Modeling soil nutrient limitations to crop production in semiarid NE of Brazil with a modified EPIC version : II. Field test of the model
The validity of the modifications in the source code of EPIC for simulation of crop development in semiarid conditions, described in de Barros et al. [de Barros, I., Williams, J.R., Gaiser, T. Modeling soil nutrient limitations to crop production in semiarid NE of Brazil with a modified EPIC version. I. Changes in the source code of the model. Ecol. Modell., submitted for publication] was tested against the results of 3 years, field experiment with nutrient omission using intercropping between maize and cowpea. The plant biomass production, grain yield, N, P, and K uptake by both crops were measured during the period of 1998-2000 and compared with simulated values. Reasonably good agreement and accuracy was obtained between the measured and simulated values for all evaluated parameters. Slight overestimation of grain yield by the simulations was attributed to high rates of floral abortion caused by dry spells during the flowering periods of 1998 and 1999, since this factor is not considered in the model yet. Improvements of the accuracy of the model may also be obtained by more comprehensive understanding of soil resource use by component plants in mixed cropping systems. Therefore, it is necessary to generate more knowledge about the partitioning of the resources by the crops before being incorporated into modeling approaches. The model was sensitive to climate variations and changes in soil chemical properties due to the application of fertilizers and/or lime under the tropical semiarid conditions of the NE of Brazil. Once calibrated and validated for different crops and soil types of this region, the model will be appropriate as tool for regional planning.