摘要
【目的】估计里岔黑猪及其与巴克夏杂交猪(巴里猪)达100 kg体重日龄(D100)的校正系数CF和A,并比较两种校正系数的校正效果,为里岔黑猪和其他地方猪种D100的估计提供参考。【方法】以554头里岔黑猪、941头巴里1代和755头巴里2代猪为研究对象,利用润农性能测定系统,收集它们肥育期的体重和日龄数据。根据始测日龄≥60d、始测体重≥25 kg等标准对数据进行质控,质控后的数据用于CF和A的估计。CF是两条直线斜率的比值,其中分子是体重对日龄回归直线的斜率,分母是经过终测点(终测日龄和终测体重)和原点(日龄和体重均为0)直线的斜率。A是体重对日龄一元回归方程的截距,通过拟合体重对日龄的一元回归方程直接得到。校正系数估计出来后,用方差分析来检验性别和群体对校正系数的影响,并把估计的CF和A分别代入各自的校正公式,分别计算D100,即CF和A的校正日龄(D100CF和D100A)。通过t检验,比较D100CF和D100A之间是否存在显著差异,并计算它们之间的相关系数。然后利用每个个体的体重和日龄,分别建立日龄对体重的回归方程,并根据各自的回归方程计算达100 kg体重的回归日龄(D100reg)。通过计算D100CF和D100A与D100reg之间的相关系数及其与D100reg差值的平均数和标准差来比较CF和A的校正效果。最后用另外2个群体的校正系数来校正某个群体,计算D100CF和D100A与该群体D100reg的相关系数来评估校正系数的通用性。【结果】由于性别和群体显著影响校正系数,因此需要分性别和分群体估计校正系数。里岔黑猪、巴里1代和巴里2代阉公猪(母猪)的校正系数CF分别为1.426(1.346)、1.340(1.389)和1.372(1.380),A分别为60.65(54.15)、49.92(57.00)和54.25(57.53)。D100CF和D100A之间差异不显著(P=0.3071),它们之间的相关系数为0.9998(P<0.0001)。D100CF和D100A与D100reg的相关系数均≥0.9917(P<0.0001),它们与D100reg差值的平均数分别为-0.1385和-0.2741d,标准差分别为2.5532和2.5446d,差异均不显著(P>0.05)。用另外2个群体的校正系数来校正某个群体,D100CF和D100A与该群体D100reg的相关系数均≥0.9793(P<0.0001)。【结论】本研究估计的D100校正系数CF和A与国家生猪产业体系使用的CF和A有较大差异。CF和A的校正效果没有显著差异,在适当的结测体重范围内,它们的校正效果都非常好。在研究的3个群体中,用另外2个群体的校正系数来校正某个群体的校正效果也非常好,说明里岔黑猪D100的校正系数可以为其他地方猪种D100的估计提供参考。
【Objective】The aims of this study were to estimate the correction coefficients CF and A of days to 100 kg(D100)and to compare their correction powers to each other in Licha black pig and its intercross with Berkshire boars,so as to provide a reference of correction coefficients for estimating D100 in Licha black pig and other Chinese indigenous pig breeds.【Method】In this study,554 Licha black pigs with pigs from its intercross with Berkshire boars,including 941 F1 pigs and 755 F2 pigs,were randomly selected,and their body weights and ages in days had been automatically collected everyday by Winland performance testing system during their fatting periods.The data quality control was carried out according to the criteria,such as age≥60 days and body weight≥25 kg at start testing.Only the data have been passed quality control,which were used to estimate CF and A.CF was the ratio of two slopes,namely,the numerator was the slope of the regression line of weight to age,and the denominator was the slope of the straight line through the end point of performance testing(age and body weight at the testing end)and the original point(both age and weight were 0).A was the intercept of the regression equation of body weight to age,which could be obtained directly by fitting a regression equation of body weight to age.After the correction coefficients were estimated,an analysis of variance was carried out to test whether gender and population had effects on them,and the estimated CF and A were substituted into their respective correction formulae to calculate D100(D100CF and D100A).A t-test was used to test whether there was a significant difference between D100CF and D100A,and their correlation coefficient was also calculated.Then,the regression equation of body weight to age for each individual was established,and regressive D100(D100reg)for each individual was estimated by its regression equation.The corrected accuracies of CF and A were compared to each other by the correlation coefficients of D100CF and D100A with D100reg,and the means and standard deviations of the differences of D100CF and D100A to D100reg.Finally,the correction coefficients estimated from other two populations were used to estimate D100 in a certain population,and the correlation coefficients of D100CF and D100A with D100reg were calculated to evaluate whether the correction coefficient estimated from one population could be used in other populations.【Result】Gender and population had significant effects on the correction coefficients,so it was necessary to estimate the correction coefficient for each gender in each population.CF of stags(gilts)in Licha black pig and the F1 and F2 of its intercross with Berkshire boars were 1.426(1.346),1.340(1.389)and 1.372(1.380),respectively,and A were 60.65(54.15),49.92(57.00)and 54.25(57.53),respectively.There was no significant difference between D100CF and D100A(P=0.3071),and their correlation coefficient was 0.9998(P<0.0001).The correlation coefficients of D100CF and D100A with D100reg were equal or greater than 0.9917(P<0.0001),there was no significant(P>0.05)between the differences of D100CF and D100A to D100reg.The means of the differences were-0.1385 and-0.2741 days,and the standard deviations were 2.5532 and 2.5446 days,respectively.The correlation coefficients of D100reg with D100CF and D100A,which were adjusted using the correction coefficients estimated from other two populations,were equal or greater than 0.9793(P<0.0001).【Conclusion】The correction coefficients CF and A of D100 estimated in this study were quite different from their counterparts using in National Swine Industry Technology System.There was no significant difference between the correction powers of CF and A,and both CF and A had a very high correction power under an appropriate range of body weight at the end of performance testing.In the three populations,the correction coefficients from other two populations had a very high correction power to a certain population,which indicated that the correction coefficient of D100 from Licha black pig could be used as a reference to other Chinese indigenous pig breeds.
作者
郑浩
涂金敏
熊秀萍
张则凯
李龙云
黄黎斌
黄智勇
肖石军
郭源梅
ZHENG Hao;TU JinMin;XIONG XiuPing;ZHANG ZeKai;LI LongYun;HUANG LiBin;HUANG ZhiYong;XIAO ShiJun;GUO YuanMei(State Key Laboratory of Pig Genetic Improvement and Production Technology,Jiangxi Agricultural University,Nanchang 330045;Jiangxi Shanxia Investment Company Limited,Dingnan 341900,Jiangxi)
出处
《中国农业科学》
CAS
CSCD
北大核心
2020年第12期2493-2501,共9页
Scientia Agricultura Sinica
基金
国家重点研发计划子课题(2018YFD0501004)
国家自然科学基金(31660303和31460590)。
