Environ. Sci. Technol. 2008, 42, 4710–4716
Mercury Deposition in a Polar Desert Ecosystem REBECCA A. WITHEROW* AND W. BERRY LYONS Byrd Polar Research Center and the School of Earth Sciences, The Ohio State University, Columbus, Ohio 43210-1002
Received January 10, 2008. Revised manuscript received March 25, 2008. Accepted April 2, 2008.
Trace metals have received considerable attention in the recent decades due to their potential toxic nature. Glacial snow and ice have been used extensively to elucidate historical changes in the atmospheric composition of trace metals and other compounds. Mercury concentrations in Antarctic ice have described changes in atmospheric mercury deposition during the transition from the Last Glacial Maximum to the Holocene, however the record of modern mercury deposition in Antarctica is limited. Here we present a record of net mercury deposition to Antarctic snow over the past two decades. Over decadal periods, mercury is conserved in the snowpack and is dependent on a regional oceanic source. Annual to subannual mercury concentrations in snow are to some extent preserved and show covariance with marine aerosols as evidenced by calcium concentrations. Aeolian inputs from exposed rock and soil also play a critical role in depositing mercurytoAntarcticsnow.Suchidentificationsalongwithprevious data illustrate that mercury transport directly from the glaciers may account for 25-65% of the total mercury concentration in proglacial streams and the surface waters of perennially ice-covered lakes.
Introduction There has been much interest in the global geochemical cycles of potentially toxicmetals, especially those introduced through anthropogenic activities (1-3). Because of the volatility of mercury (Hg), the atmosphere plays a fundamental role in global distribution, as it is the milieu that transports both natural and human introduced Hg far from its sources (2, 4). Some of the most quantitatively substantial pathways of natural emission into the atmosphere are volcanism, oceanic evasion, forest fires, and anthropogenic emissions including fossil fuel burning, waste incineration, and chlor-alkali production (2). It is critical in the understanding of the global mercury cycle to determine relationship of anthropogenic to natural fluxes by establishing background mercury levels. The Antarctic may be perhaps the only region on Earth to examine natural background levels because of its extreme remoteness (5, 6). Here we show mercury concentration variation during a recent 17-year period in Antarctic snowsthe first modern mercury time series in Antarctica.
Experimental Section Site Description. The McMurdo Dry Valleys are located in southernVictoriaLand,Antarcticaatapproximately76°30′-78°30′ S, 160°-164° E and are the largest (approximately 4800 km2) * Corresponding author e-mail:
[email protected]. 4710
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 13, 2008
ice-free region of Antarctica (Figure 1) (7). The Dry Valleys have been extensively studied since the International Geophysical Year (1958-1959), and in 1993, the United States National Science Foundation established the McMurdo Valleys Long-Term Ecological program (MCM-LTER). Taylor Valley contains three perennially ice-covered lakes, more than twenty-four ephemeral streams, and numerous coldbased glaciers. All glaciers in Taylor Valley, with the exception of the Taylor Glacier, are polar alpine glaciers flowing from the Asgard Range and the Kukri Hills with termini forming cliff faces as high as 20 m from the valley floor (8). The glaciers in the Asgard Range are typically three times larger in area than those in the Kukri Hills because they have higher and larger accumulation basins and receive less direct solar radiation (7). Local climate variation also impacts the equilibrium line altitude (ELA) such that as distance from the coast increases, so does the ELA (8). Alpine glaciers in Taylor Valley receive approximately 5 cm of water equivalent precipitation per year (9). Sampling and Analytical Procedures. Because of the ultralow concentrations of mercury in Antarctic snow, stringent protocols have been established as to avoid contamination (10, 11). Briefly, one-liter widemouth Teflon bottles and custom-made Teflon spatulas were used for sample collection. Bottles and spatulas were washed in successive acid baths of Fisher trace metal grade HNO3, Fisher trace metal grade HCl, and Optima HCl. Samples were collected from a 3 m snow pit in the accumulation zone of the Commonwealth Glacier (77°31′S, 163°01′N; elev. 678 masl) following a modification of the “clean hands/dirty hands” technique for trace metal sampling (4). After initial excavation of the snow pit, surficial layers of the sampling wall were removed using clean polypropelene shovels followed by clean Teflon spatulas. Before each sample, the spatulas were “rinsed” by placing them in adjacent snow that visually appeared to be the same layer. Samples were sent frozen to the United States within 60 days of sampling, and remained in the dark and frozen at -20 °C before melting for analysis. Samples were analyzed using a Brooks Rand Ltd. Model III cold vapor atomic fluorescence spectroscopy (CVAFS) following US EPA Method 1631. Analytical methods were similar to those described by Ferrari and others (11). Deviations from this method include (1) Analyses were not conducted in a Class 10 000 clean room. Instead, they were preformed in a room solely devoted to ultralow environmental mercury samples using Class 100 laminar flow hoods for sample