玉米主要性状与产量的回归模型及相关分析

本研究以年150份玉米杂交组合试验数据为材料,采用逐步回归建立单穗粒重与各性状的线性回归方程,方程可靠程度为95.90%;各性状对穗粒重存在极显著正相关,相关程度依次为:百粒重>行粒数>穗长>穗位高>穗粗>穗行数>株高;通径分析结果:百粒重>行粒数>穗行数>穗长,均达极显著,表明影响玉米穗粒重主要性状是百粒重和行粒数,其次是穗行数和穗长。因此,选育高产玉米组合中应将百粒重、行粒数性状放在首位,并在此基础上对穗长、穗位高、株高进行合理选择,是玉米高产育种的有效方法之一。

The regression model and related analysis between main characters and yields in maize

In this study,related,path anaylsis and regression equation of main characters of 150 maize combinations were carried out.The regression results showed that,the reliability of linear equation was 95.90 %.Related analysis indicated that all characters considered in the paper were very significantly correlated to single ear yield.The impacts of these characters were,in order,from high to low,100-seed weight,kernels per row,ear length,ear height,ear rough,rows per ear and plant height.Path analysis showed that 100-seed weight and kernels per row were the important characters that impacted the yield greatly.Rows per ear and ear length were in second place.Therefore,one effective method of high yield breeding in corn was reasonably to select the ear length,ear height and plant height based on 100-seed weight and kernels per row.