陆地棉棉铃组分性状的基因型差异及与产量的关系

为了探明陆地棉不同产量潜力品种棉铃性状的特点，于2008—2009年比较了3个大铃基因型与3个小铃基因型棉铃组分性状的差异。结果表明，大铃基因型的单铃纤维重、单铃种子重及单铃种子数显著高于小铃基因型，单粒种子重和单粒种子纤维重差异不显著。简单相关分析表明，单铃种子数与公顷皮棉产量呈极显著正相关(0.660**)，但偏相关系数或直接通径系数较小或为负值(0.179，-0.052)；而单株种子数与公顷皮棉产量的简单相关系数、偏相关系数和直接通径系数均为显著或极显著的正值(0.771*，0.650**，0.917**)。因此，单株种子数可作为皮棉产量选择的可靠指标。单粒种子纤维重和铃壳率受基因型效应的显著影响，受年份效应和年份×基因型互作效应的影响均不显著。说明单粒种子纤维重和铃壳率主要受遗传因素的影响，受环境因素的影响较小。

Genotypic Differences in Within-Boll Yield Components and Their Correlations with Lint Yield of Upland Cotton

Cotton fibers origin from the outermost layer cells of ovule epidermis. The quantity, quality and surface area of cotton seed have effects on cotton lint yield. However, little is known about how within-boll yield components differ between big-boll-weight genotype and small-boll-weight genotype and their relationship of lint yield. Three big-boll-weight genotypes and three small-boll-weight genotypes were field-tested at agricultural experiment station of Jiangxi Agricultural University in 2008 and 2009. Lint yield, yield components and cotton boll characters were evaluated in a randomized completed block design with three replications. The big-boll-weight genotypes had significantly higher lint mass per boll and seed mass per boll than the small-boll-weight genotypes. The seed number per boll of the former was significantly higher than that of the latter. Seed mass per seed did not show any significant differences among those genotypes investigated. Lint mass per seed did not consistently respond to boll weight. It was concluded that the big-boll-weight genotypes in this investigation were characterized by more seeds per boll. The pearson correlation coefficient between seed number per boll and lint yield per hm2 of land area was positive and significant (0.660**), but the partial correlation coefficient was rather small and the direct path coefficient remained negative (0.179, -0.052). The simple correlation, partial correlation and path coefficient analysis revealed that seed number per plant was positively and significantly correlated with lint yield per hm2 of land area. These results indicated that seed number per plant could be served as a reliable character for cotton yield improvement. Lint mass per seed and boll shell ratio were mainly controlled by genetic main effects. Non-significant environmental effects and genotype × environment interaction effects were detected for them.