G7 (I) Environmental Biogeochemistry: Environmental Approaches

Monday, 25 July, 2011

MG7-O1 — 8:30-8:45
Authors: HARRIS, Reed C.1, RUDD, John W.M.2, KELLY, Carol A.3, KRABBENHOFT, David P.4, ST. LOUIS, Vince5, HINTELMANN, Holger6, GILMOUR, Cynthia C. 7, HEYES, Andrew8, AMYOT, Marc9, BRANFIREUN, Brian10, BLANCHFIELD, Paul11, GRAYDON, Jennifer5, PATERSON, Michael11, SANDILANDS, Ken11, TATE, Michael T.4, DIMOCK, Brian6, BEATY, Ken12, BABIARZ, Christopher13
(1)Reed Harris Environmental Ltd., reed@reed-harris.com; (2) R&K Research Inc.; (3) R&K Research, Inc.; (4) US Geological Survey; (5) University of Alberta; (6) Trent University; (7) Smithsonian Environmental Research Center; (8) University of Maryland Center for Environmental Science; (9) Universite de Montreal; (10) University of Western Ontario; (11) Fisheries and Oceans Canada; (12) Fisheries and Oceans Canada (retired); (13) University of Wisconsin.

An understanding of the relationship between mercury (Hg) loading to ecosystems and fish methylmercury (MeHg) concentrations is critically needed by decision makers to assess the potential benefits of Hg emission controls. The dose-response relationship at contamination levels associated with global changes in atmospheric anthropogenic Hg emissions has yet to be quantified. To better understand this relationship, a whole-ecosystem experiment, METAALICUS, is being conducted in northwestern Ontario, Canada. Three stable Hg isotopes were applied to a lake, its upland catchment and adjoining wetland from 2001-2007 at a rate of ~20 µg m-2 yr-1, six fold greater than ambient wet Hg deposition in this region. Results from the loading phase (2001-2007) showed distinct differences in the rates of response of various ecosystem compartments. Hg added to the lake surface began to appear in biota within months, while Hg added to the terrestrial system was exported to the lake very slowly (see companion presentations on lake and terrestrial system by C. Gilmour and D. Krabbenhoft respectively). Analogous to the loading phase, ecosystem compartments responded at very different rates when Hg additions stopped. Concentrations of inorganic 202Hg (lake spike) declined more than 90% in the water column within a year and were below detection limits by November 2009. Inorganic lake spike concentrations in sediments also declined in the first two years of recovery, although more slowly than in the water column (~20%). Lake spike MeHg concentrations in sediments dropped more rapidly, particularly in porewater, which declined by more than half in the first two years of recovery. In contrast, upland spike concentrations in runoff increased through the first year of recovery and showed a modest decline in the second year. Wetland spike concentrations in runoff and porewater have revealed steadily increasing concentrations for two years following the cessation of Hg additions. The response of small-bodied fish in the recovery phase mirrors that of the loading phase, with rapid declines in the proportion of MeHg derived from lake spike. Large-bodied fish are showing only a minimal reduction in lake spike MeHg derived from the loading period. Overall the initial recovery period of METAALICUS indicates that fish Hg concentrations in ecosystems will exhibit fast and slow phases of response following reductions in atmospheric Hg deposition.

MG7-O2 — 8:45-9:00
Authors: TATE, Michael1, SABIN, Thomas1, DEWILD, John1, ST. LOUIS, Vince2, GRAYDON, Jennifer2, BRANFIREUN, Brian3, HARRIS, Reed4, HEYES, Andrew5, LINDBERG, Steve6, SOUTHWORTH, George6
(1)U.S. Geological Survey, mttate@usgs.gov; (2) University of Alberta; (3) University of Western Ontario; (4) Reed Harris Environmental Ltd.; (5) University of Maryland; (6) ORNL (retired);

