A Comparison of Methods to Pre-adjust Data for Systematic Environmental Effects in Genetic Evaluation of Sheep

Abstract

Sheep performance data are routinely adjusted to account for systematic environmental effects but little attention has been given to the choice between the methods available. This study compares methods to adjust liveweight measurements for individual age (within group regression vs. age intercept) and age of dam (dam age class vs. quadratic polynomial on dam age) and fat and eye muscle depth for liveweight (within group regression vs. quadratic polynomial on weight). Criteria for comparison included: (1) changes in phenotypic variance and heritability estimates; (2) predictive ability of resulting estimated breeding values (EBVs); (3) correlation between adjusted and adjustor variables; and (4) ease of computation. With the exception of dam age adjustment of liveweight, adjustment methods did not influence the phenotypic variance of the trait being adjusted. Increases in the direct heritability were observed for liveweight adjusted using the age intercept, dam age class and quadratic methods, and fat depth using quadratic adjustment for liveweight. Within group regression based adjustments generally resulted in a reduction of direct heritability. Maternal heritability and permanent environment due to dam were unaffected by the adjustment methods utilised. The regression of progeny performance on sire estimated breeding value also indicated that the age intercept produced estimated breeding values that were closer to the expectation than those using the within group regression adjustment. The age intercept and quadratic polynomial methods of adjustment appear to be slightly more appropriate for genetic evaluation purposes.