主成分分析 및 Cluster 分析을 利用한 油菜品種의 分類

Classification of Rapeseed Cultivars by the Principal Component Analysis and Cluster Analyses

Classification of Korean local and introduced rapeseed cultivars(Brassica napus L.) was carried out by using the principal component analysis and cluster analyses. Twenty six agronomic characters can be classified into two groups; one closely related to yield components, oil content, and undesirable fatty acids, the other related to maturity, plant height, and desirable fatty acids by cluster analysis with correlation coefficients among them. By using principal component analysis and clustering by taxonomic distances calculated with the first six principal components’ score of varieties, fifty six rapeseed cultivars were classified into five varietal groups. Meanwhile Korean rapeseed cultivars except “Woldongcho” were included in Group I, Japanese varieties were distributed into all varietal groups but most of them mainly including in Group I or Group IV. Most of Taiwan, North European, and Central European varieties were mainly distributed into Group V, Group I or V, and Group Ⅱ or Ⅲ, respectively. Also, these rapeseed cultivars can be classified into seven varietal groups with weighted variable group method (WVGM) using correlation coefficients between each varieties and into six varietal groups with weighted pair group method (WPGM) based on the Euclidean distances calculated from standardized data of quantitative agronomic characters. Results of the classification of rapeseed cultivars by WVGM were nearly coincident with the results by WPGM, but the results of classification using principal components analysis were slightly different from the results of the others.