Drivers of primary productivity in Antarctic coastal polynyas

The Southern Ocean displays strong latitudinal biological patterns, with the highest phytoplankton production recorded in polynyas along the Antarctic coast. Polynyas are areas of open water surrounded by sea ice that are critical for sea-ice production, deep water formation, thermohaline circulation, ecosystems and carbon sequestration. They are strongly modulated by the seasonal cycle of sea-ice formation and melt, which varies widely regionally and interannually. How changes in the cryosphere affect marine productivity is still elusive and uncertain. In this thesis, we use a range of satellite and biogeochemical Argo data to understand the close relationship between phytoplankton productivity and the environment. We first investigate this relationship at the ocean-cryosphere interface in two selected areas, and then expand to an open-water basin scale for a global perspective.
In the opening research chapter (chapter 3), we examine the impact that a large and sudden event had on biological productivity. We target the Mertz polynya, which was subject to an iceberg calving event in 2010 and highlight the differences between the pre-calving (1997- 2010) and post-calving (2010-2018) time periods. We demonstrate that this event led to significant changes in the phytoplankton bloom phenology. The pre and post-calving periods show very distinct features. After the calving, the polynya experienced an increase in productivity during summertime (+1.1 mg ch-a m^-3 ), overall higher sea-ice concentration (+ 5%) and later ice retreat time (+ 20 days), resulting in a later bloom start (+ 23 days) and shorter bloom duration (- 36 days). Regional differences in the polynya show that productivity increased significantly in the eastern part (+ 1 mg chl-a m^-3 ). These findings are linked to changes in the icescape and the repositioning of the calved glacier tongue, which was preventing iron-enriched warm circumpolar deep water from penetrating the polynya on the eastern side.
We then moved on to exploring how decadal-scale environmental processes may influence biological productivity. Chapter 4 focuses on the possible relationship between ice shelf melting and phytoplankton blooms in the Amundsen Sea polynyas, the most productive coastal area of the Southern Ocean in terms of primary productivity. This topic aims to improve our understanding of Antarctic coastal ecosystem response to predicted long-term increase in glacial melt. We find that long-term variability in ice shelf melting and fluxes are uncorrelated with phytoplankton bloom metrics in both the Amundsen Sea (ASP) and Pine Island (PIP) polynyas. We highlight a significantly higher primary production over the growing season in the ASP compared to the PIP (5.2 vs 3.7 mg m^-3 during the bloom), despite the ice shelves surrounding the PIP melting faster (-7.66 vs -20.3 m^-3 y^-1 ). We show that from 1998 to 2017 the strength and location of the Amundsen Sea Low, an atmospheric low-pressure centre, was driving coastal sea-ice patterns, which greatly affected phytoplankton growth. A bloom phenology analysis shows that blooms reaching their peak earlier last longer, implying a sustained supply of nutrients (including iron) to the area. Despite the absence of a direct relationship between the bloom and ice shelf meltwater, other indirect factors can help sustain and lengthen the bloom for a given season. Our findings highlight the intricate biological, physical and biogeochemical relationships in the Amundsen Sea, and the importance of observing platforms like remote sensing to understand long-term changes.
While chapters 3 and 4 focus on the phytoplankton bloom and its drivers, chapter 5 investigates carbon export. This is a logical step after studying primary production. To give this thesis a broader perspective, we investigate the carbon export in four zones characteristic of the Southern Ocean: the sea ice zone (SIZ), the polar Antarctic zone (PAZ), the sub-Antarctic zone (SAZ) and the subtropical zone (STZ). The use of an extended array of BGC-Argo floats in this chapter allows us to constrain a more accurate estimation of carbon biomass. Results highlight the significance of the understudied SIZ, contributing up to 31% to the total SO carbon estimation. Heterotrophic respiration is responsible for most of the carbon loss in the mesopelagic zone (up to 99% at 1000m) compared to particulate carbon export from phytoplankton. Our findings also show the spatial differences in respiration rate through the water column (100-1000m). The SIZ is characterized by strong heterotrophic respiration in the upper 300m, the PAZ by a gradual decrease in respiration with depth, the SAZ by a consistent respiration activity through the water column and the STZ by the lowest heterotrophic respiration among the four zones.
Overall, this work provides new insights and allows for a better understanding of spatial and temporal variability in primary production across the Southern Ocean. We combined numerous datasets to answer questions previously hypothesized, on the short and long-term effects of environmental changes and their impact on biological production, from local to basin scale perspectives.