Acidification in the Southern Ocean – current state and future challenges

The absorption of anthropogenic (i.e., human-made) carbon into the ocean causes “ocean acidification”, a profound change in seawater chemistry. Ocean acidification causes an increase of the acidity levels of the ocean and thus a decline in seawater pH. In addition, ocean acidification has been shown to interfere with a multitude of biological processes including reproduction, metabolism, and biomineralization. In particular, ocean acidification affects the conditions that determine how ‘easily’ marine organisms can form shells from minerals such as calcium carbonates (aragonite, calcite). Under ocean acidification, the saturation state of both aragonite and calcite decreases, so that new minerals become more difficult to form, and existing minerals may start to dissolve (with aragonite being more soluble than calcite). In this way, ocean acidification directly affects marine organisms that rely on these carbonates to build their shells or skeletons [1]. Like the Arctic Ocean, the Southern Ocean is especially sensitive to this process due to i) the prevalence of higher CO₂ concentrations in colder waters, ii) the upward transport of acidic deep waters with Southern Ocean circulation, iii) a stronger impact of anthropogenic carbon absorption on acidity due to previous uptake of anthropogenic carbon to date, which is disproportionally large in the Southern Ocean, and iv) ongoing reductions in sea-ice cover and enhanced freshwater input from melting sea ice, glaciers, and ice shelves that act to amplify ocean acidification [1, 2, 3].

The Southern Ocean contributes disproportionately to the global oceanic uptake and storage of anthropogenic carbon [4, 5]. Over the past four decades, surface ocean acidification increased rapidly in the waters around Antarctica. Based on satellite and ship observations between 1982-2021, rates of pH decline in the open ocean were estimated to be severe (15% larger) when compared to the global mean (up to 0.0189±0.001 per decade in the Southern Ocean compared to 0.0166±0.001 per decade on the total pH scale) [6]. Over the same time period, the aragonite saturation state has declined at similar or lower rates in the Southern Ocean (up to 0.067±0.005 per decade) and in the Arctic Ocean (up to 0.055±0.013 per decade) than in the global mean (0.071±0.006 per decade) due to the naturally lower carbonate ion concentrations in polar waters [6]. Since many of the factors making the Southern Ocean especially sensitive to ocean acidification equally apply to northern polar waters (cold waters, decreasing sea-ice cover), it is unsurprising that the surface acidification rates in the Southern Ocean are similar to those observed for northern high latitudes (Arctic) [6].

Obtaining observation-based estimates of interior ocean (i.e., subsurface) acidification across the Southern Ocean has been challenging. Until 2014, historical estimates of anthropogenic carbon storage and associated rates of interior ocean acidification were exclusively based on ship observations [7, 8]. Consequently, data availability only facilitates an assessment of decadal or longer changes [8]. Since preindustrial times, ocean acidification has progressed more slowly in subsurface waters than in surface waters of the open ocean regions of the Southern Ocean [8]. However, there remain uncertainties stemming from sampling sparsity, particularly at greater depths (>500m) and at higher latitudes (>60°S) [8]. Since 2014, the deployment of approximately 300 biogeochemical Argo floats equipped with pH sensors has drastically increased subsurface data availability in the open ocean [9]. These floats are transported with ocean circulation at 1000 m depth and autonomously sample the upper 2000 m of the water column every ten days. While the float-based pH record is still too short to assess longer term trends with float data alone, a recent assessment based on five years of float data combined with a data-assimilating model [9] revealed acidification trends consistent in magnitude and spatial distribution with the estimates based on historic satellite and ship data [6, 8].

Simulations with the most recent generation of Earth System Models, which have contributed to reports by the Intergovernmental Panel on Climate Change (IPCC), project global patterns of ocean acidification consistent with recent observational records. These models project future ocean conditions under different scenarios, ranging from assuming an efficient transition to the use of sustainable resources (SSP1-2.6) to a continued, intensive use of fossil fuels (SSP5-8.5 and RCP8.5). Specifically, the projected surface pH decline by 2100 in the Southern Ocean south of 30°S will exceed that in tropical and subtropical regions between 30°S-30°N by 10% for a high-emission scenario (SSP5-8.5) [10]. In open ocean waters across the Southern Ocean, surface aragonite undersaturation events, i.e., when the aragonite saturation state is low enough for aragonite to start to dissolve, are projected to become more prevalent in space and time from the 2030s onwards for a high-emission scenario (RCP8.5), occurring in >70% of Southern Ocean surface waters by 2070 [11, 12]. Importantly, many models underestimate current uptake and storage rates of anthropogenic carbon in open ocean waters of the Southern Ocean [5, 13], suggesting that these models underestimate the severity of future ocean acidification in the waters around Antarctica [13].

