Value of perennial pasture phases in dryland agricultural systems of the eastern-central wheat belt of Western Australia
The identification of the value of perennial pasture phases for mitigation of soil salinisation and herbicide resistance in Western Australia, was the research scope. Such a problem is intractable by standard methods of economic analysis without significant modification and/or simplification, both of which decrease the relevance of model output. This thesis focused on resolving that deficiency through the construction of generalised analytical and numerical frameworks for the optimal control of switching systems. The latter is based on a novel critical path algorithm incorporating Approximate Dynamic Programming (ADP) to deal with the large state space typical of switching problems. The value of these developments for the analysis of phase farming is demonstrated and discussed in relation to traditional techniques used in the study of crop sequences.
For some of the publications from my research, please click on the links at the right-hand side of the screen.
This research has numerous benefits. Firstly, producers benefit from the development of a more accurate decision-support tool. In particular, the economic and dynamic effects of strategies for the mitigation of soil salinisation and herbicide resistance may be analysed at a level of detail beyond that which can be represented using traditional solution methods. Secondly, the ability to incorporate greater detail should aid the flow of information to producers through the improved integration of models and field trials. This should also improve the cost-effectiveness of research funding. Thirdly, this application will provide a foundation for the further implementation of ADP in agricultural economics, particularly in the study of switching systems.