【摘要】 对河南省2006及2007年15个春棉区试品种的植株、产量因素和纤维品质共17个数量性状进行主成分分析,发现前4个主成分对所考察性状变异的累计贡献率达90.10%,根据各品种的主成分得分进行聚类分析,将15个参试品种从产量和主要品质的相似度聚为5大类,并对聚类结果进行判别分析验证,误判率为13.3%。根据对各类品种的产量和品质性状的多元方差分析,进行了各类品种的综合评价与性状改良目标针对性的探讨。更多还原

【Abstract】 Principal component analysis method was applied to study 17 quantitative characters of 15 cotton varieties for spring-sowing zones in He’nan Province in 2006 and 2007,including agronomic characters of plant,yield factors and fiber quality traits.The result showed that the cumulative variance contribution of the first four principal components was 90.19%.Cluster analysis was carried out according to the principal components’ scores of each variety.Fifteen tested varieties were clustered into five groups on the similarities of yield and major characters.Discriminate analysis of the cluster result was carried out,and the erroneous discriminate rate was only 13%.Special discussion had been carried out on the comprehensive evaluation on each variety and the improving targets of characters according to the result of multivariate analysis and multi-comparison of the yield and quality characters of each variety.更多还原