Aqueous Geochemistry of Mercury in Three River Systems Impacted by Mining Activities

by

Jean-Claude J. Bonzongo, Environmental Studies Program, Austin College, Suite 61553, Sherman, TX 75090--jbonzongo@austinc.edu

W.B. Lyons, Byrd Polar Research Center, Ohio State University, Columbus, OH 43210-1002; J. J. Warwick, Dpt of Environmental Engineering Sciences, University of Florida, Gainesville, FL 32611-6450; Jadran Faganelli, Marine Biological Station Piran, 33000 Slovenia; Milena Horvat, Jozef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia; M.E. Hines, Department of Biological Sciences, University of Alaska Anchorage, AK 77000; P. J. Lechler and J. Miller, University of Nevada, Reno, NV 89557.

ABSTRACT

The aqueous geochemistry of mercury (Hg) was investigated in 3 aquatic systems impacted by either Hg- or gold-mining. The first aquatic system, the Idrija River, Slovenia flows near the Idrija Hg-mine, which is the second largest Hg mine in the world with continuous mining and smelting for 5 centuries. It has been estimated that during the 500 years of mining, about 30,000 tons of Hg were introduced to the surrounding environment. The resulting contamination in this case has been associated primarily with the use of inefficient smelting technologies. In contrast to the Idrija River where cinnabar was mined, the Carson River in Nevada, USA and the Madeira River, Brazil have been contaminated by metallic-Hg (Hg⁰) used in the amalgamation process to extract gold from crude ores. The Carson River was the site of intensive gold (Au) and silver (Ag) mining for over 50 years, resulting in an estimated 7,000 tons of Hg lost to the river and its watershed. Finally, the Madeira River in the Brazilian Amazon, is a site of an ongoing use of Hg⁰ in Au-prospecting by gold miners. This study was conducted in order to determine different factors controlling aqueous methyl-Hg levels/abundance in aquatic systems with high total-Hg concentrations, but flowing on different continents/climate and different geological formations. Using ultra-clean sampling and analytical techniques, water samples were collected along longitudinal transects in the 3 river systems and analyzed for Hg levels and speciation, as well as several other key physicochemical parameters. Our results show evidence of Hg contamination in the 3 river systems, with THg concentrations up to 0.020, 0.32, and 10 µg/L (or ppb) for the Madeira, Idrija and Carson rivers, respectively. Measured physicochemical parameters and data from the literature are used to comprehend the fate of Hg in each investigated river system.

INTRODUCTION

Mercury (Hg) is a significant environmental contaminant that accumulates to toxic levels in biota. But concerns over its bio-geo-chemical cycling are mostly due to the toxicity of methyl mercury (MeHg), known for its ability to bio-accumulate and bio-magnify in food chains. The principal pathway for human exposure to this toxic metal is the consumption of Hg-contaminated fish, and such exposure could result in toxicity of the central nervous system. In fact, during the past several years, there has been a growing attention paid to Hg as a pollutant as exemplified by a number of international conferences devoted exclusively to Hg in the environment. This interest stems from the unique characteristics of Hg, which include its high vapor pressure resulting in volatilization and global dispersal via the atmosphere; its chemical speciation; the requirements for ultra-clean sampling, handling, and analytical techniques in order to measure it properly; and as mentioned above, its potential for bio-accumulation and bio-magnification.

Beside diffuse sources and industrial/municipal wastewater discharges, historic and current mining activities constitute important sources of Hg introduction into waterways. This is due to: (1) the use of metallic Hg (Hg⁰) in the almagamation process to extract Au and Ag from crude ores, and (2), the mining of cinnabar (HgS). Results published to date show that Hg released from these sources is in the inorganic form, whereas Hg bio-accumulates in food chains in the methylated form. It is important to note that, the form in which Hg is introduced into an aquatic system dictates its availability for methylation. For example, Hg⁰ released from the amalgamation process should undergo oxidation to be accounted for in the pool of Hg available for methylation. On the other hand, the combination of Hg and S in cinnabar reduces/eliminates the availability of Hg to methylating-agents. Finally, in addition to limitations related to the form of Hg, other biogeochemical parameters can either enhance or inhibit the methylation of Hg(II). Therefore, rates of environmental transformations of Hg and the accumulation of produced methyl-Hg may be site-specific, regardless of total Hg levels present in the system.

