Linking land use with riverine water quality: A multi-spatial scale analysis relating to various riparian strips

Exploring linkages between riverine water quality and land use is of great importance for catchment management and water quality conservation. Relationships between them are complex and site-specific. Therefore, land-use/landscape effects on water quality remain to be investigated. Multivariate statistics and empirical models are used to examine the influences of seasonality (wet and dry seasons), land use in various riparian strips, on water quality seasonality in the river networks of the Three Gorges Reservoir Area. Redundancy analysis (RDA) shows a comparable effect of land use on overall water quality (OWQ) during the two seasons. The variance of OWQ explained is generally higher in the 50 m riparian zone, whilst, the variance is higher along 200 m riparian in the wet season with land use composition presented as actual area. The explanatory variables of land use indices for the OWQ are quite distinct. Multiple linear regression (MLR) modeling indicates that models of water quality parameters (WQPs) depend on seasonality, riparian zone and representation of land use indices (calculated from percentage and actual area), this results in predictor variables of WQPs are highly variable. Land use within the 50 m riparian zone can predict WQPs well, and land use along the riparian zones predicts WQPs better in the dry season. We firstly compared the impacts of multi-riparian land use indices (presented as percentage and actual area) in combination with land use configuration on water quality. Our study provides critical information for efficiently land-use planning to protect water quality.