MICROSATELLITE VARIATION AND GRAIN YIELD PERFORMANCE OF FINGER MILLET (ELEUSINE CORACANA SUBSP. CORACANA) GENOTYPES EVALUATED UNDER MOISTURE STRESS ENVIRONMENT

Aims: To assess the genetic variation and grain yield performance of finger millet genotypes and, identify the available breeding opportunities for moisture stress environment.
Study Design: Randomized complete block design.
Place and Duration of Study: Field experiment was conducted at Dera Agricultural Research Station (Ethiopia) (1677 m a.s.l., 08º 19" 06.3' N and 039º 19" 0.74' E) during 2012 and 2013. DNA extraction and genotyping analysis were conducted at International Crop Research Institute for Semi Arid Tropics (ICRISAT), Nairobi.
Methodology: Nineteen finger millet genotypes and 2 improved varieties were evaluated for drought tolerance under field condition and genotyped using a set of 20 SSR markers. Field data were recorded as per finger millet descriptor guide. Analysis of variance and Pearson correlation coefficient were computed using SAS computer software. A dendrogram was built by the average linkage method of squared Euclidian distance and principal component analyses were computed using MINITAB 14 statistical package. Allelic data such as polymorphic information content (PIC), observed heterozygosity, major allele frequency, genetic distance between tested genotypes and classical F-statistics were calculated using Power Marker ver. 3.25 software. The weighted neighbor joining based clustering and relative position of genotypes on the principal component axis (PCA) was computed using DARwin v.5 software.
Results: Mean squares due to genotypes were highly significant (P=.01) for days to 50% heading, days to 50% maturity, plant height, grain yield per plant and percentage relative water content (%RWC). The principal component analysis (PCA) showed that the variations observed among the genotypes were attributed to all phenotypic traits, and considerable percentage of the variation (45.5%) among genotypes was explained by the first PCA. Pearson correlation coefficient analysis revealed most of the polygenic traits considered in the study showed positive association to grain yield per plant, but days to 50% heading, days to 50% maturity and total tiller number were negatively correlated to grain yield. The SSR markers displayed variable genetic polymorphism with an average polymorphic information content of 0.57. A total of 130 alleles were observed with an average of 6.191 alleles per locus. Weighted neighbor-joining based clustering grouped 21 finger millet materials in to three major clusters from the main node. Allelic frequency based genetic distance measure depicted relatively narrow allelic distance between Acc. AAUFM-50 and AAUFM-7 (0.33), and between AAUFM-38 and AAUFM-39 (0.33).
Conclusion: Phenotypic and SSR data based analysis in the present study revealed wider variation among finger millet genotypes, indicating the possibility to use the best performing genotypes as parental line in hybridization program and also for selection breeding. Acc.AAUFM-24 (32.49g), AAUFM-39 (29.54g) and AAUFM-7 (27.03g) were the three top yielding genotypes with more than 10% yield advantage over the standard check Tadesse (21.56g) and thus could be recommended for further evaluation and verification under moisture stress environment.