Both observations [7, 14] and models [3, 15] show that Antarctic coastal waters are a present-day sink for anthropogenic carbon due to strong biologically driven uptake in summer, and sea-ice cover preventing carbon fluxes to the atmosphere in winter. Antarctic coastal waters host several dense water formation sites which are characterized by efficient vertical mixing following sea-ice formation. The transport of newly formed dense waters from the Antarctic shelf to the abyssal open ocean (i.e., near the ocean floor) provides a pathway to spread coastal acidification to the global deep ocean. Through the strong vertical mixing in dense water formation, ocean acidification can propagate more quickly to depth on the Antarctic shelf than in open-ocean waters [3, 7]. A model specifically designed for applications in Antarctic coastal waters showed that in regions of strong dense water formation (e.g., the Weddell Sea), anthropogenic carbon entering at the surface can be transported to depths of 500m throughout the region in less than a decade. In other regions of the open ocean, however, this process can take multiple decades to centuries [3]. Based on all available ship-based observations, it was shown that since the mid-1970s, the strong vertical mixing has resulted in a pH decline in circumpolar Antarctic bottom water at 3500m depth of 0.006±0.001 per decade [7], with larger trends in the strongest dense water formation regions, i.e., the Ross Sea (0.009±0.004 per decade) and the Weddell Sea (0.007±0.004 per decade). For the highest-emission scenario, the projected change of deep-ocean transfer of newly formed denser waters in the Weddell Sea slows down acidification rates of bottom waters downstream by up to 60% in 2100 relative to the mid-century [16].

As a result of the strong coupling with physical processes such as vertical mixing, Antarctic coastal waters are projected to experience more severe ocean acidification over the 21^st century than open ocean waters (Figure 1). Acknowledging that sea ice, as long as present, slows the uptake of anthropogenic carbon and thus acidification [17] and that the spatial variability in acidification rates is large [18], pH throughout the water column of Antarctic coastal regions is projected to decline by up to 0.36 by 2100 relative to the 1990s under the highest emission scenario (SSP5-8.5) [3]. Aggravated by climate-change feedbacks (i.e., the projected decline in sea-ice cover and increase in meltwater input from sea ice and ice shelves), more than 95% of all Antarctic coastal waters south of the 2000 m depth isoline on the continental slope are projected to be undersaturated with respect to aragonite for all intermediate and high emission scenarios (SSP2-4.5, SSP3-7.0, SSP5-8.5) [3]. The extent of undersaturation in these waters is projected to be lower for calcite (up to 60%) [3].

The Southern Ocean hosts unique biodiversity, and its preservation is the main goal of ongoing efforts by the Convention for the Conservation of Antarctic Marine Living Resources to establish a network of marine protected areas in the Southern Ocean [19]. Multiple studies have documented the negative impacts of acidification on the physiological processes of many Southern Ocean species across multiple trophic levels of the food web, although it often remains unclear to what extent these effects are less or more detrimental than, e.g., ocean warming, changed nutrient availability or sea-ice decline. For example, in response to ocean acidification, studies have shown:

• negative impacts on primary productivity, growth, and microbial grazing, as well as shifts in phytoplankton community composition [20, 21, 22],
• a decline in diatom silica shell formation, with species-specific differences [23],
• diminished lipid stores in larger diatoms altering energy available to higher trophic levels [24],
• a reduction in egg hatch rate and recruitment of Antarctic krill [25],
• an increase in calcium carbonate shell dissolution in benthic (i.e, bottom-dwelling) organisms [26] and calcifying zooplankton [26, 27],
• a reduction in metabolic capacity to adapt to warming by some Antarctic fish [28], and
• increasingly malformed larvae of benthic bivalves and sea urchins [26, 29].

