Development of mechanistic computer-based models of semi-arid to arid rangeland systems, in order to understand them and to re-evaluate current rangeland theories and management practices;
Evaluation of alternative production strategies, as aids to short-and medium-term decision making, and prediction of requirements for drought relief;
Development of user-friendly software for use as an aid to decision making by rangeland managers and advisors.
We have developed a suite of mechanistic production system models based on differential equations which predict plant growth, animal forage intake, weight gains and milk yields on a daily basis, given rainfall (actual or simulated), soil type, vegetation and stocking rate. Output from these production system models is used in the development of long-term models that simulate the responses of rangeland ecosystems to different management strategies and to the variability of rainfall over 100 years or more.