Optimising recording structures of a beef cattle breeding scheme.

The recording and age structure of a commercial breeding scheme was optimised using a genetic algorithm. Genetic merit of the breeding program with selection on best linear unbiased prediction (BLUP) was estimated deterministically using a multiple-trait selection index approach. This approach accounted for the loss in variance due to selection and the build up of pedigree over generations. Traits examined were feed intake, tenderness and growth. Particular attention was paid to the recording structure for feed intake, a major contributor to genetic gain and recording costs. A multiple stage selection approach was applied to select sires for use in the nucleus. Assuming cost of central testing for feed intake is $500 per sire, it would be profitable to test the best 30 sires per year. A combination of performance and progeny testing sires would be optimal if the cost testing each of the progeny of a sire was less than $100.