Development and Validation of a Model to Simulate Phenology, Canopy Growth and Yield of Maize, Mungbean and Tomato under Tropical Farming Systems

A significant portion of crop production in the tropics occurs at subsistence level with high vulnerability to climate change. Process-based crop simulation models are effective tools to predict environmental impacts, including climate change effects, and make management decisions on crop production. The principal objective of this work was to develop crop-specific simulation models to predict phenology, canopy growth, and yield performance of maize, mungbean, and tomato crops grown in the tropics, with special relevance to locally adapted inbred germplasm. The developed models consisted of sub-models to simulate canopy growth, radiation interception and conversion, and dry matter partitioning. All sub-models were parameterized by detailed measurements of leaf growth, dry matter accumulation, and partitioning on locally adapted varieties of the three crops over two cropping seasons. The same crop varieties were established in a multi-locational field experiment spanning a sufficiently wide temperature gradient over four cropping seasons for model validation. Models were validated by comparing model predictions of phenological development and yield against field observations and additional secondary data from the literature. Overall, simulated phenological development (R2>0.5, RMSE<5 days for all three crops) and yield (R2>0.5, RMSE<68, 48 and 476 g/m2 for maize, mungbean, and tomato, respectively) of the three crops were in agreement with the observed data under a wide range of environmental conditions. These modelling approaches can be successfully applied to open-pollinated (maize) and inbred varieties (mungbean and tomato) growing under the farmer fields in tropical South Asia.