多元线性回归与相关分析在棉种SGK791高产栽培研究中的应用

为了定量研究棉花新品种SGK791产量构成因素的变化规律,以明确其高产栽培的最佳途径和主攻目标。在2003~2004年河南省棉花品种区域试验及2005年生产试验数据的基础上,应用多元线性回归、偏相关分析和通径分析方法分析SGK791产量构成因素的变化规律及各因素对产量的作用。4个产量因素与产量呈线性关系,建立了产量与产量因素间回归方程和最优回归模型。4个产量因素影响皮棉产量的综合作用为:每株铃数>衣分>单铃重>密度。皮棉产量构成因素间存在制约与促进的数量关系。4个产量因素中,每株铃数对产量的作用最大,其次是密度和衣分,铃重最小。在合理密植的基础上,SGK791高产栽培应以力争株铃数、兼顾铃重、稳定衣分为主攻目标。

Application of Multivariate Linear Regression and Correlation Analysis in the Research of High-yield Cultivation in New Cotton Variety SGK 791

The research aimed to study the change laws of yield components of new cotton variety SGK791 quantitatively and confirm the best approach and principal target for its high-yield cultivation.On basis of the data from Henan province regional test of cotton varieties in 2003～2004 and production test in 2005,the multivariate linear regression,partial correlation analysis and path analysis were used to analyze the change laws of yield components of SGK791 and the effect of its components on yield.4 yield components had linear relationships with yield,the regression equation and the optimum regression model between yield and its components were F=53.47** and F=7.368** respectively.The comprehensive effects of 4 yield components on lint yield were as follows: boll number per plant>lint percentage>single boll weight>density.Restrictive and stimulative quantity relationships existed among lint yield components.Among 4 yield components,the effect of boll number per plant was most,the next were density and lint percentage,boll weight was least.Based on reasonable close-planting,trying for more boll number per plant and boll weight and stabilizing lint percentage should be principal target in high-yield cultivation of SGK791.