1d-1: Mercury cycling, bioaccumulation and health impacts in polar regions

Elevated mercury (Hg) levels in Arctic marine biota have long been associated with atmospheric transport and deposition of anthropogenic Hg emissions from the mid-latitudes. Recent modeling studies suggest that Arctic river inputs have been a potentially overlooked source of Hg to the Arctic Ocean. Observations on Arctic river Hg fluxes, in particular from Eurasian rivers, are scarce however. Here we present seasonal observations on dissolved (DHg) and particulate Hg (PHg) concentrations and fluxes for two large Eurasian rivers, the Yenisei and the Severnaya Dvina. We observe large DHg and PHg fluxes during the spring freshet, followed by a smaller, second pulse during the fall. We compare water discharge, watershed area and Hg export relationships for Eurasian and North-American rivers. Using additional Hg freshet observations for the Great Whale river (Hudson Bay), we are able to define run-off vs Hg yield relationships for Eurasian and N-American Hg fluxes to the Arctic Ocean and for Canadian Hg fluxes into the larger Hudson Bay area. By extrapolating these DHg and PHg relationships to pan-Arctic rivers and watersheds we estimate total Hg fluxes to the Arctic Ocean of 31 Mg y-1 which is on the lower end of model-based estimates of 16-80 Mg y-1 and Hg/DOC ratio extrapolated estimates of 50-108 Mg y-1.

To assess atmospheric Hg inputs to the vast circumpolar tundra biome – covering 6% of the global land surface area –we developed a mass balance of atmospheric Hg deposition in the continental Arctic at Toolik field station in northern Alaska, USA (68°38’N, 149°36’W). Over two years, we correspondingly measured net gaseous elemental Hg0 deposition using micrometeorological techniques, deposition of wet HgII and dry HgII, and vegetation Hg inputs. We also conducted detailed Hg stable isotope measurements in the atmosphere, snowpack, vegetation, and soils to quantify sources, and performed atmosphere-snow-soil Hg0 concentration profiles to determine zones and underlying processes that drive atmospheric Hg0 sources and sinks.

We observed that gaseous Hg0 deposition of 6.5±0.7 μg m-2 yr-1 accounted for 71% of total deposition over the two years of measurement. Wet HgII deposition was 30 times lower (0.2 ± 0.1 μg m-2 yr-1) and dry deposition of HgII was 2.5 μg m-2 yr-1 (range 0.8 to 2.8 μg m-2 yr-1). Gaseous Hg0 deposition was driven by a Hg0 wintertime uptake and was enhanced in summertime due to vegetation uptake of Hg0. No photochemical re-emission of Hg0 from the snowpack is observed during the snow-covered period aside from surface re-emissions during springtime during mercury depletion events. Therefore, we conclude that during most of the year re-volatilization of Hg0 is not an important process in the Arctic tundra and that repeated deposition-emission cycles occurring at lower latitudes are tilted toward net Hg0 deposition. Deposition of gaseous Hg0, along with old soil age, explains high Hg loads in tundra soils which exceed levels at temperate sites several-fold, and explains the conundrum that watersheds with some of the lowest Hg wet deposition loads on Earth and limited impacts from AMDEs show elevated Hg in riverine runoff. Our measurements suggest that the Arctic tundra is a globally important sink for atmospheric Hg0 storing up to half of the world’s soil Hg. Anthropogenic climate change represents a risk for re-mobilization of massive tundra soil Hg.

Permafrost thaw driven by climate change in northern high latitudes is thought to play a significant role in enhancing the mobilization of previously sequestered peatland mercury (Hg) to the atmosphere and hydrosphere. Though studies have shown sub-arctic Hg dynamics to be impacted by adjacent permafrost thaw, the magnitude and long-term effects of Hg mobilization in this climate-sensitive ecosystem remain poorly constrained.

