A Dynamic Programming Model for On-Farm Decision Making in a Drought

Abstract

The model, which is based on dynamic programming, enables identification of the sequence of actions in a drought which will lead to the minimum expected cost subject to constraints. The model has been prepared as a tool which may be used to assist graziers to make decisions in a drought. The constraints consist of an upper limit on the number of a particular type of livestock which may be agisted (or droved) in a drought, an upper limit on the number sold and an upper limit on cash outlay. Actions assumed available to the grazier are sale of livestock, agistment (or droving), and supplementary feeding (or allowing mortality).