Evaluation of the Aqua‎Crop Model to Simulate Maize Yiled Response under Salinity Stress

Introduction: Limited water resources and its salinity uptrend has caused reducing water and soil quality and consequently reducing the crop production. Thus, use of saline water is the management strategies to decrease drought and water crisis. Furthermore, simulation models are valuable tools for improving on-farm water management and study about the effects of water quality and quantity on crop yield.. The AquaCrop model has recently been developed by the FAO which has the ability to check the production process under different propositions. The initial version of the model was introduced for simulation of crop yield and soil water movement in 2007, that the effect of salinity on crop yield was not considered. Version 4 of the model was released in 2012 in which also considered the effects of salinity on crop yield and simulation of solute Transmission in soil profile. Material and methods: In this project, evaluation of the AquaCrop model and its accuracy was studied in the simulating yield of maize under salt stress. This experiment was conducted in Karaj, on maize hybrid (Zea ma ys L) in a sandy soil for investigation of salinity stress on maize yield in 2011-2012. This experiment was conducted in form of randomized complete block design in four replications and five levels of salinity treatments including 0, 4.53, 9.06, 13.59 and 18.13 dS/m at the two times sampling. To evaluate the effect of different levels of salinity on the yield of maize was used Version 4 AquaCrop model and SAS ver 9.1 software .The model calibration was performed by comparing the results of the field studies and the results of simulations in the model. In calculating the yield under different scenarios of salt stress by using AquaCrop, the model needs climate data, soil data, vegetation data and information related to farm management. The effects of salinity on yield and some agronomic and physiological traits of hybrid maize (Shoot length, root length, dry weight and crop yield) under different levels of NaCl solution osmotic potential were also investigated by SAS ver 9.1 software. Data's mean comparisons were performed by Duncan's multiple range test. To assess the accuracy of AquaCrop Model for Simulation of the Maize Performance under Salt Stress used from Indicators RMSE, MAE, CRM, NSE, d and Er. Results Discussion: The results of RMSE and MAE indices showed that AquaCrop model can simulate maize yield under the salinity stress. Accuracy decreased and crop yield prediction underestimated with increasing salinity from treatment 0 to 18.13 ds/m in the first and second harvest. The highest yield related to salinity treatment of 0 dS/m and the lowest yield related to salinity treatment 18.13 dS/m. yeild simulation error increased by increasing salinity, the highest and lowest error of yield simulation in model respectively related to salinity treatments 18.13 and 0 dS/m. The highest and lowest error was in the first harvest respectively 0.56 and 13.1 percent and in the second harvest respectively 0.42 and 21.79 percent, that in the comparison with the results of studies conducted by Steduto and colleagues on maize is not much different. The results comparison in the first and second harvest showed that soil salinity was increased by increasing irrigation number in second harvest, so the error in second harvest is greater than first harvest and the maximum error is related to treatment 18.13 ds/m in the second harvest 21.79 percent.The coefficient of determination R2 for the first and second harvest is respectively 0.850 and 0.834, that indicates a high correlation between yeild values of measured and predicted by the AquaCrop model. CRM index was negative and near zero in both harvest under Salinity different scenarios. According to CRM value, AquaCrop model was overestimated and the model was simulated maize yield under the salinity stress a little more than measured yield. The d statistic index value is close to unity, indicates that yield values in model is compatible with actual values. NSE index was calculated for the first and second harvest respectively 0.81 and 0.84, that is close to one and showed that the model has suitable performance in the yield simulation. Comparison of means by Duncan's multiple range test and analysis of variance in the software SAS ver 9.1 indicated Salinity has a very significant effect on all traits including shoot length, root length, dry weight and crop yield that all traits were decreased significantly by increasing salinity. Conclusion: Comparison of the results of AquaCrop model and statistical analysis in software SAS ver 9.1 showed that maize yield was reduced with increasing salinity. According to index CRM, AquaCrop model was simulated maize yield under the salinity stress more than measured yield in farm. The results showed that the AquaCrop model simulated well maize yield in moderate and low stress, but accurately simulation slightly decreased in high stress. The results of this study was compared with other research and indicated that the error values of AquaCrop model in Karaj is not much different with the error values of other research.