Multivariate classification of cotton cultivars tolerant to salt stress

ABSTRACT Two multivariate methods were adopted to classify salt-tolerant cotton genotypes based on their growth and physiological traits. The genotypes were cultivated in a greenhouse and subjected to 45 days of irrigation with saline water from the V4 phase onwards. Irrigation was performed with saline water with electrical conductivity (ECw) of 6.0 dS m-1. A factorial-randomized block design was adopted with nine cultivars, two treatments of ECw (0.6 as the control, and 6.0 dS m-1), and four replicates. Plants were evaluated for growth, gas exchange, and photosynthesis. The data were statistically analyzed using univariate and multivariate methods. For the latter, non-hierarchical (principal component, PC) and hierarchical (UPGMA) models were used for the classification of cultivars. Significant differences were found between cultivars based on univariate analyses, and the traits that differed statistically were used for multivariate analyses. Four groups were identified with the same composition in both the PC and UPGMA methods. Among them, one contained the cultivars BRS Seridó, BRS 286, FMT 705, and BRS Rubi, which were tolerant to salt stress imposed on the plants. Photosynthesis, transpiration, and stomatal conductance data were the main contributors to the classification of cultivars using the principal component method.

摘要 本研究采用两种多元统计方法，基于生长与生理性状对耐盐棉花基因型开展分类研究。供试基因型于温室中栽培，自V4生育期起以盐水灌溉，持续45天。灌溉所用盐水的电导率（electrical conductivity, ECw）为6.0 dS·m⁻¹。试验采用析因-随机区组设计，设置9个棉花品种、2个ECw处理（以0.6 dS·m⁻¹作为对照，另一处理为6.0 dS·m⁻¹），并设置4次重复。对植株的生长性状、气体交换参数及光合作用相关指标进行测定。采用单变量与多元统计方法对试验数据进行统计分析。针对多元分析环节，分别采用非层级模型（主成分分析，principal component, PC）与层级模型（UPGMA）对供试品种进行分类。单变量分析结果显示不同品种间存在显著统计学差异，因此将具有统计差异的性状纳入后续多元分析。主成分分析与UPGMA分析均鉴定出4个组群，且两组方法得到的组群组成完全一致。其中一组包含BRS Seridó、BRS 286、FMT 705及BRS Rubi四个品种，该组品种对试验施加的盐胁迫具有耐受性。在主成分分析用于品种分类的过程中，光合作用、蒸腾速率与气孔导度相关数据为主要贡献因子。