The METAALICUS project is a whole-ecosystem experiment specifically designed to quantify the magnitude and timing of the watershed response to a change in mercury (Hg) loading. A decade ago, when the US and many other countries began to consider Hg-specific emission regulations, many doubted the effectiveness of such actions since soils and sediments were already ubiquitously contaminated, potentially fueling this contamination problem indefinitely. To address this concern, a multi-national team of scientists was formed to devise a whole-ecosystem, Hg-dosing study, whereby Hg would be deliberately added to an entire watershed – The Mercury Experiment to Assess Atmospheric Loadings in Canada and the US (METAALICUS) project, conducted at the Experimental Lakes Area (ELA), northwestern Ontario, Canada. Whole-ecosystem manipulation studies have a distinct advantage over small-scale (lab scale) studies, in that all the ecosystem components, natural processes and complexities within watersheds are accounted for.

This paper focuses on the terrestrial (upland forest and wetland) aspects of METAALICUS project. The lake, forests and wetland were dosed with different stable Hg isotopes, which allows for discrimination of the experimentally administered dose versus ambient Hg brought to the watershed via atmospheric deposition . Mercury additions to the terrestrial components of the watershed were initiated in 2001 and continued through 2006, at a rate of about 5x measured current atmospheric deposition. Since this time, METAALICUS has become a mercury reduction experiment.

During the loading phase of the project, the majority of the terrestrially applied Hg isotopes were distributed approximately equally among three major compartments: soils, plants (tree canopy and ground vegetation), and losses to emissions from soil and plant surfaces. Much less (about 1%) isotope was measured in runoff that flowed into the downstream lake. With each successive annual dose the isotope pool in soils steadily increased, while the forest canopy and emission fluxes pools had comparatively low carryover from year to year. Within two years of loading cessation, isotope concentrations in canopy and emission fluxes were negligible, and about half of the total isotope load was found in soils. Isotope concentrations in runoff gradually increased during the loading phase, and continued to increase for 1-2 years after loading ceased, suggesting significant translocation from compartments above the forest floor. Three years subsequent to cessation, isotope concentrations in soils have remained steady, and generally reflect the Hg abundance of the overall soil pool. Models calibrated to these results will be very useful for scenario testing to evaluate long-term recovery of lakes with significant watershed Hg contributions.

MG7-O3 — 9:00-9:15
Authors: GILMOUR, Cynthia1, REED, Harris2, KELLY, Carol A. Kelly3, HINTELMANN, Holger4, KRABBENHOFT, David P.5, AMYOT, Marc6, BLANCHFIELD, Paul 7, PATERSON, Michael7, RUDD, John M.W.3, TATE, Michael5, SANDILANDS, Ken7, BEATY, Ken7, LINDBERG, Steven8, SOUTHWORTH, George8, HEYES, Andrew9, ST. LOUIS, Vince10, GRAYDON, Jenny10, BABIARZ, Chris11, BRANFIREUN, Brian12, HURLEY, James P.11
(1) Smithsonian Environmental Research Center, gilmourc@si.edu; (2) Reed Harris Environmental Ltd.; (3) R&K Research, Inc.; (4) Trent U.; (5) USGS, Middleton, WI; (6) U. Montreal; (7) Canada DFO; (8) ORNL; (9) University of Maryland, Chesapeake Biological Lab; (10) U. Alberta; (11) U. Wisconsin; (12) Western Ontario;

The METAALICUS study is a whole-watershed mercury addition study designed to directly examine the relationship between Hg loading to ecosystems and MeHg bioaccumulation in food webs, at Hg loading rates representative of atmospheric deposition. Between 2001 and 2007, Lake 658 and its watershed (Experimental Lakes Area, Ontario) were loaded with enriched stable mercury isotopes, to roughly 6 times Hg wet deposition. Here we present the final budgets for Hg added directly to the surface of L658 (termed “spike”) between 2001 and 2007.

As has been presented previously, MeHg production andbioaccumulation responded rapidly to a change in Hg loading to L658. The 42 hectare L658 watershed contributes most of the natural Hg load to 8 hectare L658, mainly as complexes with dissolved organic matter. Addition of lake spike Hg roughly doubled Hg concentrations in lake water during the spike period, which was the stratified period during summer and fall. Roughly half of the Hg spike we added to L658 was lost to evasion, linked to the fact that a large fraction of the new spike remains in the epilimnion for most of the summer. Most of the remaining spike was delivered to sediments in this lake with an average water retention time of 5 years.

