Aeolian Iron and Its Contribution to Phytoplankton Production in McMurdo Sound, Southwest Ross Sea, Antarctica

Abstract

Each summer the waters in McMurdo Sound (Lat. 77.5ºS; Long. 165ºE), south-western (SW) Ross Sea encounter vast phytoplankton blooms. This phenomenon is stimulated by the addition of bio-available iron (Fe) to an environment where phytoplankton growth is otherwise Fe-limited. One possible source of such Fe is aeolian sand and dust (ASD) which accumulates on sea ice and is released into the ocean during the summer melt season. The amount of bio-available Fe (i.e. the amount of Fe immedately accessible to phytoplankton) potentially supplied to the ocean by ASD depends on a number of factors including; the ASD flux into the ocean, its particle size distribution and Fe content. However, none of these parameters are well constrained in the SW Ross Sea region and, as a result, the significance of this Fe source in the biogeochemical cycle of phytoplankton growth remains to be quantified. This study focuses on an area (7400 km²) of Southern McMurdo Sound, one of the few areas where direct sampling of ASD that has accumulated on sea ice is possible. To evaluate the flux and solubility of Fe contained in ASD into McMurdo Sound, the mass accumulation rate and particle size of 70 surface snow samples and 3 shallow (3 m) firn cores from the nearby McMurdo Ice Shelf covering the period 2000 - 2008 have been analysed. Selected samples were also measured for total and soluble Fe, Sr and Nd isotopic ratios and mineralogy as a guide to Fe-fertilisation potential and provenance, respectively. Mass and particle size data show an exponential decrease in mass accumulation rate (from 26.00 g m⁻² yr⁻¹ to 0.70 g m⁻² yr⁻¹) and a decrease in modal particle size (from 130 to 69 μm) over a distance of 120 km from Southern McMurdo Sound northwards to Granite Harbour. Both these trends are consistent with ASD being dispersed northwards across the sea ice by southerly storms from an area of the McMurdo Ice Shelf, where submarine freezing and surface ablation have resulted in a surface covered with debris from the sea floor, known as the 'dirty ice' or 'debris bands' (Lat. 77.929ºS; Long. 165.505ºE) in Southern McMurdo Sound. This assertion is further supported by the Sr and Nd isotopic signature of ASD matching local source rocks and the presence of vesicular glass of Southern McMurdo Sound in all samples which also points to the debris bands as the origin of ASD in McMurdo Sound. Bio-available Fe is extremely difficult to quantify hence Fe solubility was used as an approximation in this thesis. Analysis of both total (i.e. particulate and soluble) and the percentage of soluble Fe in the 0.4 - 10 μm dust size fraction (i.e. the fraction most likely to become bio-available) by solution ICP-MS shows a narrow range of values; 3.84 ± 1.99 wt % and 9.42 ± 0.70 % respectively. Combining these values with mass accumulation rate estimates for the particles 0.4 - 10 μm in size, gives an annual soluble Fe flux for the region 500 km² north of the debris bands in McMurdo Sound of 0.55 mg m⁻² yr⁻¹ (9.89 μmol m⁻² yr⁻¹), with spatial variability largely determined by differences in mass accumulation rate. These fluxes are at least an order of magnitude greater than predicted in global dust deposition models for the Southern Ocean and measured in snow samples from East Antarctica. Furthermore, these values exceed the Fe threshold, estimated as 0.2 nM (Boyd and Abraham, 2001), required for phytoplankton growth following the simple dust-biota model of Boyd et al. (2010) and assuming the release of captured ASD in snow is instantaneous. Whilst not constrained in the present study, ASD sourced from the debris bands may be sufficiently widely dispersed, particularly during storm years, to contribute to Fe-fertilisation up to 1200 km from Southern McMurdo Sound. Short, ~10 year long, firn core records of mass accumulation and methylsuphonate concentration, a proxy for phytoplankton productivity, shows a close correspondence between the two during particularly stormy years. Whilst not demonstrating a cause-and-effect relationship, this observation suggests coastal ice cores may contain an important record of the interplay between climate, dust supply, Fe-fertilisation of near shore waters and phytoplankton productivity on decadal and longer timescales.