CONSEQUENCES OF REMOVING THE VARIATION IN LACTATION LENGTH ON THE EVALUATION OF DAIRY-CATTLE BREEDS AND CROSSES

Three procedures commonly used for breed group comparisons were studied; (a) adjustment of milk yield for lactation length; (b) exclusion of short lactations considered abnormal; and (c) utilization of all available observations, unadjusted for lactation length. These three procedures were applied in a comparison of six Holstein-Friesian: Guzera crossbred groups at farms in the southeast region of Brazil. Genetic group effects on milk and component yields were attenuated by procedures a and b, compared to procedure c, and this resulted in underestimation of heterosis for those traits, as well as underestimation of the breed additive differences. These results were expected given the presence of genetic variation for lactation length and the high positive con-elation between this trait and milk yield. This research suggests that procedures reducing variation in lactation length are in adequate for evaluation of breed groups. Abnormal lactation data also should be reported when "normal" lactations are analysed separately.