usda logo

UNPUBLISHED
RESEARCH REPORT
1999


POSTED
SEPTEMBER 2004

Examination of Effect of Inbreeding Adjustments to Holstein Evaluations

R. L. Powell
Animal Improvement Programs Laboratory, ARS-USDA, Beltsville, MD 20705-2350
301-504-8334 (voice) ~ 301-504-8092 (fax) ~ rlaipl@aipl.arsusda.gov ~ http://aipl.arsusda.gov

Pilot Study

In an attempt to get an indication of the impact of adjustments for inbreeding before the Interbull test evaluation, a small study was conducted using only national evaluations. Wilmink conversion equations were computed for fat data from the United States and Canada, Germany, and The Netherlands. Three sets of U.S. data were from August 1999: no adjustment, past inbreeding only, or past and future inbreeding. The composition of the top 100 bulls was examined both for country representation and ranking. Contrary to expectation, the adjustment for past inbreeding in U.S. data tended to favor foreign bulls. Applying both adjustments gave rankings quite similar to no adjustments. Differences were not dramatic by any definition.

Interbull Test Evaluation

The United States submitted data for the October 1999 Interbull test evaluation that included accounting for both past and expected future inbreeding. It was understood that this was for examination purposes and not for imminent implementation. New data, and more importantly new methods for other countries, were included in the Interbull test evaluation so that interpreting the impact of the inbreeding adjustment of U.S. data was not independent.

Data from Hungary, Poland, and the Republic of South Africa were included for the first time. New national genetic evaluations were computed using a test-day model in Estonia and Switzerland and an animal model in Sweden.

New national genetic evaluations were computed in Italy based on only the first 3 lactations rather than all lactations. Additional data from a specific area in the country were newly included in the national genetic evaluation system in Belgium and submitted for the Interbull test evaluation. Data on imported Holstein bulls in France were included in the international genetic evaluation. (Again, that is not an indication that they will be included in November.)

Interbull estimates of genetic correlations tended to be lower than for the August routine evaluation. Although there were some increases, a decrease of .01 was frequent.

Averages of Bulls by Country

Variation of bulls was decreased on the U.S. scale. Changes in bull averages for predicted transmitting ability (PTA) of yield traits follow:

Country PTA milk (pounds) PTA fat (pounds) PTA protein (pounds)
Canada -16 -.3 -.3
Germany -6 .2 .1
France -29 -.5 -1.0
Italy -10 .2 -.3
New Zealand 1 .2 .7
The Netherlands -12 -.1 -.1
United States -37 -1.3 -1.2

The U.S. adjustment of national data generally lowered PTA's for bulls. Although the largest reductions were for U.S. bulls, changes were not large. There is greater interest in what happens to the top bulls.

Top 100 Bulls

Changes in average PTA's for the top 100 bulls by country follow:

Country PTA milk (pounds) PTA fat (pounds) PTA protein (pounds)
Canada -26 -1.2 -.7
Germany -25 -1.3 -1.1
France -66 -1.5 -2.5
Italy -66 -1.1 -1.6
New Zealand -32 -1.3 -1.1
The Netherlands -41 -1.1 -1.2
United States -52 -1.9 -1.7

Decreases were very similar across countries. Although the decreases for the United States were larger than average, those differences are trivial.

Country representation among the top 100 bulls in August and for the test evaluation follows:

Country Milk Fat Protein
August Test August Test August Test
United States 75 73 46 44 53 54
The Netherlands 13 13 25 26 26 26
France 1 2 5 5 9 7
Germany 2 2 9 9 2 2
Denmark 0 0 6 6 2 3
Italy 4 5 3 3 3 3
Canada 2 3 4 4 1 1
Other countries 3 3 2 3 4 4

Changes were small. The United States had 2 fewer bulls for milk and fat but 1 more for protein in the test evaluation. Implementation of inbreeding adjustments had little impact on international rankings and no apparent effect on competitive ranking.

Conclusions

Adjusting for past and expected future inbreeding in U.S. data has little impact on the relative ranking of U.S. and foreign bulls on the U.S. scale. Numbers of top bulls and relative means of top bulls in the October Interbull test evaluation were similar across countries. A previous pilot study had suggested that while changes in ranking were small, the U.S. bulls were slightly disadvantaged by application of only the adjustment for past inbreeding.