Published online by Cambridge University Press: 02 September 2010
When testing facilities are limited and the number of breeds and/or crosses to be tested is potentially large, then multibreed designs which involve a large number of breeds with only a few animals per breed give the optimum allocation of animals within and between breeds for various objectives. Multibreed designs could therefore form a useful adjunct to existing breed testing and selection programmes. How far such designs should be introduced would depend, among other things, on the degree of confidence in existing information, on the success of existing testing schemes, and on the relevance of the selection criteria currently being used. In this, the first of a series of papers on multibreed designs, the general approach is outlined, and designs for measuring between-breed variation are considered.
To estimate the extent of between-breed variation, the best overall design is to have four unrelated animals per breed and as large a number of breeds as possible. Valuable experimental work on breed variation can therefore be done without vast facilities.