S4 (II) Mercury in the Arctic

Arctic char (Salvelinus alpinus) and lake trout (Salvelinus namaycush) are the second most frequently consumed country food in Inuit regions of Canada, and can represent up to 35% of human Hg intake. Previous research has shown that concentrations of mercury ([Hg]) in Arctic fishes may vary with life history, such that landlocked fishes have higher [Hg] than anadromous (sea-run) fishes. Results from survey studies indicate that anadromous Arctic char have lower [Hg] than landlocked Arctic char. Comparisons based on survey results have often been confounded by large spatial scales and have not allowed concurrent investigations of covariates that affect [Hg], such as feeding ecology and growth rate. Also, previous comparisons of [Hg] among fish life history types in the Canadian Arctic have not included data from partially anadromous populations (where anadromous and freshwater-resident individuals occur in sympatry), and have not included comparisons among recently-described life history types of lake trout. We recently investigated how [Hg] and covariates of [Hg] (including age, size, C:N (an indicator of lipid), d¹⁵N, and d¹³C) differed among anadromous (sea-run), resident (marine access but do not migrate), and landlocked (no marine access) Arctic char and lake trout. Using data from 10 lakes in the West Kitikmeot, Nunavut, Canada, we found that [Hg] varied between species and life history types. At a standardized fork length of 500 mm, lake trout had significantly higher [Hg] (mean 0.17 µg/g wet weight) than Arctic char (0.09 µg/g). Anadromous and resident Arctic char had significantly lower [Hg] (each 0.04 µg/g) than landlocked Arctic char (0.19 µg/g). Anadromous lake trout had significantly lower [Hg] (0.12 µg/g) than resident lake trout (0.18 µg/g), but there was no significant difference in [Hg] between landlocked lake trout (0.21 µg/g) and other life history types. Significant differences in [Hg] were best explained by age-at-size and C:N ratios; [Hg] was significantly and negatively related to both of these covariates, and species-life history types that were younger-at-size and/or had higher C:N tended to have lower [Hg]. Several species of Arctic fishes have plastic life histories that respond to environmental conditions. Our results indicate that life history-based differences in growth and proximate composition result in significant intraspecific variability in [Hg]. This variability must be taken into account when modeling Hg bioaccumulation, conducting human health risk assessments, and/or developing human consumption guidance.

Seasonal snow is an active media and an important climate factor that governs nutrient transfer in Arctic ecosystems. Since the snow stores and transforms nutrients and contaminants, it is of crucial importance to gain a better understanding of the dynamics of contaminant cycling within the snowpack and its subsequent release to catchments via meltwater. Over the course two field campaigns in the spring of 2008 and 2011, we collected snow and meltwater samples from a seasonal snow pack in Ny-Ålesund, Norway (78°56’N, 11°52’E), which were analyzed for inorganic and organic chemical species, as well as total, dissolved, bioavailable (THg, DHg, BioHg, respectively) and methylmercury (MeHg) species. We observe a seasonal gradient for ion concentrations, with surface samples becoming less concentrated as the season progressed. A significant negative correlation between bioavailable Hg and MeHg (r²=0.26, p=0.0044, n=26) was observed in the snowpack. MeHg was positively and significantly correlated to methylsulfonic acid (MSA) concentrations (r²=0.45, p=0.0022, n=18). Based on these results, we propose a new model for aerobic methylation of mercury involving dimethylsulphoniopropionate (DMSP).

Mercury (Hg) is found at elevated levels in the Arctic ecosystem. It can be transformed in the atmosphere and deposited to the surface within hours. The discovery of this rapid atmospheric oxidation of gaseous elemental mercury (GEM) in the Arctic at Alert, Canada has led to considerable work to try and understand the reaction mechanisms and its implications for the Arctic environment. It is widely accepted that sunlight, frozen surfaces and sea ice are required to facilitate the mercury oxidation process. The Arctic is a vast region which limits researcher’s ability to make spatially representative long term measurements both inland and over the frozen ocean. However, the Alert GAW station has enabled the collection of continuous atmospheric speciated measurements to address seasonal and annual variations inland. Results from this long term atmospheric program show significant seasonal variation in Hg and which we are able to discern atmospheric processes leading to the deposition of Hg to the Arctic.

