|ALBUQUERQUE L G|
|KEOWN JEFFREY F|
|VAN VLECK L DALE|
Submitted to: Journal of Animal Science Supplement
Publication Type: Abstract Only
Publication Acceptance Date: 3/10/1995
Publication Date: N/A
Technical Abstract: Genetic parameters for milk, fat and protein yields were estimated using a three trait repeatability animal model and a derivative free REML algorithm. Data consisted of 110,777 records of 54,634 Holstein cows from California (CA) and 118,820 records of 55,426 Holstein cows from New York (NY). Herds in each data set were randomly assigned to 10 subsets. The model included effects of herd-year-season as fixed and animal, permanent environmental and residual effects as random. The average numbers of levels of herd-year-season as fixed and animal, permanent environmental and residual effects as random. The average numbers of levels of herd-year- seasons and animals in the relationship matrix were, respectively, 218 and 9,220 in CA and, 1,474 and 9,275 in NY. The number of mixed model equa- tions were on average, 44,703 in CA and 48,475 in NY. Convergence was assumed when the simplex variance reached less than 1.e-9. Two samples, one efrom each state, did not reach the convergence criterion. The analyses were restarted to check for local maxima. Averages of heritability esti- mates in CA were .25 plus/minus .01 for milk, .27 plus/minus .02 for fat and .27 plus/minus .01 for protein. For NY data, the heritability esti- mates were .28 plus/minus .01 for milk and fat and .26 plus/minus .01 for protein. Repetability estimates were .55 plus/minus .01, .53 plus/minus .01 and .54 plus/minus .01 for milk, fat and protein yields in CA data and .56 plus/minus .01 for milk and protein yields and .55 plus/minus .01 for fat yield in NY data. Genetic correlation estimates between milk and fat, milk and protein and fat and protein were, respectively, .63 plus/minus .02, .85 plus/minus .01 and .74 plus/minus .02 for CA data and, .44 plus/ minus .02, .81 plus/minus .01 and .62 plus/minus .02 for NY data.