Sediment and Porewater Profiles and Fluxes of Mercury and

Mercury (HgT) and methylmercury (MeHg) were measured at 1−2 cm ... The Marcell Experimental Forest, a unit of the Chippewa National Forest, was set ...
0 downloads 0 Views 232KB Size
Download PDF
Environ. Sci. Technol. 2004, 38, 6610-6617

Sediment and Porewater Profiles and Fluxes of Mercury and Methylmercury in a Small Seepage Lake in Northern Minnesota N E A L A . H I N E S , * ,† PATRICK L. BREZONIK,† AND DANIEL R. ENGSTROM‡ Department of Civil Engineering, University of Minnesota, Minneapolis, Minnesota 55404, and St. Croix Watershed Research Station, Science Museum of Minnesota, Marine on St. Croix, Minnesota 55047

Mercury (HgT) and methylmercury (MeHg) were measured at 1-2 cm resolution in sediment porewater and sediment cores from Spring Lake in the Marcell Experimental Forest of northern Minnesota. Recent sediment accumulation of HgT was 21.4 µg m-2 yr-1 (1990-2000), 2 orders of magnitude greater than the accumulation of MeHg (0.20 µg m-2 yr-1). The highest solid phase concentrations of MeHg were observed persistently at the sediment surface and declined sharply with depth. Porewater profiles showed a small diffusive flux of MeHg from sediment to water (5 ng m-2 month-1). Springtime porewater concentrations of MeHg were relatively low (∼0.5 ng L-1) and increased by late summer to early fall (1.5-2.2 ng L-1), showing distinct peaks roughly correlated with maxima in sulfate reducing activity at 5 and 15 cm. Advective transport carrying MeHg deeper into the sediment was evident in summer and fall. The percent of HgT present as MeHg was highest in the water column above the sediment (10%) and decreased with sediment depth in both the solid and porewater phases. Sediments at this study site are a net sink for MeHg, although diagenetic processes of demethylation and methylation are evident within the lake-sediment environment.

Introduction Methylmercury (MeHg) is a neurotoxin that bioaccumulates in the aquatic food chain, even in remote, undeveloped lakes. Both in-lake and watershed sources are important in considering the inputs of MeHg to lakes. The limited data on the formation of MeHg in lakes suggest that sediments and/ or near-shore wetlands can be sources of MeHg in aquatic environments (1-4). Although many studies on sediment have addressed total mercury (HgT) (5-8), less attention has been paid to MeHg. Given the general preference of Hg forms for solid phases (Kd > 103 L kg-1) (9, 10), sediment profiles are useful to determine inventories and accumulation rates of mercury in water bodies. However, formation and destruction of MeHg in sediments makes its fate more difficult to study using solid-phase profiles alone. Additional data from sediment * Corresponding author phone: (952)832-2708; fax: (952)832-2601; e-mail: [email protected]. † University of Minnesota. ‡ Science Museum of Minnesota. 6610