Importantly, these documented negative impacts are associated with different critical thresholds of ocean acidification for the different species, complicating an assessment of the ecosystem-level response to the changing acidity of the Southern Ocean [30]. Further, in lab experiments, adult krill were shown to be capable of resilience to an elevated acidity [31], demonstrating an incomplete understanding of ocean acidification impacts across life stages of the same species. Species-specific differences in ocean acidification impacts are conceivable, and species interactions and potential adaptation on climate-change relevant time scales (e.g., decadal) are not yet well understood. There are few available studies of ocean acidification impacts on Southern Ocean biota. The majority are predominantly focused on microbial communities and invertebrates, with few to no studies on mid and higher trophic levels [30]. This lack of information currently prevents a conclusive assessment of ocean acidification impacts on Southern Ocean ecosystem functioning.

On human lifespan time scales, slowing ocean acidification is only achievable through reductions of CO₂ emissions and future carbon dioxide removal because natural processes (i.e., the weathering of rocks on land) take tens of thousands of years to revert ocean acidification [32]. Further, since all of these processes start acting on the ocean surface, ocean interior acidification will persist until these waters are transported to the upper ocean. The remaining challenges to fully understand Southern Ocean acidification and its implications for ecosystems are to improve the observational data coverage and our understanding of climate-change impacts on Southern Ocean ecosystems to ultimately improve the physical and biological process representation in ocean biogeochemistry models.

Observations. Monitoring the progress of Southern Ocean acidification relies on sustaining the current observing network and expanding it. Traditional ship-based carbon observations produce high-quality data but are skewed to the summer months and restricted to few transects [33, 34]. Therefore, these must be complemented by continuous autonomous deployments, e.g., floats, gliders, sail drones, and moorings, to ensure adequate data coverage across seasons, the ocean interior, and the full ocean basin. To date, despite the severe levels of acidification projected for Antarctic coastal waters [3], relatively fewer data are available for these waters compared to those in the open ocean [33, 34]. This lower data coverage can be attributed to difficulties in accessing Antarctic coastal waters and the existing risks associated with deploying autonomous instrumentation in regions which are seasonally or permanently ice-covered and often frequented by icebergs. By continuing the development of autonomous sensors and ice-avoidance technologies and by improving our ability to locate autonomous instrumentation while under ice, overcoming these challenges will facilitate the establishment of a multi-platform observing network to monitor basin-wide Southern Ocean acidification over time.

Biological impacts. Quantifying ecosystem-wide impacts of ocean acidification in the Southern Ocean demands a better understanding of the impacts on individual species across trophic levels and life stages, including an assessment of the possibility to adapt to environmental change. Currently, multi-stressor studies, i.e., those considering ocean acidification together with other environmental stressors such as warming, are still rare [28, 35], and studies additionally including multiple trophic levels are, to our knowledge, absent. Yet, such studies would provide critical information on potential changes in ecosystem dynamics in response to environmental change. An improved understanding of direct climate-change impacts, including those from ocean acidification, on marine organisms is key for the improvement of the representation of climate-change impacts on biological processes in models. Ultimately, better understanding how ecosystem functioning might respond to climate change will facilitate an improved assessment of the efficiency of conservation measures involving a reduction of direct human impacts (e.g., from fishing), such as the proposed or adopted Southern Ocean marine protected areas.

Modeling. Ocean biogeochemistry models are indispensable for predicting potential future trajectories of acidification in the Southern Ocean, both in open ocean and in coastal waters. While the magnitude of the projected surface pH decline by 2100 exceeds the variability across Earth System Models everywhere in the ocean [10], uncertainties remain, especially for ocean interior acidification rates across the Southern Ocean. This uncertainty stems from systematic mismatches of some models with observation-based anthropogenic carbon uptake and storage over the recent past [5, 13]. Acknowledging the historically scarce, summer-skewed observational data available to evaluate ocean models, many ocean models struggle to correctly simulate the seasonal evolution of physical (e.g., circulation and vertical mixing) and biological (e.g., phytoplankton and particle dynamics) processes and their role in oceanic carbon uptake and carbon transfer to depth [5]. For example, dedicated multi-stressor laboratory experiments for Southern Ocean phytoplankton and zooplankton would facilitate the improvement of parameterizations describing how biological carbon cycling responds to climate change. In Antarctic coastal waters, the mechanisms proposed to explain the projected severe acidification are plausible based on our current understanding of coastal physical dynamics [3]. Yet, to date, the scarcity of observations in Antarctic coastal waters and the absence of an ensemble of models with similar complexity in coastal process representation prevent a thorough uncertainty assessment of coastal acidification projections in the Southern Ocean [3].