Mercury Speciation and Transport via Submarine Groundwater

Jan 27, 2012 - U.S. Geological Survey, Pacific Coastal & Marine Science Center, Santa ... As urban and agricultural development continue to expand in ...
0 downloads 0 Views 3MB Size
Article pubs.acs.org/est

Mercury Speciation and Transport via Submarine Groundwater Discharge at a Southern California Coastal Lagoon System P. M. Ganguli,†,* C. H. Conaway,‡,§ P. W. Swarzenski,§ J. A. Izbicki,∥ and A. R. Flegal‡ †

Earth and Planetary Sciences, University of California, Santa Cruz, California, United States Institute of Marine Sciences, University of California, Santa Cruz, California, United States § U.S. Geological Survey, Pacific Coastal & Marine Science Center, Santa Cruz, California, United States ∥ U.S. Geological Survey, Water Resources Division, San Diego, California, United States ‡

ABSTRACT: We measured total mercury (HgT) and monomethylmercury (MMHg) concentrations in coastal groundwater and seawater over a range of tidal conditions near Malibu Lagoon, California, and used 222Rn-derived estimates of submarine groundwater discharge (SGD) to assess the flux of mercury species to nearshore seawater. We infer a groundwater-seawater mixing scenario based on salinity and temperature trends and suggest that increased groundwater discharge to the ocean during low tide transported mercury offshore. Unfiltered HgT (U-HgT) concentrations in groundwater (2.2−5.9 pM) and seawater (3.3−5.2 pM) decreased during a falling tide, with groundwater U-HgT concentrations typically lower than seawater concentrations. Despite the low HgT in groundwater, bioaccumulative MMHg was produced in onshore sediment as evidenced by elevated MMHg concentrations in groundwater (0.2−1 pM) relative to seawater (∼0.1 pM) throughout most of the tidal cycle. During low tide, groundwater appeared to transport MMHg to the coast, resulting in a 5-fold increase in seawater MMHg (from 0.1 to 0.5 pM). Similarly, filtered HgT (F-HgT) concentrations in seawater increased approximately 7-fold during low tide (from 0.5 to 3.6 pM). These elevated seawater F-HgT concentrations exceeded those in filtered and unfiltered groundwater during low tide, but were similar to seawater U-HgT concentrations, suggesting that enhanced SGD altered mercury partitioning and/or solubilization dynamics in coastal waters. Finally, we estimate that the SGD HgT and MMHg fluxes to seawater were 0.41 and 0.15 nmol m−2 d−1, respectively − comparable in magnitude to atmospheric and benthic fluxes in similar environments.



INTRODUCTION As urban and agricultural development continue to expand in coastal areas, it has become increasingly important to quantify the concentration, transformation, and flux of contaminants at the land-sea margin. This includes mercury, a bioaccumulative neurotoxin that originates from natural and anthropogenic sources.1−3 Although several sources of inorganic mercury (Hg(II)) and organic monomethylmercury (MMHg) have been identified in ocean, estuarine, and freshwater systems, mercury biogeochemical cycling in some aqueous environments, such as the fresh water-seawater interface, remains poorly understood.4 The freshwater-seawater mixing zone typically exhibits a range of redox and organic matter gradients that affect the species, reactivity, and bioavailability of micronutrients, carbon, and metals.5 Because MMHg production is associated with the presence of anaerobic bacteria,6,7 coastal water and sediment that become oxygen depleted are potential areas of enhanced methylation. Consequently, groundwater that moves through this mixing zone before discharging to coastal lagoons or the ocean may play a role in transforming mercury species as well as transporting biologically available mercury to nearshore surface waters. © 2012 American Chemical Society