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 24, 2004

porewater are needed to determine MeHg transport and fate. Sediment-water exchange of MeHg depends on its aqueous concentration in the porewater, in-lake chemical reactions, and lake-groundwater interactions. As part of ongoing research on mercury in a forested watershed, we measured profiles of MeHg and HgT in sediments, porewater, and the overlying water of Spring Lake, a small bog lake at the Marcell Experimental Forest, Grand Rapids, MN, during the ice-free season in 2000-2002. Our objectives were to (1) quantify fluxes of HgT and MeHg to the lake sediment using multiple sediment cores, (2) determine whether sediments contribute MeHg to the lake by analyzing porewater MeHg profiles seasonally on a fine vertical scale (1 cm), and (3) measure the hydraulic gradient to estimate the direction of water movement between the sediment and lake. Site Description. The Marcell Experimental Forest, a unit of the Chippewa National Forest, was set aside by the U.S. Forest Service for research in 1960. Spring Lake, the focus of this study, is a small (8.9 ha), shallow (mean depth ) 2 m) bog lake with moderately high DOC (11 mg L-1, Table 1). The lake is well-mixed after ice-out and does not stably stratify during summer. Nonetheless, dissolved oxygen near the bottom is persistently low (18.2 MΩ). Water was collected from the bow of a boat using polyethylene gloves while it faced into the wind. Samples were acidified with 1% HCl and refrigerated until analysis. Water samples at depth were collected using an acid cleaned Teflon Niskin bottle. Seston were collected at 1.4 m depth using Teflon tubing attached to a wooden rod and a hand pump connected in series to a reusable acid-cleaned Teflon filter pack (Savillex). The lake was sampled six times (May to October) during 2003 at the same location near the center of the lake. For each sample, a 0.7 µm glass fiber filter (Whatman) was preweighed using a sensitive balance (0.001 mg). Filters were baked previously at 550 °C and handled with gloved hands. Five to seven filters were loaded with seston, and the corresponding volume of water was measured (500-550 mL). Filters were stored in individual polyethylene bags and frozen 30 min after collection. Seston-loaded filters were freezedried and weighed again to determine seston dry mass and analyzed for HgT and MeHg. Eight blank filters were compared for consistency in mass by only high-resistivity deionized water being filtered, dried, and reweighed. The average change in filter mass was 0.22%. The change in filter mass was 8.3% of the average freeze-dried seston mass (0.774 mg). Mercury Analysis. Laboratory contamination was minimized by conducting analytical work in a clean room equipped with a HEPA filter. All sampling was done using powder-free gloves, and samples were double bagged in clean, polyethylene bags. HgT refers to all forms of mercury VOL. 38, NO. 24, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6611

FIGURE 2. Percent water and organic matter in sediment from Spring Lake, April 6, 2000. Error bars are one standard deviation from five sediment cores. oxidized by bromium monochloride (BrCl) and then reduced by stannous chloride (SnCl2). This includes Hg0, mercury bound to organic matter, and mercuric and mercurous species. The analytical procedure, EPA Method 1631, involves detection and quantification of mercury as Hg0 by cold vapor atomic fluorescence spectrometry (CVAFS) (16). A fluorescence detector (254 nm) with a photomultiplier tube (Tekran 2500) was used to quantify Hg0 vapor thermally desorbed from gold traps. Standard reference material (NIST 3133; Hg(NO3)2 aq) was quantified for HgT in water. The mean concentration was within 1.1% of the certified value (1000 mg L-1). Analysis of HgT in sediment was performed using strong acid digestion following the procedure of Fleck et al. (17). Sediments were freeze-dried, and 100-300 mg aliquots were weighed and placed in acid-cleaned 30 mL Teflon digestion bombs. Ten mL of concentrated sulfuric acid and 10 mL of concentrated nitric acid were added, and the bombs were sealed tightly and placed in an oven at 45°C overnight. An aliquot (generally 0.5 mL) of the acid-digested sediment was added to an Erlenmeyer bubbler filled with high-resistivity deionized water. Nitric acid was neutralized using hydroxylamine solution. During analysis, care was taken to replace the deionized water in the Erlenmeyer bubblers because interferences and carryover have been observed (18). Analysis by CVAFS was performed as described previously. Standard reference materials from the National Research Council Canada, MESS-3 (marine sediment) and DORM-2 (dogfish muscle) were used to monitor analytical performance. The mean concentration of HgT for MESS-3 was within 4.3% of the certified value (91 ng g-1). The mean concentration of HgT DORM-2 was within 7.5% of the certified value (4640 ng g-1). Methylmercurywas measured following the methods of Horvat et al. (19). Distillation was performed using 80-100 mL of sample with KCl, H2SO4, and CuSO4 for 4-6 h in a Teflon distillation system with N2 carrier gas. Seston samples were analyzed by distillation of the freeze-dried filters with the above reagents. Distillates were kept on ice and out of direct light to prevent photodegradation of MeHg. Sodium tetraethylborate was added to convert MeHg to volatile methylethylmercury. The optimum pH for the reaction, 4.9, was attained using a sodium acetate-acetic acid buffer. Clean, Hg-free N2 was passed through the sample (65 mL min-1) for 15 min using a fritted bubbler, and the outlet of the bubbler was attached to a glass trap packed with Tenax TA (Alltech) to sorb the alkyl Hg. This trap was then placed on an analytical line with argon carrier gas (25 mL min-1) and thermally desorbed and quantified using CVAFS after being passed through a GC column (15% OV-3 on Chromasorb W-AW-DMCS, 60/80 mesh (Supelco, Inc., Bellefonte, 6612

