2 41practices, tactics within alternative strategies and competing objectives, in the back-
ground of the need for increasing productivity in highly risky climates.
Since the early 1990s, most participatory farming systems research projects have
incorporated systems analysis and a systems modelling tool to support benchmark-
ing and practical decision making. After three decades of ʻcontestable successʼ
(McCown et al. 2009 ), the role of farming systems models remains strong at quan-
tifying complex systems interactions, providing ex-ante analyses that identify best
bet options for intervention, and quantifying benefits and trade-offs from the adop-
tion of technologies particularly when land, labour and capital limit farmer options.
Clearly agricultural systems modelling techniques have played and will continue
to play a significant role in adapting the nature and extent of agriculture to fit the
challenges and opportunities from the expected increases in food and energy demand,
amid changes in climate and the environment (Rodriguez and Sadras, 2011 ).
2 Modelling Dryland Agricultural Systems
Modelling in dryland agricultural systems research has two major roles in the design
of productive and resilient farming systems: (1) to increase knowledge of the func-
tioning and dynamics of complicated, managed natural systems, where biotic pro-
cesses interact with climatic, soil and biological drivers at a range of temporal and
spatial scales, and (2) to inform (and overcome) the complexities in the system, i.e.,
those that originate as soon as the manager of the natural system is brought into the
analysis – the human dimension which has particular objectives, experiences and
perceptions, values and aspirations that also need to be met. Here we are mostly
concerned with the second more complex aim, but will also acknowledge the rich
contribution from the diversity of approaches used, from simple rules-of-thumb to
dynamic crop models, and the analysis of complicated whole-farm models.
Models and frameworks developed and applied to address different questions,
usually involve contrasting integration, temporal and spatial scales. Models also
vary in their degree of detail with regard to how the processes and relationships are
described e.g. they may have a single equation or rule or hundreds of lines of code,
they may perform a single calculation i.e. static or multiple calculations over a range
of time steps i.e. dynamic models, and they may delve into the minutiae and mecha-
nisms of phenomena or have been developed as learning tools to support the non-
specialist farmer or farm advisor.
2.1 Simple Rules-of-Thumb (Models) for Benchmarking Farm
Practice
Based on the framework developed by the celebrated Prof. deWitt ( 1958 ), the most
well-known model for informing practice change and benchmarking dryland agricul-
ture is the water use efficiency (WUE) model by French and Schulz ( 1984 ) (Fig. 1 ).
Modelling Dryland Agricultural Systems