Principal component analysis and cluster analysis of soybean yield and agronomic traits during winter nursing in Hainan Island

To determine the effect of key factors on soybean yield during winter nursing and screen out soybean materials for high-yielding，150 soybean varieties introduced from Japan，America and other countries were planted and researched by principal component analysis（ PCA） and hierarchical cluster analysis（ HCA） in Sanya，Hainan， China. The results showed that single plant yield of soybean was significantly related to 7 traits that were pods per plant，seeds per plant，plant mass，valid branch number，100-seed weight，growth period，plant height. Seven indicators reflecting soybean varieties might be represented by three principal components（ cumulative contribution rate of 86.640%）. The research results can offer reference for the use of soybean materials. Seven major agronomic characters related to single plant yield were sequenced: pods per plant ＞ seeds per plant ＞ plant mass ＞ valid branch number ＞ 100-seed weight ＞ growth period ＞ plant height，the seven traits that were related to yield were classified into 3 categories of principal components which were yield characters factor，mature period factor and branches factor. Selecting varieties with high seeds per plant，high valid branch number and high 100-seed weight was necessary in high yield breeding of soybean. 150 soybean varieties were classified into 9 clusters，which had obvious feature，that could guide high yield breeding of soybean.