The powerful isotope data suggest that most of the MeHg accumulated in the lake’s lower food web is produced in sediments and anoxic bottom waters within the lake. Spike Hg that accumulated each year in bottom waters and led to MeHg production appeared to be on average less than a few years old. Spike Hg delivered to surface sediments became less available for methylation over the same time frame. Spike MeHg production resulted in the characteristic pattern of bioaccumulation in water and biomagnification throughout the trophic levels leading to fish. By 2007, the total Hg concentration in prey and predatory fish species had increased by 50% and 40%, respectively, due to lake spike loading, roughly mimicking the change in summer and fall water column and lower food-web MeHg concentrations.

To summarize, we conclude that Hg entering L658 contributes substantively to MeHg production and bioaccumulation for only a few years. Thus the concentration of MeHg in lakes should respond on the same time scale as the rate of change of Hg entering the lake. The time to achieve benefits will depend on ecosystem conditions that control how quickly loading declines following a change in Hg deposition.

MG7-O4 — 9:15-9:30
Authors: COLEMAN WASIK, Jill K.1, ENGSTROM, Daniel R.2, MITCHELL, Carl P.J. 3, SWAIN, Edward B. 4, MONSON, Bruce A.4, BALOGH, Steven J. 5, JEREMIASON, Jeff D. 6, KOLKA, Randall K. 7, BRANFIREUN, Brian A.8, ALMENDINGER, James E. 2
(1) St. Croix Watershed Research Station/University of Minnesota, jcoleman@smm.org; (2) St. Croix Watershed Research Station; (3) University of Toronto – Scarborough; (4) Minnesota Pollution Control Agency; (5) Metropolitan Council Environmental Services; (6) Gustavus Adolphus College; (7) USDA Forest Service-Northern Research Station; (8) University of Western Ontario;

Atmospheric sulfate deposition resulting from the burning of fossil fuels is cited as a possible contributor to elevated fish mercury concentrations in lakes from acid-sensitive regions of North America and Europe. Wetlands are important sources of methylmercury to lakes and streams, and many studies have demonstrated increased methylmercury production as a result of experimental sulfate addition, particularly in sulfur-limited wetlands. However, most such studies have been generally short-lived and none have monitored mercury dynamics once sulfate addition ceased. The implementation of the 1990 Clean Air Act Amendments resulted in significant reductions of atmospheric sulfate deposition to the northeastern United States, but ecosystems in acid-impacted regions have been slow to recover. Because the relationship between methylmercury production and sulfate was discovered well after the acidifying effects of sulfate raised concerns, no data exist that allow for direct comparison of methylmercury production before and after sulfate deposition peaked.

These issues were addressed in an ecosystem-scale experiment in which sulfate loads were elevated through simulated rainfall to half of a 2-ha peatland in northern Minnesota. Between 2001 and 2008 sulfate loading to the experimental treatment was increased by four times annual ambient wet deposition. Then in the spring of 2006 sulfate addition was halted in a portion of the experimental treatment to create a new recovery treatment. Methylmercury concentrations in recovery porewaters were initially elevated relative to those in the control treatment but declined to control concentrations by mid-summer of 2008. In contrast recovery treatment sulfate concentrations returned to control levels within the first year. Peat cores collected from the recovery treatment in 2009 also showed solid-phase methylmercury concentrations returning to background levels after three years. This work demonstrates that the effects of sulfate addition on methylmercury production may be relatively short-lived due to sequestration of sulfate in recalcitrant organic sulfur fractions. These results also suggest that further controls on atmospheric sulfur emissions and deposition could lead to rapid and significant reductions in wetland methylmercury production with possible consequences for mercury accumulation in aquatic foodchains.

