The compilation of a body weight composite to predict body weight in SA Holstein cattle

This study reviews the possibility of using body measurements to predict body weight by compiling a body weight composite (BWC) for Holstein cattle in intensive management systems. A data set of 701 records from three farms that feed a total mixed ration was used to build a BWC. The BWC included the following traits and attributes with their proportional contributions: Wither height (18%), body depth (8%), angularity (-16%), rump width (11%), chest width (15%), days in milk at classification (18%) and age at classification (14%). A linear regression was fitted for the BWC against the cows' realised weights. The regression equation was y = 82.377x + 206.11 with an R2 of 0.617. This regression was used in a verification data set to establish the usefulness of the BWC to predict body weight. The correlation between the predicted and the realised weights was 59%, with the average difference between the predicted and realised weights being 3.2%. It is concluded that the BWC is a useful indicator of body weight for Holstein cows in intensive management systems.