Near-annual field measurement studies have been investigating the role of the Arctic atmosphere in mercury processes since 1997. Through these studies, it has become evident that much of the important chemical interactions occur between the snow, ice, ocean and the overlying atmosphere. For example, an intensive field campaign was undertaken to make measurements of mercury concentrations in 2009. A sled was equipped with instrumentation and brought out on the ice to the frozen ocean. GEM, RGM and PHg measurements (among other parameters) were collected near open sea ice leads for a 2-week period. Results showed that there is a significant amount of PHg measured over the sea ice that can be available for deposition to the surface. Further, it was reported that the concentration of mercury re-emitted over inland snow is greater than that re-emitted over the frozen sea ice. It is believed that the highly saline ocean and sea ice surface environment retains mercury more effectively than less saline surfaces inland. This finding has significant implications to the current understanding of mercury deposition into the Arctic Ocean.

An overview of how atmospheric oxidation of mercury in the Arctic over time will be presented. The breadth of the work that has been undertaken in this field of research will be highlighted and gaps of knowledge and the path forward in this research area will be discussed.

The atmosphere is generally the most important matrix for the transport of mercury. Globally about 2000 Tons Hg yr^-1 is emitted to the atmosphere from anthropogenic sources of which about 50% originate from coal combustion. Natural emission is likewise estimated to about 2000 Tons Hg yr^-1 as is also the case of reemission of previous deposited mercury.

The level of mercury in the environment is therefore enhanced compared to preindustrial time and levels of mercury in high predator fish is so high that EU and US have given guidelines for a maximum intake to avoid noxious effects of mercury in humans.

Therefore it is very important to determine the levels and follow the trends of mercury over time and to investigate the processes responsible for the removable of atmospheric mercury. In this way the dynamics of atmospheric mercury; the connection between sources, atmospheric transport and receptor areas can be determined. A particular area of interest is the Arctic where each spring Gaseous Elemental Mercury (GEM) is observed to be depleted during the so called Atmospheric Mercury Depletion Episodes (AMDE) which may lead to enhanced load of atmospheric mercury to the Arctic environment. The phenomenon is occurring together with ozone depletion ODE and Br is demonstrated to play a key role in both AMDE and ODE.

At Station Nord (81° 36’ N; 16° 39’ W, 24 m ASL) we have measured mercury in the form of GEM and ozone simultaneously from 1999 to 2002. Adding new measurements from 2006 – 2010 the time series is extended making an estimate of decadal trends possible. Furthermore the data will be interpreted based on chemistry and changing climatic conditions.

Intense research on mercury in Ny-Ålesund, Svalbard has revealed new aspects of importance to the fragile Arctic ecosystem.

Atmospheric mercury speciation measurements were startet up in April 2007 at the Zeppelin air monitoring station. Based on the second longest time series of composite gaseous elemental mercury (GEM), reactive gaseous mercury (RGM) and particulate mercury (PHg) measurements available from an Arctic site a clear seasonal variation was uncovered. It is well accepted that parts of GEM is converted to RGM and PHg during the annual recurring springtime phenomenon AMDE. Increased PHg concentrations occurred almost exclusively during intense AMDEs from March through April, and constituted on average 75% of the reactive mercury species during the respective period. Increased RGM concentrations occurred from March through August. The prolonged period with increased RGM concentration denotes more mercury deposited to the Ny-Ålesund environment than first anticipated.

According to what previously known increased GEM fluxes were measured by a GEM flux gradient method during springmelt in late May. Low total Hg concentrations in meltwater collected from the onset of snowmelt question the importance of meltwater in delivering a significant amount of mercury to the Arctic Ocean (i.e. ionic pulse or spring event). It seems likely that parts of the mercury accumulated in the snowpack is released to the atmosphere upon snowmelt (as indicated by the increased GEM flux), well before meltwater reach marine ecosystems

Significant GEM fluxes were observed from mid June through December. It was speculated whether the GEM flux observed during the non-AMDE period could reflect the background GEM flux under the influence of the fluctuating background GEM concentration in the northern hemisphere (1.5 - 1.7 ngm^-3).

