The effect of water and nitrogen on drip tape irrigated silage maize grown under arid conditions: Experimental and simulations

Abstract

Nutrients and nitrate move to the boundary of the wetted zone under drip-tape irrigation (DTI) of row crops, so irrigation depth, wetted width, and irrigation frequency are the most important factors in managing the wetting front and keeping N in the root zone. Many farmers in arid regions who have changed their irrigation systems to DTI, but they are still applying N in two splits as they are used for surface and sprinkler irrigation. Crop models can be used to optimize the timing and amount of nitrogen fertilizer applied to minimize both N leaching and the amount of N remaining in the soil after harvest. The objectives of this study were to calibrate the Cropping System Model (CSM)-CERES-Maize for two maize hybrids and to evaluate the model under DTI with two different soil wetting widths. The single cross Hybrid 704 (SC704) grown in two experimental fields (Exp-1 and Exp-2) and the single cross Hybrid 606 (SC606) grown in one experimental field and two farmer fields (Exp-3, Ff-1 and Ff-2) were evaluated during three years (2016, 2017 and 2019). Plant traits including leaf area index (LAI), total biomass (TB), soil moisture (SM), nitrogen uptake (NU) and water productivity (WP) were measured. Soil NO3-N was measured at three distances from the planting row and three soil depths in Exp-1 and Exp-3 during two maize growth stages. The fraction of the wetted width (fw) along drip-tape was 70% in Exp-1% and 100% in Exp-3 (referred to as fw-70 and fw-100, respectively). The results showed good performance of the model for simulating TB and LAI with NRMSE < 21.9% for the two hybrids in the applied N fertilizer treatments. The accuracy of the simulation of SM and WP in fw-100 was better than that of fw-70. Simulated NO3-N followed the observed trend for fw-100, while model performance varied for the different irrigation and N levels in fw-70. Model accuracy for soil NO3-N prediction decreased for the high N fertilizer application under deficit irrigation management. These differences were related to the wetting pattern volume, the distribution of N in the soil profile, and accumulation of soil N under DTI. Overall, the model can simulate TB and NU with high accuracy, SM with good accuracy, and soil NO3-N with an acceptable accuracy under DTI. The accuracy of the model was also higher for the unlimited wetting width, as compared to the limited one. This study showed that the CSM-CERES-Maize model can be used for evaluation of nitrogen management practices for drip irrigated maize grown under arid conditions.