Abstract Agricultural system modeling has become an effective tool for analyzing and quantifying the effects of varying management practices and environmental conditions on crop production and nutrient export from croplands. The use of such modeling tools provides useful insights in identifying the most effective management practices for enhancing productivity, sustainability, and resiliency of agricultural systems. This study focuses on testing and application of an integrated field-scale process-based model, DRAINMOD-DSSAT, for simulating hydrology, nitrate-nitrogen (NO3-N) loss, and crop growth and yield responses in artificially drained croplands. The tested model was used to evaluate the effects of different N fertilizer application rates and timings in both conventional and controlled drainage conditions on crop yield and NO3-N losses in a poorly drained Drummer-Flanagan soil in east-central Illinois. The model was calibrated and validated for a corn [Zea Mays L.] – soybean [Glycine Max (L.)] rotation using 7 years (1992–1998) of field-measured tile drainage, crop yield, and NO3-N data. The graphical and statistical evaluations indicated very good model performance. Specifically, monthly tile drainage flow was predicted with modeling efficiency (NSE), index of agreement (d), and mean absolute error (MAE) of 0.85, 0.96, and 0.69 cm, respectively. Monthly NO3-N losses were predicted with NSE, d, and MAE of 0.82, 0.95, and 1.16 kg N ha−1, respectively. Corn and soybean yields were predicted with an absolute percent error (PE) of 1.84 and 12.07, respectively. The long-term model simulation results indicated that split N application of 50 % during spring-pre plant (S) and 50 % during side-dressing (SD) could increase crop yield and reduce N leaching losses compared to other tested N application methods: spring (S) only, fall-spring split (F-S), and fall-spring-side-dressing (F-S-SD). Further, applying 10 % and 20 % reduced N rates (194 kg N ha−1 and 174 kg N ha−1, respectively) in combination with S-SD split application in controlled drainage (CD) condition could reduce N leaching losses by 30 % and 33 %, respectively compared to the conventional application method. This study explored the effects of N fertilizer management on crop yield and nitrogen losses under two drainage conditions, and underscored the importance of N fertilizer application rates and timings for achieving yield goals while minimizing nitrogen export from drained agricultural fields. The results of the model simulations will be useful for stakeholders and policy makers while making decisions regarding promotion and adoption of best management practices for drained agricultural landscapes.

中文翻译：

---

在伊利诺伊州中东部使用 DRAINMOD-DSSAT 评估减少氮损失的策略

摘要 农业系统建模已成为分析和量化不同管理实践和环境条件对作物生产和农田养分输出影响的有效工具。此类建模工具的使用为确定最有效的管理实践以提高农业系统的生产力、可持续性和弹性提供了有用的见解。本研究的重点是测试和应用基于综合田间规模过程的模型 DRAINMOD-DSSAT，以模拟人工排水农田的水文、硝酸盐 - 氮 (NO3-N) 损失以及作物生长和产量响应。测试模型用于评估在伊利诺伊州中东部排水不良的 Drummer-Flanagan 土壤中，常规和受控排水条件下不同氮肥施用率和时间对作物产量和 NO3-N 损失的影响。该模型针对玉米 [Zea Mays L.] - 大豆 [Glycine Max (L.)] 轮作使用 7 年（1992-1998）的田间测量瓷砖排水、作物产量和 NO3-N 数据进行校准和验证。图形和统计评估表明模型性能非常好。具体而言，利用建模效率 (NSE)、一致性指数 (d) 和平均绝对误差 (MAE) 分别为 0.85、0.96 和 0.69 cm 来预测每月瓷砖排水流量。预测每月 NO3-N 损失的 NSE、d 和 MAE 分别为 0.82、0.95 和 1.16 kg N ha-1。预测玉米和大豆产量的绝对百分比误差 (PE) 分别为 1.84 和 12.07。长期模型模拟结果表明，与其他测试的施氮方法相比，春季预植 (S) 期间分施 50% 和侧施 (SD) 期间分施 50% 的氮可以提高作物产量并减少氮浸出损失：仅弹簧 (S)、秋季弹簧分裂 (FS) 和秋季弹簧侧敷料 (FS-SD)。此外，在受控排水 (CD) 条件下，将 10 % 和 20 % 降低的氮量（分别为 194 kg N ha-1 和 174 kg N ha-1）与 S-SD 分流应用相结合，可以将 N 浸出损失减少 30 % 和 33 %，分别与传统应用方法相比。本研究探讨了两种排水条件下氮肥管理对作物产量和氮素流失的影响，并强调了氮肥施用率和时间对实现产量目标的重要性，同时最大限度地减少来自排水农田的氮输出。模型模拟的结果将有助于利益相关者和政策制定者，同时做出有关促进和采用排水农业景观最佳管理实践的决策。