2i: Regional and local scale stable isotope studies of mercury biogeochemical cycling and bioaccumulation

Elevated mercury (Hg) levels in the aquatic food web of the Everglades has been a public concern since the late 1980s, and remains a persistent issue confronting ecosystem restoration efforts. The formation of methylmercury (MeHg) in the system is driven by the bioavailability of inorganic Hg(II) and the metabolic activity of sulfate-reducing bacteria. These in turn are influenced by the abundance and character of dissolved organic carbon (DOC) and sulfate concentrations. Here, we report on an eight-year collaboration between the U.S. Geological Survey and the National Park Service to examine the drivers of spatial and temporal variability of Hg and MeHg within the Everglades National Park. Surface water and forage fish (Gambusia Holbrooki and Jewel Cichlid) were collected annually at 76 sites from 2008 to 2016. Water samples were analyzed for total mercury (HgT), MeHg, DOC, specific UV absorbance at 254 nm (SUVA), and major ions. Fish were analyzed for HgT, MeHg, and stable isotopes of carbon, and nitrogen. In addition, archived fish samples are currently being analyzed for stable isotopes of mercury to provide greater insight into trophic transfer.

Methylmercury concentrations in water and fish exhibited distinct regional patterns with peak levels in the Shark River Slough (SRS) a receiving area for sulfate and DOC-rich canal water. Throughout the SRS, HgT was a strong predictor of MeHg concentrations in filtered water (R-squared = 0.70), whereas in marsh sites unaffected by the canal this result does not hold true (R-squared = 0.15). Regression analysis using a gradient boosting algorithm revealed several additional drivers of MeHg production which included: dry days preceding sampling, sulfate concentration, and SUVA levels (all of which were greatly enhanced in the marsh relative to the SRS). Annual hydrologic conditions were an important driver of mercury levels in water each year. High water during an El Nio event (2009-10) corresponded to a decrease in HgT and MeHg concentrations, while low water levels during a strong La Nia event (2011-12) corresponded to a dramatic increase in concentrations due to drying and rewetting cycles.

Overall, trends of Hg in fish generally followed those of MeHg in filtered surface water with the exception of elevated Hg in fish from the marsh in 2010 (likely due to fish escaping the SRS in high water). In addition, the prevalence of exotic Jewel Fish in the Everglades has greatly expanded over the course of this study; recent data (2010-2013) show that Jewel Fish mercury concentrations (measured as HgT) are 15-20% higher than mercury concentrations in native Gambusia. Mercury sources and diet differences between Jewel Fish and Gambusia will be assessed using stable isotope techniques (Hg, C, N) to better understand their mercury uptake dynamics.

Mercury (Hg) released by mining activities can be highly dispersed in the environment and may undergo species transformations processes, which are essential to evaluate environmental risk. Hg from roaster emissions or mining-residues is accumulated in soils and sediments. In aquatic systems Hg can be transported over great distances as dissolved organic matter complexes or bound to suspended matter (SM). Hg-isotope analysis have been used to track Hg-sources in Hg-mining areas although it is mostly unclear to which extent changes in Hg-isotope ratios are attributed to different sources or Hg-species transformations. In Idrija-Slovenia, Hg was mined as Hg-sulphide (cinnabar-ore, α-HgS) for centuries. Dumping of mining-residues caused accumulation of cinnabar in river sediments, whereas atmospheric Hg deposition led to contamination of soils and enrichment of organically-bound Hg, which, in contrast to cinnabar, can potentially be methylated in aquatic systems. Hg released from soils and mine dumps is mainly transported via SM in Idrijca river to the Gulf of Trieste (GT) in Italy. So far, high δ²⁰²Hg values in the marine environment were related to cinnabar from the Idrija mine, and different values from that of cinnabar where assigned to the Adriatic background, but organically-bound Hg derived from soils have been neglected. Here, we determine Hg-species and Hg-isotopes in SM, soils and sediments from the Idrijca-catchment to decipher variability of Hg-isotope ratios related to distribution of Hg-species. Hg-isotopes were determined by means of MC-ICP-MS while Hg-species were analyzed using Hg-thermo-desorption.

In general, δ²⁰²Hg values in SM correspond to those found in soils ranging from -2.81 to 0.73‰ and from -2.45 to 0.15‰, respectively. Speciation measurements in different grain sizes reveal that smaller grain sizes (0.45-20 μm) are dominated by organically-bound Hg, while larger grain sizes mainly contain cinnabar (>20 μm). Samples dominated by cinnabar show relatively higher δ²⁰²Hg values (~50% cinnabar, δ²⁰²Hg > -1.01‰) while organically-bound Hg shows lower δ²⁰²Hg values.

