Abstract
Nitrogen (N) holds a central position in ocean biogeochemistry due to its role as a limiting nutrient for biological productivity in the ocean and its resultant influence on the marine carbon cycle. Nitrogen isotopes represent a powerful tool to investigate changes in the marine N-cycle across different timescales. However, their use in geochemical studies has been traditionally limited by the potential diagenetic artifact related to changes in organic matter preservation through geologic time. Over the past ten years, the analysis of the isotopic com- position of organic matter protected within the mineral structure of planktonic foraminifera shells (foraminifera bound, FB) has emerged as a way to circumvent diagenetic overprints of classical techniques. The objective of this thesis is to use this novel technique to study the evolution of the N-cycle over previously unexplored periods of the Cenozoic, with a particular focus on the late Pleistocene glacial cycles (Chapter 3), the Mid Miocene (Chapter 4), and the Mid Eocene Climate Optimum (Chapter 7). Foraminifera-bound N isotope measurements (FB-d15N) are complemented by the development of a new method for organic biomarker extraction (Chapter 5), which has allowed the analysis of a significant number of samples, and provided new constraints on climate evolution across the studied time periods (Chapters 6 and 8).
In Chapter 3, we show a pervasive coupling between low-latitude N-fixation and ocean circulation changes that control the supply of excess phosphorous (P) to the surface ocean in the Atlantic Ocean. Our data suggest decreased N-fixation in the North and South Atlantic oligotrophic gyres during periods when the supply of excess P by Antarctic Intermediate Water (AAIW) is suppressed by Glacial North Atlantic Intermediate Water (GNAIW). In contrast, precessional changes in the strength of equatorial upwelling - which in turn drive the supply of excess P - appear to be particularly important to N-fixation in the Caribbean Sea, noticeably weaker in the northern gyre, and negligible in the southern gyre. In Chapter 4, we report the first FB-d15N spanning the last 25 million years (Ma) from three sediment cores located in the subtropical gyres of the South Atlantic (DSDP 516), Pacific (ODP 872) and Indian Ocean (ODP 754). We find a substantial drop in FB-d15N values during the Mid Miocene Climatic Optimum (MMCO) and a significant reduction in the d15N gradient between the Pacific and Atlantic Ocean. We propose that these changes are best explained by a combination of (i) an increase in sedimentary denitrification rates triggered by the expansion of continental shelves associated with the MMCO sea level rise and enhanced continental weathering; (ii) a decrease in water column denitrification caused by tectonically-driven ocean circulation changes, which increased the supply of oxygen to the subsurface ocean and (iii) an increase in N-fixation as a response to the excess P supply from sedimentary denitrification. After the MMCO, the isotopic gradient between the Pacific and Atlantic basins increased, reaching its maximum during the Pliocene epoch. This is interpreted as the impact of ocean circulation changes associated with the closing of low-latitude oceanic gateways during the Miocene, and the subsequent influence of these changes on water column denitrification in the Pacific. Chapter 7 focuses on the Mid Eocene Climatic Optimum (MECO), a short-term episode of climate warming that occurred at ca. 40 Ma. The warming event shows a 3-4 permil decrease in FB-d15N in the South Atlantic gyre. Similar to the MMCO, this change in FB-d15N is possibly related to upper ocean stratification and increased N-fixation in the subtropical oligotrophic gyre and/or an increase in sedimentary denitrification in shallow seas and subsurface waters in shelf areas.
In parallel to the efforts above, we developed a new extraction and separation method for organic biomarkers in Chapter 5. In Chapter 6, we reconstructed biomarker-based sea surface temperatures (SST) on mid-latitude core sites across the Cenozoic. The combination of newly derived and previously published GDGT-SST data demonstrate that warm tropical SST expanded significantly towards the poles in both hemispheres, resulting in a reduced temperature gradient between mid- and low-latitudes during most of the Cenozoic. In Chapter 8, we detected an increase in methanotrophy on two Pacific sediment cores in the Mid Miocene suggesting that methane release could have potentially reinforced the CO2 greenhouse gas effect during the MMCO.
