A comparison of three empirical models for assessing cropping options in a data-sparse environment, with reference to Laos and Cambodia

In many less developed countries, there is a need to improve productivity and profitability of agricultural systems to improve rural livelihoods, particularly in areas where monocultural production has dominated the historical land use. Generally, there are a number of physical, chemical and biological soil constraints in these systems that prevent the successful production of alternative crops, often compounded by limited access to water. Rather than conduct time-consuming and expensive field trials, modelling techniques can provide a relatively inexpensive, first assessment of yield potential of cropping options under different water/nutrient regimes which can be used to inform and refine further research. However, in less developed regions, crop modelling activities are often constrained by limited institutional and technical capacity and inadequate or incomplete input datasets. In this case, complex models that require significant technical capacity and comprehensive datasets may be inappropriate, and less complex models with relatively simple input requirements may be better suited. This report discusses three relatively simple modelling options that can be applied within a data-sparse environment, and presents them within the context of Laos and Cambodia. The capabilities and limitations of each model are comprehensively reviewed to provide the reader with an understanding of the options to reliably simulate crop processes that may be useful, for example, to influence farmer practice, decision making, and to improve and optimise resource use.