Ocean deoxygenation, a paleo-proxy perspective

The oceans are losing oxygen because of global warming and the consequences of this ocean deoxygenation are far reaching, particularly for aerobic marine organisms that depend on dissolved oxygen to live. Furthermore, in oxygen-deficient waters, ocean deoxygenation promotes global warming through the microbial production of nitrous oxide, a powerful greenhouse gas. The study of the Southern Ocean deoxygenation is particularly essential because through global circulation, it supplies dissolved oxygen to all ocean basins. Moreover, this region is particularly affected by deoxygenation and accounts for one quarter of the total oxygen losses observed since 1970. However, future oxygenation trends remain unclear. To help forecast future Southern Ocean deoxygenation accurately, a precise knowledge of historical oxygen trends and their drivers is essential. One way to do this is to use sediments which record and preserve changes in the water column above them and offer a window on past conditions in the oceans. In this thesis a set of 12 sediment cores retrieved from the New Zealand region (Campbell Plateau, Challenger Plateau, Solander Trough) was used to investigate dissolved oxygen changes in the southwest Pacific sector of the Southern Ocean since the last glacial maximum (LGM) and the potential factors driving these changes. Moreover, because remineralisation by bacteria at depths constitutes the main oxygen sink in the ocean interior, export production (EP) changes since the LGM were also investigated directly. Sediment composition has to be quantified in order to identify potential changes in oxygenation and EP. This requires the complete digestion (transformation of solid material to solution) of the sediments. Previous studies have shown that microwave-assisted digestion of marine sediments using a mix of hydrochloric, nitric and hydrofluoric acids provides a complete, rapid and relatively easy digestion method. However, due to the carbonate rich composition of the New Zealand sediments, all previously published methods generated insoluble fluoride precipitate and failed to complete their digestions. Chapter 2 addresses this problem and shows that a pre-digestion step, exposing sediments to concentrated HCl at 150¬¨‚àû C before microwave digestion helps eliminate precipitate formation, leading to the complete digestion of carbonate rich sediments. This method was tested on four different certified standards with a wide range of carbonate contents. Overall, around 90% recovery across 20 elements analysed was observed. In this work, Inductively Coupled Plasma Mass Spectrometry (ICP-MS) was used as the analytical method of choice for broad screen multi-element analysis. However, for the determination of radiogenic isotopes of extremely low abundance, a special analytical methodology was introduced for the first time in our laboratories. Chapter 3 details the isotopic dilution method employed as well as the incremental analytical developments necessary to measure precisely low abundance isotopes, such as 230-Thorium \\((^{230}Th)\\), using a single collector Sector Field-ICP-MS. The importance of factors such as instrumental detector dead time on the precision of the isotopic ratio measurements was also investigated. Furthermore, the benefit of enhanced sensitivity through the activation of the instrument Platinum guard electrode in combination with a desolvating nebuliser is described. The precision of the measurements produced was assessed through intercomparison with two other techniques. Overall, measurements of 238-Uranium \\((^{238}U)\\), \\(^{232}Th\\), and \\(^{230}Th\\) from sediment digests were achieved with deviations of less than 5%. In chapter 4, \\(^{230}Th\\)-normalised opal, organic carbon, excess barium and calcium carbonate fluxes were used to investigate changes in EP in the southwest Pacific sector of the Southern Ocean since the LGM. The variations of these fluxes reveal that in Subtropical Waters (STW) and in the Subantarctic Zone (SAZ), EP largely remained unchanged since the LGM. Only one of the four sites studied shows increased EP during the deglaciation. At this site it is proposed that a shift in the position of the highly productive subtropical front to above the core site drove the increase in EP. In STW, it is suggested that nitrogen has been limiting any EP increase. In the SAZ, it is proposed that even though the increased glacial dust deposition relieved the iron (Fe)-limitation, silicic acid (`Si(OH)_4`) limited any resultant increase in EP during the LGM. This result shows that both `Si(OH)_4` and Fe co-limit EP in the SAZ around New Zealand, and is consistent with modern process studies. To understand the influence of iron fertilisation by dust since the LGM, lithogenic supply changes at the core sites were also investigated. At sites east of New Zealand an unusual lithogenic deposition pattern was observed, offset from other sites in the rest of the Southern Ocean. An explanation may be that this offset was driven by intense erosion and glacier melts, which increased the sediment discharged during the deglaciation. As EP remained mostly unchanged since the LGM in the southwest Pacific sector of the Southern Ocean, other factors are likely to have driven any observed oxygen changes. In chapter 5, authigenic Uranium and Rhenium (aU and aRe) variations in the sediments were investigated and show that intermediate depths (800-1500 m) of the southwest Pacific sector of the Southern Ocean were deoxygenated during the LGM compared to the Holocene. Additionally, data from deeper locations (‚Äöv¢‚Ä¢ 1500 m) indicate higher oxygen content during the LGM compared to the Holocene. To understand the factors driving these variations all benthic foraminiferal `˜í¬•^13C` data available in the New Zealand region were compiled. These data together with aU and aRe variations are consistent with a dramatic circulation change in the Southern Ocean, which induced a shallower Antarctic Intermediate Water (AAIW)-Upper Circumpolar Deep Water (UCDW) boundary during the LGM compared to the Holocene over the Campbell Plateau. It is proposed that this shoaling can be explained by either a decrease in the AAIW production or a northward shift of the AAIW formation region. However, aU and aRe also show that AAIW contained less oxygen during the LGM compared to the Holocene. These results are in contrast with the enhanced AAIW ventilation inferred during the LGM in the southeast Pacific sector of the Southern Ocean, and therefore highlight an asymmetry in AAIW response to climatic forcing during the LGM in the Pacific. The insights presented in this thesis improve our understanding of how circulation and EP influence the ventilation of the ocean interior on glacial-interglacial timescales. Chapter 6 details the importance of these findings for future research and also highlights the remaining gaps.