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 24, 2004

PA), which separated the Hg forms in the sequence Hg0, dimethylmercury, methylethylmercury, and diethylmercury (from ethylation of Hg2+) (20). Finally, analytes were reduced to Hg0 via combustion in a quartz-wool-packed furnace at 750 °C. Peak areas were quantified as Hg0 using peak areas from an integrator (Hewlett-Packard model 3396 series II). Sediment was analyzed for MeHg similar to the procedure described previously. A mass of freeze-dried sediment (50200 mg) was distilled with 90 mL of high-resistivity deionized water and treated like an aqueous sample. To monitor analytical precision, a reference material was prepared from a bulk sample of Spring Lake sediment collected with a Ponar dredge. The sediment was thoroughly mixed and freezedried. An aliquot of this sediment was distilled and analyzed with each set of samples. The average concentration of MeHg from the reference material was 1.06 ng g-1 (dry weight); SD ) 0.31, n ) 39. Sediment reference material from the International Atomic Energy Agency (IAEA 356) was obtained in 2003, and the measured mean concentration of MeHg was within 8.5% of the certified value (5.46 ng g-1) using the distillation technique.

Results and Discussion Sediment Cores. The sediment of Spring Lake is highly organic with small amounts of coarse debris comprised mostly of woody stems. For the five sediment cores, the average water content was 95.3 ( 0.7% (n ) 203), and the average organic content was 54.4 + 5.1% (Figure 2). Results of 210Pb analysis indicate conformable and relatively low rates of sediment accumulation at all coring sites. Profiles of total 210Pb show monotonic declines from surface activities of 2843 pCi g-1 to a constant supported (background) activity of 0.6-1.0 pCi g-1 at depths ranging from 26 to 42 cm (Figure 3). Core 5, from the deepest part of basin, had generally higher activities and greater depth to background (supported) 210Pb. Although the 210Pb-dating horizon (the depth at which supported 210Pb is reached) was fairly deep in these cores, dry-mass accumulation rates were low because of the high water content of the sediments. Among all five cores, drymass accumulation ranged from 0.004 to 0.018 g cm-2 yr-1 (Figure 4). Core 5 had generally higher rates than the other cores, and most cores showed systematic, although asynchronous changes in sediment flux over time. In three cores (1, 2, and 5), the highest sediment accumulation rates occurred within the past decade. Because there is no human development in the watershed, it seems likely that increases in sediment accumulation were driven by climatic changes and their effects on riparian wetlands surrounding the lake. Sediment core data generally are characterized by a high degree of temporal and spatial variation; therefore, several

FIGURE 3. Sediment core depth vs total

FIGURE 4.

210Pb

activity in five sediment cores from Spring Lake.

210

Pb date vs sediment accumulation rate in five sediment cores from Spring Lake.

cores and lake depths are needed to accurately characterize accumulation (14). In this study, a simple arithmetic mean of the five cores was used to estimate lake-wide mercury accumulation. This approach is justified because the cores were spaced relatively evenly about the lake and sedimentation rates varied only moderately from core to core. For example, the average sediment accumulation rate over the first 3 cm had a coefficient of variation of 31.5% among the five cores (average ) 0.013 g cm-2 yr-1; min ) 0.0084 (core 3), max ) 0.018 g cm-2 yr-1 (core 5)). Cores 1-4 are from shallow water (2-3 m). Core 5, from the deepest part of the lake (5.5 m), had a slightly higher sediment accumulation rate (0.018 g cm-2 yr-1) as a result of sediment focusing. Profiles of HgT and MeHg for the five sediment cores (Figures 5 and 6, respectively) show that mercury concentrations generally increased toward the sediment-water interface. HgT increased above ∼20 cm depth, corresponding to the late 1800s, and MeHg increased above ∼12 cm, corresponding to the late 1950s. The percent of mercury present as MeHg generally was low in the sediment and sediment porewater. At the sediment surface (0-5 cm), the MeHg solid-phase concentration is 1.0% of the HgT concentration, but the proportion declined with depth (0.25% at 21-30 cm), driven by the sharp decline of solid-phase MeHg with depth (Table 2). To examine accumulation of HgT, data for each decade were grouped to compare similar time-stratigraphic units. On the basis of the average of the five cores, the recent

