论文部分内容阅读
通过对158头黑白花奶牛初期产奶性状和头胎305d产奶量的通径分析。依据p_(yi)值的大小和P_(yi)~2的分析,在初期产奶性状的四个自变量中选择X_4(90d产奶量)、x_1(最高日产奶量出现的天数),并剔除了X_2(最高日产奶量)、X_3(最高月产奶量)。以此建立了早期预测头胎305d产奶量的多元回归方程:Y=9.8722x_1+2.6433X_4-57.7250。检验表明此回归极显著,偏回归系数b_1、b_4亦均显著。在随机抽取63头奶牛的验证中,平均误差d=7.633％,说明该回归方程可在生产中试用。作者认为,通径系数和简单相关系数的不一致性表明,通径分析的方法,可以说明各性状间的彼此相互关系,是当前研究家畜性状相关规律较为完善的方法,查明奶牛产奶量与初期产奶性状的相关规律,对奶牛选择和育种具有重要意义。在这方面的研究目前仅限于用简单相关或净相关分析方法。本文则试图采用比较完善的通径分析方法,研究奶牛头胎产奶量(305d)与初期产奶性状的相关规律。分析初期产奶性状,以及各性状的直接影响、间接影响从而建立最优回归方程。为奶牛早期选种提供参考数据。
The path analysis of the initial milk production and the first 305 days milk production of 158 Holstein dairy cows was conducted. According to the analysis of the value of p_ (yi) and the analysis of P_ (yi) ~ 2, X_4 (90d milk production) and x_1 (the maximum number of days of daily milk production) were selected among the four independent variables of initial milk production traits Excluding the X_2 (the highest daily milk production), X_3 (the highest monthly milk production). Based on this, a multiple regression equation of early prediction of 305d milk production was established: Y = 9.8722x_1 + 2.6433X_4-57.7250. The test shows that the regression is very significant, partial regression coefficients b_1, b_4 are also significant. In a random sample of 63 cows, the average error, d = 7.633%, indicates that the regression equation can be tested in production. The author believes that the inconsistency between the path coefficient and the simple correlation coefficient shows that the path analysis method can explain the mutual relationship among the traits and is a perfect method to study the related laws of livestock traits. Relevant laws of initial milk traits are of great significance to the selection and breeding of dairy cows. Research in this area is currently limited to using simple correlation or net correlation analysis methods. This article attempts to use comparatively complete path analysis method to study the correlation between the first milk yield (305d) and the first milk production traits. Analysis of the early milk traits, as well as the direct impact of each trait, indirect impact to establish the optimal regression equation. Provide reference data for the early selection of dairy cows.