S16 (I) Mercury in contaminated sites: Biogeochemistry and human health

Tuesday, 26 July, 2011

TS16-O1 — 8:30-8:45
Authors: HINES, Mark1, POITRAS, Erin L.1, COVELLI, Stefano2, FAGANELI, Jadran3, ŽIŽEK, Suzana4, HORVAT, Milena4
(1)University of Massachusetts Lowell, mark_hines@uml.edu; (2) University of Trieste; (3) Marine Biological Station, Slovenia; (4) Jožef Stefan Institute, Slovenia;

Mercury (Hg) mining results in a legacy of contamination that can occur large distances for the source and can persist for decades after mining ceases. The Idrija Hg mine in Slovenia recently closed after more than 500 years of operation. Mercury concentrations are elevated throughout the Idrija River system, downstream after the confluence with the So?a river, and well into the northern Adriatic Sea and coastal lagoons more than 100 km from the mine. Since the accumulation of methylmercury (MeHg) is a function of both its production and degradation, we used radiotracer methodology to investigate mercury methylation (203Hg and 197Hg), MeHg demethylation (14C), and sulfate reduction (SR) in sediments in the lower freshwater region of the Isonzo River, the coastal lagoons, and in the Gulf of Trieste. Mercury was readily methylated and demethylated in all sediments, but the relative activity of these processes varied greatly with location, sediment depth and time of year. Methylation activity increased from freshwater to brackish regions and was relatively similar in magnitude throughout the marine sites. However, demethylation was extremely high in the estuarine and lagoon sites. Ratios of methylation to demethylation, which were low in these coastal sites, increased further offshore in the Gulf, which agreed with increased ratios of MeHg to total Hg in Gulf sediments. Comparisons of microbial activities and experiments using microbial inhibitors and stimulants showed that SR strongly controlled both methylation and demethylation. Nitrate amendments strongly inhibited methylation, but stimulated demethylation. Iron also stimulated demethylation, but did not affect methylation. Pore water Hg and microbial activity data showed a dynamic seasonal cycle of S and Hg in which Hg is rapidly mobilized in spring/summer and Hg methylation is active. Hg availability and methylation decrease as sulfide increases in late summer and early fall. Later in fall and winter, demethylation dominates and MeHg decreases. Demethylation is predominantly oxidative, but becomes reductive in winter in surficial sediments. The freshwater to marine gradient in the Idrija/So?a /Isonzo/Adriatic region is dynamic, exhibiting seasonally and horizontally variable rates of SR and Hg transformations. Ratios of activities demonstrated that despite the mobilization of Hg that occurs in brackish waters, shallow coastal sediments are very active at demethylation, which keeps MeHg levels lower than might be expected otherwise. However, this methylation becomes more dominant offshore.

TS16-O2 — 8:45-9:00
Authors: BALDI, Franco1, MAIDA, Isabel2, MARCHETTO, Davide3, FANI, Renato2
(1) Scienze Molecolari e Nanosistemi, baldi@unive.it; (2) Biologia Evoluzionistica, Università di Firenze; (3) Scienze Molecolari e Nanosistemi, Cà Foscari University;

Mercury pollution at the Marano and Grado lagoons (Italy) is caused by Isonzo river carrying cinnabar from Hydria mines (Slovenia) and from outflows of a chloro-alkali plant. Despite of this contamination, clams (Tapes philippinarum) cultivation is one of the main economical activities in the region. Recently, a study performed by ISPRA (Istituto Superiore per la Protezione e la Ricerca Ambientale) demonstrated that clams cultivated in different stations of two lagoons did not accumulate total mercury and methylmercury as might be expected. Hg contents in soft tissues of cultivated and natural population of clams remained under values within the law. Thus, we hypothesyze that this low Hg uptake, in spite of sediments contamination should be the result of clams-bacteria symbiontic adaptation to Hg pollution. Natural clams were collected in four out of 10 stations: M-2, M-4 from Marano and GD and IGC from Grado Lagoons. Alive clams were sterile dissected in siphon, gills, and epato-pancreas. The tissue was lain carefully on solid Nelson medium amended with 5 mg.l-1 of HgCl2 and incubated at 30°C. Fourteen bacterial colonies, growing directly on soft tissue, were isolated and strains were identified by 16S DNA sequencing and analysis, revealing that isolates were representative of eight bacterial genera, four Gram-positive (Enterococcus, Bacillus, Jeotgalicoccus and Staphylococcus) and four Gram-negative (Stenotrophomonas, Vibrio, Raoultella and Enterobacter). Strains were checked for plasmids and merA genes. Biochemical tests for Hg reduction was performed with complete assay mixture amended with NADPH in order to associate bacterial activity with elemental Hg(0) production. Optical and transmission electronic microscopy observations were performed to identify bacteria which colonize soft tissue of clams. Endo-symbyontic bacteria were also observed by TEM analyses. In conclusion, Hg-resistant bacteria play a role in clam adaptation to Hg polluted in our areas and maintain Hg concentrations in tissue under threshold values within the law by enzymatic Hg reduction to volatile Hg(0).