Hg-isotopic ratios combined with speciation analyses reveal that most Hg in SM transported to GT is soil derived organically-bound Hg. During rain events SM-loads in the river drastically increase, and also the portion of cinnabar in SM increases due to erosion of larger particles from soils and mining-residues. Our results indicate that the polluted soils in the wider Idrija area are the most important source of Hg transported to GT, and that this Hg organically-bound Hg forms have a large potential for methylation and uptake in the marine food web.

Mercury cycles between atmosphere, terrestrial and aquatic reservoirs through a complex combination of transport and transformation processes, until it is buried in oceanic or terrestrial sediments. Monomethylmercury is the main species bioaccumulated and bioamplified in the food chain. Knowledge of the origin (continental, oceanic or local sediment) of this species for the littoral environnment can still be improved.

This study focuses on ~.200 samples of bivalves (mussels: Mytilus edulis and Mytilus galloprovincialis and oysters: Crassostrea gigas and Isognomon alatus) from ~80 stations along the French coast. The goals of this study were i) to describe Hg levels, speciation and isotopic composition in various bioindicator bivalves from the French coastline, and ii) to identify possible geographic, taxonomic or temporal variations in these properties.

In the bivalves, we observed that the variations in methylmercury concentrations follow linearly those of total mercury (HgT). Preliminary results of isotopic fractionation of HgT in some of the samples do show regional effects and allow to differentiate English channel,. Atlantic, and Mediterranean coastlines. These differences may come from the various trophic regimes revealed by NC stable isotopes, from different fractionnation by the three biological species of bivalves, from different mercury source isotopic signatures, or both.

At most of the study sites, HgT concentrations have not decreased since 1987, despite regulations to abate or ban mercury used by anthropic activities.

This study is part of the Trococo project, funded by Ifremer, Rgion Pays de la Loire (Pollusols), and INSU/EC2CO/Dril.

Riparian soils and sediments of the East Fork Poplar Creek (EFPC) floodplain (Oak Ridge, TN) contain large quantities of mercury from up-gradient contamination at the Y-12 National Security Complex. Important questions remain, however, on the long-term fate of legacy mercury in soils and sediments at this location. A recent study by our group observed substantial dissolved gaseous mercury (Hg(0)) evolution in saturated soil from the EFPC floodplain that contains mercury predominantly as nanoparticulate mercuric sulfide. This finding challenges the notion that authigenic mercuric sulfide is an environmental sink for mercury under all conditions. Periodic soil flooding could lead to the recycling of legacy mercury to the atmosphere. We use stable mercury isotope measurements to (1) provide insights on the process(es) responsible for the formation of Hg(0) in laboratory saturation experiments and (2) quantify the isotopic signature of mercury emitted from the EFPC floodplain to the atmosphere. A comparison of laboratory and field measurements allowed us to evaluate the environmental significance of dark Hg(II) reduction on the fate of legacy mercury in soils.

The stable mercury isotope composition of pore water samples (n = 22) from microcosm experiments was quantified over 36 days of flooding; previous work showed that the reduction of divalent mercury (Hg(II)) to Hg(0) controls mercury release dynamics. We also quantified the stable mercury isotope composition of the soil prior to the start of the experiment (t = 0 day; n = 3) and throughout the progression of the 36 day experiment (n = 6). As flooding proceeded, substantial mass-dependent fractionation (MDF) of stable mercury isotopes was observed resulting in pore water signatures becoming progressively lighter (δ202Hg = -0.52 to -1.70‰) and soil signatures progressively heavier (δ202Hg = +0.15 to +0.75‰). Mass-independent fractionation (MIF) of odd-mass isotopes in pore waters was also observed (Δ199Hg = +0.03 to +0.23). We employed a binary mixing model to explain the processes responsible for MDF and MIF of pore water and soil mercury using two endmembers: (1) the soil mercury at t = 0 day and (2) Hg(0) formed through dark Hg(II) reduction. In the field, stable mercury isotope compositions were quantified on atmospheric samples collected by gold amalgamation above the EFPC floodplain (n = 6). Stable mercury isotope measurements of atmospheric samples are compared with results from the laboratory experiments and previous investigations of the contamination site, and are discussed in context of the fate of mercury in the EFPC floodplain.

Estuarine sediments accumulate mercury (Hg) from watershed runoff and industrial point sources. However, the resuspension and transport of sediments by estuarine, tidal, and storm surge associated mixing can redistribute Hg far from where it was originally deposited. As a result, identifying the sources of Hg in complex urbanized estuaries, is challenging. Here we use Hg stable isotopes to identify the sources of Hg in sediments within the Hackensack River estuary, a tributary to the lower Hudson River estuary in New York and New Jersey, USA. Concentrations of total Hg (>99% inorganic Hg) in sediment from the Hackensack River estuary varied from 0.06 to 154 ug g-1 with the highest concentrations measured in sediment from upper Berrys Creek estuary just below the West Riser tide gate (86 ug g-1) and upper Peach Island Creek (55 ug g-1). The mean d202Hg value of mercury from Berrys Creek estuary and two small tributaries, Peach Island Creek and Ackermans Creek, was significantly higher (p < 0.001) than that from the main stem Hackensack River and its tributary Bellmans Creek. The spatial distribution of mercury and its stable isotopes indicates that mercury from Berrys Creek and its tributaries mixes with, and is partially diluted by mercury from a lower concentration source that is more depleted in 202Hg (lower d202Hg values) as it enters the Hackensack River. Mixing plots of d202Hg vs. inverse total Hg concentration in sediment for the entire system show that d202Hg values decline linearly with decreasing Hg concentrations. Using a binary mixing model, it is estimated that most (>75%) of the mercury in Bellmans Creek and perhaps Mill Creek originated from upper Berrys Creek. Surface sediments from Peach Island Creek had very high mercury concentrations indicating that they could be an important source of mercury to the system.

