Scale effects of multi-medium heavy metals in response to landscape indices in the Yuan River, China.

Heavy metals in river basins can affect the entire ecosystem through water recycling. Different heavy metals are significantly affected by riparian landscape patterns due to the difference of source and migration processes, while the intensity of impact is directly influenced by the spatial scale of landscape patterns. Here, we employ correlation analyses (CA) and redundancy analyses (RDA) to determine spatial scales for which multi-medium heavy metals (Cr, Mn, Cu, Zn, As, Cd and Pb) are most sensitive landscape indices in the riparian zone of the Yuan River. Results show that the most sensitive spatial scales for dissolved heavy metals were the 300 m line buffer in the wet season and catchments in the dry season. The most sensitive spatial scales for suspended particulate heavy metals were catchments in the wet season, and the 300 m and 500 m line buffers in the dry season. However, correlations between heavy metals in sediment and landscape indices were unclear at all spatial scales. The spatial scales of the 100 m point buffer and catchments showed the highest explained variances for all media. The most important landscape indices for dissolved heavy metals were ENN_MN (Mean Euclidean Nearest Neighbor Distance) and LPI (Largest Patch Index), and SHDI (Shannon's Diversity Index) and ED (Edge Density) for particulate heavy metals. More accurate spatial scale effects of landscape indicators can provide data and reference for similar watershed studies. [Display omitted] • The responses of heavy metals in surface water, suspended particle, and sediment on landscape indices were studied. • The point buffer 100 m and catchment spatial scales showed the higher explained variances of all media. • The most important landscape indices for dissolved heavy metals were ENN_MN and LPI. • The highest explained variances for particulate heavy metals were from the landscape indices of SHDI and ED. [ABSTRACT FROM AUTHOR]