Linking river sediment arsenic to catchment spatial attributes in agricultural landscapes.

Abstract

Understanding of the sources and processes involved in the heavy metal accumulation in river sediments is important for measuring the risks associated with human exposure. Hence, an integrated modeling approach was designed to study the linkage between landscape-related natural and anthropogenic features and high arsenic levels at the outlet of six catchments on the Ghareh-Ajagh River, central Iran. Sediment arsenic levels were measured in 8 months from October 2018 to November 2019 when the river sediment and water conditions were stable and ranged from 16.3 to 78.5 mg/kg. Monthly catchment-level agricultural areas were extracted from Landsat 8-OLI images. Predictive variables included NDVI values; area and spatial patterns of agriculture measured using four landscape metrics of NP, PD, MPS, and ENN; length and slope of the streams extended from main agricultural areas to the catchment outlet; and four hydrologic soil groups. The best-fitted multiple regression model (r2 = 0.763, p≤ 0.05) with the Akaike information criteria of 105.07 was developed using stream length, soil group C, and area and PD of agricultural areas. Results showed that sediment arsenic levels increase with increasing quantity and density of agricultural activities that were close to the river outlet and increasing proportion of silty loam or loamy soils but are relatively less dependent on agricultural structural patterns. These insights are helpful to inform policy decisions regarding the processes involved in river contamination in central Iran.