Submarine groundwater discharge (SGD), defined as all water that discharges from the seabed to coastal ocean water along the continental margin, includes terrestrially derived groundwater as well as seawater that recirculates through coastal sediment via ocean forces such as tidal pumping and wave action.5,8 Although there is growing awareness of SGD’s apparent importance in mercury biogeochemical cycling, few studies have quantified the flux of total mercury (HgT, the sum of all mercury species) in SGD.9−12 Even fewer investigations have quantified SGD-MMHg fluxes or identified processes that influence mercury bioavailability in the freshwater-seawater mixing zone.11,12 In this paper, we present mercury speciation and flux data from groundwater and seawater adjacent to Malibu Lagoon in Southern California. Coastal lagoon environments are biologically productive ecosystems that are globally abundant13 and often contain elevated nutrient and metals concentrations due to urban and agricultural runoff.14,15 As a result, these Received: Revised: Accepted: Published: 1480

August 9, 2011 December 20, 2011 December 26, 2011 January 27, 2012 dx.doi.org/10.1021/es202783u | Environ. Sci. Technol. 2012, 46, 1480−1488

Environmental Science & Technology

Article

Figure 1. Sampling location adjacent to Malibu Lagoon (Malibu, CA). Samples were collected during dry season condtions (July, 2009) when a permeable sand berm barrier physcially separated the lagoon from surface water exchange with the ocean.

lagoons may be effective ‘reactors’ for mercury transformations, with dynamic water level fluctuations driving mercury species into nearshore waters via SGD. Here we evaluate a coastal lagoon system in terms of its (1) mercury speciation and (2) ability to convey mercury species to the nearshore by SGD.

imately 3 m into the sand berm, and grab samples for seawater were collected from the adjacent surf zone and filtered onshore (Figure 2). Groundwater and seawater were sampled hourly for



MATERIALS AND METHODS Study Site. Malibu Lagoon is a 0.05 km2 brackish lagoon located in the Santa Monica Bay watershed in Southern California (Figure 1). The lagoon receives surface water runoff from Malibu Creek, which drains an approximately 280 km2 watershed with over 90 000 residents.16,17 During the dry season (summer months), a sand berm develops at the mouth of the lagoon, isolating it from direct surface water interaction with the ocean. Although waves occasionally overtop the berm during high tides and storm events, exchange between the lagoon and ocean during the dry season is predominantly by groundwater flow through the sand berm. At low tide, the water level gradient is toward the ocean and groundwater discharges to the nearshore ocean (e.g., surf zone).18 In contrast, during high tide, tidal pressure combined with changes in the water level gradient force groundwater to discharge landward, into Malibu Lagoon. During the wet season (winter months), runoff from Malibu Creek is usually sufficient to breach the sand berm. This enables tidal exchange between Malibu Lagoon and coastal ocean water, thereby reducing SGD to the surf zone. Sample Collection. Water samples from the Malibu Lagoon study area were collected from July 23 to 24, 2009 during dry season conditions (Figure 1). A well-developed sand berm physically separated the lagoon and ocean; however, the berm was briefly overtopped by incoming waves during an unusually high tide. Groundwater was sampled from a stainless steel Solinst drive-point piezometer (50-mesh cylindrical filter screen within a 20 mm drive-point body) installed approx-

Figure 2. Illustration of sampling design acrosss the sand berm that separated Malibu Lagoon from the ocean. We collected submarine groundwater from a piezometer placed approximately 3 m into the berm. The intake line was used to collected seawater samples for SGD measurements. Figure after Izbicki.27

a 10 h period (22:00 to 08:00, all times are given in Pacific Standard Time) during a falling tide (+2.0 to −0.3 m). Throughout this discussion, we define high to mid-tide elevations as being between +2.0 to approximately +0.5 m, with low tide occurring between elevations of +0.5 to −0.3 m. Unfiltered surface water grab samples from Malibu Lagoon were collected once during high tide and once during low tide conditions. Samples were collected using established trace metal clean protocols.19,20 Groundwater was extracted through a trace metal clean Teflon sample line coupled with in-line C-flex tubing placed through a peristaltic pump. Seawater samples were collected in a large trace metal clean container by wading into the surf zone to a depth of about 1 m. All samples were filtered in situ with a high-capacity Geotech 0.45 μm disposable 1481

dx.doi.org/10.1021/es202783u | Environ. Sci. Technol. 2012, 46, 1480−1488

Environmental Science & Technology



Article

RESULTS AND DISCUSSION Groundwater-Seawater Interaction. The salinity and temperature time series data at the Malibu Lagoon study site provide evidence of groundwater-seawater mixing in response to tidal pumping and changes in SGD (Figure 3). Seawater

