Estimation Of Genetic Parameters Of Production Traits In Shandong Holstein Dairy Cattle Using The Test Day Model Of Random Regression

In this study,585702 test day records of the first lactation period of 88215 dairy cows from 2005 to 2016 in Shandong province were used.Milk yield and somatic cell score traits of dairy cows were analyzed by random regression analysis.The fixed effects of random regression model include herd-year-season effect,calving season effect and calving age effect.Random effects include cubic Spline knots effect,additive genetic effect and permanent environmental effect.Additive genetic and permanent environmental effects were modeled using Legendre polynomials of order 2,3,or 4 combined with cubic Spline functions,which have 20 knots.The order of Legendre polynomials for additive genetic effects and permanent environmental effects was selected by repeatable random selection.Therefore,each trait has 9 runs per lactation.The best Legendre polynomial order of additive effect and permanent environmental effect is obtained by comparing with AIC(Akaike information criterion)BIC(Bayesian information criterion,and so on.Then the additive effect and permanent environment effect fitted with the Legendre polynomial of the best order are combined with the cubic Spline function of different knots numbers(4、5、6、7、10、15、25、30).So there are 8 models for each trait.Through the comparison of model selection criteria,the best number of knots is selected,and the model of the number of knots is the best model.The best model of milk yield is model 15-4-3,that is,the number of cubic Spline knots is 15 knots uniformly,the additive genetic effect is embedded with Legendre polynomial of order 4,and the Legendre polynomial of order 3 is embedded with permanent environmental effect.The optimal model for somatic cell score was model 20-2-4(meaning the same as milk yield model).The additive genetic variance,permanent environmental effect variance,heritability and repeatability were 8.54 ~ 15.39,17.65 ~ 31.42,0.20 ~ 0.30 and 0.43 ~ 0.54,respectively,estimated by the optimal milk yield model during the whole lactation period.The additive genetic variance,permanent environmental effect variance,heritability and repeatability of somatic cell scoring model were 0.10~0.22,1.01~1.72,0.03~0.07,0.36~0.49,respectively,estimated by the optimal milk yield model during the whole lactation period.Rank correlation between milk yield breeding value of bull with different number of daughters and the pedigree index of the bull’s parents and the pedigree index of the F-MGS(father and grandfather)of the bull were from 0.79 to 0.94,from 0.48 to 0.86,respectively.Rank correlation between somatic cell score breeding value of bull with different number of daughters and the pedigree index of the bull’s parents and the pedigree index of the F-MGS(father and grandfather)of the bull were from-0.14 to 0.29,from-0.13 to 0.30,respectively.The results of parameter estimation of random regression model showed that:(1)The Legendre Polynomial in the model has a great influence on the variance and genetic parameters,and the influence of the Spline knot was small.(2)Both the genetic and phenotypic correlations coefficients of milk yield and somatic cell score showed the same change pattern,that is,decreased with the prolongation of the interval of DIM(Day in milk).(3)The rank correlation coefficient decreases as the number of daughters in the bull increases,indicating that the amount of information of the progeny is crucial for the genetic evaluation of the bull.