The impact of volcanism on trace metal biogeochemistry in the Southern Ocean

Iron (Fe) is now recognised as a limiting, or co-limiting, micronutrient for biological production in the ocean. Areas of the ocean that are high in the macronutrients nitrogen (N) and phosphorous (P) but low in chlorophyll producing organisms are known as HNLC regions and cover approximately one third of the world's oceans, the largest of which is the Southern Ocean (SO). When coupled with the inhibition of nitrogen fixation by microbes under low Fe conditions in low-latitude regions, this means that Fe regulates biological productivity in as much as half the world's oceans. As such, Fe plays a key role in biogeochemical cycling and the drawdown of carbon from the atmosphere. Sources of Fe to the SO include resuspension of coastal and shallow sediments, glacial and iceberg melt, seasonal sea ice retreat, island wakes, vertical diffusive flux, interaction between bathymetry and currents, dust deposition and hydrothermal inputs. Hydrothermal inputs, in particular, have gained increasing recognition in recent years as a previously overlooked source of Fe to the deep ocean that may potentially affect surface productivity in some ocean basins. Furthermore, shallow hydrothermal vents and subaerial volcanic islands, while less common than deep mid-ocean ridge vents, may supply Fe directly to the surface mixed layer. This dissertation explores the impact of both hydrothermal and subaerial aspects of volcanism on Fe biogeochemistry in the SO. First, the impact of hydrothermalism on Fe cycling in the SO is explored through an extensive compilation of global hydrothermal studies and hydrothermal Fe measurements. This review highlights a dearth of hydrothermal studies conducted in the SO, with only 72 of the 631 vents discovered located south of 30° at the time of writing. A total of 31 of the 72 SO vents are located in the upper 2000 m of the water column, with 16 located in the upper 1000 m, increasing the probability of hydrothermal Fe reaching surface waters transported in upwelling SO meridional circulation. However, with only a small fraction of the ~20,000 km of tectonic plate boundaries located in the SO surveyed to date, many more observations at hydrothermal vent sites and further investigation of Fe transport mechanisms are required in order to constrain the impact of hydrothermalism on the SO and global ocean Fe cycles. Within the Indian sector of the SO lies an oasis of relatively Fe rich waters overlaying the Kerguelen Plateau. At the southern part of the central Kerguelen plateau is an active volcanic hotspot, hosting two subaerial volcanically active islands, Heard and McDonald (HIMI), the former of which is largely covered by glaciers. Waters in the region are subject to an intense mixing regime, caused by shallow bathymetry and the location of the plateau in the path of strong currents associated with the polar front. Fed by the Fe rich waters formed on the plateau, a phytoplankton bloom on the order of thousands of square kilometres forms in the lee of the plateau annually each summer. Few studies to date have focused on HIMI due to its extreme isolation. As such, the role of HIMI in the region's biogeochemical cycling is relatively unknown. In order to examine the impact of these subaerial volcanoes on biogeochemical cycling in the area, dissolved iron (DFe) and macronutrient (NO\\(_3\\)\\(^-\\), PO\\(_4\\)\\(^{3-}\\) and Si) data from waters surrounding HIMI was collected on the RV Investigator during the Heard Earth-Ocean-Biosphere Interactions (HEOBI) voyage (GEOTRACES process study GIpr05) in January to February 2016. Results show that DFe availability drives macronutrient uptake on the plateau. In late summer the majority of the plateau to the north of HIMI contains a deficit of DFe relative to macronutrients with respect to phytoplankton requirements, although dissolution of particulate Fe is expected to fulfil at least some of the Fe demand. Comparison of Fe:N and Fe:P drawdown ratios with the Redfield ratio indicate that recycling of Fe decreases near HIMI, indicating that Fe limitation is alleviated close to the islands. Comparison with data from the only previous study in the area shows that DFe distribution varies between years due to the complex oceanographic conditions on the plateau, with greatest variability observed over the rough bathymetry and strongly tidally influenced region closest to HIMI. Together these data highlight the central role of Fe and the complexity of biogeochemical cycling in the HIMI region of the Kerguelen plateau. Finally, processes supplying DFe around HIMI are further constrained by examining the DFe redox speciation and hydrogen peroxide (H\\(_2\\)O\\(_2\\)) chemistry in the region, analysed during the HEOBI voyage. Dissolved iron(II) (DFe(II)) is the reduced, short lived, potentially more bioavailable oxidation state of Fe in the ocean. Concentrations of DFe(II) in the surface ocean are inversely correlated to the concentration of reactive oxygen species such as H\\(_2\\)O\\(_2\\), which decrease DFe(II) half-life. Surface DFe(II) concentrations at the open ocean reference station were very low (<0.09 nmol L\\(^{-1}\\)), while stations near HIMI showed elevated concentrations over the entire water column (mean 0.24 nmol L\\(^{-1}\\) and 0.36 nmol L\\(^{-1}\\), respectively). At Heard Island, the greatest DFe(II) concentrations (max 0.57 nmol L\\(^{-1}\\)) were detected north of the island, and an inverse correlation of DFe(II) concentrations with salinity suggest the origin is from a marine-terminating glacier on the island. At McDonald Islands, the greatest DFe(II) concentrations (1.01 nmol L\\(^{-1}\\)) were detected east of the island which, based on water column profiles from five targeted stations and observed excess \\(^3\\)He concentrations, appears likely to originate from shallow diffuse hydrothermalism. DFe(II):DFe(total) ratios at the sites of high DFe(II) concentrations at Heard and McDonald Islands (25% and 37%, respectively) adds further evidence that strong, but different sources of reduced Fe exist at each of the islands. Furthermore, lack of correlation with H\\(_2\\)O\\(_2\\) and irradiance data suggest that over the plateau near HIMI, DFe(II) concentrations are more strongly governed by strong DFe(II) sources rather than by H\\(_2\\)O\\(_2\\) and irradiance. The findings presented in this thesis highlight the importance and sparse knowledge of the impact of hydrothermalism to Fe cycling in the SO, and significantly improve understanding of how the subaerial HIMI volcanoes influence biogeochemical cycling at a biological and volcanic hotspot in the SO.