Multi-objective simulation and optimisation of dairy sheep farms: Exploring trade-offs between economic and environmental outcomes

Abstract

A decision support tool for sheep farming systems (PASTOR-DSS) was developed to investigate trade-offs between economic and environmental objectives on Spanish dairy sheep farms. The tool combines a hierarchical stochastic simulation model at three levels with a multi-objective optimisation procedure based on genetic algorithms. The first level of simulation includes rumen, reproduction and nutrient balance submodels. These three submodels are integrated into an animal model, which constitutes the second level. The third level is the farm, which includes the flock, the feeding and reproductive management, the availability of feeding resources, and the farm economics. The multi-objective genetic algorithm applies to the farm level. The tool was validated for the different levels of simulation, with outputs having an acceptable level of accuracy and representing correctly the links between feeding and reproduction. The tool was used to optimise the Latxa breed farming systems of the Basque Country (Spain). Two farm types were simulated: a COAST farm located in low-altitude Atlantic conditions and longer grazing period, and the INLAND farm located in mountain areas with a shorter grazing period. The optimisation provided a set of optimal solutions with different economic and environmental (N excretion) performances. The optimal solutions increased the financial margin over feed costs in both farms (+24% and + 22% for COAST and INLAND, respectively). The final space of solutions showed a clear trade-off between the economic and environmental performance (nitrogen excretion). The difference in the financial margin over feed costs between the solutions could be interpreted as the opportunity cost of greening in policy design, i.e., the payment that farmers should receive to change their farming methods to reduce nitrogen pollution.