南疆海岛棉53份预试品系早熟丰产优质性状的因子-聚类分析

旨在为海岛棉新品种（系）综合评价、推广、利用以及早熟、丰产、优质亲本的选择和制定性状改良目标提供参考。利用2013—2014 年新疆维吾尔自治区海岛棉53 个预备试验品种（系）为材料，对供试材料20 个性状进行变异性和相关性的研究，并进行因子分析和聚类分析。结果表明：海岛棉品种（系）早熟、丰产、优质性状中变异系数最大的为始节位、皮棉、伸长率，变异系数最小的是霜前花率、霜前皮棉、整齐度指数。测试指标之间存在错综复杂的关系。主成分分析表明前7 个主成分的方差累积贡献率达87.25%，能够反映绝大部分性状信息；聚类分析将53 个品种（系）分为4 大类；在聚类分析的基础上对性状进行多元方差分析和综合评价，并提出性状改良方向。

Factor-Cluster Analysis of Early-Maturity, High-Yield and High-Quality Traits of Gossypium barbadense L. Test Lines in South Xinjiang

For reasonable disposal of Gossypium barbadense L. Production, comprehensive assessment ofSnewSvarieties (lines), promote, utilize, high-yield and high-effect cultivation, to provide useful reference for selections of early maturity, high yield and high-gualitySparent, Set goals for character improvement. This experiment mainly used 53Sreserve experimental varieties (lines) in 2013-2014 as materials, 20Straits of experimental materials were used for variability, correlation, factorSanalysis and cluster analysis. The results showed thatSfirst flower node, lint yield and elongation of Gossypium barbadense L. have the higherScoefficients variation, pre-frost yield, pre-frost lint cotton and uniformity Index were lower.SThe complex relationshipSexistSbetween the testSindexes. Principal component analysis indicated that the cumulative variance proportion during the first seven principalScomponents was 87.25%, could reflect the vast majority of trait information. The cluster analysis classify the 53 varieties (lines) into 4 Stypes. multivariateSanalysisSofSvariance and comprehensiveSassessment conducted on the foundation of cluster analysis, and improvement direction is presented. The result provides theoretical basis for parent selection in Gossypium barbadense L. Breeding.