manipulations. (2) Laboratory coats rather than clean garments were worn during analysis. (3) Activated charcoal was not implemented on the Class 100 laminar flow hood. Because the analytical procedure for mercury analysis determines the amount of mercury in a sample, it is independent of sample volume. Thus, all sample volumes were determined prior to mercury analysis. To derive the concentration of the samples, the total amount of mercury analyzed by CVAFS was divided by the volume for each sample. The mean of reagent blanks was 0.35 pmol Hg (n ) 10), and all data have been corrected for this value. To determine the precision of the analyses, recoveries of mercury spikes to 18 MΩ water (DI) were analyzed. This procedure was used because the limited volumes of the samples would result in consumption of the entire sample. The relative standard deviation was 15.2%. The detection limit was determined by the mercury content of two field blanks (1.3 pM; 2σ). Three 10.1021/es800022g CCC: $40.75
2008 American Chemical Society
Published on Web 05/22/2008
FIGURE 1. Taylor Valley, Antarctica. samples were sent chilled at 4 °C for external CVAFS analysis to the Brooks Rand Laboratory. Of the 34 samples analyzed at Byrd Polar Research Center, five (42, 66, 130, 170, and 250 cm) yielded very low accuracies based on mercury standard additions and poor recoveries of check standards. These samples were excluded from the data set, and data collected at the Brooks Rand Laboratory were used for three (42, 130, and 250) of the five samples. Major ions were analyzed using a Dionex DX-120 ion chromatograph (IC) by modified techniques from the work of Welch and others (12). An independent multielement standard was analyzed after creating a calibration curve to satisfy quality assurance and quality control. The recoveries for Na+, K+, Mg2+, and Ca2+ were ( 5% (n ) 3), Cl- and SO42were ( 2% (n ) 2), and NO3- was ( 17% (n ) 2). Several sample duplicates were analyzed for recovery comparisons and were used for precision analysis. The percent recoveries for the sample duplicates were ( 4% for Na+, K+, Mg2+, and Ca2+ (n ) 8), ( 5% for Cl- and SO42- (n ) 7), and ( 14% for NO3- (n ) 7).
Results and Discussion During the transition from the austral autumn to winter, the sea ice extent is low and large amounts of marine aerosols are transported from the surrounding oceans to glaciers by maritime winds from the Ross Ice Shelf (13). Samples were dated by using this seasonal association of high sodium (Na+) concentrations from this marine aerosol source. The glaciochemical determinations of mass accumulation in the snow pits were compared to annual physical measurements of snow accumulation on the Commonwealth Glacier since 1993. Annual data from the nearest stake to the snow pit (77° 33′ S, 16° 32′ E, 711.6 masl) indicate the mean accumulation has been 14.1 cm · yr-1 snow over a 10 year period at that location (14). In addition, each annual layer in the snow pit was compared against another snowpit in the accumulation zone of the Commonwealth Glacier that was collected during the 1999-2000 field season. This earlier snow pit, dated using
the same geochemical procedure, was determined to have a mean accumulation of 16 cm · yr-1 of snow for the period between 1988 and 2000. All these data are in good agreement with the mean accumulation rate calculated for the same time period in this study, 15 cm · yr-1 of snow. Using a combination of these methods, 17 annual layers from 1987 to 2004 were identified (Figure 2). Mercury was analyzed in snow samples at intervals of six centimeters in the top meter and ten centimeters in the bottom two meters (Figure 3). The data range from 1.6 to 200 pM. The data are log-normally distributed, and therefore it is appropriate to report the geometric mean of the data (G ) 9.0 pM, n ) 32). The data show two statistically significant peaks in mercury concentration from 0-6 cm and 110-130 cm (p ) 99.9%, using two-sided z-test). Post-Depositional Processes. It is thought that the primary mechanism for atmospheric elemental mercury removal in polar regions is by mercury depletion events (MDEs) during the polar sunrise (15-17). During MDEs, gaseous elemental mercury is oxidized to reactive gaseous mercury, is lost from the atmosphere, and results in large seasonal fluxes of mercury onto snow surfaces (15, 17, 18). It has been debated whether spring-time MDEs increase the mercury concentration in surficial snow and how long this signal may be preserved (15, 16, 19). Although snow may act as a sink during MDEs, mercury can be reemitted and act as a source to the atmosphere (20). It has been shown that mercury is highly mobile in snow, particularly in exposed snow (e.g., snow with no canopy cover) (21-23). Three potential post depositional process have been hypothesized (i) percolation within the snowpack, (ii) settling of aerosols, (iii) photoreduction of Hg(II) to Hg0 (22). It is thought that the primary mechanism for atmospheric elemental mercury removal in polar regions is by mercury depletion events (MDEs) during the polar sunrise (15-17). During MDEs, gaseous elemental mercury is oxidized to reactive gaseous mercury, is lost from the atmosphere, and results in large seasonal fluxes of mercury onto snow surfaces (15, 17, 18). VOL. 42, NO. 13, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4711
FIGURE 2. Sodium concentrations used to determine annual layers. The data were used in conjunction with physical measurements and sodium concentrations from a previously sampled snow pit on the Commonwealth Glacier (14).
