Our results identify distinct sources of different ages and provenance in the Bod¨¦l¨¦ Depression versus the Amazon Basin, effectively ruling out an origin for the Amazonian deposits, such as the Belterra Clay Layer, by long-term deposition of Bod¨¦l¨¦ Depression material. Similarly, no evidence for contributions from other potential source areas is provided by existing isotope data (Sr, Nd) on Saharan dusts. Instead, the composition of these Amazonian deposits is entirely consistent with derivation from in-situ weathering and erosion of the Precambrian Amazonian craton, with little, if any, Andean contribution. In the Amazon Basin, the mass accumulation rate of eolian dust is only around one-third of the vertical erosion rate in shield areas, suggesting that Saharan dust is ¡°consumed¡± by tropical weathering, contributing nutrients and stimulating plant growth, but never accumulates as such in the Amazon Basin.
The chemical and isotope compositions found in the Bod¨¦l¨¦ Depression are varied at the local scale, and have contrasting signatures in the ¡°silica-rich¡± dry lake-bed sediments and in the ¡°calcium-rich¡± mixed diatomites and surrounding sand material. This unexpected finding implies that the Bod¨¦l¨¦ Depression material is not ¡°pre-mixed¡± at the source to provide a homogeneous source of dust. Rather, different isotope signatures can be emitted depending on subtle vagaries of dust-producing events. Our characterization of the Bod¨¦l¨¦ Depression components indicate that the Bod¨¦l¨¦ ¡°calcium-rich¡± component, identified here, is most likely released via eolian processes of sand grain saltation and abrasion and may be significant in the overall global budget of dusts carried out by the Harmattan low-level jet during the winter.