Published online by Cambridge University Press: 02 September 2010
In the breeding of dairy cattle the selection of bulls to breed young bulls for progeny testing is a crucial process. This paper compares several policies for making this selection based on the criteria-selection response, inbreeding depression, loss of genetic variance and variability of response. A number called the ‘effective number of new bulls to breed bulls selected each year’ (NBBe) is defined which is closely related to the last three of these criteria. Past studies of the design of dairy cattle breeding programmes have assumed that selection is within a group of bulls progeny tested in the same year (policy I). However, modern sire evaluation methods allow comparison of sires tested in different years. To evaluate the effect of selecting bulls to breed bulls from all available bulls (policy II) a computer simulation program was used. Policy II results in an increase in the response to selection but a substantial decrease in NBBe. When compared at the same NBBe, policy II results in a smaller selection response than policy I. A policy which allows the best bulls to be used for more than 1 year but which limits the maximum number of years for which they can be used, results in the best compromise. If bulls are to be used for several years there is little advantage to be gained from making more matings within each year to more high-rated bulls or to older, more reliably evaluated bulls.