filter capsule attached to the sample line. HgT and MMHg samples were collected in acid-clean Teflon bottles which were placed on dry ice for transport back to the laboratory (University of California, Santa Cruz). A calibrated YSI 556 water quality multiprobe was used to measure conductivity, temperature, and pH. All HgT and MMHg water samples remained frozen until analysis and were preserved following protocols described by Parker and Bloom.21 HgT samples were acidified while thawing to 0.5% (v/v) BrCl and stored in the dark at room temperature until analysis. MMHg samples were acidified while thawing to 0.2% (v/v) H2SO4 (saline and brackish waters) and stored in the dark at 4 °C until analysis. Sample Analyses. HgT concentrations were determined by oxidation with BrCl, reduction with SnCl2 , gold trap amalgamation, and quantification by cold vapor atomic fluorescence spectrometry (CVAFS), following established protocols.22,23 The mean daily detection limit, calculated as three times the standard deviation of the blanks, was 0.2 pM. The mean percent difference for analytical duplicates was 3.4 ± 2.0% (mean ± one standard deviation; n = 14). Certified reference material ORMS-4 (river water spiked with inorganic mercury) was analyzed (n = 2) giving values of 95 and 97 pM, reasonably close to the certified value of 109.7 ± 8.0 pM. MMHg concentrations were determined by distillation, aqueous phase ethylation, purge and trap, and quantification by CVAFS.24−26 Forty milliliter aliquots of sample were distilled in batch sets of 18 that included at least three quality control samples (i.e., distillation blanks and/or MMHg matrix spikes). The detection limit, calculated as three times the standard deviation of the distillation blanks, was 0.1 pM MMHg. The MMHg distillation recovery was calculated by spiking Milli-Q water with 20 pg of MMHg from both MMHgCl and MMHg-OH sources. The distillation recoveries were 92 ± 21% (n = 8) for MMHg-Cl and 76 ± 18% (n = 4) for MMHg-OH. The percent difference for analytical sample duplicates was 32 ± 15% (n = 4). To assess the potential for artifact MMHg production, two groundwater samples from Malibu Lagoon were spiked with inorganic Hg(II) to approximately ten times their HgT concentration prior to distillation. MMHg concentrations in the Hg(II) spiked samples were within 2−6% of the unspiked concentrations. Submarine groundwater discharge rates were calculated by modeling time-series excess radon (222Rn, t1/2 = 3.8 days) activities.27 Following Burnett et al.,8 a mass balance was established to constrain (1) atmospheric evasion, (2) lateral exchange, (3) possible riverine sources, (4) parent 226Ra activities, and (5) a representative groundwater endmember. While a series of regional groundwater wells were sampled for 222 Rn, the groundwater endmember 222Rn activity used in these SGD determinations was obtained from groundwater within the midberm piezometer (z ≈ 3 m) adjacent to Malibu Lagoon and sampled over a multiday time series (Figure 2). Due to high surf conditions during July 2009, time-series seawater 222 Rn activities from a November 9−11, 2009 sampling campaign were used to derive a closed berm SGD rate. The groundwater discharge rate was computed in terms of m3 per m2 per time interval. A more detailed discussion on the derivation of SGD rates is provided in Swarzenski et al.28 and summarized in Izbicki et al.29

Figure 3. Salinity (3a), temperature (3b), and pH (3c) time series data for seawater (SW; ●) and groundwater (GW; ▼) adjacent to Malibu Lagoon during a falling tide, July 2009. Groundwater samples were collected from a piezometer screened approximately 3 m below ground level.

