Identification of genetic diversity and characterization of world sesame collection for Agro-Morfological and quality characters

This study was conducted to characterize 345 sesame accessions representing worldwide sesame collection and to develop a core collection. The associations among yield, yield components and oil content were also determined. Materials were grown in randomized complete blocks design with two replications in 2008 and 2009. 12 qualitative; stem hairiness, stem shape, stem flattening, stem branching, leaf hairiness, leaf arrangement, number of flowers per leaf axil, number of carpels per capsule, capsule hairiness, capsule dehiscence at ripening, corolla color, seed color and 9 quantitative characters; days to flowering date, days to 50% flowering, stem length to the first capsule, plant height, number of branches, number of capsules per plant, number of seeds per capsule, 1000 seed weight and seed yield traits were used to characterize the material. Some accessions which have better seed yield, early maturing, more branches and capsules, yellow and brown seed color, purple stem and capsule color, hairiness, non-shattering capsule and determinate growth habit were also identified and argued their potential in crop improvement programs. Relationship between yield and yield components were also determined by simple correlation coefficients and path analysis. According to path analysis, plant height showed the highest direct effect on seed yield. Number of branches and number of capsule per plant had the highest indirect effect over plant height on seed yield. Multivariate analyses of the whole collection revealed wide variation, which was mainly influenced by days to maturity, stem length to the first capsule, plant height and number of branches. A core collection has also been developed and it is highly useful for DNA-based and fatty acid composition studies as core collection allows working on less number of entries. The comparisons between core and whole collection according to qualitative and quantitative descriptors indicated that core collection sampling was optimal and represented well to whole collection.