Silicate weathering is a primary control on the carbon cycle and therefore long-term climate. Tracing silicate weathering in the geological record has been a challenge for decades, with a number of proxies proposed and their limits determined. Recently lithium isotopes in marine carbonates have emerged as a potential tracer. Bulk carbonates are increasingly being used as a Li isotope archive, though with limited tests thus far of the robustness of this approach in the modern ocean. As the bulk composition of marine pelagic carbonates has changed through time and geographically, assessing the fidelity of bulk carbonate as proxy carrier is fundamental. To address the impact of compositional variability in bulk carbonate on Li isotopes, we examine 27 Bahamian aragonitic bulk carbonates and Atlantic largely calcitic core-top sediment samples.Two core-tops only have trace (<10%) carbonate, and are analysed to test whether carbonates in such sections are still a viable archive. We selectively extract the exchangeable and carbonate fractions from the core-top samples. The exchangeable fraction contains ~2% of the total Li and has a fairly constant offset from seawater of 16.5 ± 0.8‰. When leaching silicate-containing carbonates, acetic acid buffered with sodium acetate appears a more robust method of solely attacking carbonates compared to dilute HCl, which may also liberate some silicate-bound Li. Carbonates from samples that do not contain aragonite have the isotopic fractionation of seawater of D 7 Liseawatercalcite = 6.1 ± 1.3‰ (2sd), which is not affected by latitude or the water depth the sample was deposited at. The pure aragonite bulk carbonates from the Bahamas have a fractionation of D 7 Liseawater-aragonite = 9.6 ± 0.6‰. A sediment sample from the Galician coast that mostly consists of quartz is highly offset from seawater by ~20‰ and also has relatively high Li/Ca ratios. These high values are not due to leaching of silicate material directly (Al/Ca ratios are low). We interpret this addition via cation exchange of Li from silicate during recrystallisation. Overall bulk carbonates from the open ocean are a reliable archive of seawater d 7 Li, but care must be taken with carbonate mineralogy and low-carbonate samples. Overall, therefore, any examination of the palaeo-seawater d 7 Li record must be reproduced in different global settings (e.g. multiple global cores) before it can be considered robust.

No abstract

No abstract