Data comprising 39,578 test-day records of 3,766 first lactation Holstein cows calving from 1994 to 2002 was used to estimate genetic parameters for test-day yields (TDY) and 305-day milk yield (MY305) and to compare two selection criteria. Variance components were estimated by restricted maximum likelihood using univariate and bivariate animal models. The TDY models included the additive genetic random effect and the fixed effects of contemporary groups and the regressions of calving age (linear and quadratic effects) and days in milk (linear effect). The model used to analyse MY305 included lactation length in place of days in milk effect. Contemporary groups were defined by herd, and year and month of test for TDY, and by herd and calving season for MY305. Heritability estimates for MY305 from one-trait and two-trait analyses were 0.27 and 0.25, respectively. Heritabilities for TDY estimated by univariate and bivariate (MY305 and TDY) analyses ranged from 0.11 to 0.31 and from 0.08 to 0.25, respectively. Highest heritability estimates were obtained for 4th and 5th TDY, corresponding to the 2nd and 3rd months of lactation. Genetic correlations between MY305 and TDY were positive and high, ranging between 0.83 and 1.00. The results suggest that traditional direct selection for MY305 would result in larger genetic gains in most of TDY across lactation. Direct selection for TDY would result in genetic gains for MY305, but they would be smaller than those achieved by direct selection.
genetic gain; heritability; milk production; selection criteria; variance components
