Estimates of crossbreeding parameters for defining optimal crossbreeding systems for the coastal lowland tropics

An individual animal model was fitted to data from a crossbred herd in the lowland coastal tropics of Kenya to estimated breed cross means for a total of 25 genotypes having different proportions of Ayrshire (A), Brown Swiss (B), Friesian (F) and Sahiwal (S) genes. These means were then regressed on gene proportion of breeds and on the coefficients of heterosis and recombination loss. Per lactation, the S contributed 1802 kg and 18.4 kg less milk and milk per unit of metabolic weight (MW), respectively, than the F. The performances of the A and B were intermediate. The contribution of the F breed for most traits was superior to that of the other Bos taurus breeds. The heterosis effect between B and S was large for lactation milk yield (MY) (296 kg) and calving interval (CI) (-36 days). The heterosis between A and B for most traits was small, which is consistent with other studies in the literature. The estimates of recombination loss were negative in the crosses A x B and B x S for MY, daily milk yield (DMY) and MY expressed per unit MW. Modelling of alternative crossbreeding systems of MY indicated that the performance of the F1 F x S cross for MY was not significantly different from that of the three-breed rotation and the synthetic breeds. The F x S cross was estimated to produce 745 kg more (P < 0.05) MY than the two-breed rotation. Amongst the synthetic breeds, differences in MY were small. It is concluded that each crossbreeding strategy should be considered in relation to the ecological and socio-economic characteristics of the target production systems, which vary markedly. Therefore, the generalisation that the first cross (F1) is best suited for dairying in the tropics could be misleading, and there is the need to evaluate the potential of synthetic breeds.