Sire selection with multiple objectives

The principal objective of any dairy breeding program is to increase the profitability of the herd. One way to achieve this objective is through the selection of sires that increase the genetic potential of the herd. However, selecting sires this way is likely expensive and almost surely induces inbreeding problems which may erode future profitability. A decision model for optimal dairy sire selection is developed and solved using American Jersey Association sire data. The model is an integer compromise programming model with three objectives: maximizing net merit, minimizing semen cost, and minimizing inbreeding depression. The results suggest that compromise programming may offer substantial benefits over single objective models given the ability of the technique to quantify the trade-offs associated with multiple, conflicting objectives.