FIGURE 3. Subannual to annual mercury concentration. Closed circles: Analyses performed at the Brooks Rand Laboratory. The snow in the accumulation zone of MCM glaciers is transport of Hg, but all other major ions as well, and there cold and dry and no snowmelt has been directly observed is no geochemical evidence that meltwater production has (7). Meltwater production only occurs in the ablation zones, redistributed major ions in the accumulation zone (Figure although thin ice lenses have been observed at depth in the 3). accumulation zone and are thought to be a result of buried Mercury is highly particle reactive and may be adsorbed melt-surface crusts or internal partial melt (24). In the ablation to aerosols. Settling of such aerosols would remove Hg from zone, mass is lost predominately by sublimation, accounting the uppermost snow layers, transporting particulate-bound for 40-80% (25). The 2001-2002 austral summer was notably Hg to lower layers. Such transport would be observed as an warm (26), and extensive melt occurred in the ablation zones increase in Hg concentration with depth (22). The strikingly of these glaciers. Percolation would result in not only the high mercury concentrations at 120 and 130 cm may be a 4712
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 13, 2008
TABLE 1. Calculated Source Contributions of Mercury to the Commonwealth Glacier flux (pM · yr-1)
source marine aerosols volcanic oceanic degassing aeolian inputs total calculated total measured
6 × 10 -8 4 × 10-4 0.5 0.01-0.05 0.49-0.53 0.53
result of particulate settling. There was no noted anomaly in the density of the snow layers from 120-130 cm, however an ice layer was observed from the depths of 144-160 cm (27). Other thinner ice layers were observed in the snow at depths of 50-53 cm and 230-234 cm. It is possible that the thicker ice layer has prevented transport of mercury to lower layers, but the thinner layers have allowed downward transport of aerosol-bound Hg. The high mercury concentration in the uppermost layer of snow may be the result of a strong wind event depositing abundant local dust. During September 2003, a meteorological station installed by MCM-LTER personnel on the Commonwealth Glacier recorded maximum wind speeds of 8.0 m · s-1, the highest speeds recorded in over a 10 year period. This intense katabatic wind event is corroborated by high concentrations of Ca2+, a geochemical indicator of dust, in the uppermost snow layers. It is possible that during this intense event, mercury was directly deposited onto the glacier and/or dust deposited on the snow surface contained a significant amount of mercury. Another possible explanation for the high concentration of mercury in the uppermost layer may be due to the relative age of the snow and exposure to sunlight. Assuming a pseudozero-order kinetics and reduction from Hg(II) to Hg0 only occurs in the presence of unobstructed sunlight, Lalonde and co-workers (21) observed a reduction rate of 1.6 pM/h in in situ incubations of solid snow, whereas Poulain and co-workers (23) observed a Hg0 production rate of 0.27 pM/ h. It has been suggested that the production of Hg0 is limited by the amount of reductants rather than photoreducible Hg(II) (22). The aforementioned production rate of Hg0, r, can be used in the equation: Co ) Ct - rt
(1)
where C0 is the background concentration, Ct is the concentration at a specified time, t, and r is the rate of Hg0 production. Assuming the same factors influencing photoreduction are ubiquitous, all Hg0 is removed upon production, and the site received constant irradiation, the time in which it will take this fresh surface layer to revolatilize mercury to background concentrations can be estimated. Using a background concentration of 7.0 pM (the mean mercury concentration excluding outliers at 6 cm, 120 cm, and 130 cm), the uppermost snow layer (49 pM) will reach background concentrations within 26 to 156 h. If the snow layer is rapidly buried, it is likely that the mercury concentration will remain high. Thus, some variability in mercury concentration in the snow pit may be a result of different burial rates and snowfall events. Sources of Mercury. Unlike glaciers further inland, discerning primary sources of chemicals to glaciers in the MCM is difficult (28). This is due in part to the local ice-free area, which can contribute salt and/or dust, as well as the proximity to the Ross Sea and Mt. Erebus. During the months of January and February into the austral autumn, McMurdo Sound can be largely ice-free, and this close proximity of open water may greatly affect the geochemistry of glacier