The Hg speciation measurements as well as the GEM flux gradient measurements in Ny-Ålesund are still ongoing.

Gaseous elemental mercury (GEM) is a ubiquitous atmospheric contaminant derived from both natural and anthropogenic sources. Due to its long atmospheric lifetime in mid-latitudes, elemental mercury undergoes long-range transport, including to remote polar regions. It is now well established that GEM is depleted from the atmosphere following polar sunrise in close correlation with ozone depletion events. The main driver of these atmospheric mercury depletion events (AMDEs) is hypothesized to be halogen radical species, causing oxidation to reactive gaseous mercury forms that are easily deposited to the snow and ice. Bromine radicals, including Br and/or BrO, are believed to be the primary oxidants driving AMDEs, and numerous kinetic and thermodynamic laboratory studies in the literature suggest the greater importance of Br. Chlorine has been generally dismissed as an unimportant player, due to the relatively low estimates of atomic chlorine concentrations in the Arctic. During the OASIS (Ocean-Atmosphere-Sea Ice-Snowpack) campaign in spring of 2009 in Barrow, AK we conducted measurements of GEM, speciated mercury, ozone and numerous halogenated and organic species. Calculated (steady-state) time-resolved Cl and Br atom concentrations allow us to pursue the hypothesis that Br atom reaction represents the principal Hg oxidation pathway through examination of the dependence of the observed rate of decay of Hg on [Br], [BrO, and [Cl]. Here we will discuss the results of this analysis, including the extent to which chlorine can represent an important Hg sink. A box model is used to investigate the effect on AMDEs of the surprisingly high chlorine levels observed during this campaign. We also address the relative importance of each radical species for different events and the factors that could cause switching between the dominant oxidation pathways.

We investigate the chemistry and deposition of atmospheric mercury to the Arctic surface in spring using observationally-constrained bromine emission from sea ice as input to a global mercury model. Bromine explosion chemistry plays a key role in driving atmospheric mercury depletion events (AMDEs) in Arctic spring. High bromine concentrations are frequently observed in the Arctic boundary layer and are thought to arise from photochemical debromination of sea salt accumulated over sea ice. This elevated bromine induces rapid oxidation of Hg(0) to Hg(II), which can subsequently deposit to the ice-covered Arctic Ocean where it may become available for methylation and biological uptake upon ice melt. A major source of uncertainty in this system is the bromine source. We use observations from aircraft and surface sites to constrain the emission of bromine from Arctic sea ice leads in spring. With our improved estimate of the bromine source, we simulate mercury chemistry and AMDEs in the Arctic using the GEOS-Chem global chemical transport model, which includes an atmospheric component coupled to land reservoirs and a dynamic surface ocean. We evaluate our Hg simulation using a range of data from aircraft (NASA ARCTAS) and surface sites (Alert, Zeppelin, and Amderma) across the Arctic. Using GEOS-Chem, we quantify deposition to snow and sea ice and assess the fate of mercury deposited during AMDEs. We find significant reemission of Hg(0) from the snowpack under sunlit conditions in spring and late summer, with 60% of deposited mercury eventually reemitted to the atmosphere. We evaluate the sensitivity of deposited mercury to bromine emission, sea ice fraction, and anthropogenic mercury emissions. Implications for input of mercury to the Arctic Ocean upon sea ice melting will be discussed.

Mercury (Hg) exhibits long range transport due to its persistence in the atmosphere and bio-accumulates in aquatic ecosystems in the form of monomethylmercury (MMHg), the toxic form of mercury. There are however uncertainties regarding the sources and fate of MMHg in the ecosystem. Even if the atmosphere is the major pathway of Hg contamination in the arctic, the direct atmospheric MMHg contribution to the aquatic ecosystem is not well understood due to the lack of known reliable measurement methods.

MMHg is formed by methylation of reactive mercury (Hg²⁺), or demethylation of dimethylmercury (DMHg) in the aquatic ecosystem. Various studies suggest that DMHg can be volatilized from surface water. We hypothesized that atmospheric DMHg quickly degrades to MMHg in the atmosphere and is then deposited to snow packs on ice fields. In this fashion, DMHg would be a major source of MMHg to snow or the open water itself.

