小麦多数量性状相关与回归分析

The Muli-quantity Characteristic Correlation and Regression Analysis on Wheat

本文从小麦新品种多数量性状分析,得到如下结果:株穗数和主穗粒数两性状对株粒重直接影响较大。生物产量,不实小穗数和株粒数3性状对株粒重直接负影响较大或影响很小,而通过其他性状的间接影响很大。百粒重对株粒重直接或间接有一定影响。穗长、株蘖数和株高对株粒重直接影响和间接负影响很小。从回归分析,7个显著性状建立的最优回归方程为:y′=0.520+0.097X_2+0.032X_3+0.120X_4-0.109X_5+0.377X_7+0.131X_8-0.302X_(10)。

The Multi-quantity Characteristic analysis on a new species of wheat we got the results as follow. Effective spike number per plant and grain number of a main spike, these two characteristics have a considercble direct effect on grain weight per plant;Living yield, nonbearing spikelet number and grand grain number per plant, these three characteristics have little direct effect but a larger direct negative effect, and passing over other characteristics the indirect effect is fairly large; Weight of 100 kernels has a cerlain direct of indirect negative effect on the grain weight per plant; Direct effect or indirect negative effects-of spike's length, nonbearing tillering number and plant height on grain weight per plant are considerably small. The optimum regression equation upon the seven characteristics is Y' = 0.520 + 0,097x2 + 0.032x3 + 0. l20x4- 0.109x5 + 0.377x7-0.131x8-0.302x10