Modeling P Dynamics and Crop Responses in Contrasting Soils of the Tropics.
Robert J. Delve, TSBF-CIAT Zimbabwe, P.O. Box MP228, Mt Pleasant, Harare, Zimbabwe and Merv Probert, CSIRO Sustainable Ecosystems, St Lucia, Australia.
The development and application of crop simulation models has focused on water and nitrogen as the main constraints to crop growth (Probert and Keating 2000). Such models have been useful for evaluating alternative management strategies and the effects of climatic conditions. However, this assumes that nutrients, other than N, are not limiting. In the low-input systems of tropical farming systems this assumption is incorrect as phosphorus is a major limiting nutrient that frequently affects crop growth, thereby reducing the usefulness of such models. This paper describes progress in developing and validating a unique capability in modelling soil phosphorus (P) dynamics and crop response to different sources of P in the Agricultural Production Systems Simulation (APSIM) modelling framework (Keating et al., 2003; web site www.apsim.info). APSIM now has functionality that can capture the release of N and P from various organic inputs, including manures (as described by Probert and Dimes 2004), on differing soil types and predict the growth of crops in situations where N and/or P is limiting. The APSIM SoilP module and the necessary modifications to the Maize module to provide a capability to simulate P-constrained maize crops have been described previously (Probert, 2004). More recently this ‘P-aware' capability has been added to the APSIM Plant module, which is the basis for simulation of most crops in APSIM. The validation of this P-aware APSIM model was conducted on three datasets - an Oxisol in western Kenya, an Alfisol and Vertisol in India, and an Andisol in central Colombia. These datasets were on contrasting soil types, with different degrees of P-fixation, P responsiveness and cropping systems to the soil in semi-arid Kenya on which the model was first developed. Data will be shown that demonstrates that across the different datasets there is close agreement between observed and predicted data and that more importantly, there is very little that needs to be changed within the model to parameterize it for simulating P responses, beyond the expected changes in soil water and crop parameters. The model testing and validation shows that it is possible to capture soil and crop responses to additions of different amounts, qualities and sources of P and N, across soil types, seasons and cropping systems and provides strong evidence of the robustness of the APSIM model in simulating P responses.