Evaluation of the capability of CERES-Maize model in estimating corn yield in different water and nitrogen levels

Background and objectives: With increasing population growth and limited expansion of agricultural land, it is necessary to look for a way to increase agricultural productivity. Today, the need for a sustainable agricultural system that fully utilizes environmental resources and at the same time does not cause any damage to the environment seems very necessary. In recent years, much attention has been paid to soil quality aspects and increasing crop production by using plant residues, organic and green fertilizers as sources of soil organic matter and plant nutrients. Therefore, replacing or combining appropriate systems of organic sources with chemical fertilizers in order to reduce the consumption of chemical inputs and achieve sustainable agriculture goals can be considered as an appropriate solution. Since conducting field experiments is time-consuming and costly, plant simulation models are suitable options for evaluating different crop management strategies to optimize the use of available resources. The DSSAT model is a comprehensive tool for simulation and decision-making in agricultural crop management, which integrates climatic, soil, genetic, and management data to help researchers and managers assess and improve the efficiency and sustainability of agricultural production systems. In this study, in addition to evaluating the ability of the CERES-Maize model to simulate crop yield and growth and development stages, management of optimal water and nitrogen consumption along with the use of biochar and hydrochar in corn cultivation has been done and by estimating the amounts of nitrogen losses and absorption by the model, nitrogen leaching and the amount of effluent were also estimated.
Materials and methods: In order to calibrate and evaluate the CERES-Maize model in simulating the production of SC-704 and SC-611 corn cultivars under two spring and summer cropping systems in the climatic conditions of Ahvaz, field experiments were conducted in a lysimeter as a factorial experimental design in a randomized complete block design with three replications. To collect and create the database required for the model, field experiments were conducted, including measurements of dry matter weight at each stage, total biomass nitrogen, grain yield, effluent drainage rate, total drainage nitrogen, and leaf area. In this study, in addition to urea fertilizer, the effect of biochar and hydrochar derived from sugarcane bagasse on growth and development, grain yield, effluent drainage rate, and nitrogen leaching was also investigated. To quantify the plant reaction and calculate the efficiency of water and fertilizer consumption, two levels of irrigation, full (I1) and dificit up to 30% (I2) were considered as the first factor and two levels of nitrogen, 200 (N1) and 160 (N2) kg/ha were considered as the second factor. The indices of normalized root mean square error (NRMSE), Wilmot's agreement index (d), coefficient of determination (R2), and mean absolute error (MAE) were used to evaluate the effectiveness of the model in simulating the desired data and comparing them with the measured data.
Results: Based on the results of control treatments (optimal conditions and no stress) with the application of all three nitrogen sources (urea, biochar, and hydrochar), genetic coefficients were determined for both SC-611 and SC-704 cultivars in the Ahvaz region. The results of the model calibration for the use of each of these sources showed that the model is able to simulate the growth and development characteristics of corn cultivars in the climatic conditions of Ahvaz with minimal differences, which indicated the high accuracy of the calculated genetic coefficients.
The simulation steps for other experimental treatments (except those used in the calibration phase) were repeated for all three nitrogen sources in both cropping seasons and compared with the results of observations from field experiments to validate the model. From the information of grain yield, maximum leaf area index, total biomass weight, total biomass nitrogen, total drainage water and total drainage nitrogen that were extracted from other experimental treatments, It was used to determine the validity of the model. The model validation results showed that the model has a high ability to predict dry matter weight, biomass nitrogen, grain yield, and maximum leaf area index. Although the model accurately simulated the trend of effluent drainage changes in different water and nitrogen treatments but the model performed poorly in predicting effluent drainage. Both in field and simulated conditions, the use of biochar and hydrochar instead of urea fertilizer increased grain yield, but leaf area index, total dry weight, and effluent drainage decreased.
The results of fitting the linear regression between the observed and simulated data and comparing it with the 1:1 line also showed that by using different sources of nitrogen in both cropping seasons, the model can predict more than 90% of the changes in grain yield, nitrogen leaching and the maximum observed leaf area index. Under optimal and no stress conditions, combining biochar and hydrochar with urea fertilizer (I1B1 and I1H1) significantly reduced runoff in spring and summer crops. This decrease in comparison to the application of urea alone (I1N1), was 15.7 and 13.1 percent in spring cultivation and 17.4 and 14.4 percent in summer cultivation, respectively. In spring and summer cultivation using all three sources of urea, biochar, and hydrochar, the MAE for nitrogen leaching is between 0.53-1.73 and 0.31-2.43, respectively. The amount of nitrogen leaching is a function of the amount of drainage and only the movement of nitrate and urea is considered in the simulation of nitrogen leaching from the soil layers by the model. In general, in the experimental treatments in both cropping seasons, the model simulates the nitrogen leaching values slightly more than the measured values.
Conclusion: The CERES-Maize model was able to predict the growth and development characteristics of maize cultivars grown in the Ahvaz region with relatively high accuracy and minimal error. However, the results indicated that the model’s accuracy in estimating drainage during both spring and summer cropping seasons was limited, highlighting the need for calibration or adjustment of parameters related to water and soil management. Despite this limitation, CERES-Maize can serve as an effective and reliable tool to support managerial decision-making and agricultural planning under the climatic conditions of Ahvaz. Utilizing this model can significantly reduce the time and costs associated with field studies, enabling farmers and planners to develop optimal cultivation, irrigation, and soil management strategies based on model simulations. Therefore, although improvements in drainage estimation are needed, the model holds substantial potential to enhance the efficiency and effectiveness of regional agricultural planning.