Source versus weathering processes as controls on the Mackenzie river uranium isotope signature

Abstract

Uranium (U) isotope signatures (δ238U) recorded in sedimentary archives provide insight into the paleo redox state of the ocean. But the robust interpretation of these sedimentary U isotope records requires characterisation of the U isotope signature of rivers, the main source of U to the ocean. The main controlling factors on riverine δ238U remain poorly constrained. Here, we investigate the elemental and isotope signatures of uranium in the dissolved and solid loads of a well-characterised river, the Mackenzie Basin (Canada).In the Mackenzie Basin, the solid load δ238U shows a positive relationship with U and vanadium (V) contents, consistent with the suggestion that particulate δ238U are explained by variable contributions via erosion of silicate and black shale. The δ238U of the dissolved and solid loads are correlated which, at first sight, suggests no U isotope fractionation during chemical weathering, and a purely lithological control on both the river dissolved and solid δ238U. Moreover, relationships between dissolved U and δ238U and major elements such as calcium and sulfate, also support the idea of a lithological control. However, the δ238U of end members inferred from mixing relationships are not consistent with binary mixing of two sources, suggesting some potential U isotope fractionation during weathering. In fact, the abundance of U in the river dissolved load is always lower than that predicted by silicate rock weathering. This suggests that 1) the weathering of silicate only can explain the abundance of U in the river dissolved load and 2) secondary weathering processes scavenge a proportion of the U released by primary mineral breakdown. The broad negative relationship between δ238U and the depletion of dissolved U is also consistent with the control of dissolved δ238U by secondary weathering processes following silicate mineral breakdown. The relationships observed between dissolved U, δ238U and the large-scale environmental controls on weathering processes (such as weathering intensity or runoff) support this hypothesis.Overall, our interpretations of the variation in the river dissolved δ238U challenge the common assumption of the control of dissolved U by black shale and carbonate weathering. In addition, we suggest that the extent of secondary weathering processes can imprint on the U isotope signature of rivers, now and in the past.