Evaluation of AquaCrop for Yield and Water Use Efficiency Simulation of Corn with Different Irrigation Management under Salinity Stress

Corn is one of the most important crops that is well adapted to arid and semi-arid regions. Amount and quality of irrigation water has an important effect on its yield. Therefore, different researchers have tried to study the effect of different irrigation methods and qualities on corn’s yield and biomass. The Iran\s arid and semi-arid climate confirms that it is necessary to study the best irrigation methods along with different qualities of irrigation water on the growth of this crop. In order to save time and money, crop models have been proposed to simulate the response of plants to different field conditions. AquaCrop, provided by the Food and Agriculture Organization (FAO), is one of the best crop models due to the simplicity, low input data, user-friendliness, high accuracy and acceptable proximity of modeling conditions. The present study was conducted using data collected by Minaei (2014) in a research farm in Ahvaz during two cropping seasons (2012-2013) on corn crop. In this research, irrigation types (D: sprinkler irrigation with saline water, F: sprinkler irrigation with saline and fresh water, and S: surface irrigation) and water quality (S1: 2.5, S2: 3.2, S3: 3.9, S4: 4.6 and S5: 5.1 dS.m-1) were examined. The minimum salinity was selected equal to the available water source as S1 treatment, which was the Karun River. On the other hand, the maximum salinity was also considered as S5, to reduce the corn yield by 50%. The S2, S3 and S4 treatments were also the intermediate between these two treatments. The highest and lowest differences between the simulated and observed yields were 0.7 and 0.1 ton.ha-1, obtained from SS5 and FS1 treatments, respectively. The average of these values was equal to 0.3 ton.ha-1. It can be concluded from the results that the model had an average proximate error of 300 kg. Based on the results, the maximum and minimum values of the biomass differences were 2 and 0.1 ton.ha-1, obtained from SS3 and FS3 treatments, respectively. The average difference was 0.7 ton.ha-1. Comparison of biomass results and yield showed that the differences in results between yields were less than those in biomass. If these results are expressed as a percentage; the highest and lowest differences between the observed and simulated yield values were 223 and 31%, respectively. The highest and lowest percentages of these differences in biomass were 252 and 12.6%, respectively. By increasing water salinity, the yield and biomass decreased linearly. These results were observed in all three irrigation treatments. The canopy cover results simulated by the AquaCrop confirm these results (Figure 3). As it can be seen in the figures, the increase in salinity caused a rapid decrease at the end of the growing season. The maximum and minimum differences between the simulated and observed water use efficiency were 0.08 and 0.001 kg.m-3, respectively. The average difference was 0.03 kg.m-3. Based on these results, the AquaCrop had an underestimation error in simulating corn yield in all three irrigation methods. The accuracy of this model was almost the same in all three irrigation methods and was in the excellent category. However, according to Table (5), the efficiency of this model based on EF statistics for each of the used irrigation methods was less than the determined value. The AquaCrop in simulating sprinkler irrigation with saline water (D) had an underestimation error to simulate corn biomass. This model had an overestimation error in simulating corn biomass in the two F and S irrigation methods. The accuracy of this model in simulating corn biomass in S method was less than sprinkler irrigation. The efficiency of AquaCrop in biomass simulation was desirable. The accuracy and efficiency of this model in simulating the water use efficiency of corn in all three irrigation methods were almost the same and acceptable. Comparison of two NRMSE statistics for two parameters of performance and water use efficiency represented that the model was more accurate in simulating performance. According to the results, AquaCrop had acceptable efficiencies for simulation of yield, biomass and soil salinity. However, its efficiency for water use efficiency was not acceptable. In addition, the AquaCrop’s results were the same for all three irrigation types. Then, irrigation type had not any effect on AquaCrop’s accuracy and efficiency.