snow and ice in the Dry Valleys region (28). Comparisons of MCM glaciers to ice from Taylor Dome and South Pole, areas completely surrounded by ice, indicate significantly higher geochemical inputs to glaciers in the MCM (28). Identifying mercury sources in the Dry Valleys may put mercury concentrations found on the Commonwealth Glacier into a global context. Commonwealth Glacier snow has higher mercury concentrations than snow from the Greenland ice sheet (mean ) 3.0 pM) (29). The lower concentrations in the Greenland snow may be the result of the inland location and therefore insignificant marine and aeolian dust contributions of mercury. Commonwealth Glacier snow typically has lower concentrations of mercury than coastal Arctic locations. Mercury in snow from Svalbard ranges from 6.0 to 149 pM (16, 30). Snow from coastal locations in the Canadian Arctic ranges from 9.0 to 796.0 pM (15). The higher concentration in these northern hemisphere locations may be the result of regional marine sources of mercury, concentration magnification during mercury depletion events, or hemispheric anthropogenic sources. Sources of mercury to the Commonwealth Glacier were assessed using the techniques of Vandal and others (31) by comparing well understood molar ratios of mercury to ions associated with various geochemical sources. Although it is clear that mercury is removed from the snow by postdepositional processes, this technique can be applied to assess the relative contributions of various sources. Generically, this can be determined by the equation: C × R × A-1 ) F
(2)
where, C is the mean concentration of a major ion in the Commonwealth Glacier snow, R is a well established mol to mol ratio of mercury to a major ion of an associated mercury source, A is the number of years represented in the snow profile, and F is the expected mean annual flux of mercury from the atmosphere to the snow. Marine aerosols appear to be an insignificant source of mercury to the Commonwealth Glacier. The Hg/Na+ ratio of seawater has been established as 1.1 × 10-13 (M:M) (32). Assuming all sodium (Na+) is from seawater aerosols (G ) 9.79 µM), the mercury contribution from McMurdo Sound aerosols is determined by: CNa+ × RHg⁄Na+ × A-1 ) FHg Marine
(3)
Aerosols
resulting in a mercury flux of 6 × 10-5 fM · y-1. Globally, approximately 13% of nonseasalt sulfate (NSS SO42-) is a result of volcanic emissions (33). Using the geometric mean of NSS SO in the snow analyzed (G ) 1.84 µM), ratio of Hg/SO from Mt. Erebus (2.9 × 10-8 M:M) (34) the expected annual atmospheric flux to the Commonwealth Glacier can be determined by slightly modifying the aforementioned equation: 0.13CNSS SO42- × R Erebus × A-1 ) FHg Erebus
(4)
This application yields a mean annual flux of 0.4 fM · y-1. A more substantial contribution from McMurdo Sound is likely due to gaseous evasion of mercury. It has been suggested that dimethylmercury (DMHg) emitted from ocean water could decompose and be deposited on Antarctic glaciers (35). As dimethylsulfide (DMS) is biologically produced and evaded from the ocean, it is oxidized in the atmosphere and deposited as NSS SO Approximately 80% of the NSS SO in Antarctic snow and ice is due to this process (31, 36). Air measurements from the Equatorial Pacific yield a Hg/SO42- ratio of 5.6 × 10-6 (M:M) (32) thus, the equation 0.80C NSS SO42- × R Pacific Ocean Air × A-1 ) FHg Oceanic Evasion (5) VOL. 42, NO. 13, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4713
FIGURE 4. Mercury concentration and marine aerosol calcium concentration by depth.
TABLE 2. Mercury Concentrations in Snow at Other Antarctic locations location
latitude, longitude
D40 (31) D47 (31) D80 (31) South Pole (31) Windless Bight (44) Victoria Land (45) Coats Land (30) Commonwealth Glacier
66° 42′ S, 139° 62° 23′ S, 154° 70° S, 167° 40′ 90° S 77° 51′ S, 167° 77° 32′ S, 159° not indicated 77° 33′ S, 163°
inland distance (km)
elevation (masl)
mean concentration (pM)
33 103 433 1274 30 100 not indicated 10
848 1500 2525 2880 0 not indicated not indicated 678
2.5 1.0 0.65 1.2 16 3.1 0.1-80 9.0
57′E 03′ E E 40′ E 50′ E 01′ E
TABLE 3. Total and Dissolved Mercury Concentration in Taylor Valley streams
Canada (35) Von Guerard (35) Canada (6) Aiken (6) Lost Seala (6) McKnighta (6) a
total Hg (pM)
dissolved Hg (pM)
dissolved Hg (%)
4.5 2.7 1.3 3.0 9.5 8.6
0.7 0.8 1.7 1.5
52 28 18 17
Streams that drain the Commonwealth Glacier.
