Silica biomineralization first appeared in the Cambrian with the appearances of siliceous sponges and polycystinean radiolarians, yet most siliceous taxa with a fossil record did not appear until the Cretaceous, about three hundred million years later. Contemporaneous seawater chemistry may have influenced the acquisition of a siliceous skeleton during the Cretaceous. Cretaceous seawater may have had elevated silica concentrations due to enhanced hydrothermal fluxes from fast spreading rates at mid-ocean ridges and the emplacement of voluminous ocean basin flood basalt provinces on the seafloor. In addition, heightened riverine runoff and warm temperatures may have also increased the delivery of silica to seawater by enhancing continental weathering. A proxy for tracing silica fluxes from continental weathering was undertaken by performing linear regression analyses on dissolved silica with other dissolved cations and polyanions reported in riverine runoff.

No proxy for tracing global riverine silica fluxes to the ocean was identified. For rivers draining predominantly one type of lithology, however, dissolved Na+ was found to strongly correlate with silica in rivers draining granitoid rocks Linear regression analyses were performed on hydrothermal vent fluid datasets and a strong correlation between silica and strontium concentrations was found. When coupled with the 87Sr/86Sr seawater record, the correlation in hydrothermal vent fluids has great potential as a silica proxy. The 87Sr/ 86Sr record was therefore explored as a potential silica proxy. Two models are presented that attempt to explain the seawater 87Sr/ 86Sr curve: the first is a simple mixing model from Faure et al. (1965) that describes the 87Sr/86Sr value of the ocean as a result from the mixing of weathering products from three sources, each with a distinct 87Sr/86Sr value.

The second model is from Brass (1976), which shows the rate of change in the 87Sr/86Sr seawater record is caused from variations in the fluxes of strontium from the weathering of carbonates and non-carbonates, each with a distinct 87Sr/86Sr value. A third, hydrothermal component was added to this model. Neither model is able to accurately explain the 87Sr/86Sr curve through time and therefore the 87Sr/86Sr is not a good candidate as a seawater silica proxy. Based on all available evidence, the hypothesis that high silica availability in Cretaceous seawater was a trigger for the first appearances of several siliceous taxa and the diversification of diatoms cannot be rejected. Further testing on the release of silica from ocean basin magmatic processes is required to further corroborate the link between silica biomineralization and heightened silica fluxes to Cretaceous seas.