To investigate the coupling of different export pathways of Hg, along with the effects on Hg cycling resulting from the evolution of this mire as climate change reshapes the landscape to more fully thawed zones, we measured total gaseous Hg (TGM) fluxes across a permafrost thaw gradient and compared these fluxes to the total Hg measured in peat and lake sediment cores collected at the Stordalen Mire, Abisko, Sweden (68°21'N). In two of the peat core sites we also measured pore water methyl mercury (MeHg) contents.

Mercury flux measurements were estimated using a Tekran 2537 ambient air mercury analyzer integrated into a dynamic chamber system. The nine chamber array is divided into three sites, three chambers per site: (1) Palsa: dwarf-shrub dominated hummocks overlying permafrost, (2) Sphagnum: semi-wet hollows with 100% sphagnum cover with minor Eriophorum vaginatum, and (3) Fen: wet hollows dominated by Carex rostrata and Eriophorum angustifolium. All three sites show variability in the relative magnitude of deposition/evasion throughout the day but consistently show a diel pattern characterized by Hg deposition during lows of photosynthetically active radiation (PAR) and ground temperature and a release of Hg during peak PAR hours and ground temperature periods (13:00-15:00h).

Consideration of the Hg flux measurements along with total Hg concentrations of the peat suggests that palsa has the highest amount of stored Hg. The Sphagnum- dominated sites have less abundant Hg in the core (1-20 ng/g) than the palsa (10-60 ng/g), and the fen sites have intermediate values (1-70 ng/g), at certain depths comparable to the peat cored from the palsa. All core profiles have the highest total Hg concentrations near the surface and decrease with depth. The fen site contains elevated Hg near the surface relative to the Sphagnum while also containing significantly higher levels of pore water MeHg.

Taken together our results suggest that during initial stages of permafrost thaw export of gas-phase mercury may be an important pathway, but as thawing continues mercury export into the hydrosphere becomes significant and environmental conditions for MeHg become more favorable.

Climate change is rapidly altering the Arctic's freshwater cycle. Glaciers and ice sheets are melting, permafrost is thawing and precipitation is increasing throughout the north. In fact, conservative climate models predict that in 50 years, river discharges to the Arctic Ocean may increase by 25-50% with yet unknown consequences on water quality, productivity and physical structure of nearshore marine environments. Our ability to predict the future health and productivity of these marine arctic ecosystems thus ultimately depends on our understanding of arctic freshwater resources discharging to them. Here we present the results of detailed water quality surveys of two arctic watersheds in northern Canada: one in the High Arctic dominated by glacier melt, and one in the Low Arctic defined by extensive retrogressive thaw slump activity. The Lake Hazen watershed (82°N) is anchored by Lake Hazen (544 km2, 267 m deep), which collects meltwater from the Grant Land Ice Cap before discharging into the Ruggles River and then Chandler Fjord and Nares Strait between northern Ellesmere Island and Greenland. The Peel River watershed (68°N) is a tributary of the Mackenzie River, a major inflow to the Beaufort Sea. Transects extending from glaciers/permafrost slumps downstream were completed in 2015 and 2016 to assess the impacts of changes in the terrestrial cryosphere to mercury exports from catchments. In the Lake Hazen watershed, average total mercury (THg) and methylmercury (MeHg) concentrations in 7 glacial rivers were 9.7 and 0.048 ng L-1, respectively. High density turbidity currents transported these waters to the bottom of Lake Hazen, making this ultra-oligotrophic lake a sink for predominately particulate-bound Hg. Although Lake Hazen itself was not an important contributor of Hg to the Ruggles River, slump activity downstream of the lake increased THg and MeHg exports to the marine environment by an order of magnitude relative to the outflow of Lake Hazen. In the Peel Plateau, THg and MeHg concentrations, averaged over 8 slumps, increased from 4.5 and 0.296 ng L-1 upstream of the slumps to 296 and 1.68 ng L-1 downstream, with maximum THg and MeHg concentrations reaching 2.05 µg L-1 and 10.1 ng L-1. Flux estimates for each catchment will be presented. These results highlight the mobilization of mercury previously locked in various forms of ice to aquatic ecosystems with potential consequences for mercury cycling in downstream receiving systems.

