Review of mathematical models for sow herd management

This paper is a survey of the different sow models described in literature, which made use of different mathematical methodologies, and were intended for sow herd management. Models were discussed under a wide classification, that is, simulation and optimisation. The latter included linear programming and dynamic programming with Markov decision models and optimal control as major representative models. In a first stage we recalled general traits and modelling foundations of herd management models and later, different aspects of sow herd models published up to now were reviewed. Special attention is paid to main variables, source of parameters, validation, output and intended use. Most of such models have been developed as research tools and teaching aids. Actually, the increasing ability to represent complex systems is not corresponded with an augmentation of decision support tools including such complex models in field conditions. Thus, a need of new proposals dealing with transient situations and non-time homogeneous parameters was detected. The inclusion of variability-risk features and multicriteria decision methods was also of interest for practical purposes. Actual changes in the pig sector lead to expect new management herd models, in particular considering more than one herd at a time.