GENETIC-PARAMETERS OF MILK, FAT, AND PROTEIN YIELDS IN THE FIRST 3 LACTATIONS, USING AN ANIMAL-MODEL AND RESTRICTED MAXIMUM-LIKELIHOOD

Milk, fat, and protein yields of Holstein cows from the States of New York and California in the United States were used to estimate (co)var iances among yields in the first three lactations, using an animal mod el and a derivative-free restricted maximum likelihood (REML) algorith m, and to verify if yields in different lactations are the same trait. The data were split in 20 samples, 10 from each state, with means of 5463 and 5543 cows per sample from California and New York. Mean herit ability estimates for milk, fat, and protein yields for California dat a were, respectively, 0.34, 0.35, and 0.40 for first; 0.31, 0.33, and 0.39 for second; and 0.28, 0.31, and 0.37 for third lactations. For Ne w York data, estimates were 0.35, 0.40, and 0.34 for first; 0.34, 0.44 , and 0.38 for second; and 0.32, 0.43, and 0.38 for third lactations. Means of estimates of genetic correlations between first and second, f irst and third, and second and third lactations for California data we re 0.86, 0.77, and 0.96 for milk; 0.89, 0.84, and 0.97 for fat; and 0. 90, 0.84, and 0.97 for protein yields. Mean estimates for New York dat a were 0.87, 0.81, and 0.97 for milk; 0.91, 0.86, and 0.98 for fat; an d 0.88, 0.82, and 0.98 for protein yields. Environmental correlations varied from 0.30 to 0.50 and were larger between second and third lact ations. Phenotypic correlations were similar for both states and varie d from 0.52 to 0.66 for milk, fat and protein yields. These estimates are consistent with previous estimates obtained with animal models. Yi elds in different lactations are not statistically the same trait but for selection programs such yields can be modelled as the same trait b ecause of the high genetic correlations.