用随机回归模型估计宁夏地区荷斯坦牛头胎测定日产奶量遗传参数  

Genetic Parameters Estimation of Test Day Milk Yield in Holstein Heifers in Ningxia Using a Random Regression Test-Day Model

作　　者：任小丽[1,2] 刘澳星 李想[1] 张旭[1] 王雅春[1] 邵怀峰 秦春华 王瑜 温万 张胜利[1] 

机构地区：[1]中国农业大学动物科技学院,北京100193 [2]河南省奶牛生产性能测定中心,郑州450046 [3]宁夏回族自治区畜牧工作站,银川750001 [4]宁夏四正生物技术工程研究中心,银川750200

出　　处：《中国农业科学》2017年第10期1885-1892,共8页Scientia Agricultura Sinica

基　　金：宁夏自治区农业育种专项-<优质高产奶牛育种>项目;现代农业(奶牛)产业技术体系建设专项资金(CARS-37)

摘　　要：【目的】通过拟合泌乳曲线对影响宁夏地区荷斯坦牛测定日产奶量的诸多因素进行分析,并采用随机回归测定日模型估计遗传参数,为后续估计乳成分和体细胞数遗传参数及个体育种值估计奠定基础,同时也为制定适合该群体的优化育种方案提供参数。【方法】收集了2009—2013年间,宁夏地区24个牛场38 592头荷斯坦牛原始DHI数据,共包含550 078条测定日产奶量记录。质控标准为:产犊月龄为22—36 mo、泌乳天数为5—305 d及日产奶量为5.9—53.3 kg。质控后,共得到14 320头母牛的127 478条测定日记录(占原始记录的23.14%)。将有表型的母牛至少向上追溯三代(父母、祖父母、外祖父母),得到遗传评估所用的系谱文件,共包含24 272头奶牛。使用Excel 2013计算各测定日的平均产奶量,利用SAS 9.1软件的非线性回归过程(NLIN)结合Wood模型拟合群体水平的泌乳曲线。采用随机回归测定日动物模型,通过DMU 5.2软件对头胎产奶量进行遗传参数估计。随机回归测定日模型中包括了一般固定效应、固定回归项及随机回归项三类效应。研究中使用场-测定日作为固定效应、产犊年季效应作为固定回归项,并将加性遗传效应和永久环境效应作为随机回归项。测定日子模型采用4阶勒项德多项式(Legendre polynomials)。根据宁夏地区气候特点,将产犊季节划分为春(3月11日至5月20日)、夏(5月21日至8月25日)、秋(8月26日至10月15日)和冬(10月16日至3月10日)四季。【结果】宁夏地区荷斯坦牛头胎测定日产奶量群体均值为29.66 kg,在泌乳90 d时达到泌乳高峰,高峰产奶量为31.84 kg。通过不同产犊年份、产犊季节和牧场效应组合下的泌乳曲线拟合效果分析,以上因素通过影响泌乳曲线拟合效果进而影响头胎产奶量。在使用测定日模型估计头胎遗传参数时均应通过子模型拟合考虑其泌乳曲线的差异,在产奶量遗传评估模�[Objective] This experiment was conducted to quantify the effect of environmental factors on milk yield of first lactation Holsteins in Ningxia, and to estimate genetic parameters of test day milk yield using test-day model, to provide a theoretical foundation for genetic parameter and breeding value estimation of milk components and somatic cells, and provide basic parameters for optimal breeding scheme which is suited to the Holstein population in Ningxia. [Method] A total of 550 078 test-day milk yield records from Holstein in Ningxia were collected, and the standards of calving month between 22-36 mo, the milk days between 5-305 d, and test day milk yield between 5.9-53.3kg were used to edit data, and finally a total of 127478 test-day milk yield records from 14320 Holstein heifers distributed in 24 herds between 2009 and 2013 in Ningxia were used in lactation curve mimicking and genetic analysis. Pedigree information of three generations were collected （father, mother, grandfather and grandmother from both father side and mother side） to form the pedigree file consisting 24 272 individuals. Microsoft Excel 2013 was used to manage the data to derive average milk production for each test-day, and NL1N procedure of SAS 9.1 was used to fit the Wood model and used to mimic the lactation curve to derive the population characters of milk yield. A random regression test-day animal model was employed and DMU 5.2 software was used for parameter estimation. The model included general fixed effect and fixed regression, random regression. In the present study, herd-test-day was the fixed effects, and a fixed regression were fitted for calving year and calving month combination effects, direct additive genetic, permanent environment were the random effects. Regression curves were modeled using Legendre polynomials of order 4. Based on climate characteristics in Ningxia, four calving seasons were categorized, spring （Mar. llth to May 20th）, summer （May 21st to Aug. 25th）, autumn （Aug. 26th to Oct. 15th） and

关 键 词：测定日产奶量 遗传参数 随机回归模型 宁夏 荷斯坦牛 

分 类 号：S823[农业科学—畜牧学]