In this study, we have investigated the aqueous geochemistry of Hg in river systems with very high Hg concentrations in order to identify factors that could limit MeHg abundance in each system. The following is a brief description of the 3 study sites.

(1)--IDRIJA and the IDRIJCA RIVER: The Idrija Hg mine is located at the southwestern edge of the Alps in the northwestern Slovenia, about 30 Km east of the Italian border. It is the world second largest Hg deposit, surpassed only by the Almaden in Spain. The principal ore mineral in Idrija has always been cinnabar (HgS), but metallic Hg did also occur in a few ore bodies (Bancroft et al. 1991). Five hundred years of Hg extraction from the Idrija ore deposit have resulted in a widespread contamination of the environment. It is estimated that during the active mining period, about 30,000 tons of Hg got introduced into the environment, primarily due to poor smelting technologies. Mercury was released into the environment through both emission from ore smelters and mine tailings, resulting in the contamination of the atmosphere, and terrestrial- and aquatic-systems. A recent study of Hg speciation in tailings materials by Biester et al., (1999) showed that older tailings (i.e. those produced before the 20^th century) contained mostly Hg as HgS. They attributed that left over to inefficient roasting techniques used in the past. In younger tailing materials, however, they found that Hg⁰, either sorbed to mineral matrix or free, as well as traces of oxidized Hg as HgO became predominant. Finally, in a leaching experiment on younger tailing materials, Biester et al. (1999) found that soluble Hg existed in reactive forms. These findings suggest that the introduction of Hg-contaminated younger tailing materials in waterways could bring about Hg species other than HgS, and the above information on Hg speciation could find its importance in comparative studies between the Idrija River and aquatic systems contaminated with Hg species other than HgS.

The Idrija Hg mine is located in the Idrijca River's watershed. A recent investigation of Hg concentrations in sedimentary materials in the Idrijca River watershed showed that riverine, as well as river overbank sediments were highly contaminated (Gosar et al., 1997). Gosar et al. found that beside the upstream section of the river, which had an average Hg concentration of about 2 ppm, sections downstream of the Idrija mine had Hg levels in hundreds of ppm.

(2)--Carson River: In 1859, the Comstock Lode was discovered in western Nevada, USA. Metallic-Hg was then used for nearly over 50 years to extract Au and Ag from crude ores. These mining operations conducted in the Carson River watershed generated extensive hazardous wastes, resulting in significant Hg contamination of both the abiotic and biotic compartments in the watershed. Here, contaminated mining wastes previously accumulated in mill tailings were later dispersed throughout the drainage basin, in large part due to fluvial processes. Hg levels in river-bank materials reach hundreds of ppm, while concentrations in reverine sediments are usually less than 20 ppm. The Carson River flows in a very arid environment in the eastern Sierra Nevada, where evapo-concentration processes affect the biogeochemistry of the river.

(3)--Madeira River: The Madeira River basin is located in the southwestern Amazonian region and drains a very large watershed, mostly covered by tropical rain forest ecosystems (Pfeiffer et al., 1991). In most fluvial systems in the Amazon, artisanal mining became an island of prosperity in a sea of poverty. Metallic Hg is used in the almagamation process to recover fine alluvial gold. In the Madeira River, most of the gold mining occurs in the river itself along a 300Km longitudinal transect. While Hg mining in Idrija and Hg use in the Carson River Drainage basin can be classified as historic, Au-extraction in the Madeira River is an ongoing operation. The mining process uses boats and divers, as well as mechanical dredges to remove riverine-sediment. Heavy particles are then concentrated gravimetically and amalgamated with Hg⁰ to form Au-Hg complexes. The latter is then burned releasing Hg vapor in to the atmosphere. The entire process releases also a significant amount of Hg directly into waterways. Pfeiffer et al. (1991) report that 100 tons of Hg were lost into the Madeira River from 1979 to 1985, and THg concentrations in sediment range from 0.05 to 2.62 mg/Kg (or ppm).