We investigated monomethylmercury (MMHg) bioaccumulation in lakes across a 30° latitudinal gradient in eastern Canada to test the hypothesis that climate-related environmental conditions affect the sensitivity of Arctic lakes to atmospheric mercury pollution. Aquatic invertebrates (chironomid larvae and bulk zooplankton) were used as indicators of MMHg bioaccumulation near the base of benthic and planktonic food chains. Previously published estimates of atmospheric mercury deposition in Canada showed this mercury flux declines with latitude. In step with that trend, we observed lower concentrations of THg in both water and sediment in higher latitude lakes. Despite latitudinal declines of inorganic mercury exposure, MMHg bioaccumulation in aquatic invertebrates did not concomitantly decline. The highest MMHg concentrations in zooplankton and chironomids were observed in sub-Arctic (55N) and polar desert lakes (75N). Lakes with greater MMHg in aquatic invertebrates either had higher water MMHg concentrations (reflecting ecosystem production of MMHg) or less MMHg in water but also low water concentrations of DOC, chlorophyll and total nitrogen (reflecting ecosystem sensitivity). The MMHg:DOC ratio of surface water was a highly significant explanatory variable that may be useful to predict lake sensitivity to mercury pollution. Bioaccumulation factors of basal organic matter sources (rock biofilms, seston) showed more efficient uptake of MMHg in low DOC lakes. Our findings demonstrate that Arctic lakes are more sensitive to mercury pollution and this sensitivity is related to the low amount of organic matter in their surface waters. Climate-related controls on lake productivity and DOC influence the fate of mercury deposition on a broad geographic scale.

Elemental mercury, Hg(0) is emitted to the atmosphere from natural and anthropogenic sources and deposited in the oceans as inorganic divalent Hg, Hg(II). Around 80% of the Hg(II) deposited into oceans has been estimated to be reduced and re-emitted to the atmosphere. Oceans cover ~60% of the worlds surface and the re-emission of Hg(0) from sea surfaces is an important source of Hg to the atmosphere. The air-seawater flux of Hg(0) is commonly estimated using gas exchange models based on the calculated gas transfer velocity, the Henry´s law coefficient and the concentration gradient of Hg(0) between air and surface seawater. Geographical variations in the air-seawater flux of Hg(0), especially in areas partly covered with ice such as the arctic, is a major contributor to the uncertainty in current global Hg models.

During the SWEDARCTIC 2016 expedition (8th of August to 20th of September, 2016) in the Arctic Ocean arranged by the Swedish Polar Research Secretariat (SPRS), gaseous elemental mercury (GEM), dissolved gaseous mercury (DGM) in surface seawater and Hg(II) species in air were continuously measured. The expedition started and ended in the East Greenland Rift Basin crossing the Eurasian Basin, the North Pole and the Canadian Basin. GEM and Hg(II) species in air was measured using a Tekran 2537B instrument coupled to an 1130/35 mercury speciation system and the average GEM concentration during the campaign was 1.4 ± 0.2 ng m-3. The average concentrations of Hg(II) and particulate Hg(II) in air were found to be 1.6 ± 2.4 and 2.4 ± 2.0 pg m-3, respectively.

DGM was measured using an automated method and the seawater was sampled from the bow water system of the ship having a water intake at an approximate depth of 8 m. The extracted equilibrium concentration of DGM in the outgoing air was measured using a Tekran 2537A instrument. The average DGM concentration was 40 ± 19 pg L-1 with higher concentrations found when passing through sea ice (48 ± 15 pg L-1) than in open water (19 ± 4 pg L-1). The higher surface DGM concentrations found under sea ice was due to a capsuling effect.

