Rapid Release of Mercury from Intertidal Sediments Exposed to Solar

Environ. Sci. Technol. 2004, 38, 3901-3907

Rapid Release of Mercury from Intertidal Sediments Exposed to Solar Radiation: A Field Experiment JOA ˜ O C A N AÄ R I O * A N D C A R L O S V A L E National Institute for Agronomy and Fisheries Research, IPIMAR, Av. Brası´lia, 1449-006 Lisboa, Portugal

There is increasing evidence of the primary importance of photochemical reactions and transfer of gaseous mercury to the atmosphere. Although mercury in aquatic sediments is efficiently retained, resuspension and bioturbation in intertidal sediments may expose temporarily anoxic sediments to solar radiation. Field experiments were performed to investigate these processes. Anoxic sediments from two areas in the Tagus estuary with different degrees of Hg contamination (experiments I and II) were homogenized and distributed into two sets of 36 uncovered Petri dishes. The samples were placed on the intertidal sediments and exposed to direct solar radiation and kept under dark (control) for 6-8 h. The decrease rates of acid volatile sulfides (abrupt in the first 3 h) and of pyrite (linear) were the same in sediments under solar radiation and dark. The total Hg concentrations were relatively constant in sediments kept in dark, but decreased from 17.6 to 7.65 and 3.45 to 1.35 nmol g-1 in experiments I and II, respectively. In those exposed to solar radiation during the period of higher UV intensity. Similar evolutions were found in nonreactive Hg in pore waters (3.00-2.59 and 0.725-0.105 nM). On the contrary, reactive Hg was higher in pore waters of the sediments exposed to solar radiation and increased with time, from 424 to 845 pM and 53 to 193 pM. These results indicate that most mercury released in pore waters was photochemically reduced in a short period of time and escaped rapidly to the atmosphere. Episodes of bottom resuspension and bioturbation in the intertidal sediments enhance the transfer of gaseous mercury to the atmosphere.

Introduction Mercury entered the environmental history because it was the first metal with a direct connection between concentrations in water, bioaccumulation up the food chain, and a serious impact on human population (1). Unlike most others heavy metals, the global cycle of natural and anthropogenic mercury is dominated by atmospheric transport (2). Metallic mercury has the highest vapor pressure of all metals (3). Significant quantities are released by volcanic activity (4), vegetation (5), volatilization from land (6, 7) and ocean surfaces (8, 9), smelting of minerals, and burning of fossil fuels (10, 11). Among the volatile forms of mercury, Hg0 is the major constituent of the dissolved gaseous mercury in open ocean waters (8), and several works have shown its transfer to the * Corresponding author phone: +351.21.3027000; +351.21.3015948; e-mail: [email protected]. 10.1021/es035429f CCC: $27.50 Published on Web 06/10/2004

 2004 American Chemical Society

Fax:

atmosphere (9-11). In the past decade there was increasing evidence of the primary importance of photochemical reactions in the mercury reduction processes (9, 12-14). The ability of many Hg(II) compounds to absorb part of the solar radiation (15) and the importance of Fe(III) in the reduction processes (16) are mechanisms proposed for the reduction of Hg(II). Even solid surfaces such as HgS can also be photochemically reduced (17). Soils naturally enriched in Hg have been considered important sources contributing to the atmospheric Hg load (18-21). Recent field investigations indicate that these natural sources may be comparable to the anthropogenic sources in their impacts on regional and global atmospheric Hg pools (22). Anthropogenic Hg is retained in sediments nearby the sources, recording the historical evolution of the anthropogenic inputs (23, 24). Mercury in sediments is mainly precipitated as sulfides, associated with iron sulfides, or incorporated in organic matter (25-27). In many coastal zones with high tidal amplitudes, extensive areas of sediments are periodically inundated and exposed to the atmosphere. The topmost layer, which in many cases coincides with the oxic layer of the sediments, is frequently resuspended by the waves and tidal currents (28) in macro and mesotidal systems and mixed with suboxic and anoxic layers by macrofauna activity (29). When resuspension or bioturbation is stronger, anoxic sediments from the intertidal zone are exposed to the atmosphere during the ebb tide. Moreover, dredging operations in harbors and navigational channels cause exposure of anoxic sediments to the atmosphere. This paper reports the effect of exposing anoxic sediments to the atmosphere, under direct solar radiation and dark conditions, on the alterations of mercury concentrations in solids and pore waters. The field experiments were performed during the atmospheric exposure of intertidal sediments in two areas of the Tagus estuary with contrasting degree of contamination.

