Modeling spatial partitioning of light and nitrogen resources in banana cover-cropping systems

In banana cropping systems, cover crops are introduced mainly to manage weeds and mitigate the use of herbicides. But this introduction modifies the structure of the field, its biophysical functioning and then farmers' practices. We designed the SIMBA-IC model to simulate nitrogen and light partitioning and crop management, which can differ across the different zones of the field (banana row, small and large inter-rows), and to assess agronomic and environmental performances of banana cover cropping systems under scenarios of different spatial arrangement. We calibrated and validated the model using independent data sets from a fertilizer and an intercropping experiment, respectively. Results showed that SIMBA-IC realistically simulated the differences between treatments in terms of crop cycle duration (for flowering date, RMSE = 2.3 weeks in the calibration and validation steps) and biomass produced (RMSE = 0.67 and 0.94 kgDM ha(-1) at flowering in calibration and validation steps, respectively). We simulated different management options related to the four field zones to optimize fertilization and cover crop management and ensure a tradeoff between agronomic (banana yield) and environmental (N leaching mitigation) performances. Simulations showed that yield was maximal and N teaching was reduced when fertilization was applied in the banana row. When cover crops were mowed according to the banana N stress, agronomic and environmental performances were higher than when mowing was based on the cover crop leaf area index, but the former approach led to very frequent mowing. Future studies should consider the impacts of these cropping systems from a socio-economic point of view to assess their feasibility and ability to be adopted. (C) 2012 Elsevier B.V. All rights reserved.