B14B-01 INVITED
Estimating Ecosystem Responses to Changes in Mercury Loading: Lessons From the METAALICUS Project
The Mercury Experiment to Assess Atmospheric Loadings in Canada and the US (METAALICUS) project is a whole-ecosystem, mercury (Hg) loading experiment specifically designed to examine the relation between atmospheric mercury deposition and fish Hg concentrations. This project was prompted by the observation that we lacked clear evidence whether a changes atmospheric Hg deposition might lead to a corresponding change in fish Hg, and at what time scales. To address this information need, a multi-national team of scientists was formed to devise a whole-ecosystem, Hg-dosing study, whereby mercury would be deliberately added to an entire watershed. The study is being conducted at the Experimental Lakes Area (ELA), which is located in northwestern Ontario, Canada. Whole-ecosystem manipulation studies have a distinct advantage over small-scale (lab scale) studies, in that natural processes and complexities that are present in watersheds are accounted for in the scientific results. Starting in the spring of 2001, the METAALICUS team been dosing the entire Lake 658 watershed with about 20 ug/m2/y (about 4-5 times the current ambient load). However, the applied Hg is in the form of enriched stable isotopes that can be analytically distinguished from previously existing ambient Hg, or currently depositing Hg. Thus, using the applied isotope as a tracer allows for improved insights into process rates occurring in watersheds that have not been possible heretofore. One of the greatest areas of uncertainty for making reliable predictions of the environmental response to changes in atmospheric Hg deposition is quantifying the flux of Hg from terrestrial systems to downstream aquatic ecosystems. This is especially problematic for settings where terrestrial inputs are similar in scale to atmospheric deposition or may exceed it. In these cases, it is critical to understand the details of how Hg is delivered from watersheds and the relative bioavailability of this Hg compared to that from atmospheric deposition. With this information in hand, improved predictive capability of response times of Lake 658 (and other similar lakes) can be achieved. Net accumulation of the applied isotope is reflected in the steadily increasing isotope concentration in terrestrial soils with each year of application (2001- 2004). However, the total soil-mass accumulation of the Hg isotope represents only about a quarter of that which has been applied. The remaining isotope mass is primarily accounted for by reemission from soils and plants, interception and storage in the forest canopy, and runoff. The concentration and overall flux rate of the isotope in runoff has increased about linearly with each year of application, but runoff mass fluxes are small compared to the isotope application rate (about 1-3 percent). By calibrating a simple numerical soil-box model, we can estimate the response time of Lake 658 to changes in Hg load, and if the processes operating in this basin are applicable elsewhere the model will have utility for estimating response times for a wider range of watershed conditions.
B14B-02 INVITED
Biogeochemical Controls on Mercury Methylation: A compilation of Data Across Fresh and Saltwater Wetlands
Over the past decade, we have examined the biogeochemical controls on net methylmercury production across a number of wetland ecosystems, including salt marshes in Chesapeake Bay, the freshwater and estuarine Everglades, and a variety of boreal freshwater wetlands in Ontario. The balance between sulfate and sulfide is key for understanding Hg methylation rates among these ecosystems. Sulfate stimulates Hg-methylating sulfate- reducing bacteria (SRB) while sulfide creates charged mercury-sulfide complexes that are unavailable for uptake by SRB. Sulfate-stimulation of methylation has been demonstrated in experimental studies that range from pure culture, to sediment and soil amendments, to large-scale field additions. Stimulation of methylation by sulfate has also been demonstrated in freshwater ecosystems impacted by sulfur pollution derived from atmospheric deposition, agriculture and mining. This presentation will present a compilation of field and laboratory studies on the impact of sulfate and sulfide on MeHg production to create a simple, general model for the control of net Hg methylation in surfaces sediments and wetland soils that includes microbial activity (sulfate reduction rate), dissolved sulfide, dissolved organic matter and and soil organic matter. In particular, the model focuses on the balance between sulfate and sulfide, and the optimal concentrations of each for methylation across studies and ecosystems. Data to be presented will include new information from high sulfate and sulfide coastal ecosystems in Chesapeake Bay. Optimal sulfate concentrations for methylation appear to range widely among ecosystems, while the optimal sulfide concentrations are more constant, and often quite low, often in the low micromolar range. However, recent data from estuarine and marine systems suggest that net methylation can proceed at somewhat higher sulfide concentrations when microbial activity is particularly high. By compiling these data, we can begin to predict the magnitude of net MeHg production across different wetland types, the sensitivity of different types of wetlands to mercury inputs, the role of wetlands in MeHg budgets for different types of aquatic ecosystems, and impact of wetland reconstructions and mitigations on MeHg budgets.