As stated above, the form in which Hg is released into the environment determines its reactivity and transformation rates. For river systems considered in this study, Hg is primarily introduced into waterways as metallic Hg in the Madeira and Carson rivers, and also as cinnabar in the Idrija River. However, from differences in their geographical locations and climate, bedrock geology, and water biogeochemistry, one could anticipate differences in MeHg levels.

MATERIALS AND METHODS

Surface water samples were collected along longitudinal transects from headwaters to river deltas in each aquatic system, using ultra-clean sampling techniques. Due to fact that all samples were analyzed in the US, in both the Madeira and Idrija rivers, samples for the determination of total concentrations (THg and TMeHg) were collected directly into acid cleaned and pre-acidified Teflon bottles (in situ preservation). In addition to the in situ acidified samples, water samples were also collected without acidification. The latter were earmarked for the determination of dissolved fractions of Hg after filtration through acid-cleaned 0.45µm membrane. Our research group has been working on Hg in human impacted rivers on a worldwide scale. Data obtained to date show that for water samples collected as far as Russia and Slovenia, concentrations obtained from both pre-acidified and non-acidified samples do fall nearly on the 1:1 line on a graphical representation, suggesting that no significant change or loss occurs in non-acidified samples during long-distance transportation. During transportation and/or shipping, samples were kept chilled in the dark and analyzed within 2 to 3 weeks from collection date.

The accurate quantification of often very small quantity (ppt) of Hg generally present in both polluted and non-polluted natural waters requires an extremely sensitive, low background, noise free analytical technique. The method of choice is cold vapor atomic fluorescence spectrometry (CV-AFS). This technique was used to obtain information on the chemical speciation of Hg as its solution forms include species out of thermodynamic equilibrium due to biological interactions. Total-Hg, including both dissolved and particle-associated forms, were analyzed by subjecting whole water samples to BrCl/SnCl₂, followed by gas-phase sparging with Hg-free N₂ and trapping of Hg⁰ onto gold-coated sand (Bloom and Creciulus, 1983; Gill and Fitzgerald; 1987). The Hg⁰ was then thermally desorbed from the gold trap in a stream of Hg-free helium and quantified by CV-AFS. Methyl-Hg in water samples was first separated from its original matrix by distillation or extraction in an organic solvent (Bloom, 1989; Horvat et al. 1993a; 1993b) and then ethylated using sodium tetraethylborate (Bloom, 1989), followed by CV-AFS detection after GC-separation and thermal decomposition of alkyl-Hg.

RESULTS AND DISCUSSION

Overall, our results show evidence of contamination in the 3 river systems, with Hg concentrations spanning a wide range of values (Table 1). Absolute values of THg increase in the order MadeiraàIdrijaàCarson, while the absolute values of methyl-Hg increase in the order IdrijaàMadeiraàCarson. The longitudinal distribution of THg concentrations in the 3 river systems shows that Idrija and Madeira Rivers have their peak values near mining operation sites, followed by a decrease in concentrations due to either the removal via sedimentation of contaminated particles or dilution by cleaner waters from tributaries, or both. In contrast, the Carson River flows in an arid environment, where evapo-concentration and Hg affinity to fine clay particles sustain high THg concentrations downstream up to the terminal artificial lake; Lahontan Reservoir.

Table 1: Concentration ranges (in ng Hg/L) of THg and MeHg in surface waters of the 3 river systems

(*DL = Detection Limit: 0.008ng/100ml for THg, and 0.0042ng/100ml for MeHg)