shows that approximately 0.5 pM · y-1 of mercury can be attributed to oceanic evasion. The K+/Ca2+ ratio of snow at the South Pole and Taylor Dome ranges from 0.47 to 0.60 (M/M) and has been hypothesized to be a result of background dust (37, 38). The mean molar ratio of K+ to excess Ca2+ (e.g., Ca2+ not associated with marine aerosols) is 0.22. Such a value indicates that the dust in the Commonwealth Glacier is not background dust but rather maybe a unique signature of dust from the surrounding glacier-free areas (i.e., the dry valleys). This ratio 4714
may be a result of a combination of ferromagensium rocks and the soils derived from them, which typically have K+/ Ca2+ ratios less than 0.44, and the widely distributed salts in the McMurdo Dry Valleys such as calcite and gypsum (39, 40). The input of such local dust is transported onto glaciers by aeolian processes, especially katabatic wind events during the winter (28). By using the geometric mean concentrations of excess Ca2+ and K+ (2.2 and 0.48 µM, respectively) and the ratios of mercury to these elements in the mean upper continental crust (3.8 × 10-7 for both Hg/Ca2+ and Hg/K+) (41) in the equations
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 13, 2008
C Excess × R UpperContinental Crust × A-1 ) F Aeolian Material
(6)
C Excess × R UpperContinental Crust × A-1 ) F Aeolian Material
(7)
and
yields 0.05 pM · y-1 and 0.01 pM · y-1, respectively. The sum of these source contributions ranges from 0.49 to 0.53 pM · y-1 depending on the method to quantify the aeolian source (i.e., using either excess K+ or excess Ca2+ as the geochemical indicator for dust). This range is in good agreement with the geometric mean annual flux determined by dividing the geometric mean mercury concentration by
the total number of years identified in the snow pit (0.53 pM · y-1) (Table 1). This agreement can be interpreted that snow acts as both a local source and sink of mercury. Once mercury is deposited in snowfall, post depositional processes, e.g. revolitization, partial melting, and the settling of particulates, does not significantly affect the net flux of mercury to a snowpack. Photoreduced mercury is probably emitted from recently deposited snow only to be incorporated into precipitation and redeposited during subsequent snow events. Mercury transport due to partial melting does not appear to be significant in the accumulation basin of the Commonwealth Glacier due to temperature insufficient for snowmelt. Particulate settling may remove mercury from discrete annual layers, but over decadal intervals, mercury is not permanently lost from the system. Mercury in the Antarctic Environment. According to these calculations, evasion from the ocean is the dominant source of mercury to the Commonwealth Glacier. Analysis of the Dome C ice core noted a covariation of Hg with SO42concentrations over the past 30 000 years implying an oceanic degassing signal as far inland as 1070 km (5). Despite such theoretical calculations, no such pattern emerges from the Hg and excess (NSS) SO42- concentrations in Commonwealth Glacier snow (not shown). This lack of relationship suggests either that mercury associated with ocean degassing is not conservative in snow, or the ratio of the flux is not constant with time at least on interannual timescales. A more substantial correlation emerges from the interannual variability or mercury concentrations and calcium associated with marine aerosols (Figure 4). In many instances, the concentrations of Hg and Ca from marine aerosols rise and fall in phase, although the magnitudes of these changes are dissimilar. This observation is counter to predications based on well constrained ratios of mercury associated with various sources. Such a pattern indicates that attributing mercury sources to glacial snow and ice using the approach of Vandal and others (5) may not be appropriate over seasonal and even annual time scales. It appears that mercury is either being directly deposited from marine aerosols or gaseous mercury evaded from the ocean is being scavenged by these aerosols. It is hypothesized that coastal snow is a more important source of halogens than frost flowers (42). Bromide is a primary reactant in mercury deposition chemistry and is thought to be scavenged by marine aerosols throughout the polar winter (43). A dominant marine source is further corroborated by comparison to mercury concentrations in snow at other Antarctic locations (Table 2). Generally, mercury concentrations are higher at low elevations closer to the coast. Such low elevation, coastal glaciers are predominately influenced by marine chemistry (9, 28). Furthermore, the proximity to a local dust source may play a significant role in the range of mercury concentrations in snow. Recent work on the Tibetan Plateau indicates that significant amounts of mercury are transported to glacial surfaces during dust storms (46). The three locations with the highest concentrations of mercury are located within 30 km of exposed rock (44, 45). Perhaps dust from the Shackleton Range and Theron Mountains may contribute to the mercury concentrations in the snow (30). The two main sources of solutes to the McMurdo Dry Valley streams are from chemical weathering and/or dissolution of marine aerosols (47). Vandal and others (35) suggest that a crustal component may account for 35-75% of the total mercury content in dry valley streams. This crustal contribution may be evidenced by the relatively large amounts of particulate mercury as chemical weathering in the stream channel is probably a minor contributor of dissolved mercury (6). This crustal mercury may enter streams either as sediment within the channel or as aeolian material
deposited on the glacier surfaces. Crustal derived mercury may account for nearly 10% of the total mercury flux to the Commonwealth Glacier (Table 1.). The majority of the dissolved mercury load in the streams is probably from the dissolution of marine salts and/or from direct glacier melt as the average mercury concentration in Commonwealth Glacier snow (9.0 pM) is similar to the total mercury concentration of streams that drain the Commonwealth Glacier: Lost Seal Stream and McKnight Creek (Table 3). The mercury concentration of the glacial snow is over five times higher than the concentration of dissolved mercury, so it is not unreasonable to expect that glacial melt may be the primary source of dissolved mercury.
