Considerations on models to predict crossbred performance in cattle

Abstract

Crossbreeding may rapidly and substantially increase the economic efficiency of beef and dairy cattle production. The customary genetic models to describe crossbred performance are presented and the assumptions on which they are based are discussed, along with the methods to validate them and to estimate the necessary parameters. The additivedominance model, assuming that heterosis is caused solely by dominance, is exemplified with results of Friesian/Jersey crosses in New Zealand. Under that model, crossbred performance is described as a function of two parameter classes, heterosis and breed additive differences, expressed in different proportions in each crossbred group, and the methods to calculate those proportions are also presented. Other models consider the loss in crosses of epistatic non-allelic combinations, which have been important in crosses of Bos taurus x Bos indicus but not in crosses of European Bos taurus breeds. Examples of models for class traits based on logistic regression are presented as well as methods to design experiments to estimate crossbreeding parameters.