IDRIJA RIVER, SLOVENIA			MADEIRA RIVER, BRAZIL			CARSON RIVER, NEVADA
THg	MeHg	%MeHg	THg	MeHg	%MeHg	THg	MeHg	%MeHg
2.77-322	DL–0.613	0 to 4.52	2.25-20.05	0.61-1.825	0 - 24	4 - 7585	0.3-7.2	0 - 12
Dissolved THg	Dissolved MeHg	%Dissolved MeHg	Dissolved THg	Dissolved MeHg	%Dissolved MeHg	Dissolved THg	Dissolved MeHg	%Dissolved MeHg
0.43-39.23	*DL-0.094	0 -- 7.70	0.28-8.06	0.31-0.647	0 -- 100	0.2 - 88	0.08-3.06	0 - 10

In these 3 river systems, Hg enters waterways mostly in the inorganic form. Since most, if not all, of Hg in fish tissues is methyl-Hg, it is important to understand the factors governing the production and bio-availability of methyl-Hg. Ideally, experiments determining potential rates of Hg(II)-methylation and MeHg demethylation should be conducted to investigate factors controlling Hg transformations in the systems. However, from a simple observational approach such as the one used in this study, the determination of bio-geo-chemical parameters known from laboratory experiments to either enhance or inhibit methyl-Hg production can be used in linear regressions to assess their potential effects on ambient levels of methyl-Hg. For example, it is known that in most tested estuarine and freshwater sediments, sulfate reducing bacteria (SRB) are key players in the production of methyl-Hg (Compeau and Bartha, 1985; Gilmour and Henry, 1991; Gilmour et al., 1992; Bonzongo et al., 1996; Chen et al. 1997). This ties methyl-Hg production to anoxic conditions, which in the river systems under consideration in this study are found only in bed sediments, and seasonally, in bottom waters of reservoirs in the Carson and Idrija River systems. Other parameters such as dissolved organic carbon (DOC), pH, major ions, etc may have significant impact on the fate of Hg in aquatic systems.

The above-described approach shows that when methyl-Hg concentrations is normalized to total-Hg levels in surface waters of the three river systems, an overall decreasing trend of the fraction of Hg present as MeHg is observed with the increase in water sulfate concentrations. While the decrease is observed over a wide range of sulfate concentrations (from 3 to 200mg/L) for Idrija River, a steep decrease is observed in the Madeira River over a very small range of sulfate concentrations (i.e. 3 to 6mg/L). In anoxic conditions, high sulfate concentrations may result in the removal of Hg available for methylation due to HgS formation. In 1991, Gilmour and Henry hypothesized that sulfate concentration far above10 mg/L could result in the inhibition of Hg-methylation, while lower values would enhance Hg-methylation. Apparently, this hypothesis doesn’t hold true for the Madeira and Idrija rivers, based on water data. In both the Madeira and Idrija rivers, the abundance of methyl-Hg decreases significantly in river sections with sulfate concentrations far below 10 mg/L. This observation may suggest that the complexity of natural environments may exacerbate the negative effect of sulfate on Hg-methylation, even in the presence of sulfate concentrations where one would expect significant methyl-Hg production. On the other hand, SRB may not be the major methylating microbial population in these systems. Finally, the above tentative explanations can be confirmed only through studies of sedimentary transformation of Hg in each system.

In contrast to the Idrija and Madeira Rivers, the Carson River shows an apparent increasing trend of MeHg/THg ratios with increasing sulfate concentrations up to values as high as 35 mg SO₄^2-/L before showing a decrease for greater sulfate concentrations. The determination of potential rates of Hg methylation in sediment samples collected from several locations in the Carson River suggests that SRB are the main MeHg producers. Also, SRB's methylating activity is negatively impacted by increasing concentrations of either sulfate or any other group-VI oxyanion (Bonzongo et a., 1996; Chen et al., 1997). The evapo-concentration in this arid environment favors the build up of major ions, including those ions with inhibitor potential for Hg methylation by microbial populations.

