Skip to main content
ARS Home » Research » Publications at this Location » Publication #176907
Title: COUNTRY BIAS IN INTERNATIONAL DAIRY BULL EVALUATIONS

Author
item POWELL, REX - 1201-01-00
item Sanders, Ashley
item Norman, H

Submitted to: Journal of Dairy Science
Publication Type: Abstract Only
Publication Acceptance Date: 2/28/2005
Publication Date: 7/24/2005
Citation: Powell, R.L., Sanders, A.H., Norman, H.D. 2005. Country bias in international dairy bull evaluations [abstract]. Journal of Dairy Science. 88(Suppl. 1):228-229.

Interpretive Summary:

Technical Abstract: The International Bull Evaluation service combines national dairy bull evaluations and provides results on each participating country scale. Theoretically, this process is designed to avoid favoring one country relative to another, but this concern has been raised frequently by international marketers of bull semen. Existence of a bias is difficult to assess; one approach is to compare evaluation results for full brothers from different countries. On average, these full brothers have the same genetics and should have similar evaluations. Over 12,000 Holstein bulls in 4336 full-brother families linked yield evaluations from 20 countries having bulls in at least 25 multi-country families. Slightly fewer bulls and families were in 16 countries with SCS data. The model analyzed with SAS® GLM included fixed effects of full-brother family (absorbed) and home country, where home country was the country of most daughters. To improve estimates of within-family variation, 6761 single-country families were also included. Primary analyses were on the US scale but results were similar on other scales. Full brothers from several countries had significantly higher evaluations than their US sibs for milk (Australia, Czech Republic, Germany, France, Great Britain, Japan, and South Africa), fat (Australia, New Zealand, and South Africa), and protein (Australia, Czech Republic, Germany, Japan, and South Africa). In contrast, bulls from the United States had significantly favorable evaluations for milk relative to Italy and SCS relative to South Africa. Largest biases involved bulls from South Africa where only 8 to 9 families were in common with the United States (thus giving indirect data greater importance), but other significant differences were based on hundreds of direct ties. The reason for these inequities is unknown but elimination of biases is important to maintaining confidence in international evaluations.