TS16-O3 — 9:00-9:15
Authors: BRATKI?, Arne1, HORVAT, Milena1, OGRINC, Nives 1, FAJON, Vesna1, SNOJ TRATNIK, Janja1, KOTNIK, Jože1, KOCMAN, David1
(1) Jožef Stefan Institute, arne.bratkic@ijs.si

Mercury (Hg) represents one of the most potent toxics and its environmental cycle is not yet clearly established. The shifting between several species with various toxicity is, however, strongly linked to physico-chemical settings in the environment and also dependent upon biological activity. As seasonal changes greatly affect influencing parameters, a seasonal study was conducted in order to better understand the controls on Hg water cycling and thus its potential to enter biota and the food chain.

The investigated area was Gulf of Trieste at the river So?a estuary, which is under high human-induced environmental stress due to water traffic, intensive mariculture and gas terminal construction. Mercury is imported into this environment by river So?a, which transports Hg originating from cinnabar deposits in the region of closed mercury mine in Idrija. During four samplings (September 2009, May, August and October 2010) Hg species were determined in sediments and in water column on a relatively short space scale to asses the influence of seasonal changes throughout the year. Possible influences of elevated freshwater input were also considered. This study is particularly unique, since the last seasonal investigation occured in 1995 and 1997 (Horvat et al, 1999). The measurements of sedimentary Hg indicate that its input has not decreased during recent years, which is consistent with findings of Horvat et al. (1999). Furthermore, decrease of total Hg (THg) with depth was only evident in September 2009, whereas in other samplings was not very distinct. In contrast, methylmercury (MeHg) concentrations were generally higher at sediment-water interface. MeHg was never more than 0,15% of total Hg but was comparatively higher in May and August 2010 (sampling occured after rain events). In September 2009 and August 2010 were the concentrations of THg and MeHg higher closer to the river mouth. Results also show that up to half of water column Hg is present as dissolved gaseous species or can be easily reduced. The rest is mostly bound to particulates larger than 0,45 µm. The water column appears to be uniformly mixed as no vertical gradient was observed, except that gaseous Hg concentrations are higher closer to surface. To conclude, the seasonal changes show influence on Hg concentration in upper sediment, but not in the water column, which has an important capability to reduce Hg. Hg input to the area may be significantly increased after heavy rain or snowmelt.

TS16-O4 — 9:15-9:30
(1) USGS, mmarvin@usgs.gov; (2) UC Davis.

The San Francisco Bay (SFB) estuary and its watershed are currently undergoing tremendous change in the form of numerous wetland restoration projects, the largest of which involves converting approximately 6,500 hectares of former salt-production ponds in South SFB back to tidal marsh, as well as enhancing existing ponds for wildlife habitat. With a 50-year time horizon for completion, this represents one of the largest on-going wetland restoration efforts in the world. Apart from its very large scale, one of the biggest challenges faced by project managers is legacy mercury (Hg) buried in primary slough channels and former salt ponds within the restoration area. The source of this Hg is largely associated with drainage from the New Almaden mining district, the largest cinnabar (HgS) mine in North America with approximately 38,000,000 kg of elemental Hg(0) produced during its operation. Recent data indicates that fish and birds in the South SFB region have among the highest levels of Hg contamination within the whole SFB estuary and its watershed. The reopening of many of the hydrologically isolated salt ponds to natural tidal flushing has raised concerns that the resulting increased tidal prism will extensively scour adjacent slough channel sediment, subsequently remobilizing legacy Hg and enhancing Hg bioaccumulation. Further, concerns have been raised regarding the ultimate increase in wetland habitat, an ecotype known to be particularly important for methylmercury (MeHg) production. Scientists are working with agencies responsible for the wetland restoration project to better understand the linkages between restoration actions (e.g. levee breaches) and the mobilization and potential bioaccumulation of Hg in both the restoration area and the larger South SFB region. Preliminary results examine how the existing strong gradients in salinity, primary production, and dissolved organic matter in the slough/pond complex affect Hg speciation, methylation, bioavailability and bioaccumulation. These results indicate that the partitioning of both total Hg and MeHg is strongly affected by both salinity and dissolved organic carbon concentrations in surface water across the slough/pond complex. Further, the availability of inorganic Hg(II) for methylation appears to be largely controlled by sediment redox conditions across the restoration area. This research has significant value for all habitats and ecosystems where the remobilization of legacy contaminants of any type is an important restoration management consideration.

