Uso del modelo epic para estimar rendimientos de maiz con base en variables fisiotecnicas en el oriente del estado de Mexico | Use of the EPIC model for estimating corn yield based upon crop-physiological variables in east region of state of Mexico

In Mexico, estimation of corn (Zea mays L.) yields has been performed through field methods which, besides costly, are time consuming. Simulation models stand as a good option, once they are calibrated with data from field experiments, because they are faster and less expensive. The objective of this study was to apply both calibration and validation phases for the EPIC (Erosion Productivity Impact Calculator) Model to estimate corn yields in the DDR 003-Texcoco. In the calibration phase, local information on corn yields, crop practices, general characteristics of 56 sites grown with this cereal (geographical position, slope, and land area), climate, and physical and chemical properties of soils was gathered. Furthermore, for the first time in Mexico, crop physiological variables (as biomass-energy relationship, leaf area index potential, and harvest index) were incorporated to the model. Simulated and observed yields were compared; then crop variables were adjusted. The model was rerun and a new comparison between observed and simulated yields was made. The model was validated by comparing yields of 65 production plots, reported by Secretaría de Desarrollo Agropecuario (SEDAGRO), with simulated yields obtained once the adjustment of variables was done. The results of the calibration phase, before the adjustment of crop physiological variables, indicated that sum of squares for the error (SEE) were 1.65 and 1.68 for irrigated and rain-fed plots, respectively. After the adjustment of variables, corresponding SEE decreased to 0.05 and 0.22. In the validation phase, SEE values of 9.66 were found, possibly because farmers do not report to SEDAGRO all the crop practices applied during the growing season precisely. For better model validation, it is necessary to register crop physiological variables in the common field formats.