Skip to main content
ARS Home » Research » Publications at this Location » Publication #88405

Title: DEVELOPMENT OF MULTITRAIT EVALUATION FOR LINEAR TRAITS AND SINGLE-TRAIT EVALUATION FOR FINAL SCORE OF SIX DAIRY CATTLE BREEDS

Author
item GENGLER, NICOLAS - FNRS, GEMBLOUX, BELGIUM
item Wiggans, George
item Wright, Janice

Submitted to: American Dairy Science Association Abstracts
Publication Type: Abstract Only
Publication Acceptance Date: 7/27/1999
Publication Date: N/A
Citation: N/A

Interpretive Summary:

Technical Abstract: A multitrait animal model was developed for genetic evaluation of final score and 15 linear type traits of US Ayrshires, Brown Swiss, Guernseys, Jerseys, and Milking Shorthorns; 14 linear traits were included for Red and Whites. An appraisal during first lactation was required. Variance components were estimated from appraisals since 1988 that had all linear traits scored. The model included fixed effects for interactions of herd, appraisal date, and parity group (first or later); parity and appraisal age groups; and parity group and lactation stage. Random effects were included for permanent environment, animal, and residual. A canonical transformation was used with approximate diagonalization. Solutions for variance components did not converge for linear traits of Red and Whites; therefore, previously estimated Holstein correlations among traits were used for computation of evaluations. Data for estimating breeding values included appraisals during 1980 and later, but data for some traits were missing. Effects for appraisal age group and lactation stage were defined within time interval (before 1988 or 1988 and later). A random effect of interaction between herd and sire was added to the model and assigned 40% of estimated (co)variance for permanent environment. Solutions for appraisal age groups from a preliminary analysis were smoothed with a quadratic curve to generate additive age adjustments by month of age at appraisal, parity group, and time interval (before 1988 or 1988 and later). Solutions from a previous evaluation were used as starting values. Fewer than 50 iterations were required to reach convergence. For Jersey bulls with 20 daughters or more, most correlations of solutions with those from the current sire model ranged from .85 to .90.