Joint Evaluations for Sires and Cows using Field Data of Hokkaido Holstein Population

Abstract

A joint sire and cow evaluation of milk production for every lactation was carried out for the Hokkaido Holstein population, and evaluation values were compared with those of present official proofs of Hokkaido Milk Recording Association (HMRA). The data contained milk and fat yields for cows calving from 1975 to March 1987 during HMRA dairy herd testing. The data screenings were processed in the same way as those conducted at HMRA. A total of 1, 672, 032 lactations from 615, 024 cows in 10, 482 herds were used in the study. A mixed linear model included Herd-year effects as fixed and additive genetic effects, nonadditive genetic plus permanent environmental effects of cows and residual errors as random variables. An inverse of numerator relationships matrix, computed by ignoring the relationships between females across herds, was used to obtain the BLUP solutions via the animal model. Under this model, calculations were performed with a three-stage absorbing approach in the following order; nonadditive genetic plus permanent environmental effects, herd-year effects, and cow's additive genetic effects. Assumed heritabilities and repeatabilities were 0.25 and 0.40 in both milk and fat yields, respectively.
4, 525 sires were evaluated simultaneously, about 2.3 times the number of the present official proofs by BLUP under the sire model with first lactation daughters. Correlations between the estimated breeding values of sire by simultaneous and those of the official evaluation were 0.795 for milk yield and 0.757 for fat yield. Correlations from 81 sires with 1, 000 or more daughters increased to 0.933 for milk yield and 0.866 for fat yield. Correlation coefficients between the breeding values of cows from the joint evaluation and those of the intra-herd cow evaluation of HMRA were 0.959 and 0.941 for milk and fat yields, respectively. But, correlation coefficients from 1, 462 elite cows which were the 1% upper bound on both milk and fat proof decreased to 0.909 for milk yield and 0.908 for fat yield, respectively.
It was concluded that the simultaneous evaluation of sire and cow can be applied to improving dairy cattle without extra computing time and resource requirements.