Application of the first canonical variable in the evaluation of animal production trials