Increase of water erosion is one of the main problems in watersheds all over the world and directly affects soil and water resources, causing economic and environmental losses. We proposed a combination of Genetic Algorithm and Erosion/Deposition modelling to allocate vegetation zones in order to reduce soil losses within watersheds. In this study, we have used Unit Stream Power Based Erosion Deposition (USPED) model to predict the spatial distribution of erosion and sediment deposition. Based on USPED estimates, Genetic Algorithm optimize site as well size of forest restoration zones in order to reduce soil losses. Proposed approach considered watershed entirely, in a cell-level optimization, to find the best possible arrangement of forest zones. For initial scenario, soil loss estimate was 2.045,55 ton. ha⁻¹. year⁻¹ and after optimization we achieved 52% reduction. Optimized forest zones formed connected and compacted patches, which enable field implementation. Methodology is also applicable to other watersheds around world.

水蚀的加剧是全世界流域的主要问题之一，直接影响土壤和水资源，造成经济和环境损失。为了减少流域内的土壤流失，我们提出了遗传算法和侵蚀/沉积模型相结合的方法来分配植被区。在这项研究中，我们使用了基于单位流功率的侵蚀沉积（USPED）模型来预测侵蚀和沉积物的空间分布。基于USPED的估算，遗传算法可优化森林恢复区的面积和大小，以减少土壤流失。提议的方法在单元级优化中完全考虑了分水岭，以找到最佳的森林区域布置。对于初始方案，土壤流失估计为2.045.55吨。ha ^-1。- ^1年，经过优化，我们减少了52％。经过优化的林区形成了相互连接并压实的斑块，从而可以进行现场实施。方法学也适用于世界其他地区。