甜高粱成熟期抗旱性指标筛选与评价方法

Screening and evaluation methods of drought resistance index of sweet sorghum at maturity

设置正常灌水处理(T1)、重度干旱处理(T2)2种处理,测定株高、穗重、单株粒重、千粒重和产量(单位面积产量)等性状,运用相关性分析、主成分分析、抗旱系数(DRC)、综合抗旱系数(CDRC)、加权抗旱系数(WDC)、隶属函数值(D)和聚类分析等方法,对成熟期不同甜高粱品系的抗旱性进行鉴定及指标筛选。结果表明:干旱胁迫对各性状均有显著影响(P<0.05);抗旱系数频次分析表明,各性状抗旱性敏感程度由大到小顺序为穗粒数>分枝数>株高>茎粗>千粒重>穗茎粗>穗重>穗长>单株粒重>产量;相关性分析表明:株高与穗重呈显著负相关(P<0.05),产量与穗粒数、穗重、单株粒重、千粒重呈极显著正相关(P<0.01),穗粒数与穗重、单株粒重呈极显著正相关(P<0.01),单株粒重与千粒重呈极显著正相关(P<0.01);通过主成分分析,将10个性状转化成5项新的互相独立综合指标,可代替88.087%原始所有性状抗旱信息;灰色关联度分析表明,各性状DRC与D的密切程度大小顺序为单株粒重>千粒重>穗粒数>产量>穗茎粗>分枝数>穗重>穗长>株高>茎粗。对D值进行聚类分析,将22个不同品系甜高粱分为5个抗旱级别,其中Ⅰ级3份、Ⅱ级7份、Ⅲ级6份、Ⅳ级4份、Ⅴ级2份。对聚类分析结果的抗旱系数进行抗旱性等级统计分析表明,除株高、茎粗和穗长以外,千粒重、单株粒重、穗重、穗粒数、D值、分枝数、穗茎粗、CDRC值和WDC值随着抗旱级别的升高而增加。逐步线性回归方程表明,千粒重、单株粒重、穗粒数、穗茎粗与D值密切相关。综上所述,干旱胁迫对甜高粱成熟期各性状均会产生显著影响,千粒重、单株粒重、穗粒数和穗茎粗可作为甜高粱成熟期直观、准确和简单的抗旱鉴定评价指标,筛选出成熟期抗旱性强的甜高粱品系为F417、F438、F6137。

In order to screen out the drought-tolerant sweet sorghum cultivars(lines) and identification indicators, an experiment was conducted using 22 sweet sorghum cultivars(lines), normal irrigation(T1), severe drought(T2),plant height, spike weight, spike weight per plant, 1000-grain weight, and yield(yield per unit area) were measured. By using correlation analysis, principal component analysis, drought resistance coefficient(DRC),comprehensive drought resistance coefficient(CDRC value), weighted drought resistance coefficient(WDC value), membership function value(D value), cluster analysis, and other analytical methods, we did identification and index screening of drought resistance of different sweet sorghum varieties at maturity. The results showed that drought effected indicators remarkably(P>0.05). The frequency analysis of drought resistance coefficient showed that the drought resistance of each index was in the order of the grain number per spike>number of branches>plant height>culm diameter>1000-grain weight>spike culm diameter>spike weight>spike length>spike weight>yield. Correlation analysis showed that plant height was significantly negatively correlated with ear weight(P<0.05);yield was extremely significantly positively correlated with ear grain number, ear weight, kernel weight per plant, and thousand kernel weight(P<0.01);Ear weight and grain weight per plant were extremely significantly positively correlated(P<0.01);ear weight was extremely significantly positively correlated with grain number per ear(P<0.01);grain weight per plant was extremely significantly positively correlated with thousand-grain weight(P<0.01). By using principal component analysis, those indicators were transformed into 5 new independent comprehensive indicators, which could replace 88.087% information of all the original indicators. The grey correlation analysis showed that the correlation level of indicators’ DRC and D values ranked as spike weight per plant>1000-grain weight>grain number per spike>yield>spike culm diameter>branch of number>spike weight>spike length>plant height>culm diameter. According to D value clustering analysis, tested sorghum materials were divided into 5 drought resistance grades, 3 belonged to grade I, 7 belonged to grade II, 6 belonged to grade III, 4 belonged to grade IV, and 2 belonged to grade V. The result of cluster analysis showed that except for the plant height, culm diameter, spike length, the 1000-grain weight, spike weight per plant, spike weight, grain number per spike, D value, number of branches, spike culm diameter, CDRC value, and WDC value went up following the index of drought tolerance. Stepwise linear regression indicated that the 1000-grain weight, spike weight per plant, grain number per spike, spike culm diameter and D value were strongly linked. Drought stress had significant effects on various indicators of sweet sorghum maturity. The 1000-grain weight, spike weight per plant, grain number per spike, and spike culm diameter were regarded as visual, precise, and simple evaluation indexes of drought-tolerance. F417, F438 and F6137 were the most resistant sweet sorghum cultivar(lines).