The surface of the Arctic Ocean was found to be supersaturated (438 ± 196 %) and an average air-sea flux rate of 1.52 ± 2.27 ng m-2 h-1 was calculated, indicating a net evasion of Hg(0) from open sea surfaces in the Arctic Ocean.

Extreme warming events are currently being observed in the Arctic with November air temperature being 20°C higher than historic normal and the sea ice coverage in the early winter month being at record lows. As the sea ice extent decreases, new areas, including the central Arctic Ocean, are made more accessible and will become attractive for the commercial fishing industry. Concerns are now being raised for a future “gold-rush” as international regulations are inadequate to protect the ecosystem and human health. One major concern is methylmercury (MMeHg). Human health concerns related to the pollution of the pristine environment of the Arctic by Mercury (Hg) and MMeHg may be triggered by climate change. However, fundamental understanding of the cycling of methylated mercury species in these environments, including the central Arctic Ocean, is lacking and more study warranted. The central Arctic Ocean covers around 40% of the Arctic Ocean area but remains largely unexplored when it comes to the occurrence and cycling of methylated mercury species. In this presentation, we will present data on mono- and dimethylmercury (MMeHg and DMeHg, respectively) in ice, brine, meltwater and sea water collected in 2016 during the SWEDARCTIC 2016 expedition, and for comparison, total methylmercury (MeHgT; i.e. MMeHg + DMeHg) collected in 2007 during the LOMROG expedition. Both expeditions were performed onboard the Swedish IB Oden and coordinated by the Swedish Polar Research Secretariat (SPRS). Our studies cover the central Arctic Ocean from the Canada Basin to the Nansen Basin. The concentrations of DMeHg in the polar mixed layer were unexpectedly low and calculated fluxes of DMeHg from and to the polar mixed layer suggest it to be a sink for DMeHg. We further suggest the degradation of DMeHg to be a major source of MMeHg in the surface waters. The concentrations of MeHgT in seawater in 2007 and 2016 were similar to concentrations previously reported from the central basin but lower than the concentrations typically reported from the Canadian Arctic Archipelago. One contributing factors to the lower MMeHg concentration in the central basin could be an overall lower primary production and thus lower re-mineralization and in situ formation of MMeHg in the deeper waters of this region.

We will present the combined results of the 2014 French GEOTRACES GA01 GEOVIDE cruise (North Atlantic), the 2015 German GEOTRACES GN04 TransArc II cruise (Central Arctic Ocean) and the 2016 German GEOTRACES GN05 GRIFF (Fram Strait) cruises. Full water column profiles were sampled using ultra-trace clean rosettes and total mercury (Hg) was determined directly on board. We find consistent surface depleted profiles in the North Atlantic Ocean, averaging overall 0.57 ± 0.2 pM (n = 527), while we observe surface enrichments in the Arctic Ocean (1.01 ± 0.48 pM, n = 54, 0 – 20 m) and Fram Strait (1.44 ± 0.39, n = 30, 0 – 20 m). Fram Strait is the only deep connection between the North Atlantic and the Arctic Ocean, were warm saline Atlantic waters transit into the Arctic as the West Spitzbergen Current. At the same time, a great part of the immense Arctic fresh water excess is exiting as the East Greenland Current and sea ice. We find that Hg concentrations (0.65 ± 0.19 pM, n = 280) of the Atlantic-sourced water mass present in the Arctic Ocean below the halocline (>200 m) matches well the North Atlantic average. The East Greenland Current transports the Hg enriched surface waters southwards, out of the Arctic Ocean. Hg inputs to the Arctic Ocean were until now based on measurements from the North Atlantic and modeled seawater flows. Our refined measurements will assist in narrowing down the relative contributions of Hg fluxes to and from the Arctic Ocean.

We will make use of the combined data sets of all three cruises to refine the previously published Arctic Ocean Hg budgets using numerical model evaluations based on a coupled 3-D atmosphere-ocean simulation (GEOS-Chem coupled to the MITgcm).