Skip to main content
ARS Home » Research » Publications at this Location » Publication #84372

Title: CHANGES IN USDA-DHIA GENETIC EVALUATIONS (AUGUST 1997)

Author
item Van Tassell, Curtis - Curt
item Wiggans, George
item Vanraden, Paul
item Norman, H

Submitted to: AIPL Research Reports
Publication Type: Government Publication
Publication Acceptance Date: 8/11/1997
Publication Date: N/A
Citation: N/A

Interpretive Summary: Changes in August 1997 USDA-DHIA genetic evaluations for dairy cattle were documented. Heritability, the proportion of variation among records that results from genetics, was increased to from .25 to .30 for yield traits (milk, fat, and protein) after research revealed that heritability estimates for first-lactation yield traits were considerably higher than .25. The proportion of overall variation in yield attributable to other factors was reduced, and yield deviations were restricted to limit the impact of large deviations on genetic evaluations. Predicted genetic merit from the new and previous systems were similar for most artificial- insemination (AI) bulls. Accuracy of Holstein AI bull evaluations increased by .04 from May to August. To lower costs for distribution of evaluations, the minimum sum of fat and protein merit to have an update released for bulls with daughters in only one herd was raised from 10 to 15 lb. August 1997 USDA-DHIA evaluations included records from 10 US daughters or more of each of 5 Netherlands Holstein bulls. USDA-DHIA evaluations for these bulls had low reliability and are unofficial. Any assessment of how well INTERBULL evaluations of Netherlands bulls predict daughter performance in the U.S. should be considered preliminary. Future USDA-DHIA evaluations will include additional U.S. daughters of more non- U.S. bulls and lead to more accurate global rankings and conversion assessments for use by breeders in making selection decisions.

Technical Abstract: Changes in USDA-DHIA genetic evaluations in August 1997 were documented. Heritability for yield now ranges from .25 to 35, with a midpoint of .30 (previously .25). Proportion of overall variation was reduced to .10 for herd-sire interaction and .15 for permanent environment. Yield deviations were restricted to +/-4 herd-year standard deviations with a lower limit of half of management group mean. Predicted transmitting abilities (PTA) from the new and previous systems were similar for most artificial- insemination (AI) bulls. Correlations between 1st- and 2nd-crop PTA were larger for the new system. Reliability of Holstein AI bull evaluations increased by .04 from May to August. For bulls born after 1989, standard deviations of milk, fat, and protein PTA for bulls with 100 daughters or fewer increased by 5% from May to August. For bulls in active AI service after May 1997, mean PTA in August declined slightly: 7.8 lb milk, .62 lb fat, and .51 lb protein. For bulls with daughters in only 1 herd, sum of PTA fat and PTA protein had to change by more than 15 lb (previously 10 lb) to have an update released. August 1997 USDA-DHIA evaluations included records from 10 US daughters or more of each of 5 Netherlands Holstein bulls. USDA-DHIA evaluations for these bulls had low reliability and are unofficial. Mean daughter yield deviations of the 5 bulls based on early in-progress records of only 85 cows were 1245 lb milk, 42 lb fat, and 65 lb protein. Any assessment of how well International Bull Evaluation Service evaluations of Netherlands bulls predict daughter performance in the US should be considered preliminary. Future USDA-DHIA evaluations will include more US daughters of more non-US bulls and lead to more accurate global rankings and conversion assessments.