MG7-O5 — 9:30-9:45
Authors: ENGSTROM, Daniel R.1, COLEMAN WASIK, Jill2, SWAIN, Edward B.3, MONSON, Bruce A.3, MITCHELL, Carl P. J.4, ALMENDINGER, James E.5, BALOGH, Steven J.6, BRANFIREUN, Brian A.7, KOLKA, Randy K.8, JEREMIASON, Jeff D.9
(1) Science Museum of Minnesota, dre@umn.edu; (2) University of Minnesota; (3) Minnesota Pollution Control Agency; (4) University of Toronto - Scarborough; (5) Science Musuem of Minnesota; (6) Metropolitan Council Environmental Services; (7) University of Western Ontario; (8) USDA Forest Service; (9) Gustavus Adolphus College.

Boreal wetlands are important sites for mercury methylation by sulfate-reducing bacteria and are widely thought to be sulfate-limited based on laboratory and field-mesocosm experiments. These observations have been reinforced by an ecosystem-scale study in which sulfate deposition to a boreal wetland was increased in order to examine the effects of atmospheric sulfate on production and export of methylmercury. A whole-wetland irrigation system for amending sulfate through simulated rainfall was established in the fall of 2001 at the Marcell Experimental Forest in northern Minnesota. The 2-ha wetland was divided into control and experimental halves, and the experimental half was treated with a sulfate solution meant to increase annual sulfate loading by four times the current rate of atmospheric deposition. Beginning in 2006 sulfate application was suspended on a portion of the experimental half to evaluate ecosystem response to a reduction in sulfate loading. We focus here on the final three years of the study (2006-2008) and address in particular the effects of water-table fluctuations on mercury methylation in a wetland enriched by elevated sulfate deposition. Pore-water MeHg and sulfate concentrations show very clear trends that reflect both recovery of the peatland from chronic sulfate enrichment and the added effect of water-table drawdown and peat oxidation during several years of recurring mid-summer droughts. Extended periods of pre-addition sampling provide strong evidence for sulfate regeneration by oxidation of organic-sulfur stores in the peat during periods of water-table depression and subsequent release to pore-waters during rebound. The released sulfate appeared to stimulate mercury methylation, raising pore-water MeHg concentrations in a manner similar to that following experimental sulfate addition. This natural methylation response was observed in control and recovery treatments as well as in the experimental section, although the magnitude of the MeHg increases were significantly lower in both. Experimental sulfate additions following prolonged water-table fluctuations, especially those in fall, produced only a muted response in MeHg levels, presumably because sulfate was not limiting microbial methylation at these times. The role of hydrologic fluctuations in stimulating sulfate regeneration and MeHg production was confirmed by water-table manipulations in experimental mesocosms. Climate change models for this region predict an increase in precipitation variability and intensity, which in addition to stimulating mercury methylation, could also enhance hydrologic connectivity within and from wetlands and thereby increase MeHg flux to downstream lakes.

MG7-O6 — 9:45-10:00
(1) SCHOOL OF ENVIRONMENTAL STUDIES, das.reshmi@gmail.com; (2) University of South Carolina;

Salt marshes are an important sea - land interface, and a source of nutrients to coastal waters, but little is known about the mercury cycling in these environments. Here we report on the mercury concentration and isotopic composition of surface and cored sediments from the salt marsh and adjoining seaward sand dune from Cabretta Island, Georgia, a Holocene barrier island. Core TT-4 from the sand dune is composed of fine sand. Core TT-2 from the salt marsh is composed of interbedded sands and clays. Mercury isotopes were determined on a NEPTUNE MC-ICPMS using the Tl addition technique at USC. Total Hg concentrations are lower in the sand dune than in the marsh core sediments (TT-4~ 3-6 ppb Hg and TT-2 ~ 3.4-15.7 ppb Hg, respectively). d202Hg values range from -0.74‰ to -2.03‰ in the dune and -1.33‰ to -2.19‰ in the marsh. Sediments from both cores show significant MIF with ?199Hg ranging between +0.46‰ to -0.51‰, and ?199Hg broadly correlates with lithology and mercury concentration. Well sorted sand layers have lower Hg concentrations and zero to slightly negative ?199Hg, whereas clay-rich layers have higher Hg concentrations and positive ?199Hg values. This pattern matches with surface sediment compositions, where dune sands have zero to slightly negative ?199Hg (-0.07‰ to -0.15‰) and the marsh clay-rich surface has significantly positive ?199Hg (0.48‰ to 0.79‰). ?199Hg and ?201Hg of a Spanish moss sample collected from the beach front was -0.06‰ and -0.15‰ which is possibly the local atmospheric composition of mercury. The significant ?199Hg difference between the dune and marsh surface and core sediments rules out atmospheric deposition as the only source of Hg in this environment. We suggest that for the marsh a significant proportion of Hg is adsorbed from the seawater enriched in 199Hg and 201Hg due to photoreduction of Hg+2 and MeHg. In contrast the ombrotrophic sand dune ?199Hg and ?201Hg composition matches that of the Spanish moss in the area, suggesting slightly negative ?199Hg and ?201Hg for the local atmosphere. Our data shows that Hg isotopes can delineate the different sources of Hg in a pristine coastal environment.