Acknowledgments This work was supported by NSF grants OPP-0096250 and ANT-0423595. Dr. K. Kreutz and other members of the University of Maine Climate Research Group generously assisted with their field and scientific expertise. K. Welch and T. Fitzgibbon from the Byrd Polar Research Center assisted in sample collection and analysis. We also appreciate two anonymous reviewers who greatly improved this paper with their constructive critiques.
Literature Cited (1) Boutron, C. F.; Bolshov, M. A.; Candelone, J. P.; Delmas, R. J.; Goerlach, U. Decrease in anthropogenic lead, cadmium and zinc in Greenland snows since the late 1960s. Nature 1991, 353, 153–156. (2) Mason, R. P.; Fitzgerald, W. F.; Morel, F. The biogeochemical cycling of elemental mercury: Anthropogenic influences. Geochim. Cosmochim. Acta 1994, 58, 3191–3198. (3) Barbante, C.; Veysseyre, A.; Ferrari, C.; Van de Velde, K.; Morel, C.; Capodaglio, G.; Cescon, P.; Scarponi, G.; Boutron, C. Greenland Snow Evidence of Large Scale Atmospheric Contamination for Platinum, Palladium, and Rhodium. Environ. Sci. Technol. 2001, 35, 835–839. (4) Fitzgerald, W. F.; Engstrom, D. R.; Mason, R. P.; Nater, E. A. The case for atmospheric mercury contamination in remote areas. Environ. Sci. Technol. 1998, 32, 1–7. (5) Vandal, G. M.; Fitzgerald, W. F.; Boutron, C. F.; Candelone, J. P. Variations in mercury deposition to Antarctica over the past 34,000 years. Nature 1993, 362, 621–623. (6) Lyons, W.; Welch, K.; Bonzongo, J. Mercury in aquatic systems in Antarctica. Geophys. Res. Lett. 1999, 26, 2235–2238. (7) Fountain, A.; Lyons, W.; Burkins, M.; Dana, G.; Doran, P.; Lewis, K.; McKnight, D.; Moorhead, D.; Parsons, A.; Priscu, J.; Wall, D.; Warton, R.; Virginia, R. Physical controls on the Taylor Valley ecosystem, Antarctica. Bioscience 1999, 49, 961–971. (8) Fountain, A.; Lewis, K.; Doran, P. Spatial climatic variation and its control on glacier equlibrium line altitude in Taylor Valley, Antarctica. Global Planet. Change 1999a, 22, 1–10. (9) Witherow, R. A.; Lyons, W. B.; Bertler, N. A. N.; Welch, K. A.; Mayewski, P. A.; Sneed, S. B.; Nylen, T. H.; Handley, M. J.; Fountain, A. G. The aeolian flux of calcium, chloride and nitrate to the McMurdo Dry Valleys landscape: evidence from snow pit analysis. Antarct. Sci. 2006, 18, 497–505. (10) Fitzgerald, W. F. In Clean Hands: Clair Patterson’s crusade against environmental lead contamination; Davidson, C. I., Ed.; Nova Science Publishers: Commack, NY, 1999; pp 119-137. (11) Ferrari, C. P.; Moreau, A. L.; Boutron, C. F. Clean conditions for the determination of ultra-low levels of mercury in ice and snow samples. Fresenius J. Anal. Chem. 2000, 366, 433–437. (12) Welch, K. A.; Lyons, W. B.; Graham, E.; Neumann, K.; Thomas, J.; Mikesell, D. Determination of major element chemistry in terrestrial waters from Antarctica by ion chromatography. J. Ion Chromatogr., A 1996, 739, 257–263. (13) Legrand, M.; Mayewski, P. Glaciochemistry of polar ice cores: A review. Rev. Geophys. 1997, 35, 219–243. (14) Witherow, R. A. Mercury concentrations in snow and the modern mercury flux to Taylor Valley, Antarctica, M.S., Thesis, The Ohio State University: Columbus, OH, 2005. (15) Lu, J.; Schroeder, W. H.; Barrie, L.; Steffen, A.; Welch, H.; Martin, K.; Lockhart, L.; Hunt, R.; Boila, G.; Richter, A. Magnification of atmospheric mercury deposition to polar regions in springtime: The link to tropospheric ozone depletion chemistry. Geophys. Res. Lett. 2001, 28, 3219–3222. VOL. 42, NO. 13, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4715