B14B-03
Mercury Cycling in Agricultural and Non-agricultural Wetlands of the Yolo Bypass Wildlife Area, California: Sediment Biogeochemistry
The Yolo Bypass Wildlife Area (YBWA) is part of the larger Yolo Bypass floodwater protection zone associated with the Sacramento River and the Sacramento–San Joaquin Delta, California. Land use in the YBWA consists of white and wild rice fields, seasonally flooded fallow agricultural fields, and permanently and seasonally flooded non-agricultural wetlands used for resident and migratory waterfowl. A recent assessment of mercury (Hg) and methylmercury (MeHg) loads indicates that the Yolo Bypass is responsible for a high proportion of the aqueous MeHg entering the Delta, and that biota from the Yolo Bypass are considerably elevated in MeHg. The current study examines benthic MeHg production and biogeochemical controls on this process, as a function of YBWA land use, wetland management, and agricultural practices during the 2007 rice growing season (June to October). Preliminary results indicate that in the week following initial flooding of agricultural fields, prior to the establishment of rice plants, the microbial community in the 0–2 cm surface sediment zone exhibited very little potential Hg(II)-methylation activity compared to the permanent wetland habitat (as assessed via the 203Hg(II)- methylation assay). Approximately 1 month after flooding, rice plants were established and the activity of the resident Hg(II)-methylating microbial community had increased substantially in all agricultural fields, although the observed rates of MeHg production were still much lower than those observed in the permanent wetland setting. Ongoing field sampling includes analysis of reactive Hg(II) in sediments and of iron and sulfur redox species in sediments and pore waters.
B14B-04
Direct and Indirect Effects of Vegetation on Methylmercury Production in Wetlands as Assessed by Experimental Plant Removal
Although vegetated wetlands are among the most active habitats for microbial methylmercury (MeHg) production, the relative influence of wetland vegetation itself is poorly understood. Plant physiology and biomass (live and dead) can modify both microbial populations and inorganic mercury (Hg(II)) bioavailability through a number of soil, water and atmospheric interactions. Alternatively, plant activity and structure can be simply a response to geochemical conditions that also favor Hg(II)-methylation. Linked studies within the San Francisco Bay watershed have demonstrated that habitat-specific biogeochemical characteristics are the dominant factors controlling MeHg production, and that differences in wetland plant biomass, root density, decomposition rates, can directly influence sediment mercury cycling. A vegetated:de-vegetated paired plot approach was used to directly assess the influence of live plant activities on surface sediment mercury dynamics and associated biogeochemistry in differing wetland settings: salt marshes, permanent and seasonal freshwater wetlands, a freshwater floodplain, and agricultural rice fields. Although results from several of these subhabitats are pending, the data thus far have illustrated linkages between wetland plants and microbial Hg(II)-methylation. De- vegetation strongly influenced sediment biogeochemistry (e.g. redox, dissolved organic content, and reduced sulfur pools) in high interior pickleweed (Sarcocornia pacifica) dominated saltmarshes, where the high rates of MeHg production (up to 1 ng g-1dry sed d-1) observed in vegetated plots were reduced to <10 pg g-1dry sed d-1 in de-vegetated plots. Further, plant root densities were positively correlated with the activity of Hg(II)-methylating bacteria in these interior saltmarsh settings. The pool size of mercury available for methylation ("reactive mercury") was not measurably influenced by this short-term de-vegetation experiment, but across field studies, rhizosphere biomass was often negatively correlated with reactive mercury concentration due to a corresponding increase in solid-phase reduced-sulfur compounds associated with this zone. Because mercury methylation is controlled by both the reactive mercury pool size and the microbial Hg(II)-methylation activity, the direct influence of wetland plants on both of these terms can be profound and reflect multiple, and potentially contrasting, mercury cycling pathways. Experimental field manipulations, in conjunction with comparative habitat and process studies, represent essential tools to elucidate the influence of wetland plant communities on Hg cycling.
