Adjusting for age can lead to biased genetic evaluation for body weight in cattle

The aim of this study was to investigate the relationship between age (Age; 24.2±8.7days) and BW (61.0±8.2kg) of Brown Swiss male calves (n=6826) sold at auction between 2003 and 2007. Animals were progeny of 330 sires mated to 6826 heifers. Preliminary least-squares analysis showed that herd of origin and date of auction significantly influenced Age and BW, along with age of dam at calving and age of calf for BW. Single-trait animal models highlighted that both Age and BW were heritable (0.075±0.016 and 0.153±0.026, respectively); therefore, under these conditions, Age cannot be regarded as a managerial factor. When included as an explanatory variable, age of calf increased the heritability of BW by 5 or 9% depending on whether the effect was introduced as discrete factor or linear covariate, respectively. Bivariate analyses with or without the inclusion of age of calf in the model for BW provided estimates of heritability similar to those from single-trait analyses (0.079±0.014 for Age, and 0.136±0.021 to 0.192±0.031 for BW). The re-ranking of sires for BW of their calves at auction was very limited comparing bivariate models, while it was more pronounced comparing the results from single-trait models, and from single-trait and bivariate models; this is due to the strong and negative genetic correlation (−0.926±0.081) between Age and BW, which reflects the tendency of farmers to sell first the fast growing and then the slow growing calves. Hence, the inclusion of age of calf as an explanatory variable in single-trait models for BW leads to biases in the estimation of breeding values. On the contrary, combining Age and BW in bivariate analyses leads to unbiased results, suggesting that both traits can be used as good predictors of growth potential in calves.