Application of linear programming in bull selection for a dairy herd.

Approximately 70% of dairy cows are inseminated with semen from bulls provided by commercial artificial insemination centers. To help producers in the selection process, artificial insemination centers publish information regarding production and type traits associated with each bull. Producers must decide from these lists which bulls to use, how often they are to be used, and to which cows they should be allocated. The bulls selected have a direct effect on cost of semen for the herd and genetic improvement in production of future generations. In general, producers' goals are to replace poor genes through strategic culling (not breeding certain cows) with better genes through semen purchases in an economically efficient manner. The cost of semen, however, is realized immediately, whereas economic returns from the traits selected are dependent on the realization of future cash flows. Higher production returns are needed to justify using more expensive bulls. The purposes of the study reported here were to examine the bull selection portfolio of a 450-cow commercial dairy herd in southeastern Pennsylvania and to use linear optimizing programming to select a genetically similar group of bulls at a much lower cost than that the producer had been using. This approach of dominant decision-making can be generalized and implemented for various dairy from conditions.