Overall, decreasing trends are obtained when the dissolved methyl-Hg expressed as a fraction of total dissolved-Hg is plotted versus DOC. In both the Idrija and Madeira rivers, the impact of DOC on the fraction of Hg occurring as methyl-Hg is observed within a very small range of DOC concentrations (i.e. 0.5 to ~1mgC/L). In fact, Hg has very high stability constants with organic ligands forming stable bonds to carbon, hence true organo-metallic compounds (Andrea, 1986). Accordingly, the increase in DOC concentrations often results in reduced availability of inorganic Hg to methylating agents (Jackson, 1989). Our data showed that increasing DOC concentrations are accompanied with decreasing methyl-Hg to THg ratios in the Idrija and Madeira rivers. In the Carson River, a wider range of DOC values seems not to affect MeHg/THg ratios, suggesting that in the latter, other factors play more important role in controlling MeHg formation than does DOC. One such factor would be total suspended solid concentrations.

Overall, the geology of the watershed always controls the background chemistry of a river. A parameter such as the pH of river is greatly influenced by the bedrock composition. In addition, vegetation and rainfall play a role in the chemical composition of the river water. With regard to pH, the Madeira River has acidic pHs values (~5.6), while the Carson and Idrija Rivers have neutral to alkaline pHs. For the Madeira (lowest THg concentrations) and Idrija rivers, the difference in pH values could, at least partly, explain the fact that absolute MeHg concenttrations are greater in the former than in the latter. Alkaline pHs tend to reduce MeHg accumulation as they favor higher demethylation rates (Bonzongo et al., 1996).

Other information related to Hg contamination, transformation, and bioaccumulation in the above 3 systems will be discussed in the presentation, based on data available in the literature. Meanwhile, the key points of this study can be summarized as follows.

· Despite the fact that Hg concentrations in Idrija riverine-sediments are much higher than in the two other investigated systems, it has the lowest levels of MeHg in water. Also, Hg concentrations in waters of the Carson River System reach values 1 to 2 orders of magnitude higher than those measured in both the Idrija and Madeira rivers. In addition to other differences among the three systems, the lack of vegetation in the Carson River watershed allows Hg-contaminated particles accumulated in riverbanks and flood plain to be easily introduced into the river via erosion processes, while the aridity of the climate favors the evapo-concentration in terminal basins.

· The Madeira and Carson rivers are comparable, in that Hg is introduced into waterways primarily as metallic Hg, while most of Hg introduced into the Idrija is certainly bound to ligands. Certainly the combination of the following: initial form and amount of Hg injected into the system; the time of mining operations (past versus present); the conditions governing transformations of inorganic Hg to org-Hg, and other factors specific to each river system may explain the difference in methyl-Hg levels.

· Sulfate and DOC seem to have negative effects on methyl-Hg accumulation in the Madeira and Idrija rivers, but not in the Carson River.

· These few examples show the complexity of Hg-methylation and MeHg accumulation processes in natural environments, as well as the difficulty, which may result from the extrapolation of the study of one aquatic system to another. Our data sets for the Madeira and Idrija Rivers are very small as compared to the well-studied Carson Rive. The examination of the data indicates the necessity of further investigations in order to assess differences and similarities in factors controlling the fate of Hg in these aquatic systems.

SELECTED REFERENCES

Andrea, M.O. 1986. In the importance of chemical speciation in environmental processes. Bernhard, M., F.E. Brinkman, and P.J. Sadler, Eds. Springer Verlag, berlin. pp 301-335.

Bancroft et al., 1991. The Mineralogical Record, 22: 201-208.

Biester et al., 1999. J. Geochemical Exploration, 65: 195-204

Bonzongo J.C. et al., 1996. Environ. Poll. 92: 193-201.

Bonzongo J.C. et al., 1996. Toxicol. Chem. 15: 677-683.

Chen, Y. et al., 1997. Environ. Toxicol. Chem. 16(8): 1568-1574.

Compeau, G.C. and R.Bartha, 1985. Appl. Environ. Microbiology. 50: 498-502.

Gosar et al., 1997. . J. Geochemical Exploration, 58: 125-131

Gilmour, C.C. and Henry, A.E., 1991. Environ. Pollut. 71: 131-169.

Gilmour, C.C. et al., 1992. Environ. Sci. Technol. 26: 2261-2287.

Jackson, T.A., 1989. Appl. Organomet. Chem. 3: 1-30.

Pfeiffer et al., 1991. Forest Ecology & Management, 38: 239-245.