accumulation of HgT is 21.4 µg m-2 yr-1 ( 7.9 (SD) (Figure 7) for the decade 1990-2000. This rate is 2 orders of magnitude greater than that for MeHg accumulation. The background flux of HgT was estimated using a time-weighted mean of samples dated between 1800 and 1850, yielding a value of 5.0 µg m-2 yr-1 + 3.1 (SD). The ratio of these two numbers (modern/background) provides a measure of how much Hg accumulation has increased as compared to preindustrial times. The range of these flux ratios from the five cores is 2.9-7.2, and the average is 4.3. This is slightly higher than that in studies from other lake sediment cores from northern North America, which generally show 2-4fold increases (21-26). The sediment profiles from Spring Lake do not show a decline in modern HgT accumulation but rather a leveling off of HgT over the last 1-2 decades. Previous research has shown that recent declines in HgT accumulation are evident in regions of Minnesota closer to point sources, such as the metropolitan areas of Minneapolis and St. Paul, but in more remote sites, such as western Minnesota and southeast Alaska, long-range atmospheric transport supports increasing HgT accumulation at the sediment surface (23). Another study of HgT accumulation in sediment cores from five lakes near Grand Rapids, MN, showed declines of 15-20% in Hg deposition since the mid-1970s (27). The declines in HgT accumulation in the Grand Rapids lakes are tentatively attributed to historical use of mercuric fungicides by local paper mills (1940s-1970s) and the emission and near-source VOL. 38, NO. 24, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6613

FIGURE 5. Profile of sediment depth vs HgT concentration for five sediment cores from Spring Lake on April 6, 2000.

FIGURE 6. Profile of sediment depth versus MeHg concentration for five sediment cores from Spring Lake on April 6, 2000.

TABLE 2. Concentrations of MeHg in Lake Bottom Water, Porewater, Seston (0.7 µm), and Sedimenta

depth cm

environmental matrix

140 100 6-11 1-5 (-) 1-5 (-) 6-10 (-) 11-15 (-) 16-20 (-) 21-30 (-) 31-40 (-) 41-50

Seston lake bottom water porewater equilibrator porewater equilibrator porewater equilibrator or sediment

porewater or bottom water MeHg May 2001 (ng L-1)

porewater or bottom water MeHg/HgT May 2001 (%)

0.20 0.32 0.39 0.62 0.69 0.56 0.45 0.33 0.31 nab

10 3.4 7.5 2.8 3.5 2.0 1.7 1.6 1.5 na

porewater MeHg August 2002 (ng L-1)

av sediment or seston MeHg (ng g-1) (dw) 48

0.10 0.19 0.59 0.83 0.68 1.07 0.60 0.26 0.21

1.6 0.85 0.51 0.30 0.22 0.10 0.059

sediment or seston MeHg/HgT (%) 10

1.0 0.57 0.38 0.27 0.25 0.14 0.085

a Lake bottom water was collected spring to fall 2001. Porewater concentrations are shown from May 2001 and August 2002 because the greatest number of MeHg data was available above the sediment-water interface on these dates. Porewater HgT was available only for May 2001. Seston were collected May-October 2003. Sediment MeHg is averaged from five sediment cores, April 6, 2000. The sediment-water interface is at depth ) 0 b na, not available because of lack of porewater phase data.

deposition of reactive Hg species. The absence of such a signal in Spring Lake is consistent with the lake’s greater distance from the Grand Rapids mills (31 km) as compared with the five local lakes (6-11 km). In contrast to the HgT profiles, which largely record historical trends in mercury deposition, the interpretation of MeHg profiles is complicated by active methylation and 6614

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 24, 2004

demethylation within the sediment column. In the case of Spring Lake, the sharp down-core decline in MeHg concentration and accumulation most likely reflects diagenetic processes rather than a historic change (increase) in MeHg production within the lake. The declines are largely a function of sediment depth rather than age, and their shape suggests a first-order decay process. More important, there is no