TS16-O5 — 9:30-9:45
Authors: NEFF, Margaret1, BHAVSAR, Satyendra2, ARHONDITSIS, George1, JACKSON, Donald1
(1) University of Toronto; (2) Ontario Ministry of the Environment;

The English-Wabigoon River system in northwest Ontario, Canada, is a mercury-contaminated waterway. In the 1960s, effluent from a chlor-alkali plant in Dryden, Ontario, released approximately nine metric tonnes of mercury into Wabigoon Lake, which was subsequently transported downstream through the Wabigoon and English Rivers. Following the complete stoppage of mercury release in 1975, several studies have documented the dynamics of bioavailable mercury in this system, as well as the impacts on aquatic biota and human consumers. Early studies documented initial rapid declines in mercury concentrations, but by the early 1980s, declines showed signs of slowing. The Ontario Ministry of Environment, in collaboration with Ontario Ministry of Natural Resources, has monitored levels of mercury in several species of fish from four representative lakes in the English-Wabigoon River system since the early 1970s. This study compiles these long-term data to assess temporal trends in mercury contamination in walleye, northern pike and lake whitefish, three species important for sport and subsistence fishing in this region. We used a variety of statistical techniques, including linear regression, piecewise regression, Mann-Kendall Test, standardized length concentrations and dynamic linear modeling to assess change in mercury concentrations in these three species from four lakes within the river system. We found that for all lakes and species, there is a significant decline in mercury contamination through time; however, there is evidence that in most recent years, this decline is either slowing or levels remain constant. Restrictive fish consumption advisories are in place for these lakes. We also compared most recent mercury levels with data from nearby unaffected lakes and 200+ locations in northwest Ontario. These results will have important impacts on the continued monitoring of mercury levels in these systems, and their relationship to human health concerns in this region.

TS16-O6 — 9:45-10:00
Authors: PTACEK, Carol J.1, BLOWES, David W.1, DAUGHERTY, Shannon D.1, DESROCHERS, Krista A.N.1, GIBSON, Blair D. 1, WANG, Ou1, TORDIFF, James1, LIU, Peng1, LINDSAY, Matthey B.J.1, LANDIS, Richard C.2, DYER, James A.2, GROSSO, Nancy R.2, BERTI, William R. 2
(1) University of Waterloo, ptacek@uwaterloo.ca; (2) DuPont Co.;

The addition of solid-phase reactive materials has the potential to modify (bio)geochemical conditions within watersheds and thereby reduce contaminant fluxes and lessen biotic exposure. This study evaluates the effectiveness of solid-phase reactive media for stabilizing Hg in fluvial environments. Potential applications include use of reactive materials in sediment caps, as floodplain amendments, in permeable reactive barriers and in flow-through treatment systems. The reactive media selected for evaluation included a range of carbon-based adsorbents, clays and organically-modified clays, complexing agents and strong reductants. Laboratory batch-style studies were conducted to evaluate Hg uptake from river water by reaction with the solid-phase materials under aerobic and anaerobic conditions. The observed removal of Hg was variable, with reductions in concentrations ranging from 70 to > 99%. Sediment stabilization experiments were conducted by blending Hg-containing river sediment with reactive media. Aqueous concentrations of Hg were observed to decrease by 0 to > 99% relative to unamended controls. Aqueous Hg concentrations could be maintained in the low ng L-1 range for extended times within mixtures containing carbon-based adsorbents, organically modified clay and strong reductants. The influence of dynamic flow conditions on Hg treatment was evaluated through a series of column experiments. Reactive media was blended with sediment and the column effluent was monitored to determine changes in water chemistry as column flow rates varied. Effluent Hg concentrations were consistently < 50 ng L-1 in the columns containing sediment amended with organically modified clays and strong reductants, independent of variations in flow velocity. In the unamended sediment, effluent concentrations of Hg increased after periods of stagnation. Field implementation of reactive media in a river environment requires consideration of the interactions among (bio)geochemical and hydrological processes.

TS16-O7 — 10:00-10:15
Authors: GHOSH, Upal1, GILMOUR, Cynthia C.2, KWON , Seokjoon1, HENRY, Elizabeth3, MENZIE, Charlie3, BROWN, Steven4, LANDIS, Richard 5
(1) U. Maryland Baltimore County, ughosh@umbc.edu; (2) Smithsonian Environmental Research Center; (3) Exponent; (4) Dow Chemical; (5) DuPont.