Experimental Section Description of the Field Experiment. The field experiments consisted of exposing anoxic sediments to the atmosphere during 6-8 h, under direct solar radiation and dark conditions. The experiments were performed in the two intertidal areas in the Tagus estuary (A1 and A2), during the period between two high tides when sediments were air-exposed, in May and July 2003, respectively (Figure 1). The area A1 is in the vicinity of an anthropogenic Hg source, and sediments are heavily contaminated (30), while the area A2 is in the Tagus Natural Park with sediments containing low Hg contamination (31). Approximately 5 kg of suboxic to anoxic sediments (between 5 and 10 cm depth) were collected in each area, transferred into a decontaminated recipient, homogenized, and distributed into 36 glass uncovered Petri dishes (diameter 9.5 cm). Transferred sediment covered the entire surface of the dish, and the height was such (between 0.5 and 1 cm) as to facilitate the exposure of the sediment to solar radiation. At each site 18 dishes were exposed to direct solar radiation and another 18 were covered with a black box in order to maintain the darkness inside. Every hour two dishes exposed to solar radiation plus two kept in the dark were removed and immediately frozen. The sediment temperature and the intensity of UV radiation (UV pronde, Macam, model UV113E) were measured at the same rhythm. The experiments ran between (GMT) 10:00 and 16:00 (A1, experiment I) and between 9:00 and 17:00 (A2, experiment II) under similar conditions of clear sky and sunshine. The intensity of UV radiation increased pronouncedly by 11:00 VOL. 38, NO. 14, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3901

FIGURE 1. Map of the Tagus estuary with the site location of the field experiments (A1 and A2). (654 times in experiment I and 440 in experiment II) and then started decreasing by 15:00 (Figure 2). In both experiments sediment temperature followed the solar semidiurnal cycle, reaching a maximum of 39 °C in experiment I and 27 °C in experiment II around 14:00. As a result of these conditions, the water content of sediments in the Petri dish decreased gradually during the experiments. The evaporation was more accentuated in sediments exposed to solar radiation (e.g., water content decreased from 73% to 13% in experiment I). The temperature of light exposed sediments never exceeded the values registered in dark sediments by 2 °C.

Analytical Methods. In the laboratory, samples were centrifuged at 6000 rpm for 20 min (4 °C) and the supernatant filtered through 0.45 µm polycarbonate membranes (MSI, Micron Separations Inc.) in a N2 chamber. This procedure was used in previous works, and errors associated with iron and sulfur determinations were less than 6% (32). The obtained pore waters were acidified to pH < 2 with HNO3 (Merck, mercury-free) and used for dissolved mercury determinations. The solid fraction was oven dried at 40 °C, disaggregated, homogenized, and stored in polyethylene bottles for future analysis. Fresh portions of each sample were frozen before centrifugation for acid volatile sulfide (AVS) determinations. Duplicates of the fresh sediment were used for determination of the water content by weight loss at 105 °C. Total Analysis of Solid Sediment. Total determinations of Al, Si, Ca, Mg, Fe, and Mn were performed by mineralization of the sediment samples with a mixture of acids (HF, HNO3, and HCl) according to the method described by Rantala and Loring (33). Metal concentrations were obtained by flameAAS (Perkin-Elmer Aanalist 100) using direct aspiration into N2O-acetylene flame (Al, Si, Ca, and Mg) or air-acetylene flame (Fe and Mn). For total Hg determinations, sediment was digested overnight at room temperature with 4 M HNO3 in borosilicate glass Erlenmeyer flasks and then heated (60-70 °C) for 2 h in a sand bath (21). Mercury was determined by cold vapor AAS (Perkin-Elmer FIMS-FIAS100) using SnCl2‚2H2O as reduction agent. The precision expressed as relative standard deviation was less then 4% for all metal investigate (p < 0.05). International certified standards were used to ensure the accuracy of our procedure, and precision was determined by analyzing replicate samples. Extraction with Hydroxylamine Chloride. Each sediment sample was stirred for 6 h with an NH2OH‚HCl (0.04 M) solution in CH3COOH (25%) according to the method described by Chester and Hughes (34). The extraction was performed in leak-proof tubes sealed with Parafilm to minimize possible losses of Hg0. The supernatant solution was removed by centrifugation at 3000 rpm for 10 min and filtered through 0.45 µm membranes. Iron and Mn in the overlying solution were determined by flame-AAS and Hg by CV-AAS as described above. Detection limits for Fe, Mn, and Hg were 0.80 and 0.60 µmol g-1 and 0.02 nmol g-1, and precision errors were 7, 6, and 3% (p < 0.05), respectively.

