Seasonal Hydrology and Projected Landscape Changes Jointly Shape Future Water Quality in a Typical Watershed of the Three Gorges Reservoir Area

Land use and topography jointly influence pollutant dynamics and landscape functionality, yet the spatiotemporal scale dependence of water quality responses to landscape patterns remains largely unexplored. To address this, we used the Binan River as a case study, integrating redundancy analysis, Bayesian regression, and the FLUS model to assess water quality across hydrological periods and quantify the spatiotemporal scale effects under future scenarios. Results showed that most pollutants were significantly higher in the dry season due to reduced dilution capacity from low runoff. A 1500-m riparian buffer zone was identified as the optimal analytical scale, where the model consistently explained over 90% of water quality variations in all hydrological periods. The explanatory power followed the order: landscape composition (51.20%–94.27%) > landscape configuration (38.22%–65.98%) > physiographic factors (0.67%–31.18%). Notably, key landscape metrics such as the Largest Patch Index (LPI), forestland area percentage (For), and gardenland area percentage (Gar) had contrasting effects on water quality during the wet and dry seasons. During the dry season, the key landscape metrics controlling total nitrogen (TN) were identified as Edge Density (ED) and Number of Patches (NP). Under future scenarios of aggravated landscape fragmentation, TN concentrations decreased while TOC and CODMn concentrations significantly increased compared to the 2024 dry season. The study suggests that future water quality management should focus on the dry season by prioritizing the mitigation of landscape fragmentation and the scientific planning of natural landscape types.