(16) Berg, T.; Sekkesaeter, S.; Steinnes, E.; Valdal, A. K.; Wibetoe, G. Springtime depletion of mercury in the European Arctic as observed in Svalbard. Sci. Total Environ. 2003, 304, 43–51. (17) Temme, C.; Einax, J. W.; Ebinghaus, R.; Schroeder, W. H. Measurements of atmospheric mercury species at a coastal site in the Antarctic and over the south Atlantic Ocean during polar summer. Environ. Sci. Technol. 2003, 37, 22–31. (18) Ebinghaus, R.; Kock, H. H.; Temme, C.; Einax, A. G.; Lowe, A. G.; Richter, A.; Burrows, J. P.; Schroeder, W. H. Antarctic springtime depletion of atmospheric mercury. Environ. Sci. Technol. 2002, 36, 1238–1244. (19) Ferrari, C. P.; Dommergue, A.; Boutron, C. P.; Jitaru, P.; Adams, F. C. Profiles of mercury in the snow pack at Station Nord, Greenland shortly after polar sunrise Geophys. Res. Lett. 2004, 31, DOI: 10.1029/2003GL018961. (20) Dommergue, A.; Ferrari, C. F.; Poissant, L.; Boutron, C. F. Chemical and photochemical processes at the origin of the diurnal cycle of gaseous mercury within the snow-pack at Kuujjuarapik, Whapmagoostui, Quebec. Environ. Sci. Technol. 2003, 37. doi: 10.1021/es026242b. (21) Lalonde, J. D.; Amyot, M.; Doyon, M.; Auclair, J. Photo-induced Hg(II) reduction in snow from the remote and temperate Experimental Lakes Area (Ontario, Canada). J. Geophys. Res. Atmos. 2003, 108, DOI: 10.1029/2001 JD 001534. (22) Lalonde, J. D.; Poulain, A. J.; Amyot, M. The role of mercury redox reactions in snow on snow-to-air mercury transfer. Environ. Sci. Technol. 2002, 36, 174–178. (23) Poulain, A. J.; Orihel, D. M.; Amyot, M.; Patterson, M. J.; Hintelmann, H.; Southworth, G. R. Relationship to aquatic between the loading rate of inorganic mercury ecosystems and dissolved gaseous mercury production and evasion. Chemosphere 2006, 65, 2199–2207. (24) Fountain, A.; Nylen, T. H.; MacClune, K. L.; Dana, G. Glacier mass balances (1993-2001), Taylor Valley, McMurdo Dry Valleys, Antarctica. J. Glaciol. 2006, 52, 451–462. (25) Lewis, K.; Fountain, A.; Dana, G. How important is terminus cliff melt?: a study of the Canada Glacier terminus, Taylor Valley, Antarctica. Global Planet. Change 1999, 22, 105–115. (26) Foreman, C. M.; Wolf, C. F.; Priscu, J. C. Impact of episodic warming events on the physical, chemical and biological relationships of lakes in the McMurdo Dry Valleys, Antarctica. Aquat. Geochem. 2004, 10, 239–268. (27) Williamson, B. R. Sources and deposition processes linking atmospheric chemistry and firn records from four glacier accumulation zones in the McMurdo Dry Valleys, Antarctica. M.S. Thesis, University of Maine, Orono, ME, 2006. (28) Lyons, W.; Welch, K.; Fountain, A.; Dana, G.; Vaughn, B.; McKnight, D. Surface glaciochemistry of Taylor Valley, southern Victoria Land, Antarctica and its relationship to stream chemistry. Hydrol. Process. 2003, 17, 115–130. (29) Boutron, C.; Vandal, G. M.; Fitzgerald, W. F.; Ferrari, C. A forty year record of mercury in central Greenland snow. Geophys. Res. Lett. 1998, 25, 3315–3318. (30) Planchon, F. A. M.; Gabrelli, P.; Gauchard, P. A.; Dommergue, A.; Barbante, C.; Cairns, W. R. L.; Cozzi, G.; Nagorski, S. A.; Ferrari, C.; Boutron, C.; Caposaglio, G.; Cescon, P.; Varga, A.; Wolff, E. W. Direct determination of mercury at the sub-
