Abstract The Annualized Agricultural Non-point Source (AnnAGNPS) model can be used to analyze the effects of management practices on sediment loads in agricultural watersheds. The study was performed in a 506 km2 Mediterranean watershed located in Apulia, Southern Italy, planted with mostly winter wheat (83%) where runoff and sediment loads have been monitored at an in-stream gage. The AnnAGNPS model was used to predict runoff and sediment load without calibration during a 5-year period. On an annual scale, the model showed good prediction capability for runoff (R2 = 0.8, NSE [Nash and Sutcliffe coefficient of efficiency] = 0.7) and satisfactory results for sediment load (NSE = 0.5, R2 = 0.5). Based on the current conditions of the watershed as a baseline scenario, the effectiveness of alternative conservation practices applied within the watershed was also evaluated. No-tillage practices applied to the entire cropland area reduced soil erosion within fields by 44% and sediment yield from fields to streams by 20%. Reduced tillage decreased soil erosion, sediment yield, and sediment load at the gage location by 12%, 7%, and 4% respectively. Limiting the placement of alternative practices to the cropland sub-watersheds with the most erosion (Scenarios G and H) proved to be a promising and viable approach to sediment erosion reduction throughout the watershed. In this perspective, areas that produce the most sediment were identified and targeted for replacement of varying levels of cropland with forest, and consequently sediment loads were reduced from 5% to 97%. The effect of vegetated streams and riparian buffers as natural traps that can increase the in situ sediment deposition was also considered. Most, but not all of the scenarios discussed herein could realistically be implemented within the watershed, particularly if there are incentive policies. Even considering the intrinsic uncertainty of modeling results, evaluating these systems with the aid of AnnAGNPS serves as a means to provide reference information and allows watershed conservation planners to compare the impacts of different management scenarios with sustainable agriculture guidelines.

中文翻译：

---

在 Carapelle 流域中使用 AnnAGNPS 模型评估替代管理实践

摘要 年化农业非点源（AnnAGNPS）模型可用于分析管理实践对农业流域泥沙负荷的影响。该研究是在位于意大利南部普利亚的一个 506 平方公里的地中海流域中进行的，该流域主要种植冬小麦 (83%)，其中径流和沉积物负荷已在河内测量仪进行监测。AnnAGNPS 模型用于预测 5 年期间未经校准的径流和沉积物负荷。在年度尺度上，该模型显示出良好的径流预测能力（R2 = 0.8，NSE [纳什和萨特克利夫效率系数] = 0.7）和令人满意的泥沙负荷结果（NSE = 0.5，R2 = 0.5）。以流域现状为基准情景，还评估了流域内采用的替代保护做法的有效性。整个农田区域采用免耕做法，使田间土壤侵蚀减少了 44%，从田间到溪流的沉积物产量减少了 20%。减少耕作使测井位置的土壤侵蚀、沉积物产量和沉积物负荷分别降低了 12%、7% 和 4%。将替代做法限制在侵蚀最严重的农田子流域（情景 G 和 H）被证明是减少整个流域沉积物侵蚀的一种有前途和可行的方法。从这个角度来看，确定了产生最多沉积物的区域，并将不同水平的农田替换为森林，因此沉积物负荷从 5% 减少到 97%。还考虑了植被河流和河岸缓冲区作为可以增加原位沉积物沉积的天然陷阱的影响。大多数（但不是所有）这里讨论的情景都可以在流域内实际实施，特别是如果有激励政策。即使考虑到建模结果的内在不确定性，在 AnnAGNPS 的帮助下评估这些系统也可以作为提供参考信息的一种手段，并允许流域保护规划者将不同管理方案的影响与可持续农业指南进行比较。