An analytical method was developed to establish the presence of organic Hg species in arctic air. The method is based on species specific Hg isotopic dilution and online ethylation of MMHg from air samples and trapping of ethylated MMHg and DMHg on Tenax traps. This study presents for the first time concentrations of organic mercury species (monomethylmercury and dimethylmercury) in the arctic lower atmosphere measured during the CGCS Amundsen expedition in summer 2010.

The Hudson Bay airshed is dominated by MMHg ranging from 3.9 – 8.1 (mean = 5.5 ± 2.0, n=6) pg/m³, while concentrations of DMHg ranged from < LOD to 1.6 (mean = 2.8 ± 3.6, n = 5) pg/m³. In the high Arctic however, DMHg concentrations are highest ranging from 1.8 to 9.6 (mean = 4.1 ± 2.3, n = 13) pg/m³. MMHg levels were significantly lower than those measured in Hudson Bay ranging from < LOD to 5.2 (mean = 2.3 ± 1.8, n = 8) pg/m³.This supports the possibility that organic mercury species are volatilized from the ocean and contribute to MMHg bio-accumulating in the arctic ecosystem. The concentrations of organic Hg in water as well as the extent of DMHg photodegradation may be important factors affecting the organic Hg species concentrations. The potential sources and factors affecting organic Hg species concentrations and compositions in arctic troposphere will be further discussed in the presentation.

High mono-methylmercury (MMHg) concentrations have been reported in Arctic Marine mammals creating concern in northern communities that rely on these animals as a food source. Budgets indicate that the vast pool of mercury (Hg) in the Arctic Ocean cannot respond quickly to changes in anthropogenic Hg emissions and that rapid changes mercury concentrations observed in food webs cannot be explained by atmospheric Hg dynamics alone. It has thus been hypothesized that physical, biogeochemical and ecological factors influence Hg cycling within the Arctic Ocean, likely changing the bioavailability of Hg to food webs. However, the role and relative influence of many of these factors in relationship to the Hg cycle remains poorly understood. In this study, we have adapted a 1D sea ice-ocean-biological (NPZD type) model for a column of water in the Beaufort Sea to provide the platform for a Hg fate model. The model is seasonal and calculates nutrient, phytoplankton, zooplankton, and suspended particulate concentrations for 22 water layers with a combined depth of 120m. Total Hg and total MMHg are partitioned into particulate, phytoplankton, zooplankton, and chloride-associated pools and the transformation and transfer of Hg between pools is based on speciation rate constants and partition coefficients previously measured in the Arctic Ocean and elsewhere. The model considers photolytic, chemical, physical and organic transformations of Hg, as well as uptake by zooplankton via phytoplankton grazing. Assumptions made using the current version of the model include that 1) all Hg(0) in the water column is dissolved gaseous Hg and; 2) the only evasion of Hg from the water column to the atmosphere occurs as Hg(0). Here we present simulated seasonal Hg concentrations of all Hg species included in the model and compare them to Hg concentrations measured over the past decade in the Beaufort Sea.

Increasing Hg emissions from tropical and temperate regions should have increasing impact on total Hg (THg) accumulations in lake and stream sediments in Polar Regions. The analysis of the Geological Survey of Canada (GSC) files on THg in the lake and stream sediments across northern Canada revealed that THg in sediments is strongly affected by local metallogenic sources. Factors pertaining to local climate, vegetation cover and topography also play a role, with colder generally frozen areas keeping the THg accumulations fairly low, while vegetated catchments and biologically active streams and lakes with enhanced organic matter accumulations display a greater degree of Hg sequestration from geogenic as well as atmospheric sources. The role of topography varies from exposing trace-metal bearing bedrock along steep slopes to trapping water-mobilized Hg through sedimentation processes in lakes and slow flowing streams. Hence, THg concentrations were found to be lower at lowland sampling locations than at upland sampling locations. In addition, despite the already reported higher lake sediment THg than streams in national scale, the close-up of northern survey revealed the opposite trend (p-value < 0.0001). The THg concentrations decreased with increasing stream order and increasing wet-area coverage per basin.