玉米种质资源的抗倒伏性评价及鉴定指标筛选

Evaluation of Lodging Resistance and Selection of Identification Indexes of Maize Germplasm Resources

倒伏是影响玉米高产稳产的主要问题之一。本研究以181份玉米自交系为试验材料,对22个与抗倒伏性紧密相关的性状进行测定,采用相关性分析、主成分分析、聚类分析、逐步判别分析和岭回归分析等方法,综合评价玉米自交系的抗倒伏性。相关性分析的结果表明,22个性状间存在不同程度的相关性。主成分分析的结果表明,前7个主成分代表了181份玉米自交系的22个性状74.460%的信息,其贡献率分别为25.700%、12.369%、9.782%、8.159%、7.782%、5.490%和5.177%。聚类分析的结果表明,当遗传距离为3.5时,181份玉米自交系被聚为4大类,第Ⅰ类群有35份自交系、第Ⅱ类群有47份自交系、第Ⅲ类群有49份自交系、第Ⅳ类群有50份自交系。逐步判别分析的结果表明,175份玉米自交系被正确判别,判对概率为96.69%;6份玉米自交系被误判,误判率为3.31%,这说明聚类分析的结果是准确可靠的。进一步利用岭回归方法筛选出半纤维素含量、穿刺强度、第3节间长度、第3节间直径、穗位高和维管束总面积6个性状,建立了可靠的玉米自交系抗倒伏性评价回归模型。在181份玉米自交系中,以AHU24、e220、7026B等35份自交系的抗倒伏能力最强。本研究结果为玉米抗倒伏种质资源的遗传改良和组配抗倒伏杂交种提供借鉴。

Lodging is one of the main problems affecting the maize production and extension.In this study,181 maize inbred lines were used for determining 22 characters associating to lodging resistance.Statistical analysis using the correlation analysis,the principal component analysis,the cluster analysis,the stepwise discriminate analysis and the ridge regression analysis was deployed for data interpretation.The results of the correlation analysis showed that 22 characters existed in varying degrees of correlation.The first seven principal components explained 74.460%of the phenotypic variation with the contribution rates of 25.700%,12.369%,9.782%,8.159%,7.782%,5.490% and 5.177%.Four categories were clustered in the 181 maize inbred lines at a genetic distance of 3.5.Among them,the first classⅠhas 35 inbred lines,the classⅡhas 47 inbred lines,the classⅢhas 49 inbred lines,the classⅣhas 50 inbred lines.The result of stepwise discriminant analysis showed that 175 maize inbred lines were correctly discriminated and the identification rate was 96.69%.Six maize inbred lines were incorrectly discriminated and the identification rate was 3.31%,which means that the result of cluster analysis is accurate and reliable.The hemicellulose content,rind penetrometer strength,third internode length,third internode diameter,ear height and area of total vascular bundles were selected and the regression model of lodging resistance of maize inbred line was established by using the ridge regression method.35 inbred lines,such as AHU 24,e220 and 7026B,showed the highest lodging resistance.Thus,the results provided a reference of the maize lodging resistant germplasm resources,which become valuable in selection of elite germplasm resources in breeding for the lodging resistance hybrids.