Economic evaluation of crossbreeding for dairy production in a pasture based production system in Kenya

Data on accumulated life performance of crosses of Ayrshire (A), Brown Swis s (B), Friesian (F) and Sahiwal (S) cattle collected over a 21-year period from a dairy ranch in the lowland tropics of Kenya were analysed to estimat e additive and non-additive genetic effects on economic traits. These were used to predict and compare, first, performance of cows under nine crossbre eding strategies and, second, the performance of the production systems whe n applying the strategies that included maintaining the purebred S dams and bulls required for the production of crossbreds. Performance was predicted from parameters of a genetic model based on additive-dominance and additiv e X additive interaction effects for the following strategies: first cross (F x S), two-breed rotation (AS)(Rot), three-breed rotation (BFS)(Rot) and two- (F and S), three- (B, F and S) and four- (A, B, F and S) breed synthet ic ((Syn)) breeds based on equal and unequal contributions of the foundatio n breeds. The sensitivity of predicted performance at the cow level to vari ation in economic parameters was also investigated. For profit per day of p roductive herdlife (PLD), the B and F additive breed effects were not signi ficantly different from that of A. The additive breed effect for S was nega tive and significant (P < 0.01) indicating that it was inferior to the Bos taurus for PLD. Dominance effects for PLD in the crosses A x S and B x S we re positive and significant (P<0.05). The additive X additive interactions were negative and significant in all the crosses. At the individual cow lev el, predicted PLD would be lowest in (ABFS)(Syn) and highest in F x S. The (3/4F 1/4S)(Syn) would be the second-best strategy giving 90% of the expect ed F x S profit, while (FS)(Syn) would dye 87%. At me production system lev el, systems based on F x S cows were superior to those based on the rotatio ns at all level of number of calvings (NL) and would be superior to those b ased on the two-breed synthetics only at a NL higher than 4. Variation of c osts and prices greatly affected predicted economic benefit but not the ran king of strategies. The absence of significant differences in the additive breed effects of the B. taurus breeds for PLD showed that comparable econom ic benefits were derived by use of any of the three breeds for continuous c rossbreeding with the S in a production system with management achieving 30 00 kg lactation yields. It is also expected that the economic benefits from the development of two-breed synthetic breeds based on A, B or F would als o be comparable in production systems achieving lower yields (e.g., in many smallholder units). (C) 2000 Elsevier Science BN. All rights reserved.