Application of improved risk assessment methods to best management practices research for non-point source pollution: From a hilly mountainous-dominated region of Southern China.

[Display omitted] • NPSP in typical, hilly mountainous areas was evaluated at annual and monthly scales. • Critical source areas were determined based on the multifactor method. • NPSP reduction and cost efficiency in different BMPs were simulated. • BMPs for the control of NPSP in typical, hilly mountainous areas were proposed. Non-point source pollution (NPSP) significantly affects regional water environments worldwide. Identifying critical source areas (CSAs) and implementing best management practices (BMPs) are essential to prevent and control NPSP. This study used the Soil and Water Assessment Tool model to simulate and evaluate the characteristics of NPSP in the Gongjiang Watershed, a typical, hilly mountainous area in Southern China. The study identified CSAs of NPSP using total nitrogen (TN), total phosphorus (TP), water quality level, and population density factors. The reduction effect of different BMPs on the NPSP load was evaluated, and the cost-effectiveness (CE) of various BMPs was evaluated for the control of NPSP in the watershed. The interannual variation in NPSP load was high from 2013 to 2018, with the largest pollution load in 2016 (wet year) and the smallest in 2018 (dry year). The pollution load fluctuated throughout the year and was related to fertilization and precipitation runoff. The spatial heterogeneity of NPSP was large, with TN and TP losses showing a spatial pattern of high in the south, low in the north, and highest in the center. The CSAs of the NPSP covered 1566.66 km2, representing 9.77 % of the total watershed area. However, TN and TP loads accounted for 19.39 % and 23.53 % of total watershed production. The order of effectiveness of individual management measures in reducing TN was stubble coverage (SC) > vegetative buffer strips (VBS) > terrace engineering (TE) > grassed waterways (GW) > 20 % fertilizer reduction (FR20) > 10 % fertilizer reduction (FR10). The order of effectiveness of TP reduction was SC > VBS > TE > FR20 > GW > FR10. Combined management measures were much more effective in reducing TN and TP than individual measures, with an average increase in TN and TP reduction of 20.59 % and 25.23 %, respectively. Based on the results of the CE evaluation of BMPs, a combination of FR20, GW, and VBS was more effective than individual measures. We suggest a combination of FR20, GW, and VBS measures for effective NPSP prevention and control in Southern China's typical, hilly mountainous areas. [ABSTRACT FROM AUTHOR]