Rivers are an important mediator between human activities and the natural environment. They provide multiple functions, including irrigation, transportation, food supply, recreation, and water supply. Therefore, evaluations of water quality and pollution sources are of great significance for ecological restoration and management of rivers. In this study, the improved “vušekriterijumska optimizacija i kompromisno rješenje” (VIKOR in Serbian; in English: Multicriteria Optimization and Compromise Solution), and a geodetector were used to analyze the water quality characteristics and pollution sources of the Maotiao River Basin (Gizhou province, SW China). The results showed that the water quality of the Maotiao River Basin deteriorated significantly during the summer drought period, as was evident in the reservoirs and lakes. It improved in the wet season (i.e., during the summer period) due to runoff dilution. Water quality decreased along the river’s course, from upstream to downstream sections. The results of the geographic detector analysis showed that agricultural areas were the primary factor affecting the spatial distribution of water quality in the river basin. In July, August, and November 2020, the influence of agricultural land was 0.72, 0.60, or 0.80, respectively, and the interactions among urban, industrial, agricultural, and forested areas explained 99.2%, 83.2%, or 99.9% of the spatial differentiation of water quality, respectively. Due to the influence of spatial scale, settlements have a small influence on the spatial distribution of water quality. Their impact factors were 0.38, −0.24, and −0.05, respectively. Notably, the negative relationship of water quality and forested areas reflects that topography, types of landscapes, and soil thickness have considerable influences on the Maotiao River Basin’s water quality. Based on the findings, we infer that good farmland water conservancy projects and comprehensive management of different types of landscapes, such as forests, agriculture, and urban area and water bodies, are of great significance for improving water quality.

Understanding the accumulation, ecological risk, and source interactions of potentially toxic elements (PTEs) in reservoir sediments is essential for protecting drinking water safety, yet such processes remain insufficiently understood in karst tea-plantation watersheds influenced by mixed anthropogenic activities. In this study, sediment cores collected from four sites (CY-1 to CY-4) during 2022–2024 were analyzed, and an integrated framework combining the Potential Ecological Risk Index (RI), Spearman correlation analysis, Principal Component Analysis (PCA), and Positive Matrix Factorization (PMF) was applied to evaluate contamination characteristics and quantify source contributions. The results revealed significant spatial–vertical heterogeneity of PTEs, with Zn (up to 153 mg/kg) and Cr (up to 64.6 mg/kg) showing the greatest variability, and strong co-enrichment among Cu, Zn, Pb, and Ni (r > 0.85, p < 0.01). Although the overall ecological risk was low (RI = 83.15–106.69), As contributed the highest proportion of risk (28–35%). PCA indicated distinct grouping patterns among elements, while PMF resolved three major sources: domestic sewage and agricultural runoff, agricultural and coal-combustion inputs, and industrial–traffic emissions. Notably, physicochemical parameters (TP, TN, and COD) played important roles in regulating the mobility and partitioning of PTEs by influencing nutrient-associated adsorption processes, organic matter complexation, and redox-related transformations. These findings highlight the multi-source-driven accumulation mechanisms of PTEs in karst reservoirs and provide a scientific basis for targeted pollution control and watershed management in agriculturally impacted regions.