JOINT PUREBRED AND CROSSBRED (CO)VARIANCE COMPONENT ESTIMATION WITH APSEUDO MULTIPLE-TRAIT MODEL - LOSS IN EFFICIENCY

Based on simulated data sets, (co)variance component estimation was ca rried our in a purebred-crossbred situation using a common multiple tr ait animal model where purebred and crossbred performances were treate d as separate traits. The inability of such a model to reflect the spe cial (co)variance structure of purebred-crossbred data leads to an imp ortant bias of the purebred-crossbred covariance; however, the loss in selection response was small.