Author
 |
Powell, Rex |
|
Submitted to: Interbull Annual Meeting Proceedings
Publication Type: Proceedings Publication Acceptance Date: 1/22/1998 Publication Date: N/A Citation: N/A Interpretive Summary: The ability of national and international evaluations to predict later national evaluations for milk, fat, and protein yields was compared for Holstein bulls evaluated in Canada, Germany, Italy, The Netherlands, and the United States. Correlations and standard deviations of differences of most recent 1997 national evaluations from 1995 national evaluations and February 1995 international evaluations were used to assess the usefulness of international evaluations as predictors of later national evaluations. Overall for bulls with international evaluations that included national evaluation data from an individual country and at least one other country, correlations with 1997 national evaluations generally were higher for 1995 national evaluations than for 1995 international evaluations. This result is attributed to the strong part-whole relationship for national evaluations over time. Although standard deviations of evaluation differences generally indicated better predictions from international evaluations than from national evaluations, improvements were small for Canada and the United States. The benefit of including international data when predicting national evaluations was more evident for bulls with large increases in daughter data. Improvement from including foreign data is more useful for imported rather than exported bulls and for importing rather than exporting countries. Technical Abstract: The ability of national and international evaluations to predict later national evaluations for milk, fat, and protein yields was compared for Holstein bulls evaluated in Canada, Germany, Italy, The Netherlands, and the United States. For bulls with International Bull Evaluation Service (INTERBULL) evaluations that included daughter information from a country and at least one other country, data included national evaluations (N95) that were input data for February 1995 INTERBULL evaluations, February 1995 INTERBULL evaluations (I95) on each country's scale recalculated using August 1997 methodology and August 1995 genetic parameters, and recent 1997 national (N97) evaluations. Evaluations of bulls with the largest increases in daughter data also were examined. Correlations and standard deviations (SD) of differences of N97 with N95 and I95 evaluations were used to assess the usefulness of INTERBULL evaluations as predictors of later national evaluations. Correlations with N97 evaluations generally were higher for N95 evaluations than for I95 evaluations. However, for Italy, correlations with N97 evaluations were higher for I95 evaluations than for N95 evaluations. Although SD of evaluation differences generally indicated better predictions from I95 evaluations than from N95 evaluations, improvements were small for Canada and the United States. Requiring that I95 and N97 evaluations contain minimum increases in data from N95 evaluations showed more clearly the benefit of including foreign data from INTERBULL evaluations when predicting national evaluations. Improvement from including foreign data is more useful for imported rather than exported bulls and for importing rather than exporting countries. |