FIGURE 7. Accumulation of HgT and MeHg in Spring Lake sediments using a decadal time-weighted average. Error bars are the standard deviation from five sediment cores. historical context or explanation for such recent and sharp increases in MeHg production. Rather, the sharp down-core decline in sediment MeHg implies production and/or deposition of MeHg at the sediment surface and subsequent demethylation with burial such that sediments are ultimately net sinks for MeHg. If there is a historical signal in MeHg accumulation, it cannot be seen because of the high turnover (production and degradation) of MeHg in the near-surface layers. Sediment Porewater. Several methods have been used to measure sediment porewater profiles of MeHg, including centrifugation, sediment filtration (by gravity or vacuum), whole core squeezing, and porewater equilibrators. A critical review of these procedures was performed by Mason et al. (28), who showed that centrifugation under an inert atmosphere was the most reliable method for quantifying MeHg. However, the method is time-consuming and requires a high degree of operator training. The porewater equilibrators used in that study were constructed from Teflon, which absorbs oxygen and possibly can alter the redox environment within the sediment. In contrast, our equilibrators were constructed of acrylic, which has a low permeability to O2. Profiles of sediment porewater were measured five times: May 2-June 1, 2001; September 12-October 17, 2001; March 19-April 29, 2002; July 16-September 4, 2002; and September 5-October 17, 2002 (Figure 8). Ice melted from Spring Lake on May 3, 2001. The highest concentrations of MeHg in May 2001 were found near the sediment-water interface and declined from sediment to lake bottom water. The slope of the profile at the interface in spring 2001 (Figure 8A) indicates that sediments were a source of MeHg to the lake at that time. A diffusive flux of 5 ng m-2 month-1 was estimated using Fick’s first law, Dw(∂C/∂x) and the following assumptions and estimates: negligible vertical advective transport, Dw ) 0.5 × 10-5 cm2 s-1 and a concentration gradient ) 0.04 ng L-1 cm-1. This flux, if it continued over the whole year, would be equivalent to one-fourth to one-third of the annual accumulation rate of MeHg at the sediment surface (Figure 7, 0.20 µg m-2 yr-1). The porewater profile from September to October 2001 had sharp peaks of MeHg centered at depths of 19 and 31 cm in the sediment (Figure 8A). These peaks may represent in situ generation of MeHg, possibly from sulfate-reducing bacteria. Many species of sulfate-reducing bacteria methylate mercury, and a positive correlation has been found between MeHg formation and sulfate-reducing bacterial activity (29, 30). Sulfate reduction rates (SRR) in Spring Lake littoral sediments from August 13, 2001, and February 23, 2002,

showed that the maximum SRR occurred at 5 cm depth in the sediment (31). These SRR profiles covered four equally spaced depths spanning the top 20 cm of sediment. A more detailed analysis of SRR using nine sediment cores collected across the lakebed on June 28, 2000 showed sulfate-reducing activity up to a depth of 23 cm (31). As was found in the profiles from August 13 and February 23, SRR peaked at 5 cm (1.2 nmol SO42- mL-1 h-1), but a larger peak occurred at 15 cm (1.6 nmol SO42- mL-1 h-1). In September 2001, the peak in porewater MeHg at 31 cm (Figure 8A) is well below the region where sulfate reduction was measured in the June 28 transect (0-23 cm). This suggests that sulfate-reducing activity may extend deeper in the sediments later in the season. Because Spring Lake is shallow and the water column is nearly isothermal, sediment temperatures increase toward late summer and may contribute to higher SRB activity. A study at small headwater lakes in southern Finland found the highest rates of methylation and demethylation with maximum sediment temperature (32). It is also possible that MeHg was formed higher in the sediment and then advected downward, although one would expect broad peaks from dispersion instead of the sharp peaks observed. For a hydraulic conductivity of 3.0 × 10-6 cm s-1 and the observed hydraulic head gradients from summer and fall 2002, advection 10 cm into the sediment could be achieved in 100 d. Porewater equilibrators were deployed in the littoral zone during ice cover in late March 2002 to examine whether MeHg diffuses out of the sediment during winter. Unfortunately, the tops of the equilibrators were deployed just below the sediment-water interface because the ice cover prevented accurate observation of the lake bottom. Thus, data are not available from the overlying water, and the assigned depths in the profile (Figure 8B) are an estimate. The shape and magnitude of MeHg concentrations in this profile are roughly similar to the May, 2001 profile, which suggests that although porewater concentrations of MeHg are low (∼0.5 ng L-1) during spring, maxima are at or near the sediment surface. The littoral porewater MeHg profile from August 2002 shows increases in MeHg concentrations with sediment depth but without well-defined peaks (Figure 8B). A diffusive flux was calculated from the concentration gradient (0.04 ng L-1 cm-1) and was equal to the sediment to water flux from May 2001 (5 ng m-2 month-1). Porewater profiles of MeHg were measured in two areas of the lake from September 5 to October 17, 2002: (i) the littoral area sampled during 2001 (z ) 1.0 m) and (ii) near the deepest part of the lake (Figure 1). This was accomplished by separating the two equilibrators VOL. 38, NO. 24, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6615

