Potential Uses and Limitations of Crop Models

Abstract

Crop models have many current and potential uses for answering questions in research, crop management, and policy. Models can assist in synthesis of research understanding about the interactions of genetics, physiology, and the environment, integration across disciplines, and organization of data. They can assist in preseason and in‐season management decisions on cultural practices, fertilization, irrigation, and pesticide use. Crop models can assist policy makers by predicting soil erosion, leaching of agrichemicals, effects of climatic change, and large‐area yield forecasts. Cautions and limitations in model uses are suggested, because appropriate use for a particular purpose depends on whether the model complexity is appropriate to the question being asked and whether the model has been tested in diverse environments. There is a need for both complex and simple models. In some cases, simple models are not appropriate because they are not programmed to address a particular phenomenon. In other cases, complex models are not appropriate because they may require inputs that are not practical to obtain in a field situation. Modelers need to be forthright in model description and promotion. For example, what does a given model respond to? What are the limitations of the model? What factors does the model not address? What are the limitations of inputs to run the models? Examples are given of model use to evaluate genetic improvement in photosynthesis and seed‐filling duration, yield response to planting date and row spacing, and effects of change in seasonal temperature. We believe that use of crop growth models will play an increasingly important role in research understanding, crop management, and policy questions.