FIGURE 2. Evolutions of UV radiation intensity (mW cm-2), sediment temperature (°C), and water content (%) during experiments I and II. 3902

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 14, 2004

TABLE 1. Mean and Standard Deviations of Water Content (%); Concentrations of Al (mmol g-1), AVS, and Pyrite (µmol g-1 S); and Si/Al, Fe/Al and Mn(×10-4)/Al Molar Ratios of Sediments from Experiments I and II

experiment I (n)6) experiment II (n)6)

water content (%)

Al (mmol g-1)

AVS (µmol g-1)

pyrite (µmol g-1)

Si/Al

Fe/Al

Mn 10-4/Al

75 ( 2.2 84 ( 1.3

2.90 ( 0.02 2.30 ( 0.04

663 ( 12 2.41 ( 0.34

236 ( 3.34 0.91 ( 0.01

2.15 ( 0.07 2.59 ( 0.06

2.06 ( 0.04 1.89 ( 0.03

2.91 ( 0.05 2.30 ( 0.06

TABLE 2. Mean Concentrations and Standard Deviations of Fe and Mn (µmol g-1) Simultaneously Extracted with a Hydroxylamine Solution in Experiments I and II under Light (solar radiation) and Dark Conditions Fe (µmol g-1)

experiment I (n)6) experiment II (n)6)

Mn (µmol g-1)

solar radiation

dark

solar radiation

dark

384 ( 30.4 67.1 ( 13.3

412 ( 32.6 81.8 ( 11.6

5.64 ( 0.16 1.32 ( 0.18

5.41 ( 0.17 1.44 ( 0.19

Extraction with 1 M HCl. Acid volatile sulfides, mainly amorphous iron sulfides and poorly crystallized iron oxides were extracted with 1 M HCl (35, 36). Sulfide was trapped in a deaerated NaOH solution and analyzed by differential pulse polarography (DPP) using a Metrohm apparatus equipped with a 693 VA processor and a 694 VA stand. Fe, Mn, and Hg were determined in the extracted solution by flame-AAS (Fe and Mn) and CV-AAS (Hg) as described above. Recovery of standard sulfide solutions was 97% (32). Detection limits for sulfide, Fe, Mn, and Hg were 0.01, 0.14, and 0.36 µmol g-1 and 0.04 nmol g-1 and precision errors were 5, 2, 3, and 6% (p < 0.05), respectively. Determination of Pyrite and Elemental Sulfur. Pyrite was determined in a 100 mg sediment sample using the chromium reduced sulfur (CRS) method described by Canfield (37). This method has proven to be specific for inorganic sulfur (AVS + FeS2 + S0) with an accuracy that is not affected by the presence of organic sulfur. Before analysis, S0 was extracted from the sample by 16 h of stirring with 20 mL of acetone followed by centrifugation (3000 rpm/10 min) and filtration through 0.45-µm membranes. The residue was then placed in the reaction vessel with 10 mL of 1 M HCl and purged with N2 for 20 min to release AVS. Finally, the CRS method was used in the last residue to analyze the pyrite content. Elemental sulfur was determined using the same CRS method in the acetone extracts. The measurements of the released H2S were made by DPP. Dissolved Mercury in Pore Waters. Reactive dissolved mercury was measured directly from the filtered acidified solutions by cold vapor atomic fluorescence spectroscopy (CV-AFS) using a cold vapor generator (PSA, model 10.003) associated with a fluorescence detector (PSA, Merlin 10.023). This mercury fraction includes forms of inorganic complexes and weakly bound organic complexes easily reducible by SnCl2 (38). Total dissolved mercury was also measured in the same equipment after a UV oxidation step with a 1000 W UV lamp following the method described by Mucci et al. (39). Nonreactive dissolved mercury was calculated by the difference between total and reactive mercury. The detection limits and precision errors for Hgdiss were 0.01 nM and 4.0% (p < 0.05), respectively. Statistical Analysis. Statistical tests were performed using the computer software Statistic. The normality of all data was assessed by a Kolmogorov-Smirnoff test. Mean concentrations of the chemical parameters used to characterize the sediment samples were compared with ANOVA test (oneway). For all other analysis, nonparametric tests MannWhitney and Krustal-Wallis were performed.

