Suitability analysis of nitrogen fertilizer management on dryland of Loess Plateau based on root zone water quality model

Abstract: Luxurious and improper application of nitrogen (N) is an obvious problem on the Loess Plateau. Employing a crop model to explore an optimal N fertilizer management for local farmers is an effective research method. In the present study, we firstly selected a continuous eight-year winter wheat-spring maize experiment (from 2004 to 2011) and a two-year winter wheat N fertilizer experiment (from 2010 to 2012) to calibrate and validate root zone water quality model (version 2, RZWQM2) aiming to assess the adaptability of this model on the Loess Plateau. We set up the following 8 scenarios of N application rate of base fertilizer for winter wheat in the RZWQM2: 0, 30, 60, 90, 120, 150, 180 and 210 kg/hm2 aiming to search for proper N application rate of the base fertilizer. Based on the proper N application rate (90 kg/hm2 in this study) of base fertilizer, we also set up the following 5 scenarios of N application rate with the ratio of base to topdressing in the RZWQM2: 5:1, 5:2, 5:3, 5:4 and 5:5 in wheat turning green, jointing and heading stages, respectively. In this way, we aimed to search for the best N topdressing stage and the proper N topdressing rate for winter wheat. The calibrated and validated RZWQM2 was used to characterize the respond of winter wheat yield, water use efficiency (WUE), N uptake of grain and N harvest index to the different N application rates and timings based on the long historical meteorological data (from 1956 to 2012). The results demonstrated the capability of the RZWQM2 to efficiently simulate crop growth, water movement (100 cm depth), and final crop yield in the different rainfall year on the Loess Plateau; meanwhile, RZWQM2 effectively simulated the yield and N uptake of winter wheat under different N application method. Both of the results provided a solid basis for the application of RZWQM2 to study the N fertilizer management on the Loess Plateau. In the simulations, the wheat yield, WUE and N uptake of grain increased fast before the base N fertilizer of 90 kg/hm2 and slowly afterwards. The N harvest index in the wet year increased with the N application rate of base fertilizer and was stable after the N application rate of 180 kg/hm2; it maintained a high level when the N application rate of base fertilizer was between 90 and 180 kg/hm2 in the medium year, as well as between 60 and 150 kg/hm2 in the dry year. Thus, the single base N fertilizer of 120-150 kg/hm2 could adequately satisfy the need of wheat growth in the dryland of the Loess Plateau in different rainfall years. The optimal topdressing stage of N fertilizer was between turning green and jointing stage because of the high wheat yield. On the basis of base N fertilizer of 90 kg/hm2, the wheat yield, WUE and N uptake of grain increased with the topdressing amount in the turning green stage; the N harvest index in the wet year always increased with the N topdressing rate; the N harvest index peaked in the medium year on the topdressing N rate of 36 kg/hm2 and in the dry year on the topdressing N rate of 18 kg/hm2. We found topdressing N rate of 54-72, 36-54, and 18-36 kg/hm2 respectively in the wet, medium and dry year could satisfy the need of high wheat yield and maintain high level of N harvest index. These findings provide some useful guidance for local farmers to optimize their N fertilizer management. More field experiments need to be conducted to verify the conclusions of this study in the future.