in recombination breeding to develop cultivars with desirable combination of traits. The SSR markers detected a total of 1,356 alleles with an average of 16.14 alleles per locus. Of these, 368 were rare alleles; 906 common alleles; and 82 the most frequent alleles. Sixty one unique alleles which were specific to a particular accession and useful for germplasm identification were also detected. The genetic diversity of foxtail millet in this study was correlated well with racial classification and the race indica showed greater genetic distance from the maxima and moharia. Ten pairs of genetically most diverse accessions were identified. Large molecular variation observed in core collection could be utilized effectively for selection of diverse parents for breeding cultivars and development of mapping populations. Mantel test showed significant correlation between phenotypic and molecular dissimilarity matrix. The STRUCTURE analysis provided the evidence for the presence of four subpopulations. The mixed linear model (MLM) was used and the number of significant marker trait association was 130 in E1, 69 in E2 and 106 in E3 at P≤0.05, whereas only 49 in E1, 23 in E2 and 61 in E3 were found to be highly significant MTAs at P≤0.01. In pooled BLUPs of three environments, a total of 108 MTAs were detected at P≤0.05. Of these 18 SSR markers showed 37 significant associations at P≤0.01 with yield and yield contributing traits. Fifteen MTAs, that occurred in all three environments and overall in pooled data were identified as stable. Our research provided a first report of association study for yield and yield contributing traits in foxtail millet using SSR markers. The results from this research also demonstrated the use of core collection as association mapping panel to disclose marker-trait associations in foxtail millet for yield traits that could lead to effective utilization of ex-situ conserved genetic resources