Results Sediment Characteristics. Visual inspection of the sediments revealed that material used in the two experiments consisted of fine material with an absence of macrofauna. The mean values of Al concentrations and of Si/Al, Fe/Al, and Mn/Al ratios showed no significant differences (p < 0.05) between the two sediments (Table 1). Low Si/Al ratios confirm that sediments consist of fine particles. Sediments used in experiment I were extremely rich in sulfur, the maximum concentrations of AVS (663 µmol g-1) and pyrite (236 µmol g-1) being 2 orders of magnitude higher than those in experiment II. Evolution of the Parameters during the Experiments. Extractable Fe and Mn. The concentrations of Fe and Mn extracted from the sediments with the hydroxylamine solution did not vary significantly (p < 0.05) during the experiments, although slight differences were observed between solar radiation and dark conditions (Table 2). Acid Volatile Sulfide, Pyrite, and Elemental S. The AVS and pyrite concentrations in the sediments are presented in Figure 3. Despite the differences in sulfur content, the timecourse evolutions of AVS were practically the same in the sediments of the two experiments. The AVS decreased abruptly in the first 3 h, the evolutions being independent of the light conditions and described by polynomial curves [AVS] ) 103(18t2 - 21t + 6) (r2 ) 0.93) and [AVS] ) 45t2 55t + 17 (r2 ) 0.88) in experiments I and II, respectively. Pyrite concentration decreased linearly with time (r2 better than 0.91) in both experiments, although values were lower in solar-radiation-exposed sediments. The slope was more accentuated in experiment I (11.2 µmol g-1 h-1) than in experiment II (0.06 µmol g-1 h-1). Elemental sulfur showed no differences between light and dark conditions, both in experiment I (34.7-36.8 µmol S g-1) and in experiment II (0.38-0.41 µmol S g-1). Total Hg in Solid Sediments. The evolutions of total Hg in the two experiments are presented in Figure 4: concentrations were relatively uniform in sediments kept in dark, but decreased considerably in those exposed to solar radiation. This contrast was observed in the two experiments, despite the differences on Hg concentrations. In experiment I, Hg decreased abruptly from 17.6 to 10.0 nmol g-1 during the first 3 h and to 7.65 nmol g-1 at the end. The decrease in experiment II was more gradual, from 3.45 to 1.35 nmol g-1 in 8 h. Reactive and Nonreactive Hg in Sediment Pore Waters. Evaporation differed in the sediments maintained at the two experimental conditions. Mercury concentrations in sediment pore waters were thus corrected to the decrease of VOL. 38, NO. 14, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3903

FIGURE 3. Evolutions of acid volatile sulfides (AVS) and pyrite concentrations (µmol g-1 S) in sediments under light (solar radiation) and dark conditions in experiments I and II.

