Marine silicon cycle through the Cenozoic

Sammanfattning: Silicon (Si) cycle is one of Earth's major biogeochemical cycles. Furthermore, the dissolved form of Si (DSi) is an essential nutrient for both terrestrial and marine ecosystems. DSi ultimately derives from the slow process of chemical weathering of silicate minerals, a mechanism that consumes carbon dioxide, and therefore participates in regulating Earth's climate over geologic timescales. Si is delivered to the ocean mostly by rivers and will be used by a variety of organisms (e.g. diatoms, siliceous sponges, radiolarians, and silicoflagellates) that precipitate DSi into an amorphous form (biogenic silica, BSi) and control the export of Si out of seawaters. While the modern Si cycle and the processes controlling it are becoming better and better understood, its evolution through Earth's history are still poorly constrained.Hence, this thesis aims at shedding more light on the evolution of the marine Si cycle on millennial to million-years timescales. To do so, we investigated the Si isotopic composition (expressed as δ30Si) of siliceous microfossils recovered from marine sediments. The analysis of δ30Si from the remains of marine diatoms, radiolarians, and siliceous sponges is a powerful tool to reconstruct several facets of the oceanic Si cycle in the past.On millennial timescales, the marine Si cycle is mostly dominated by variations in biologic productivity in the surface ocean and riverine inputs of DSi. On the other hand, on million-years time scales, the marine Si cycle appears to be mostly controlled by oceanic circulation. Further, the analysis of sponge δ30Si, performed during this thesis, allowed us to reconstruct the concentrations of DSi in the bottom waters in the North Atlantic and Equatorial Pacific. Our results indicate that contrary to previous hypotheses, the ocean did not experience a rapid decline in oceanic DSi content during the Paleogene (65.5 to 23.03 Ma). Conversely, we show that the North Atlantic already had low DSi concentrations, similar to today, during the early Cenozoic, whereas the Equatorial Pacific has become progressively enriched in DSi since at least 35 Ma.Overall, although research into the evolution of the ocean Si cycle is still at an early stage, the work carried out in this thesis fills some of the existing knowledge gaps regarding the development of the marine Si cycle through geologic times.