Despite the significant impacts of natural factors such as rainfall, topography, soil type, and river network as well as agricultural activities on the environmental water quality, little is known about the influence of their temporal and spatial variations in a fluvial-lacustrine watershed. In this study, a whole process accounting method based the export coefficient model (WP-ECM) was first developed to quantify how natural factors and agricultural activities distribution influenced water quality. A case study was performed in a typical fluvial-lacustrine area –Dongting basin, China. The simulated results indicated that the natural factors can promote the migration and transformation of agricultural pollutants generated from the watershed and the spatial distribution of the natural factors displayed high variability. It should be priority to monitor the areas with greater natural impact in the basin. Moreover, the cultivated land area and the number of pig-breeding were positively correlated with the pollutant discharge, and it is an important measures to reduce and control the anthropological influence in the agricultural high-impact areas. From the perspective of the spatial distribution of comprehensive influence, the comprehensive high-impact areas are mainly distributed in the Dongting Lake district in 2005–2010 and in Xiang River watershed in 2010–2020. A key strategy for controlling or reducing the cultivated land area and the intensity of livestock breeding in these high-impacts areas, especially in Dongting Lake district and Xiang River watershed, is to reduce the impact of the environmental water quality for the entire basin.