Skip to main content
ARS Home » Research » Publications at this Location » Publication #85408
Title: USE OF MULTI-COUNTRY DATA TO IMPROVE UPON NATIONAL BULL EVALUATIONS

Author
item POWELL, REX
item NORMAN, H

Submitted to: Journal of Dairy Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/15/1998
Publication Date: N/A
Citation: N/A

Interpretive Summary: The use of multinational data through the International Bull Evaluation Service (INTERBULL) is designed to provide improved prediction of the genetic merit of a dairy bull on a particular country's scale even though that country has limited or no data for daughters of the bull. However, not all countries accept INTERBULL evaluations as official. This study evaluated the usefulness of international data for better predicting a bull's true but unknown genetic merit. Added data from other countries were found to be most useful for evaluations with the largest increases in data. Combined U.S.-Canadian evaluations were an improvement over U.S.- only evaluations if sufficient data were added. The INTERBULL evaluations on Canadian, German, and U.S. scales appeared generally to provide useful additional information beyond the particular national evaluation. Agreement of differences from the added national and international data was not consistent but did support the conclusion that INTERBULL evaluations provide useful information beyond the more limited national data available at the same time. Greater acceptance and use of international evaluations, which generally are more accurate than national evaluations with limited data, should allow dairy cattle breeders worldwide to make better informed breeding decisions.

Technical Abstract: Estimates of bull genetic merit should improve with additional data. The added data could come from later national data or from other countries. Two studies assessed the usefulness of international data. The combined U.S.-Canadian and U.S.-only yield evaluations for 222 Holstein bulls from January 1993 were compared with August 1997 U.S.-only or maximum data (either U.S. or international). Standard deviations of differences in evaluation from added data and correlations of those differences generally indicated that the combined evaluations were improved predictors of future evaluations for 35 bulls increasing in data sufficiently to provide a good test; for all 222 bulls, relationships favored U.S.-only results because of the part-whole relationship. A second analysis involved three data sets of July 1996 national yield evaluations, August 1996 international evaluations, and August 1997 national evaluations for each of three countries: Canada, Germany, and the United States. The latter two sources of evaluations resulted from addition of data to the first. Agreement in differences from the added national and international data were not always consistent but did support the conclusion that the international evaluations provide useful information beyond the more limited national data available at the same time.