GENETIC AND PHENOTYPIC PARAMETERS FOR YIELD, FOOD-INTAKE AND EFFICIENCY OF DAIRY-COWS FED AD-LIBITUM

Records on milk yield, fat plus protein yield, food intake, food efficiency, calving live weight and mean live weight, up to 26 and 38 weeks of lactation, were obtained from dairy cows, fed ad libitum, in the Edinburgh School of Agriculture's Langhill herd. The data were divided into first and later lactations and restricted maximum likelihood analyses carried out on heifer, cow and pooled data, fitting an animal model, with repeat lactations as an additional random effect. Univariate analyses were done after canonical transformation of heifer data and approximate canonical transformation of cow and pooled data. Heritability estimates for food efficiency and food intake, from pooled data, were 0.13 (s.e. 0.09) and 0.37 (s.e. 0.11) for 26-week and 0.13 (s.e. 0.12) and 0.52 (s.e. 0.14) for 38-week lactation periods, respectively. Over the same periods, estimates for milk yield were 0.20 (s.e. 0.08) and 0.20 (s.e. 0.11), respectively. Estimates from the analyses of cow and heifer data separately were higher, as were their standard errors. Genetic correlations between milk production traits and efficiency, from the pooled data analysis, ranged from 0.44 to 0.61 and those between milk production traits and food intake from 0.32 to 0.74. Genetic correlations between live-weight traits and efficiency ranged from -0.81 to -0.99, and those between food intake and live-weight traits from 0.28 to 0.46. The results indicate that when selection is on yield, the correlated responses in efficiency may be smaller under ad libitum feeding, compared with published values where cows were given food according to yield. Including live weight in the selection criterion may give higher responses in efficiency compared with selection on yield alone. In MOET nucleus schemes it may be worthwhile to include food intake or efficiency directly in the selection criteria.