Abstract
Land use impacts from agriculture, industrialization, and human population should be considered in surface water quality management. In this study, we utilized an integrated statistical analysis approach mainly including a seasonal Mann-Kendall test, clustering analysis, self-organizing map, Boruta algorithm, and positive matrix factorization to the assessment of the interactions between land use types and water quality in a typical catchment in the Huai River Basin, China, over seven years (2012-2019). Spatially, water quality was clustered into three groups: upstream, mid-stream, and downstream/mainstream areas. The water quality of upstream sites was better than of mid-, down-, and mainstream. Temporally, water quality did not change significantly during the study period. However, the temporal var-iation in water quality of up-, down-, and mainstream areas was more stable than in the midstream. The interactions be-tween land use types and water quality parameters at the sub-basin scale varied with seasons. Increasing forest/grassland areas could substantially improve the water quality during the wet season, while nutrients such as phosphorus from crop-land and developed land was a driver for water quality deterioration in the dry season. Water area was not a significant factor influencing the variations of ammonia nitrogen (NH3-N) and total phosphorus (TP) in the wet or dry season, due to the intensive dams and sluices in study area. The parameters TP, and total nitrogen (TN) were principally linked with agricultural sources in the wet and dry seasons. The parameters NH3-N in the dry season, and chemical oxygen demand (CODCr) in the wet season were mainly associated with point source discharges. Agricultural source, and urban point source discharges were the main causes of water quality deterioration in the study area. Collectively, these results high-lighted the impacts of land use types on variations of water quality parameters in the regulated basin.