内蒙古绒山羊早期生长性状协方差函数估计的研究.docVIP

内蒙古绒山羊早期生长性状协方差函数估计的研究 　　摘要：为了研究内蒙古绒山羊生长发育的特点，试验采用内蒙古绒山羊1998～2011年的体重记录（出生重到2岁体重）资料。通过建立随机回归模型，采用单变量分析，嵌套3阶legendre模型，筛选固定效应。固定效应包括群、出生类型、性别，出生年和测定年。通过综合比较对数似然函数值、残差方差、AIC和BIC信息标准，选择拟合生长性状的固定效应。最后分别对固定效应和随机效应（加性遗传效应和永久环境效应）配合不同阶数的Legendre多项式，应用似然比及残差方差等作为综合评价指标，选择用于生长性状遗传评估的最优模型。研究结果表明，群、出生类型、性别、出生年和测定年均对所有体重影响极显著；生长性状最优遗传评估模型为LP8-8-7。 　　关键词：内蒙古绒山羊；随机回归模型；协方差函数；Legendre多项式 　　中图分类号：S827；S813 文献标识码：A 文章编号：0439-8114（2016）20-5304-06 　　DOI：10.14088/j.cnki.issn0439-8114.2016.20.034 　　Abstract： In order to study the growth characteristics of Inner Mongolia White Cashmere goats，the data from 1998 to 2011 weight performance records of Inner Mongolia White Cashmere goats（weight from birth to age of 2 years） adopted to establish the random regression model. The fixed effects were screened by using univariate analysis nested third-order Legendre polynomial and included groups，types of birth，gender，birth years and measure years. By comprehensive comparison of logarithmic likelihood ratio，residual variance，AIC and BIC information，the fixed effects fitted growth traits were selected. Through fitting the fixed and random effects（additive genetic effects and permanent environmental effects） with different orders of Legendre polynomial，the optimal genetic evaluation model for growth traits was screened by using the comprehensive evaluation of likelihood ratio and residual variance from different model. The results showed that the fixed effects such as groups，types of birth，gender，birth years and measure years all had extremely significant effects on the weights of Inner Mongolia White Cashmere goats. The optimal genetic evaluation model for growth traits were LP8-8-7（LP8-7）. 　　Key words： Inner Mongolia white cashmere goats； random regression models； covariance function； legendre polynomials 　　?S机回归模型（Random regression model，RRM）最早是由Henderson[1]提出，Schaeffer等[2]最早应用随机回归模型对奶牛育种中测定日产奶量进行了分析。如今，随机回归模型已成为家畜育种改良实施计划中对经度数据资料进行遗传研究的一个有效方法[3]。Jamrozik等[4，5]对荷斯坦奶牛第一泌乳期产奶量应用基于随机回归分析的测定日模型