Records from 2.440 first lactations of Gir cows, calving between 1990 and 2005, were used for estimating genetic parameters of test-day milk yield (TDMY) and accumulated 305 day milk yield (M305) in univariate and bivariate animal models analyses. The model for TDMY included the contemporary group (CG) defined by the combination of herd, year and season of test day, the class of age of cow at calving and the linear covariate of days in milk (linear regression) as fixed effects and the additive genetic as random effect. For M305 the model included the CG (herd, year and season of calving), age of cow as covariate (linear and quadratic regression) and the additive genetic as random effect. Classes of CG with less than three or five animals respectively for TDMY or M305 were eliminated. The variance components were estimated by Restricted Maximum Likelihood. The highest values of phenotypic variances were observed in middle lactation, but in early lactation for the additive genetic variances. Heritability estimates ranged from 0.13 to 0.36 in univariate analyses and from 0.14 to 0.34 in bivariate analyses and were higher at the beginning and lower at the end of lactation in both analyses. Genetic correlations between TDMY were positive and high between adjacent test days. These results suggest that TDMY may be used as selection criteria for improving milk yield of Gyr cattle.
dairy cattle; genetic gain; milk production; variance components
