The utilization of prediction models to optimize farm animal production systems: the case of a growing pig model.

Abstract

In the last 25 years, mathematical models for growing animals have been developed to predict growth from knowledge on animal biology. The increasing precision of these models allows them to be used to evaluate economic return and/or environmental impact of the simulated production system. It is also possible to incorporate these models into optimization routines to determine feeding programmes or slaughter management policies. To evaluate the usefulness of this approach, a simplified growing pig model was developed to determine the feeding management system that maximizes net return in commercial growing/finishing production systems. The model is mechanistic, deterministic, dynamic and aggregated at the whole-animal level. The overall model has four different submodels. The first estimates the growing pig's maximal energy and protein requirements for the entire growing period. The second submodel determines the least-cost diet meeting those requirements. The third submodel simulates the animal's growth in terms of protein and fat deposition. The last submodel calculates the net economic return from the simulated production system. Finally, a non-linear optimization algorithm finds the feeding programme that maximizes net return. Model results indicate that maximal net returns are obtained by restricting protein intake by 6 to 30% of the maximal requirements. Model results also show that the payment grid used in several countries is the most important factor to be taken into account to maximize revenue. Also, our results indicate that carcass reference and feedstuff prices have only a limited impact on the feeding programme that achieves maximal net return. This study shows that growth models may help to determine optimal production strategies from an economic and an environmental point of view.