Discussion

FIGURE 4. Evolutions of total mercury concentrations (nmol g-1, dry weight) in sediments under light (solar radiation) and dark conditions in experiment I (heavily contaminated sediment) and experiment II (moderately contaminated sediment). water content and reported to the initial moisture. The corrected concentrations of reactive mercury (HgR) and nonreactive mercury (HgNR) are presented in Figure 5. A similar pattern was observed in the two experiments: HgR was higher in pore waters of the sediments exposed to solar radiation and increased with the time, 424-845 pM (experiment I) and 53-193 pM (experiments II). On the contrary, HgNR concentrations decreased from 3.00 to 2.59 nM in experiment I and from 0.725 to 0.105 nM in experiment II. Extractable Hg in Solids. The mercury extracted by the hydroxylamine solution was only detectable in the highly contaminated sediments (experiment I) kept in dark (Figure 6). The Hg concentrations increased from below the detection limit (0.02 nmol g-1) to 0.75 nmol g-1 after 4 h. Mercury was undetected ( 50%) of mercury that escapes from low- to high-contaminated sediments used in the field experiments. To our knowledge, this is the first time that release of mercury from intertidal sediments during their temporary exposure to direct solar radiation is reported. However, these observations are in concordance with the atmospheric emission of mercury recorded in contaminated soils (40), wetlands (6, 7), vegetation (5), and air-sea interface exchanges on ocean surface. These emissions influence the mercury concentration in the atmosphere, which is the most important compartment in the global cycle of mercury (8,9). The Mercury-Sulfur Association. The pronounced decreases of AVS and pyrite concentrations observed in all sediments (Figure 3) indicate a rapid oxidation of sulfur compounds during the experiments. The oxidation rates apparently result from anoxic sediments being in contact with the air, and no significant differences (p < 0.05) were found between light and dark conditions. The decrease of AVS was more rapid than that of pyrite, as would be expected from the different oxidation kinetics (41, 44). Mercury is associated with sulfur in sediments, either forming insoluble Hg sulfides (26, 27) or being incorporated into iron sulfides including pyrite (43). Since Hg sulfides are difficult to quantify due to their poor solubility in 1 M HCl (44, 45), the undetectable Hg concentration recorded in this solution does not exclude the existence of Hg in those forms. On the other hand, it indicates that the amount of Hg incorporated into iron sulfides, quantified as AVS, was negligible. Mercury could also be associated with pyrite, as observed in other environ-

FIGURE 5. Evolutions of reactive (HgR) and nonreactive mercury (HgNR) concentrations (pM and nM) in sediment pore waters under light (solar radiation) and dark conditions in experiments I and II.

FIGURE 6. Evolution of Hg concentrations (nmol g-1) extracted with a hydroxylamine solution from sediments under light (solar radiation) and dark conditions in experiment 1. ments (43), since the degree of trace metal pyritization (DTMP) for mercury is one of the highest for all trace metals (27). Pyrite Oxidation and Hg Released to Pore Waters. Under dark conditions, the production rates of reactive and nonreactive mercury in pore waters (expressed on dry weight basis) were relatively constant: 0.021 and 0.138 pmol g-1 h-1 in experiment I and 0.002 and 0.004 pmol g-1 h-1 in experiment II. These linear evolutions over the time had no relation to the drastic decrease of AVS occurring in the first 2 h of the experiments. Indeed, mercury production in pore waters persisted beyond the consumption of the entire AVS in experiment II. The oxidation rate of pyrite was also constant in experiments I and II: 11.2 and 0.06 µmol g-1 h-1. This suggests that mercury is progressively released to pore waters as pyrite is oxidized. The proportion of the oxidation rate of pyrite and the production rate of mercury differed in the two experiments, reflecting the different incorporation of mercury in pyrite in the two sediments. Although oxidation of pyrite results in the increase of reactive mercury, its variation was accomplished by similar changes on nonreactive mercury in pore waters (Figure 5). This similarity probably reflects an equilibrium between dissolved species included in the two operational groups. Resemblance of reactive and nonreactive mercury was also observed in vertical profiles of sediment pore waters in previous studies (30, 46). Hg in Pore Waters Exposed to Solar Radiation. Undoubtedly, solar irradiation of sediments led to additional

