【摘要】 为筛选高产优质的棉花杂交组合，以15个陆地棉推广品种为亲本配置的192个F₁杂交种为试验材料，通过相关性分析、主成分分析和回归分析探讨主要农艺性状、产量和纤维品质之间的关系。结果表明，单铃重、籽棉产量和皮棉产量等产量指标都与始节数呈负相关，棉花纤维品质中上半部平均长度和整齐度这2个指标与主要农艺性状间的相互作用是正向的，马克隆值和纤维成熟度与主要农艺性状间具有反向相互作用。对16个指标降维后提取出5个主成分，分别命名为成铃因子、产量因子、品质因子、生长因子和衣分因子，共解释群体81.63%的变异。分别以产量性状和品质性状为因变量，以主要农艺性状为自变量构建了8个最佳回归模型，为适应不同需求的快速准确选择群体材料提供依据。研究结果为棉花品种选育提供了理论依据及材料支持。更多还原

【Abstract】 To screen high yield and high quality cotton hybrid combinations, 192 F₁ hybrids with 15 experimental materials Gossypium hirsutum L. cultivars as parental configurations were used as experimental materials.Correlation, principal component and regression analyses were used to explore the relationships among main agronomic traits, yield and fiber quality. The results showed that yield indexes such as single boll weight, seed cotton yield and lint cotton yield were negatively correlated with the number of the first branch. The fibre upper half mean length and fiber uniformity fiber of cotton fiber quality had a positive interaction with the main agronomic traits, while the macronaire value and fiber maturity had a negative interaction with the main agronomic traits. After dimensionality reduction of 16 indexes, 5 principal components were extracted, which were named as bolling factor, yield factor, quality factor, growth factor and clothing factor, explaining 81.63% of the population variation. 8 optimal regression models were constructed with yield and quality traits as dependent variables and main agronomic traits as independent variables to provide a basis for rapid and accurate selection of population materials to meet different needs. Above results provided theoretical basis and material support for cotton variety breeding.更多还原