INTERPRETATION OF SILICON ISOTOPE RECORDS IN MARINE SEDIMENTS

The inventory, distribution and utilisation of dissolved silicate (DSi) in the global ocean reflects the complex interplay between DSi inputs, biogeochemical activity, and ocean circulation. Reconstruction of the ocean Si cycle in the past therefore has potential to provide insight to many issues besides Si biogeochemistry. Specific applications can include the connection between marine siliceous productivity and the biological pump, the relationship between terrestrial silicate weathering rates and global climate or the utility of DSi concentrations as a water mass tracer (e.g. Fontorbe et al., in press). To this end, the silicon isotope composition (expressed as δ³⁰Si) of the remains of siliceous organisms in marine sediments is increasingly being used as a paleoceanographic proxy to address questions across a range of temporal and spatial scales.

Yet there are many unknowns that hamper our ability to predict and interpret δ³⁰Si, even in the modern ocean. These relate to the magnitudes and rates of the individual biological and geochemical fluxes constituting the ocean Si cycle, and even more so to the changes in isotope ratios associated with these processes. In particular, the fractionations associated with biotic and abiotic processes are variable and poorly constrained, there is a largely unexplored potential for diagenetic alteration of δ³⁰Si, and characteristic timescales of ocean δ³⁰Si response to external forcing are unclear. The extent to which any palaeoceanographic interpretations are robust to these uncertainties is unclear.

Here we present a compilation of new and published δ³⁰Si data and a series of simple box-models and associated sensitivity tests to test the extent to which we can confidently make various assertions about the past ocean Si cycle. We focus on the role of an evolving terrestrial Si cycle, and how these changes may be manifest in ocean sediments. These simple numerical models are used to provide an interpretative framework for palaeoceanographic δ³⁰Si records across transient events and at steady-state.