Regional geochemical baseline based pollution assessment and dominant factors influencing metals in riparian sediment of Yangtze River, China.

• Regional geochemical baseline values can well model the background values of studied metals. • Cd was the main pollutant in the riparian sediment. • The contamination and risk of Cd was high in upper downstream and site of 24. • Influencing factors explained variation of metals better at 1 km buffer scale. • Land use composition contributed most to the high explanation at 1 km scale. In this study, we assessed pollution status and potential risk of metals in riparian sediment of Yangtze River with geo-accumulation index (I geo), ecological risk index (RI) and health risk index (THQ) by calculated regional geochemical baseline (RGB) values. We explored the influence of natural factor (sediment property) and anthropogenic factors (land use composition and landscape configuration) at different buffers (500 m, 1 km and 5 km) on metals in sediment using redundancy analysis (RDA) and variation partitioning analysis (VPA). Results indicated that RGB values can effectively model background values of metals (p < 0.01). I geo and RI assessment showed that Cd was the primary pollutant, which showed no contamination to moderate contamination (I geo > 0) and moderate ecological risk (ecological risk factor > 40) in upper downstream of the river. Besides, Cd showed moderate to strong contamination (I geo > 2) and considerable ecological risk (ecological risk factor > 160) at site 24, which is located in the intersection of Poyang Lake and Yangtze River and strongly affected by natural erosion and human activities of Poyang Lake drainage. THQ assessment indicated low health risk of metals from consuming fish influenced by suspended sediment. RDA and VPA analysis indicated that explanation rate of land use composition, sediment property and landscape configuration on metals was 27.01–55.1 %, 36.66 %, 9.77 %-12.52 %, respectively. The total explanation rate was significantly higher at 1 km scale (82.5 %), which was mainly contributed by land use composition and its interaction with sediment property and landscape configuration (55.11 %). Specifically, percentage of farmland, grassland, wetland and forest, content of TN, TOC, silt and clay in soil and landscape configuration index of NumP were positively correlated with metals, and content of sand in soil and landscape index of MSI negatively correlated with metals in riparian sediment at different buffer scales. Our results offer valuable insights for safeguarding riparian zone. [ABSTRACT FROM AUTHOR]