FIGURE 8. Sediment porewater profiles of MeHg at Spring Lake. The sediment-water interface is at 0 on the y-axis. (A) May 2-June 1, 2001, and September 12-October 17, 2001. (B) March 19-April 29, 2002, and July 16-September 4, 2002. (C) September 5-October 17, 2002. and installing one at each location. MeHg concentrations in porewater were relatively low (range of 0.1-0.5 ng L-1) at both locations (Figure 8C). The profile from the deeper site showed that MeHg in the overlying lake water increased toward the sediment water interface, but a decline was evident just above the interface. A similar trend was found in the littoral zone, but porewater concentrations were more variable than those in the deepwater profile. A potentially important phase not quantified in this study is colloidal MeHg and HgT. Stordal et al. (33) suggested that colloidal Hg is chemically more similar to the sediment phase. Colloidal MeHg would pass through the large pores of the polyethersulfone ultrafiltration membrane (500 kDa) used in this study. A critical evaluation of ultrafiltration membranes for HgT and MeHg showed that sorption and charge rejection was evident with polyethersulfone membranes (34). A combination of colloidal-phase MeHg, episodic production of MeHg in the sediments, and preferential binding to DOM may maintain porewater MeHg despite the sharp decline in 6616

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 24, 2004

solid-phase MeHg. Studies have shown that water-particle equilibrium of MeHg is attained within minutes (10, 35). Piezometer readings from the northwest shore of the lake showed the least vertical hydraulic gradient in June (-0.10), with increasing values toward October (-0.48). The negative value indicates a lower water level inside the piezometer relative to the lake surface. Thus, there appears to be downward seepage from the lake to groundwater throughout the open-water season with flow rates generally increasing from spring to fall. Using the measured head differences from June to October, a lake-bed surface area of 8.9 ha (assumed equal to the lake surface area) and a hydraulic conductivity (Ksat) of 1.1-8.6 × 10-6 cm s-1 estimated from nearby wetlands (36), flux calculations based on Darcy’s Law produce hydraulic discharges from lake to groundwater ranging from 8 to 320 m3 d-1 or 0.2 to 3.6 mm d-1. This is comparable to diffusion-controlled flow. (Equation 1 solved for σ using Dw ) 1 × 10-6 cm2 s-1 yields 4 mm for 1 d of diffusion.) Early in the year, when advection into the sediment is small, a diffusive flux of MeHg to the lake may be possible, but the larger hydraulic gradient later in the year likely leads to advective-controlled transport downward in the sediment. Bioturbation also could result in higher MeHg fluxes across the sediment-water interface than those estimated for diffusion-controlled flow. However, 210Pb profiles are largely exponential near the core tops (Figure 3) and lack the characteristic flattening (near-constant values) that severe bioturbation induces. Some mixing may be occurring nonetheless, and the observation of occasional chironomid (midge) larvae in the cores during sectioning suggests that limited mixing or irrigation of the surface sediments by benthic fauna is likely. The degree to which MeHg diffusion and advection are enhanced by this mechanism is unknown. Profiles for solid-phase MeHg (Figure 6) all have the highest concentrations at the sediment surface, in contrast with the porewater MeHg profiles, which reveal peak concentrations up to 31 cm into the sediment. Although some peaks in porewater MeHg near the sediment-water interface are evident for 2001 profiles (Figure 8A), sediment porewater does not appear to be a strong source of MeHg to the lake for the following reasons. When the gradient (∂C/∂x) supports a diffusive flux across the sediment-water interface, porewater MeHg concentrations near the sediment-water interface are relatively low (