Nitrogen (N) from farm fields is a source of pollution to fresh and marine waters. Modifying N fertilizer application rate and timing to consider the spatial and temporal variability in plant N requirements could reduce N losses from farmlands, resulting in improvements to surface water quality. In this study, the field-scale hydrologic and N simulation model DRAINMOD-NII was used to predict nitrate–N losses from fields planted in a corn-soybean rotation at Waseca, Minnesota, USA, over a 15-year period (2003–2017) for two fertilizer application treatments. The N fertilizer treatments simulated included a single uniform fertilizer application in the spring before planting and a variable rate N practice (VRN) where fertilizer was applied as a split pre-plant, side-dress application, based on in-season monitoring of plant N requirements to determine fertilizer rate. Measured discharge (2003–2008) and nitrate–N concentrations in subsurface drainage (2003–2008 and 2016–2017) at the site were used to calibrate discharge and nitrate–N losses in model simulations and validate model performance for uniform vs VRN fertilizer management. Measured nitrate–N concentrations in weekly samples were 13% lower for fields utilizing VRN versus a single spring application in 2016, and 18% lower in 2017. Model predictions of nitrate concentrations based on daily predictions of discharge accurately matched observed data for these years, predicting reductions of 23% and 19% for the years 2016 and 2017, respectively. The results of model simulation for the 15-year period indicated that changing the timing of fertilizer application from a single application to a VRN application could reduce annual N loads lost in drainage by 40%.

中文翻译：

---

施肥时间和可变氮肥对使用 DRAINMOD-NII 模拟的瓷砖排水玉米-大豆轮作中硝酸盐-N 损失的影响

农田中的氮 (N) 是淡水和海水的污染源。修改氮肥施用率和时间以考虑植物氮需求的时空变异性，可以减少农田的氮损失，从而改善地表水质量。在这项研究中，田间规模的水文和 N 模拟模型 DRAINMOD-NII 用于预测美国明尼苏达州 Waseca 15 年（2003-2017 年）玉米 - 大豆轮作种植田的硝酸盐 - N 损失) 进行两次施肥处理。模拟的氮肥处理包括在种植前的春季单一均匀施肥和可变氮肥实践 (VRN)，其中施肥作为分开的种植前施肥、侧施施肥、基于对植物氮需求的季节性监测来确定施肥量。现场测量的排放量（2003-2008）和地下排水（2003-2008 和 2016-2017）中的硝酸盐-N 浓度用于校准模型模拟中的排放量和硝酸盐-N 损失，并验证模型在均匀施肥与 VRN 施肥管理中的性能. 与 2016 年使用 VRN 的单一春季应用相比，每周样本中测量的硝酸盐-N 浓度降低了 13%，2017 年降低了 18%。基于每日排放预测的硝酸盐浓度模型预测与这些年的观测数据准确匹配，预计 2016 年和 2017 年分别减少 23% 和 19%。