多元分析在黄淮冬麦区小麦良种区试中的应用分析

Analysis of Multiple Regression for Yield of Winter Wheat in Huang-Huai Plain

本文将多元分析法应用于小麦良种区域试验,并与Eberhart和Russell法及G.C.C.Tai(1971)法进行了比较。结果表明,多元分析法不仅提供了基因型各生育阶段对环境的反应敏感性强弱的参数V_1、V_2和V_3,而且提供了各环境地区的环境资源优劣的估计值r_1、r_2及r_3。据V_1r_1+V_2r_2+V_3r_3值最大来选各环境地区适宜品种,在生产上更具有实际指导意义。Tai(1975)多元分析法各基因型的复相关系数月均大于0.8,说明该分析是可靠的。并从生产实际中得到了验证。相比之下,Eberhart和Russell法及Tai(1971)法显得较粗放。

Twelve varieties of wheat were studied on the redomized block design of 4 replicates at sites in the Huang-Huai Plain during the 1986-1987 crop season. Three methods were used to study genotype-environment interaction in winter wheat yield. The. results of this study showed that the multiple correlation coefficients of the genotypes were bigger than 0.8. This and other data from production confirmed that this method was better than the Eberhart and Russell methods. This method of multiple regression nanlysis was considered tobe more accurate.