北疆杂交棉主要性状相关性及主成分分析

以北疆主推的14个杂交棉品种为供试材料,采用多年多点栽培试验,对17个性状的相关性及主成分进行分析。结果表明,籽棉产量和皮棉产量与现蕾期、吐絮期、生育期、始节高、黄萎病病指极显著负相关,与开花期显著负相关,并且籽棉产量与铃质量、株高显著正相关,皮棉产量与籽棉产量极显著正相关;黄萎病病指与现蕾期、生育期极显著正相关,与吐絮期、始节高显著正相关;始节高与现蕾期、吐絮期、生育期极显著正相关。产量性状因子、铃部性状因子、枯萎病因子、衣分因子和开花期因子5个主成分的特征向量累积贡献率达87.469%;根据主成分因子得分,可将供试材料聚为3类,其中金垦杂1064、七杂5号、新陆早43号的综合表现较好。

Correlation and Principal Component Analysis of Main Characters of Hybrid Cotton in Northern Xinjiang

Selected 14 mainly planted hybrid cotton varieties in northern Xinjiang for test material, multiple years and multiple locations cultivation test was employed to analyze the correlation and principal component of 17 main characters. The results indicated that yield and lint yield were negatively relative to squaring period, boll opening period, growth period, first fruit branch height, and verticillium index at 0.01 level, respectively, and were negatively relative to flowering period at 0.05 level. Yield was positive relative to boll mass and plant height at 0.05 level, lint yield was positive relative to yield at 0.01 level. Verticillium index was positive relative to squaring period and growth period at 0.01 level, and was positive relative to boll opening period and first fruit branch height at 0.05 level. First fruit branch height was positive relative to squaring period, boll opening period, and growth period at 0.01 level, respectively. Principal component analysis indicated that feature vector cumulative contribution rate of 5 principal components of yield characters factor, boll characters factor, wilt factor, lint percentage, and flowering period factor was 87.469%. Based on principal components scores, tested material could be clustered into three categories, and Jinkenza 1064, Qiza 5, Xinluzao 43 were superior varieties with good comprehensive performance.