Simulation of Maize Yield with Different Levels of Nitrogen by Using DSSAT Model

Decision Support System for Agrotechnology Transfer (DSSAT) model is able to simulate plant growth, development, and yield that are grown on a uniform surface under simulated management conditions, including changes in soil water, soil carbon, soil nitrogen contents and nitrogen leaching. This study was aimed to investigate the effects of nitrogen on yield and yield components of maize variety SC704 by using this model, and to calibrate CERES-Maize model under 4 levels of nitrogen fertilizer: N1: 25% less than the recommended level, N2: recommended level (200 kg/ha), N3: 50% less than recommended level (260 kg/ha), and N4: 50% more than the recommended level (310 kg/ha). To evaluate the applicability of this model an experiment based on randomized complete block design with three replications was conducted during 2013 at the Research Field of Agriculture Faculty of Islamic Azad University – Karaj Branch. The measured traits, and their simulated values for ear and biomass yields, leaf area index (LAI) and stem dry matter content were compared. The results of the biomass simulation showed that Root Mean Square Error (RMSE) of the four fertilizer levels ranged 2496.48, 2159.24, 2302.43, and 3289.19 kg/ha respectively. For the ear yield, the highest coefficient of determination (R2 = 0.98) was obtained by N4. In fact, this treatment provided highest accuracy for predicting the yield of maize by the model. For leaf area index, the Willmott Agreement Index (d) varied between 0.77-0.94. This indicates that the model has successfully predicted the variation of leaf area index. Therefore, the model is considered appropriate for simulating growth, development and yield of maize under 4 levels of nitrogen fertilizer. In this case, it is recommended that the model is calibrated and verified, and then, it is applied for research purposes in Karaj climatic conditions.