Every year, the oceans absorb about 30% of anthropogenic carbon dioxide (CO2) leading to a re-equilibration of the marine carbonate system and decreasing seawater pH. Today, there is increasing awareness that these changes–summarized by the term ocean acidification (OA)–could differentially affect the competitive ability of marine organisms, thereby provoking a restructuring of marine ecosystems and biogeochemical element cycles. In winter 2013, we deployed ten pelagic mesocosms in the Gullmar Fjord at the Swedish west coast in order to study the effect of OA on plankton ecology and biogeochemistry under close to natural conditions. Five of the ten mesocosms were left unperturbed and served as controls (~380 μatm pCO2), whereas the others were enriched with CO2-saturated water to simulate realistic end-of-the-century carbonate chemistry conditions (~760 μatm pCO2). We ran the experiment for 113 days which allowed us to study the influence of high CO2 on an entire winter-to-summer plankton succession and to investigate the potential of some plankton organisms for evolutionary adaptation to OA in their natural environment. This paper is the first in a PLOS collection and provides a detailed overview on the experimental design, important events, and the key complexities of such a “long-term mesocosm” approach. Furthermore, we analyzed whether simulated end-of-the-century carbonate chemistry conditions could lead to a significant restructuring of the plankton community in the course of the succession. At the level of detail analyzed in this overview paper we found that CO2-induced differences in plankton community composition were non-detectable during most of the succession except for a period where a phytoplankton bloom was fueled by remineralized nutrients. These results indicate: (1) Long-term studies with pelagic ecosystems are necessary to uncover OA-sensitive stages of succession. (2) Plankton communities fueled by regenerated nutrients may be more responsive to changing carbonate chemistry than those having access to high inorganic nutrient concentrations and may deserve particular attention in future studies.

The marine CO2 system in Tempelfjorden (Svalbard) was investigated between August 2015 and December 2017 using total alkalinity, pH, temperature, salinity, oxygen isotopic ratio, and nutrient data.Primary production resulted in the largest changes that were observed in the partial pressure of CO2 (pCO2, 140 µatm) and the saturation state of aragonite (ΩAr, 0.9). Over the period of peak freshwater discharge (June to August), the freshwater addition and air-sea CO2 uptake (on average 15.5 mmol m -2 day -1 in 2017) governed the surface pCO2. About one fourth of the uptake was driven by the freshening.The sensitivity of ΩAr to the freshwater addition was investigated using robust regressions. If the effect of air-sea CO2 exchange was removed from ΩAr, a freshwater fraction larger than 50% (lower range of uncertainty) was needed to provide aragonite undersaturated waters. This study shows that ΩAr and freshwater fraction relationships that are derived from regression techniques and the interpretation thereof are sensitive to the effect of air-sea CO2 exchange. Since the freshening in itself only drives a fraction of the air-sea CO2 uptake, studies that do not account for this exchange will overestimate the impact of freshwater on ΩAr. Finally, in the summer an excess in the salinity normalized dissolved inorganic carbon, corrected for aerobic primary production/respiration, of on average 86 µmol kg -1 was found in the deepest water of the fjord. This excess is suggested to be a result of enhanced CO2 uptake and brine release during the period of sea ice growth. Dates CTD Stations Bottom depths (m) Max sampling depths (m) TA/pH δ 18 O Nutrients

Concentrations of dissolved inorganic carbon (DIC), total alkalinity (TA), nutrients, and oxygen in subsurface waters of the central Arctic Ocean have been investigated for conceivable time trends over the last two decades. Data from six cruises that cover the Nansen, Amundsen, and Makarov Basins were included in this analysis. In waters deeper than 2000 m, no statistically significant trend could be observed for DIC, TA, phosphate, or nitrate, but a small rate of increase in apparent oxygen utilization (AOU) was noticeable. For the individual stations, differences in concentration of each property were computed between the mean concentrations in the Arctic Atlantic Water (AAW) or the upper Polar Deep Water (uPDW), i.e., between about 150 and 1400 m depth, and in the deep water (assumed invariable over time). In these shallower water layers, we observe significant above-zero time trends for DIC, in the range of 0.6-0.9 lmol kg 21 yr 21 (for AAW) and 0.4-0.6 mmol kg 21 yr 21 (for uPDW). No time trend in nutrients could be observed, indicating no change in the rate of organic matter mineralization within this depth range. Consequently, the buildup of DIC is attributed to increasing concentrations of anthropogenic carbon in the waters flowing into these depth layers of the Arctic Ocean. The resulting rate of increase of the column inventory of anthropogenic CO 2 is estimated to be between 0.6 and 0.9 mol C m 22 yr 21 , with distinct differences between basins.

Seasonal and interannual variability in surface water partial pressure of CO 2 (pCO 2 ) and air-sea CO 2 fluxes from a West Spitsbergen fjord (IsA Station, Adventfjorden) are presented, and the associated driving forces are evaluated. Marine CO 2 system data together with temperature, salinity, and nutrients, were collected at the IsA Station between March 2015 and June 2017. The surface waters were undersaturated in pCO 2 with respect to atmospheric pCO 2 all year round. The effects of biological activity (primary production/respiration) followed by thermal forcing on pCO 2 were the most important drivers on a seasonal scale. The ocean was a sink for atmospheric CO 2 with annual air-sea CO 2 fluxes of À36 ± 2 and À31 ± 2 g C·m À2 ·year À1 for 2015-2016 and 2016-2017, respectively, as estimated from the month of April. Waters of an Arctic origin dominated in 2015 and were replaced in 2016 by waters of a transformed Atlantic source. The CO 2 uptake rates over the period of Arctic origin waters were significantly higher (2 mmol C·m À2 ·day À1 ) than the rates of the Atlantic origin waters of the following year.In waters north of Svalbard, the effect of sea ice processes on surface water pCO 2 , mainly through the dissolution of calcium carbonate (CaCO 3 ) in the surface waters, dominates changes (i.e., undersaturation) in surface water pCO 2 on a 6 months' time scale (January to June, Fransson et al., 2017). Further south, in Isfjorden, the largest fjord along the West Spitsbergen coast, the amount of sea ice has decreased substantially over the last decade, both in terms of maximum sea ice cover and days of fast ice (Isaksen et al., 2016; ERICSON ET AL.4888

Abstract. The Siberian shelf seas are areas of extensive biogeochemical transformation of organic matter, both of marine and terrestrial origin. This in combination with brine production from sea ice formation results in a cold bottom water of relative high salinity and partial pressure of carbon dioxide (pCO 2 ). Data from the SWERUS-C3 expedition compiled on the icebreaker Oden in July to September 2014 show the distribution of such waters at the outer shelf, as well as their export into the deep central Arctic basins. Very high pCO 2 water, up to ∼ 1000 µatm, was observed associated with high nutrients and low oxygen concentrations. Consequently, this water had low saturation state with respect to calcium carbonate down to less than 0.8 for calcite and 0.5 for aragonite. Waters undersaturated in aragonite were also observed in the surface in waters at equilibrium with atmospheric CO 2 ; however, at these conditions the cause of undersaturation was low salinity from river runoff and/or sea ice melt. The calcium carbonate corrosive water was observed all along the continental margin and well out into the deep Makarov and Canada basins at a depth from about 50 m depth in the west to about 150 m in the east. These waters of low aragonite saturation state are traced in historic data to the Canada Basin and in the waters flowing out of the Arctic Ocean north of Greenland and in the western Fram Strait, thus potentially impacting the marine life in the North Atlantic Ocean.