The 14-km Grand Portage Trail (GRPO), located in extreme northeastern Minnesota, was a primary route of transport between Lake Superior and inland waters for the French-Canadian Voyageur fur trade during the 17th-19th centuries. Recent work has indicated elevated levels of total mercury, methylmercury, and % methylmercury in components of the lower food web (water, sediment, soil, benthic invertebrates), compared to other national parks in the western Laurentian Great Lakes region. Surficial soil and sediment samples at GRPO yielded 3 to 4-fold enrichment in total Hg relative to organic matter content when compared to other regional forested sites. We sampled soil cores at five locations along the GRPO trail in June 2015 to define the spatial extent of total Hg contamination, and to determine whether the Hg stable isotopic signature differed between contaminated and non-contaminated sites. Two locations (Fort Charlotte and the Stockade) were located at each terminus of the trail, and were known historic locations of vermilion (a bright red HgS-containing pigment) storage and use. These terminal sites exhibited total Hg concentrations (Ft. Charlotte: 6.5±5.0 µg/g dw; Stockade: 0.58±0.36 µg/g) far exceeding locations between them along the trail (0.052±0.021 µg/g). Analysis of the Hg stable isotopes 199Hg and 202Hg by multi-collector ICP-MS indicated that the contaminated sites were indeed isotopically similar to each other (Ft. Charlotte and Stockade data combined: δ202Hg(II): -0.49±0.14‰, ∆199Hg(II): 0.00±0.03 ‰), but different from both the low-total Hg locations mid-trail (δ202Hg(II): -1.24±0.17 ‰, ∆199Hg(II): -0.15±0.07 ‰) and reference soils collected at the Experimental Lakes Area (ELA) in northwestern Ontario, Canada (δ202Hg(II): -1.67±0.17 ‰, ∆199Hg(II): -0.24±0.07 ‰). Samples formed a mixing line between the contaminated and non-contaminated sites, with the mid-trail sites falling between the GRPO contaminated sites and that of ELA. Two archived vermilion samples from the fur trade era were analyzed isotopically for comparison, one each from GRPO and Fort Michilimackinac (Mackinaw City, MI). The GRPO vermilion exhibited an isotopic signature that was markedly similar to the GRPO contaminated sites, while the Michilimackinac sample did not. Our results suggest that anthropogenic contamination associated with historic fur trade commerce is likely responsible for elevated total and methylmercury concentrations observed in the contemporary GRPO food web. We plan to analyze additional regional archived vermilion samples to aid in source attribution.

Source apportionments of mercury (Hg) using stable Hg isotopes have been applied in many previous studies, providing unique clues of typical emission sources and processes in the environmental circulation of Hg. However the fingerprint of atmospheric Hg isotopes, especially in the background area, has been rarely investigated limited by the extremely low concentrations and complicated transformation among three species of atmospheric Hg (gaseous elemental Hg, gaseous oxidized Hg and particle-bound Hg). In this study, total gaseous Hg (TGM, composed by gaseous elemental Hg and gaseous oxidized Hg) samples were collected using chlorized active carbon trap in marine boundary layer of the Yellow Sea and Bohai Sea in summer, 2016. Negative δ202Hg (-1.26±0.69‰, 1σ, n = 7) and Δ199Hg (-0.09±0.05‰, 1σ, n = 7) signatures were observed in TGM samples collected in the Yellow Sea, and samples collected in the Bohai Sea shows more negative δ202Hg (-2.02±0.69‰, 1σ, n = 9) and near-zero Δ199Hg (-0.02±0.04‰, 1σ, n = 9) signatures. The isotopic compositions and ratio of Δ199Hg to Δ201Hg (~1.0) suggesting different mixing ratio of gaseous Hg from both natural background pool and anthropogenic emissions transported from continent, with influences from environmental processes, i.e. the photochemical reactions and evaporation of Hg in the marine surface. In addition, negative Δ200Hg signatures (-0.04±0.03‰, 1σ, n = 7) observed in TGM samples collected in the Yellow Sea implying the active oxidation of elemental Hg and adsorption/desorption on the particle surface.