processes besides oxidation of reduced-sulfur forms in solids. Furthermore, the considerable changes of reactive and nonreactive mercury concentrations in pore waters exposed to solar radiation with respect to those kept under dark (Figure 5) reflect the reactions that took place under those conditions. These changes are better illustrated by the evolution of the ratios between concentrations of reactive dissolved mercury ([HgR]PW) and of total mercury in solids ([HgT]S), and of the total dissolved mercury ([HgT]PW) presented in Figure 7. The ratio [HgR]PW/[HgT]S, which represents the fraction of Hg that is transferred from solids to pore waters, was greater under light conditions, both in lowly and highly contaminated sediments, and extended beyond the oxidation of most AVS. The increase of the ratio [HgR]PW/[HgT]PW represents the destruction of mercury complexes in pore waters and the increase of simpler forms of Hg. Possible Mechanisms for Hg Escape. The substantial decrease of Hg in sediments coincided with the period of higher UV radiation around 14:00 (Figures 2 and 4). Since radiation had no effect on the pyrite oxidation, other processes had to induce additional mercury released to pore waters, namely the oxidation of particulate organic matter by UV radiation. The direct photoreduction from the solid sediments is not to be excluded, since HgS can be photochemically reduced and semiconductor compounds (e.g. metal oxides) can promote Hg photoreduction in providing charge carriers and valence band holes (17). The increases of reactive mercury in pore waters at the end of experiments I and II, under light conditions (0.29 and 0.05 pmol g-1, respectively), can be considered negligible compared to the decrease in solids (10 000 and 2100 pmol g-1, respectively). The unbalanced situations mean that most mercury released from solids to pore waters should have been photochemically reduced in short periods of time and escaped rapidly to the atmosphere. In fact, mercury species such as complexes with OH-, HS-, Cl- and organic ligands (including CH3Hg0, (CH3)2Hg, CH3Hg+, CH3HgCl, CH3HgOH, CH3HgSH) can absorb the highly energetic UV of the solar spectrum (17, 47, 48) and consequently be reduced in its excited state. Besides Hg(0), the production and subsequent release of dimethylmercury (DMM) (49) may also contribute to the unbalanced situations. Mercury evasion was shown to correlate with temperature (50, 51), and thus it could have accelerated the reactions occurring in the exposed sediments during the periods of VOL. 38, NO. 14, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3905

FIGURE 7. Evolution of the ratios between concentrations of reactive dissolved mercury and total mercury in solids ([HgR]PW/([HgT]S) and of reactive dissolved mercury and total dissolved mercury ([HgR]PW/[HgT]PW) in pore waters from sediments under light (solar radiation) and dark conditions in experiments I and II.

SCHEME 1. Schematic Representation of the Solar Radiation Effect on the Release of Hg from Intertidal Sediments to the Atmosphere

higher temperature. Photoreduction by H2O2, which is produced under sunlight conditions, should not be excluded, since it is good reducer for mercury (17). Moreover, the effect of DOC (16) could also be considered in experiment I due to the high Fe(III) concentrations. According to Nriagu (17), the photolysis rate is rather fast, occurring on the order of seconds to hours. This explains the rapid release of Hg from sediments observed in the experiments. Apparently reactive mercury (HgR) in pore waters corresponds to an intermediate step of the following simplified scheme (Scheme 1). Interestingly, by the second hour of experiment I in the dark, mercury was trapped in iron oxides and extracted by the hydroxylamine solution (Figure 6). The incorporation has been documented in sediment cores due to the high concentration of iron oxides (30). This trapping was not observed under light conditions. Two reasons may be invoked: the rapid photoreduction of Hg(II) by the energetic UV radiation (17) and the use of Fe(III) in the reduction process (16) offering no possibilities to incorporate mercury. The trapping effect was not detected in sediments used in experiment II that contained lower Fe concentrations (Table 2). Environmental Implications. The results obtained in these field experiments suggest that the transfer of gaseous mercury to the atmosphere may occur frequently in physically dominated estuaries and coastal lagoons with extensive intertidal areas. The thickness of oxic sediment in these coastal environments is generally of millimeter scale (52), and stronger currents (28), winds, or benthic organisms (29) can resuspend this layer or transfer anoxic sediments to the 3906

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 14, 2004

surface, exposing fresh surface of intertidal sediments to solar radiation. This transfer may occur in intertidal sediments with different mercury contamination and moves forward until Hg complexes susceptible of being destroyed by UV radiation are present. Simple calculations permit us to compare the magnitude of this transfer with mercury concentration in the topmost sediments. In heavily contaminated sediments (e.g., area A1) with [Hg] ) 10 nmol g-1 and a flux of 235 nmol m-2 h-1, the proportion of mercury transferred to the atmosphere when exposed to solar radiation during 6 h is 56% of the amount existing in 0.5-cm thickness layer. In moderately contaminated sediments (area A2, [Hg] ) 3.5 nmol g-1 and a flux of 37.5 nmol m-2 h-1), virtually all mercury is transferred to the atmosphere. The release of mercury may also take place in dredging operations during the period that anoxic sediments remained exposed to direct solar radiation. All these processes should be taken into consideration when trying to examine the regional and global mercury budget.

Acknowledgments The authors which to thank colleagues Eduarda Pereira, Mo´nica Va´lega, Pedro Brito, and Joa˜o Lavrado for the help on the analytical and field work. The manuscript has benefited from the comments of the anonymous reviewers. This work was funded by the Portuguese Ministry of Science and Higher Education through the restructuring program MACAC, MLEM12/2000.

