Where is basalt in river sediments, and why does it matter?

International audiencethose processes affect the geochemistry of sediments is however not yet fully understood. Here, we report trace element data as well as Nd, Hf and Pb isotopic compositions of sediments sampled at different water depths in the Ganges, Yamuna and Chambal Rivers draining the Deccan Trap basalts and the crystalline and sedimentary rocks from the Himalayan mountain range and the northern Indian shield. Isotopic differences between surface and bed sediments sampled at the same location reach 6 εNd and about 15 εHf units, suggesting that suspended loads and bedloads do not carry similar provenance information. Such differences are explained by the combined effects of differential erosion and mineralogical sorting processes during fluvial transport. Materials eroded from basalts are preferentially transported in suspension near the river surface while materials eroded from more crystalline precursors are transported near the bottom of the river. This depth-dependent provenance within the river channel leads to an overrepresentation of basaltic materials in fine39 grained suspended loads that are finally delivered into the ocean and become part of the oceanic terrigenous clays. By contrast, the proportion of basaltic materials in coarser sediments such as bedloads or turbidites is underestimated.Our results have important consequences on the use of Nd, Hf and Pb isotopic compositions of sediments as provenance proxies because they indicate that all grain-size fractions must be taken into account to properly trace source compositions. They also suggest that upper continental crust estimates derived from fine-grained sediments, such as suspended loads, may be biased towards basaltic compositions if basaltic outcrops are present in the drainage area