Skip to main content
ARS Home » Research » Publications at this Location » Publication #47193
Title: ESTIMATES OF (CO)VARIANCES AMONG YIELDS OF MILK, FAT AND PROTEIN IN THE FIRST LACTATION FOR CALIFORNIA AND NEW YORK HOLSTEIN COWS USING AN ANIMAL MODEL

Author
item ALBUQUERQUE L G - UNIV OF NEBRASKA
item DIMOV G - UNIV OF NEBRASKA
item KEOWN JEFFERY F - UNIV OF NEBRASKA
item VAN VLECK L DALE - 5438-01-30

Submitted to: Journal of Dairy Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 3/15/1995
Publication Date: N/A
Citation: N/A

Interpretive Summary: Genetic correlations between milk and fat, and fat and protein yields were larger for California than for New York which confirms results for milk and fat yields reported previously for small samples from California and New York. Reasons for a difference in genetic correlations between New York and California are not apparent but may be associated with yields of milk or management systems. Estimates of environmental and phenotypic correla- tions were similar for California and New York. Heritability estimates for milk, fat, and protein yields were moderate and slightly larger in New York than California. These estimates are similar to previous estimates obtained using an animal model obtained with smaller data sets, and more limited rounds of iteration than in the present study. These estimates of correlations with protein yields are from larger samples of data and should be considered for use when USDA-AIPL changes from single trait to multiple trait genetic evaluations of dairy cattle for yields of milk, fat and pro- tein. Although the difference in the milk with fat yield correlations between California and New York herds merits further study, the differences are not likely to be important for multiple trait genetic evaluations.

Technical Abstract: First lactation milk, fat, and protein yields from New York and California Holstein cows, were used to obtain REML estimates of (co)variances for yield traits using a multitrait animal model. Data from each state were split randomly into 10 samples, with an average of 5,504 and 5,078 cows per sample, from California and New York, respectively. Averages of herit- ability estimates for milk, fat, and protein yields were, .30 plus/minus .02, .31 plus/minus .01, and .29 plus/minus .01 for California data, and .33 plus/minus .01, .35 plus/minus .01, and .30 plus/minus .01 for New York data. Averages of genetic correlation estimates for California and New York were .63 plus/minus .01 and .52 plus/minus .02 between milk and fat, .84 plus/minus .01 and .83 plus/minus .01 between milk and protein, and .73 plus/minus .01 and .68 plus/minus .01 between fat and protein. Estimates of environmental correlations were larger than genetic correla- tions. Averages of estimates of phenotypic correlations for California an New York were .75 plus/minus .01 and .72 plus/minus .01 between milk and fat, .92 plus/minus .01 and .91 plus/minus .01 between milk and protein, and .81 plus/minus .01 and .79 plus/minus .01 between fat and protein yields. On average, these estimates agree with those obtained from animal models with limited rounds of iteration for small data sets.