Abstract
Responses to individual and sire selection on indexes of carrier, fat, protein, and lactose were examined for various sets of economic weights, population parameters, and indexed, omitting protein and lactose. Twenty-one sets of economic weights represented several Canadian pricing structures with different methods of derivation, which had all been previously applied or proposed in the literature. Population parameters were either literature averages or estimates for Holsteins in Canada with several sets of assumed parameters involving lactose. Alternative economic weights led to reductions of economic responses to individual and sire selection of up to 41 and 34%, indicating the importance of obtaining the correct economic weights. Use of literature averages versus Canadian estimates of population parameters reduced responses to individual and sire selection by 14 and 1%, while use of different parameters for lactose affected responses by up to 10 and .5%. Dropping lactose from the index reduced individual and site selection responses by up to 8 and 2%. Generally, protein was slightly more important in the index than lactose, but in specific situations the reverse was true. Collection of information of lactose is probably not worthwhile when the genetic merit of selected animals is estimated with high accuracy, such as current sire selection, but might be worthwhile when accuracies are lower, such as in multiple ovulation and embryo transfer nucleus breeding schemes.