Abstract: Critical source areas (CSAs), characterized by severe soil erosion and high sediment yield, are considered to have a high priority for conservation. How to identify CSAs and assess the effectiveness of conservation practices is a key issue in site-specific watershed management. The Soil and Water Assessment Tool (SWAT) model is a useful tool for site-specific conservation practices design and several studies have attempted to identify CSAs based on watershed models. However, limited research has reported about the effectiveness of conservation practices targeting CSAs. The aim of this study was to assess the effectiveness of conservation pracrices targeted on CSAs using the SWAT model. CSA was firstly identified based on the 4-year average yearly erosion of each HRU. Appropriate soil conservation practices were then designed for the CSAs. A scenario with conservation practices for the whole watershed was also established as the contrasting counter parts scheme and then compared to the outcome of CSA-targeted conservation pracrices. The result shows that SWAT can accurately simulate sediment yield in the study area. CSAs were mainly located in slope farmland areas and steep gullies, coinciding with the distribution of land use and slope. The identified CSA covered 20% of the HRUs and contributed on average 44% of sediment yield. Conservation practices targeting CSAs had higher sediment reduction effectiveness (24 115 t km^-2 y^-1) than conservation practice covering the whole watershed (20 290 t km^-2 y^-1). Thus conservation practices targeting CSAs are more effective than broad conservation practices. We conclude that soil conservation practices focusing on CSAs do increase sediment reduction effectiveness. Targeting the placement of soil conservation practices based on the CSAs concept will assist water quality control in watersheds.

Key words: effectiveness assessment, critical source areas, SWAT, HRUs, soil conservation practice