4716
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 13, 2008
(31) (32) (33) (34) (35)
(36) (37)
(38)
(39)
(40) (41) (42)
(43)
(44) (45) (46) (47)
picogram per gram level in polar snow and ice by ICP-SFMS. J. Anal. At. Spectrosc. 2004, 19, 823–830. VandalG. M. FitzgeraldW. F. BoutronC. F. CandeloneJ. P. In Ice Core Studies of Global Biogeochemical Cycles; Delmas, R. J., Ed.; Springer-Verlag: Berlin, Germany, 1995; 401-415. Kim, J. P.; Fitzgerald, W. F. Sea-air partitioning of mercury in the equatorial Pacific Ocean. Science 1986, 231, 1131–1133. Boutron, C. F.; Patterson, C. C. Lead concentration changes in Antarctic ice during the Wisconsin Holocene transition. Nature 1986, 323, 222–225. Nriagu, J. O.; Becker, C. Volcanic emissions of mercury to the atmosphere: global and regional inventories. Sci. Total Environ. 2003, 304, 3–12. Vandal, G. M.; Mason, R. P.; McKnight, D. M.; Fitzgerald, W. F. Mercury speciation and distribution in a polar desert lake (Lake Hoare, Antarctica) and two glacial meltwater streams. Sci. Total Environ. 1998, 213, 229–227. Prospero, J. M.; Savois, D L.; Saltzman, E. S.; R, L. Impact of oceanic sources of biogenic sulfur on slufate aeroson concentrations at Mawson, Antarctica. Nature 1991, 350, 221–223. Hinkley, T. K.; Matsumoto, A. Atmospheric regime of dust and salt through 75,000 years of Taylor Dome ice core: Refinement by measurement of major, minor, and trace metal suites. J. Geophys. Res. 2001, 106, 18487–18493. Whitlow, S.; Mayewski, P. A.; Dibb, J. E. A comparison of major chemical species, seasonal concentration,a nd accumulation at the South Pole and Summit, Greenland. Atmos. Environ. 1992, 26, 2045–2054. Hinkley, T. K.; Pertsiger, F.; Zavjalova, L. The modern atmospheric background dust load: Recognition in Central Asian snowpack, and compositional constraints. Geophys. Res. Lett. 1997, 24, 1607–1610. Keys, J.; Williams, K. Origin of crystalline, cold desert salts in the McMurdo region, Antarctica. Geochim. Cosmochim. Acta 1981, 45, 2299–2309. Wedepohl, K. M. The composition of the continental crust. Geochim. Cosmochim. Acta 1995, 59, 1217–1232. Simpson, W. R.; Alvarez-Aviles, L.; Douglas, T. A.; Sturm, M.; Domine, F. Halogens in the coastal snow pack near Barrow, Alaska: Evidence for active bromine air-snow chemistry during springtime. Geophys. Res. Lett. 2005, 32 Art. No. L04811. McConnell, J.; Henderson, G.; Barrie, L.; Bottenheim, J.; Niki, H.; Templeton, E. Photochemical bromine production implicated in Arctis boundary-layer ozone depletion. Nature 1992, 355, 150–152. Dick, A. L.; Sheppard, D. S.; Patterson, J. E. Mercury content of Antarctic surface snowsinitial results. Atmos. Environ. 1990, 24, 973–978. Sheppard, D. S.; Patterson, J. E.; McAdam, M. K. Mercury conternt of Antarctic ice and snowsfurther results. Atmos. Environ. A 1991, 25, 1657–1660. Loewen, M.; Kang, S.; Armstrong, D.; Zhang, Q.; Tomy, G.; Wang, F. Atmospheric transport of mercury to the Tibetan Plateau. Environ. Sci. Technol. 2007, 41, 7632–7638. Green, W. J.; Gardner, T. J.; Ferdelman, T. G.; Angle, M.; Varner, L.; Nixson, P. Geochemical processes in the Lake Fryxell basin (Victoria Land Antarctica). Hydrobiologia 1989, 172, 129–148.
ES800022G