Literature Cited (1) Jackson, T. A. In Metal metabolism in aquatic environment; Langstron, W. J., Bebianno, M. J., Eds.; Chapman & Hall Lda. Publishers: London, UK, 1998; pp 77-138. (2) Schroeder, W. H.; Munthe, J. Atmos. Environ. 1998, 32, 809822. (3) D’itri, F. M. In The Environmental Mercury Problem; CRC Press: Cleveland, OH, 1972. (4) Coolbaugh, M. F.; Gustin, M. S.; Rytuba, J. J. Environ. Geol. 2002, 42, 338-349. (5) Lindberg, S. E.; Meyers, T. P. Wetlands Ecol. Management 2001, 9, 333-347.

(6) Wallschla¨ger, D.; Kock, H. H.; Schroeder, W. H.; Lindberg, S. E.; Ebinghaus, R.; Wilken, R. D. Water Air Soil Pollut. 2002, 135, 39-54. (7) Zhang, H.; Lindberg, S. E.; Barnette, M. O.; Vette, A. F.; Gustin, M. S. Atmos. Environ. 2002, 36, 835-846. (8) Mason, R. P.; Lawson, N. M.; Sheu, G.-R. Deep-Sea Res. II 2001, 48, 829-2853. (9) Beucher, C.; Wong-Wah-Chung, P.; Richard, C.; Mailhot, G.; Bolte, M.; Cossa, D. Sci. Total Environ. 2002, 131-138. (10) Hudson R. J. M.; Gherini, S. A.; Fitzgerald, W. F.; Porcella, D. B. Water Air Soil Pollut. 1995, 80, 265-272. (11) Fitzgerald, W. F.; Mason, R. P. In Mercury and Its Effects in the Environment Biology; Siegel, A., Siegel, H., Eds.; Marcel Decker, New York. USA. 1997; pp. 53-111. (12) Schroeder, W. H.; Yarwood, G.; Niki, H. Water Air Soil Pollut. 1991, 56, 653-666. (13) Munthe, J.; McElroy, W. J. Atmos. Environ. 1992, 26A, 553-557. (14) Xiao, Z. F.; Munthe, J.; Stromberg, D.; Lindqvist, O. In Mercury Pollution: integration and synthesis; Watras, C. J., Huckabee, J. W., Eds.; Lewis Publishers: Boca Raton, FL, 1994; pp 581592. (15) Lin, C. J.; Pehkonen, S. O. Atmos. Environ. 1999, 33, 2067-2079. (16) Zhang, H.; Lindberg, S. E. Environ. Sci. Technol. 2001, 35, 928935. (17) Nriagu, J. O. Sci. Total Environ. 1994, 154, 1-8. (18) Rasmussen, P. E. Environ. Sci. Technol. 1994, 28, 2233-2241. (19) Gustin, M. S.; Tayler, G. E., Jr.; Leonard, T. L.; Keislar, R. E. Environ. Sci. Technol. 1996, 30, 2572-2579. (20) Lindberg, S. E. In Global and Regional Mercury Cycles: sources, fluxes and mass balances; Baeyens, W., Ebinghaus, R., Vasiliev, O., Eds.; Kluwer Academic Publishers: Dordrecht, Netherlands, 1996; pp. 359-380. (21) Zhang, H.; Lindberg, S. E.; Marsik, F. J.; Keeler, G. J. Water Air Soil Pollut. 2001, 126, 151-169. (22) Gustin, M. S.; Lindberg, S. E.; Marsik, F.; Casimir, A.; Ebinghaus, R.; Edwards, G.; Fitzgerald-Huble, C.; Kemp, R.; Koch, H.; Leonard, T.; London, J.; Majewski, M.; Owens, J.; Pilote, M.; Poissant, L.; Rasmussen, P.; Schneeberger, D.; Schroeder, W.; Sommar, J.; Turner, R.; Vette, A.; Wallschla¨eger, D.; Xiao, Z. F.; Zhang, H. J. Geophys. Res. 1999, 104, D17, 21831-21844. (23) Gobeil, C.; Cossa, D. Can. J. Fish. Aquat. Sci. 1993, 50, 17941800. (24) Pereira, M. E.; Duarte, A. C.; Millward, G. E.; Abreu, S.; Vale, C. Water Sci. Technol. 1998, 36, 6-7, 125-130. (25) Mantoura, R. F. C.; Dickson, A.; Riley, J. P. Est. Coast. Mar. Sci. 1978, 6, 387-408. (26) Drobner, E.; Huber, H.; Wachterhauser, G.; Rose, D.; Steller, K. Nature 1990, 546, 742-744. (27) Morse, J. W.; Luther, G. W. Geochim. Cosmochim. Acta 1999, 56, 3373-3378. (28) Vale, C.; Sundby, B. Estuar. Coast. Shelf. Sci. 1987, 25, 495-508.

