多變量 解釋法에 依한 참깨의 品種群 分類

Varietal Classification on the Basis of Multivariate Analysis in Sesame (Sesamum Indicum L.)

The purpose of this experiment was to classify the varietal group and to help to choose parents for hybridization in sesame varieties. An analysis of genetic divergence as measured by Mahalanobis’ D² statistics in 1225 possible comparisons among 50 varieties of sesame for ten characters and an average link cluster analysis based on the D²’s were attempted. Classification was carried out by using principal component analysis and clustering by taxonomic distances calculated with the scores of the first six principal components. Fourteen varietal groups were classified by the average link cluster analysis based on the D² computed in ten characters. The averages of intra-group D² except for varietal group IV, VII, XIII were lower than those of inter-group D², and yield per plant and oleic acid content among the ten characters were the greatest contributorsin the inter-groups The results of the classification by the Mahalanobis’ D² were nearly coincident with those of the classification by taxonomic distance calculated with the scores of the first six principal components. In the results of principal component analysis for ten characters, 75.07 % and 83.7 % of total variation could be appreciated by the first five principal components and the first six principal components, respectively. The first principal component was highly loaded with linoleic acid content and oleic acid content, the second principal component with yield per plant, 1000-grain weight and no. of branch, and the third component with plant height and stem diameter.