Variations in GDGT flux and TEX86 thermometry in three distinct oceanic regimes of the Atlantic Ocean: a sediment trap study

The global climate change and global warming have been observed since the pre industrial times. This phenomenon calls for immediate actions to fight climate change. The climate change of the Earth´s history is preserved in the ocean, covering ~70% of the Earth, and the understanding of this history is the significant way to predict the future climate. The temperature changes in the past ocean can be estimated using several proxies. One of the most extensively applied temperature proxy to the paleoclimate is the TEX86 (tetraether index of tetraethers consisting of 8 carbons). It is based on the relative compositions of thaumarchaeotal membrane lipids, glycerol dialkyl glycerol tetraethers (GDGTs), in marine surface sediments, allowing us to estimate sea surface temperature (SST). The ubiquities of Thaumarchaeota, one phylum of the Archaea domain, and their dominant abundance in the ocean have made them a novel proxy. However, it has been recognized that the composition of GDGTs can be altered by non thermal factors, leading the unconventional relationship between TEX8 and SST. This thesis shall contribute to a better understanding of TEX86 thermometry and the controlling environmental factors in various oceanic provinces by evaluating the GDGT flux and TEX86 related temperature estimate in sinking particles. The sinking particles collected using a time series sediment trap system can be a great value to understand the export mechanism of lipids and to estimate the seasonal variability of the proxy signal. Three different ocean regimes from the upwelling (the Guinea Basin and Lüderitz off Namibia), high latitude (the eastern Fram Strait and the Antarctic Polar Front), and the oligotrophic oceans (the central Brazil Basin) in the Atlantic Ocean are investigated. In the first part of the thesis, TEX86 values are converted to temperatures using the TEX86H calibration. The results in the eastern equatorial Guinea Basin (GBN3) show that GDGTs are mainly transported by particles containing opal. The TEX86H-derived temperatures at both different depth traps correspond to the subsurface water depth (~50 m), where the nutricline exists, implying the favorable habit depth of thaumarchaeotal communities and thus the record of the water temperature. In the coastal upwelling area off Lüderitz (LZ), the export mechanism of GDGT is yet unclear due to the shortage of the sampling period. The results show that the TEX8 H derived temperatures resemble the satellite derived SSTs with a delay of 2 days during the warmer season while the warm biased estimates occur during the colder season. As relatively higher TEX86 values were found in oxygen depleted waters, it explains the warm bias of TEX86H temperature in the Lüderitz, where the oxygen minimum zone is pronounced. This result is in agreement with the finding from off Cape Blanc, Northwest Africa. The second part of the thesis focuses on the high latitude cold regions. The TEX86 values are converted to temperatures using the TEX86L calibration. In the eastern Fram Strait (79° N; FEVI1 ), GDGT fluxes are correlated with biological and non biological component fluxes. The best correlations are found between GDGT and opal/ carbonate fluxes, implying that opal and carbonate are the major transport of GDGTs to the deep ocean. The TEX86L derived temperatures display the strong variability due to temporal variability of transporting materials and their different sinking velocities for particles carrying the lipids. The TEX86L signal corresponds to water temperature at 30 to 80 m depth, where nitrification might occur, and the almost identical is found in the underlying surface sediment. In the Antarctica Polar Front (50° S; PF3), the TEX86L derived temperatures at the shallow trap display the cold and warm biases relative to the SSTs and the latter has a tendency to occur during periods of relatively low GDGT flux, which may be more dominant in the deep trap. The warm biased TEX8 L signal (~7 °C) compared to the SST at the deep trap and in the underlying surface sediment might be due to the contribution of Euryarchaeota or the non liner relationship of TEX86L with SST in the Southern Ocean. In both polar regions, hydroxylated GDGT related temperature proxy seems applicable. The third part of the thesis contributes to the oligotrophic regions, where usually attract less attention compared to two previous regimes despite its size and the impact on global biogeochemical budgets. The TEX8 values at two depths from two sites are examined (WAB1 and WA9). Good correlations between GDGT and opal/carbonate at both shallow traps reveal the preferential incorporation of GDGTs in opal and carbonate rich aggregates. At WAB1, which was located in the fringes of the gyre system, the TEX86H derived temperatures of the shallow trap resemble the SSTs. The WA9 trap was collected in the more nutrient depleted location of the central oligotrophic gyre relative to the site WAB. The warm biased TEX86H derived temperature of the shallow trap relative to the SST can be caused by the response of Thaumarchaeota under the energy stressed condition. In the deep traps of both sites, the TEX8 H derived temperatures record the deep subsurface temperature. It is assumed that these signals are caused by the relatively dominant contribution of colder signal from deep in situ production and the smaller contribution of warmer signal from shallow waters. The last part of the thesis investigates the alkenone based temperatures (U or U ) in all three oceanic regimes, where GDGT based temperatures are discussed in previous chapters. Most of UK'37 derived temperatures display the SSTs of the tropical regions (i.e. the equatorial Guinea Basin, Lüderitz, the central Brazil Basin). It implies the regional geochemical characteristics (e.g., availability of nitrogen or oxygen), which probably affect the TEX86 thermometry, would not have an impact on the U thermometry. In the shallow trap of the Antarctic Polar Front, the UK'37 derived temperatures record the clear SST seasonality with a delay time of 10 weeks. However, the mixture of cold and warm biased temperature estimates of the deep trap is underconstrained. Samples from the eastern Fram Strait yield UK'37 and UK'37 derived temperature estimates fall outside the expected temperature range with the warm biases. It is in line with other studies that the applicability of alkenone proxies is limited in low temperature regions that disfavor alkenone producers. This chapter shows that the UK'37 proxy provides an additional temperature record complementary to the TEX86 proxy because of different responses of each source organism to environmental conditions in the tropical oceans.