(29) Rhoads, D. C. Oceanographic Marine Biology Annual Review; Barnes, H., Ed.; 1974; Vol. 99, pp 263-300. (30) Cana´rio, J.; Vale, C.; Caetano, M.; Madureira, M. J. Environ. Pollut. 2003, 126, 425-433. (31) Lavrado, J. Distribuic¸ a˜o e reactividade de mercu ´ rio em sedimentos do Estua´rio doTejo. Thesis for degree in Technological Chemistry. University of Lisbon, Lisbon, Portugal. 2003. (32) Madureira, M. J.; Vale, C.; Simo˜es Gonc¸ alves, M. L. Mar. Chem. 1997, 58, 27-37. (33) Rantala, R.; Loring, D. Atom. Absorp. Newsletter. 1975, 14, 117120. (34) Chester, R.; Huges, M. J. Chem. Geol. 1967, 2, 249-262. (35) Luther, G. W.; Ferdelman, T. G.; Kostka, J. E.; Tsamakis, E. J.; Church, T. M. Biogeochemistry 1991, 14, 57-88. (36) Henneke, E.; Luther, G. W.; De Lange, G. J. Mar. Geol. 1991, 100, 115-123. (37) Canfield, D. E.; Raiswell, R.; Westrich, J. T.; Reaves, C. M.; Berner, R. A. Chem. Geol. 1986, 54, 149-155. (38) NOAA. In Contamination in aquatic habitats at hazardous waste sites: Mercury. Technical Memoradum; NOS ORCA 100. 1996. (39) Mucci, A.; Lucotte, M.; Montgomery, S.; Plourde, Y.; Pichet, P.; Tra, H. Can. J. Fish. Aquat. Sci. 1995, 52, 2507-2517. (40) Lindberg, S. E.; Kim, K. H.; Meyers, T. P.; Owens, J. G. Environ. Sci. Technol. 1995, 29, 126-135. (41) Kamei, G.; Ohmoto, H. Geochim. Cosmochim. Acta 2000, 15, 2585-2601. (42) Schippers, A.; Jorgesen, B. B. Geochim. Cosmochim. Acta 2002, 15, 85-92. (43) Morse, J. W. Mar. Chem. 1994, 46, 1-6. (44) Cooper D. C.; Morse J. W. Environ. Sci. Technol. 1998, 32, 10761078. (45) Cooper D. C.; Morse J. W. Aquat. Geochem. 1998, 5, 87-97. (46) Ramalhosa, E. Mercu ´ rio na Ria de Aveiro: Associac¸ o˜es, reactividade e especiac¸ a˜o. Ph.D. Thesis, University of Aveiro, Aveiro, Portugal, 2002. (47) Fujita, S.; Horri, H.; Taniguchi, S. J. Phys. Chem. 1973, 77, 28682871. (48) Stromberg, D.; Stromberg, A.; Wahlgren, U. Water Air Soil Pollut. 1991, 56, 681-695. (49) Wallschla¨ger, D.; Hintelmann, H.; Evans, R.-D.; Wilken, R. D. Water Air Soil Pollut. 1995, 80, 1325-1329. (50) Lindberg, S. E.; Meyers T. P.; Munthe J. Water Air Soil Pollut. 1995, 85, 725-730. (51) Poissant, L.; Casimir, A. Atmos. Environ. 1998, 32, 883-893. (52) Baillie, P. W. Estuar. Coast. Shelf Sci. 1986, 22, 143-149.

Received for review December 19, 2003. Revised manuscript received May 5, 2004. Accepted May 10, 2004. ES035429F

VOL. 38, NO. 14, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3907