B14B-05
Mercury Emission From Plants Depends on Reduction by Ascorbate
The importance of vegetation for the ecological Hg cycle has been recognized recently. One step in this cycle is the poorly understood phytogenic reduction of dissolved Hg(II) to volatile Hg(0) which had initially been reported for common reed growing on Hg-contaminated sediments. The hitherto unknown mechanism of this reduction was the objective of our investigations. Young barley and European-beech plants were cultivated for 24 h and 2 days, respectively, on a sterile hydroponic medium containing 20-40 µM HgCl2. Within 10 min after seclusion in a closed exposure system, the Hg(0) emission from the encapsulated aerial part of the plants reached 10 times the control value in a plant-free system and was proportional to the Hg(II) concentration in the medium. At 20 µM Hg(II) in the medium, a flux of 12.8 µg Hg(0)/m2/h was estimated for beech leaves. The phytogenic Hg(II) reduction was further examined by addition of powderized homogenates from deep-frozen leaves (barley, beech, Arabidopsis thaliana) or from needles (Norway spruce) to solutions of 1-5 µM Hg(II). These samples consistently produced a strong transient Hg(0) release at neutral pH that was even reinforced in alkaline medium and vanished at acidic pH. The very same pH dependence was observed after addition of pure L(+)-ascorbate (AA) instead of plant material to the HgCl2 solutions, whereas the reductants NADPH and GSH produced only little or no Hg(0), respectively. At neutral and alkaline pH, the Hg(II)-reducing capacity of spruce needle homogenates was 2 - 4 times that of beech leaves, which paralleled a 6-fold difference in AA concentrations. Homogenates from whole wildtype-plants of Arabidopsis reduced 8-times more Hg(II) than those from the AA-deficient mutant vtc1-1 (AA concentration 30% of wild type). A comparison of literature data on AA concentrations revealed for wetland plants a range from 0.3 µmol/g DW (Phragmites communis) over 15.0 (Typha latifolia) to < 34.1 (Spartina altiflora), and for trees between 2 and 18 for beech leaves and spruce needles, respectively. CONCLUSIONS The comprehensive model of metabolic mercury volatilization by plants is based on the transpiration flow carrying Hg(II) ions to the leaves for phytogenic reduction in the apoplastic space from where Hg(0) diffuses outward under stomatal control. The reductive regeneration of ascorbate proceeds stepwise in the so called antioxidative defense pathway of plants. Considering the ubiquitous presence of AA in plants, the findings contribute to the basic understanding of the soil-plant-air mercury exchange. Furthermeore, our data support the preliminary conclusion that the mercury reducing capacity of wetland plants is not inferior to that of land-bound vegetation.