salinity was approximately 33 between high and mid-tide, and then fell to about 25 during low tide. There was a corresponding increase in seawater temperature, from about 18.5 to 20 °C. Groundwater was fresher and warmer than seawater, with a salinity of about 23 and a temperature of 22 °C, suggesting that the observed changes in seawater salinity and temperature during low tide were caused by an increase in groundwater discharge (i.e., SGD). Our model of groundwater-seawater interaction during the course of the study is based on a consideration of our salinity and temperature data, previously described seasonal groundwater gradients,18 and an understanding of SGD drivers documented in other systems.5 During high tide, when our 1482

dx.doi.org/10.1021/es202783u | Environ. Sci. Technol. 2012, 46, 1480−1488

Environmental Science & Technology

Article

ten-hour sampling event began, the water level gradient was landward. As a result, SGD to the ocean was diminished and seawater recirculation within the shallow aquifer sediment was enhanced. Coastal groundwater therefore exhibited a strong seawater geochemical signal. As the tide fell and the water level gradient shifted toward the ocean, seawater was flushed from the shallow aquifer and the groundwater geochemical signal shifted toward the terrestrial endmember, resulting in a decrease in nearshore seawater salinity and an increase in seawater temperature. Simultaneously, greater SGD rates along the coast appeared to alter the geochemistry of nearshore seawater, as evidenced by the mercury speciation data described below. Total Mercury (HgT) Concentrations in Groundwater and Seawater. HgT concentrations in groundwater and seawater adjacent to Malibu Lagoon (Figure 4) were low relative to typical marine surface waters.4 Filtered HgT (F-HgT) concentrations in groundwater ranged from 1.1 to 1.9 pM and were generally constant throughout the tidal cycle. In contrast, unfiltered HgT (U-HgT) in groundwater decreased from 5.8 to 2.2 pM as the tide began to fall, and remained around 2 pM during low tide conditions. Remarkably, U-HgT concentrations in groundwater were almost always less than concentrations observed in seawater. Particle-bound mercury concentrations in groundwater (inferred from the difference between filtered and unfiltered mercury concentrations) were higher than filtered mercury concentrations, which is consistent with studies that show mercury has a high partition coefficient.1 Trends in U-HgT in seawater between high and mid-tide were similar to those observed in groundwater, with concentrations decreasing from about 5 to 3 pM (Figure 4). However, during low tide conditions, U-HgT in seawater was variable. F-HgT in seawater remained relatively constant as the tide fell, ranging from 0.5 to 1.2 pM, and then began to increase linearly from 0.5 to 3.6 pM during low tide when there was presumably enhanced groundwater discharge. The seawater FHgT concentration then dropped to 1.4 pM after the tide began to rise. The maximum HgT concentration in filtered seawater (3.6 pM) was similar unfiltered seawater (3.9 pM) collected during the same time interval. We attribute the similarity in U-HgT concentrations in groundwater and seawater during high to mid-tide (Figure 4) to the mixing of these water masses when seawater recirculated through the shallow subsurface sediment. As the tide began to fall, upgradient terrestrial groundwater with a lower HgT concentration moved through the aquifer, causing U-HgT in coastal groundwater and seawater to decline. The variability in seawater U-HgT during low tide may have been the result of particle resuspension brought on by south swell waves, which were the largest to reach this beach in over 10 years. The upgradient source of groundwater may have been from (1) the coastal aquifer, (2) Malibu Lagoon surface water flowing through the permeable sand berm, or (3) a combination of these sources. The unfiltered surface water sample collected from Malibu Lagoon during low tide had a UHgT concentration of 7.3 pM. While this concentration is more than three times higher than the average low tide groundwater U-HgT concentration (∼2.4 pM), mercury’s high particle reactivity could cause it to adsorb onto aquifer sediment as water moved through the subsurface. Therefore, we cannot discount Malibu Lagoon surface water as a potential source of HgT in groundwater.

Figure 4. Total mercury (HgT) (4a) and monomethylmercury (MMHg) (4b) concentrations in coastal seawater (SW; ●) and groundwater (GW; ▼) adjacent to Malibu Lagoon during a falling tide, July 2009. Groundwater samples were collected from a piezometer screened approximately 3 m below ground level. Paired F-MMHg and U-MMHg concentrations converge when the variability of replicate analyses (±32%) is applied (the average MMHg value is shown as a dashed line in Figure 4b). MMHg samples below the detection limit (