There are few available in situ remediation options for Hg contaminated sediments, short of capping. Since Hg risk in moderately-contaminated sediments and soils derives mainly from the bioaccumulation of methylmercury (MeHg), a reasonable remediation goal in these areas can be to reduce the production and bioaccumulation of MeHg, rather than to reduce total Hg concentrations per se.

Activated carbon (AC) and other sorbents have been successfully used to remediate PCBs and other organic contaminants in situ. The mechanism is sequestration of these contaminants in a solid phase that is less mobile and less bioavailable to organisms. Theoretically, activated carbon should act in the same way to reduce MeHg bioavailability, and perhaps also reduce Hg bioavailability for microbial methylation.

We tested the ability of AC, plus a small set of other amendments, to reduce Hg bioavailability in contaminated sediments using sediment/water microcosms. Sediments from four contaminated sites were tested, including 2 freshwater and two tidal sites. Total Hg concentrations at the sites ranged from ~5-50 ppm, and MeHg from 5 to 25 ppb. We followed sediment biogeochemistry, net MeHg production, Hg and MeHg partitioning, and Hg and MeHg accumulation in Lumbriculus over 14 days. Amendments were mixed into sediments at concentrations of 1- 5% of wet weight, or between 0.3 and 1 g of amendment per g of native organic matter in the sediments tested.

Activated carbon was effective in reducing MeHg concentrations in worms in all of the sediments tested. AC reduced MeHg concentrations in porewaters, relative to unamended controls, by 50-95%, a result of increased MeHg partitioning to the solid phase. Pore water total Hg concentrations were also lower that controls, but the effect was less pronounced. Bioaccumulation of MeHg by Lumbriculus was reduced between 30 and 90%. MeHg bioaccumulation across treatments and sites was strongly correlated with porewater MeHg concentration; thus porewater MeHg concentrations may serve as a good surrogate for MeHg bioaccumulation.

Activated carbon can be physically mixed into sediments or added to capping materials. These materials can also be added with little physical disturbance by delivering them as pellets that will naturally mix into sediments over time via bioturbation and advective water flows. SediMite® is a patented technology for this delivery method. We are currently testing AC-based SediMite in a number of field trials.

TS16-O8 — 10:15-10:30
Authors: MILLER, Carrie1, DAVID, Watson1, LESTER, Brian1, LOWE, Kenneth1, LIANG, Liyuan1
(1) Oak Ridge National Laboratory, millercl@ornl.gov

The Y-12 National Security Complex (Y-12 NSC), a US Department of Energy facility in Oak Ridge, TN, USA, used 11 million kg of elemental mercury (Hg0) between 1950 and 1963 for lithium isotope separation processes. Mercury released during this period still remains in soils, groundwater, and in and under buildings at the facility. Mercury field characterization was conducted in 2010 using soil-gas analyses at contaminated sites and mercury contaminated cores were recovered from the surface down to 9 meters. More than 75 soil sub-samples were processed to quantify and characterize Hg using analytical methods, including total Hg digest/atomic absorbance (AA) detection, sequential extraction, Hg(0) headspace analysis and SEM. Hg concentrations, determined by AA detection following soil digestion, ranged from 0.2 to 19000 ppm. Analytical challenges exist in quantifying total Hg concentrations in samples that contain elemental Hg due to the difficulty of obtaining homogeneous samples. Multiple analyses (n=5) were conducted on several samples containing Hg0 and the resulting percent relative standard deviation in total Hg concentration were as high as 76%. Headspace analysis for Hg0 conducted in the laboratory proved to be an effective and rapid technique to identify samples containing beads of Hg0 and was used to prescreen samples before total Hg analysis. Interestingly, headspace analysis conducted on some samples containing Hg0 showed that the gas phase was not saturated with Hg0, as would be expected if Hg0 was present in the sample. This could be a result of the formation of coatings around the Hg0 in the soils inhibiting the volatilization of the Hg. The presence of a coating around small beads of Hg0 is supported by SEM images obtained on the samples. These coatings have the potential to reduce the solubilization and oxidation of Hg in contaminated soils and therefore, understanding these coatings has important implications on the remediation of soils containing Hg0. Sequential extractions were also conducted on samples with and without Hg0 to understand Hg speciation and to examine the reactivity of different Hg fractions within contaminated soils collected from the site.

Tuesday, 26 July, 2011