B14B-06
The Influence of Wetland Cover and Dissolved Organic Carbon on Mercury Export in Forest Landscapes, Northeastern USA
Mercury (Hg) contamination is widespread in remote areas of the northeastern USA. Atmospheric Hg is deposited on terrestrial uplands and subsequently mobilized to downstream aquatic ecosystems. We are investigating the fate of Hg deposited in forested watersheds by quantifying stream transport of Hg, and the interactions with dissolved and particulate organic matter. We hypothesize that the landscape characteristics controlling the production and mobility of organic matter will likewise control the mobility of Hg. This research was conducted at three sites in the Northeast that represent a range of hydrochemical conditions and span a range of wetland cover. Most stream export of Hg occurs at high flow; therefore we collected samples during snowmelt and storms. Mercury concentrations increase with discharge at all three sites; however the partitioning of Hg fractions (dissolved vs. particulate) differs among sites during high flow events. At the Hubbard Brook Experimental Forest, NH (watershed 6), there are no true wetlands and dissolved organic carbon (DOC) and total Hg (THg) concentrations, and suspended sediment concentration (SSC) (mean DOC = 3.1 mg C L-1; THg = 1.5 ng L-1; SSC < 50 mg L-1) are low even during the highest of flow events. At Sleepers River, VT (watershed 9), SSC can be elevated during events (SSC > 500 mg L-1), consequently the particulate Hg fraction can range as high as 95% of the THg concentration (mean particulate Hg concentration = 10.2 ng L- 1). At Archer Creek (Huntington Forest, NY), which has the greatest percent wetland cover (10%) of our three sites, DOC concentrations are high (mean DOC = 7.5 mg C L-1), while SSC are low (SSC < 10 mg L- 1). At Archer Creek, Hg is largely in the dissolved form (~75% of Hg) and strongly correlated with DOC (r2 = 0.90). The hydrophobic organic acid (HPOA) fraction of DOC is most effective at mobilizing Hg and is strongly correlated to Hg concentrations in stream water. This research suggests that high concentrations of Hg can be present in stream water from forest watersheds during high flow events, and that wetlands and suspended sediment favor Hg export. However, it is not clear to what degree this Hg is bioavailable for potential uptake by organisms in aquatic ecosystems following mobilization.
B14B-07
The Influence of Hurricanes and Other Biogeochemical Factors on Net Mercury Methylation and Mercury Cycling in the Gulf of Mexico
Methylation of mercury (Hg) in coastal region sediments is a potentially important source of methylmercury (MeHg) to ocean food webs. Sediment Hg studies in the coastal zone have focused mostly on biogeochemical relationships but have not studied in detail the impact of extreme events, such as hurricanes, on Hg dynamics. As a result of two funded studies, samples were collected at a number of stations in July and October 2005 and March and July 2006 in the Gulf region, and covering locations impacted by both Hurricanes Katrina and Rita. Some locations are also impacted by seasonal hypoxia. Sediment cores were analyzed for total Hg, MeHg, and ancillary variables, and were also dated using various proxies. Rates of methylation (km) and demethylation (kd) were also estimated using stable isotopes of Hg and MeHg. Typical relationships were found for bulk sediment parameters, and for dissolved-particulate partitioning and these relationships will be discussed. The impact of the hurricane disturbance was clearly evident from large differences before and after in the sediment profiles of Hg, MeHg, sediment organic matter content (TOC) and other bulk elemental concentrations, as measured by XRF. Additionally, measurements of radiotracer distributions (7Be, 210Pb), from this and other studies confirm the substantial redistribution of sediments as a result of the hurricanes. Hurricane disturbances appear to have stimulated net Hg methylation by increasing the methylation rate and depressing the demethylation rate, and as evidenced by the higher %MeHg, at the highly disturbed sites. Our results also suggest there has been a further redistribution of sediments post-hurricanes. In total, we estimate that the amount of Hg moved and redistributed as a result of the hurricanes was substantially greater than the yearly input of Hg from the Mississippi/Atchalfalaya rivers and the atmosphere. This redistribution suggests that historically deposited Hg needs to be considered when evaluating inputs of Hg in terms of the potential availability of Hg within the ecosystem to Hg methylation. Overall, large physical disturbances have the potential to be important drivers of net Hg methylation, likely through providing fresh organic matter to the methylating sulfate reducing organisms, through the overall oxygenation of the sediment column, and by redistributing buried Hg into bioactive sediment layers. The implications of these results will be discussed.