威宁牛公牛体重与体尺性状的相关、通径分析及最优回归方程的建立

Correlation Analysis and Path Analysis and Establish of the Best Multiple Linear Regression Equation between Body Weight andBody Size Traits of the Bull of Weining Cattle

为了研究威宁牛公牛体重与体尺性状的相关关系,阐明体尺性状对体重的直接影响和间接影响,该试验使用2022年11月至2023年11月测定的29头24月龄以上、生长发育正常、健康且具有典型表型特征的威宁牛种公牛的体重和体尺数据,并对威宁牛种公牛体重与体尺性状进行相关分析、通径分析以及逐步回归分析,建立最优回归方程。结果表明:威宁牛公牛体重(y)与体高(x_(1))、体斜长(x_(2))、十字部高(x_(3))、胸围(x_(4))、腹围(x_(5))、管围(x_(6))以及各体尺性状间均呈极显著正相关(P<0.01),腹围(x_(5))与体重(y)的相关系数最大(0.9546)。腹围(x_(5))对体重(y)的直接影响最大,直接通径系数为0.6558,胸围(x_(4))通过腹围(x_(5))对体重(y)的间接影响最大,间接通径系数为0.5518。采用逐步回归法建立的最优回归方程为:y=-795.5392+2.4444x_(3)+1.6745x_(4)+2.8519x_(5),该模型的F值为180.98,决定系数R^(2)=0.9560(P<0.01)。可见,腹围、胸围和十字部高是影响威宁牛公牛体重的主要性状,后续育种中可将其作为威宁牛公牛体重选择的重要参考性状。

To study the correlation between body weight and body size traits of Weining bulls,and to clarify the direct and indirect effects of body size traits on body weight,the body weight and body size data of 29 Weining bulls over 24 months old,with normal growth and development,health and typical phenotypic characteristics were measured from November 2022 to November 2023.The correlation analysis,path analysis,and stepwise regression analysis of body weight and body size traits of Weining bulls were carried out to establish the optimal regression equation..The results showed that:there was a highly significant positive correlation (P<0.01) between body weight (y) and body height (x_(1)), body length(x_(2)), hip height (x_(3)), chest circumference (x_(4)), abdominal circumference (x_(5)), circumference of cannon bone(x_(6)), as well as various body size traits of the bull of Weining cattle. The correlation coefficient between abdominal circumference (x_(5)) and body weight (y) was the highest (0.9546). Abdominal circumference (x_(5)) has the greatest direct effect on body weight (y), with a direct path coefficient of 0.6558. Chest circumference (x_(4)) has the greatest indirect effect on body weight (y) through abdominal circumference (x_(5)), with an indirect path coefficientof 0.5518. The best multiple linear regression equation was established by stepwise regression method, and the equation is: y=-795.5392+2.4444x_(3)+1.6745x_(4)+2.8519x_(5). The F-value of the model is 180.98, and the determination coefficient R2=0.9560(P<0.01). It can be seen that abdominal circumference, chest circumference and cross height are the main traits affecting the weight of Weining bulls, which can be used as an important reference trait for weight selection of Weining bulls in subsequent breeding.