Available methods for the optimization of agricultural systems vary wi
dely, in terms of derivation, applicability and performance. A whole-f
arm dairying model with 16 separate, interacting managerial options wa
s optimized by the hill-climbing (quasi-Newton), direct search (simple
x), genetic algorithm (GENESIS) and simulated annealing (VFSR) techniq
ues. The latter two clearly out-performed the former, with simulated a
nnealing always identifying the global optimum.