MG7-O7 — 10:00-10:15
(1) Universidad de Castilla-La Mancha, rosacarmen.rodriguez@uclm.es; (2) Universidad Castilla-La Mancha;

The biogeochemical cycling of mercury (Hg) has received considerable attention because of the toxicity of methylmercury (MeHg), its accumulation in biota and its biomagnification along food chains. Earthworms constitute a major part of the animal biomass in terrestrial ecosystems and are considered suitable for studying and monitoring Hg bioaccumulation in the soil ecosystems. Little is known about the toxicity of Hg to terrestrial invertebrates and information on the effects of speciation on Hg uptake will be of interest for assessing the risk associated with Hg pollution in terrestrial ecosystems. In recent years, many studies have reported that Hg species can be easily assimilated by earthworms. However, it is still unclear whether MeHg can be directly absorbed from soil by earthworms or it is transformed by inorganic Hg (IHg) in earthworm bodies.

Thus, the aim of this work has been to use the potential of species-specific Hg enriched stable isotopes to assess Hg methylation and demethylation pathways in earthworms exposed to historically Hg contaminated soils spiked with isotopically enriched inorganic Hg (199IHg) during a 28-day uptake period followed by a 14-day depuration period. Hg species in earthworms and soils were simultaneously extracted using a closed-vessel microwave system and analysed after ethylation by species-specific-isotope dilution and gas chromatography coupled to inductively coupled plasma mass spectrometry (GC-ICP-MS).

The results indicated that Hg bioavailability in soil had important influence on Hg uptake and methylation mechanisms. Thus, kinetics controlling MeHg uptakes were dependent on Hg bioavailability more than Hg concentrations in soils. Furthermore, these experiments showed different methylation/demethylation pathways during the depuration period for Hg species coming for natural or isotopically enriched IHg. Therefore, these observations can be related to the different mechanism involved in methylation of IHg to MeHg such as methylation via bacteria in the digestive tract of the earthworm or methylation induced by the microorganisms in the soil.

MG7-O8 — 10:15-10:30
Authors: TSUI, Martin Tsz Ki1, FINLAY, Jacques C.2, BLUM, Joel D.1
(1) Department of Geological Sciences, University of Michigan, mtktsui@umich.edu; (2) Department of Ecology, Evolution & Behavior, University of Minnesota;

In stream networks, differences in sunlight penetration through the canopy cover drive spatial variation in stream productivity, leading to generally increasing primary productivity downstream. In the northern California Coast Range, we examined mercury (Hg) cycling during the summer baseflow period over multiple years in a stream network with the overall goal of studying the mediation of Hg cycling by in-stream processes. Along a stream size gradient (drainage area from 0.5-150 km2), we observed that methylmercury (MeHg) concentration in water and biota increased, in parallel to increasing primary productivity. Within a productive channel, we observed substantial in-stream production of MeHg leading to relatively high levels of MeHg in the resident food webs. Because of the higher sunlight availability downstream we hypothesize that there may also be enhanced photochemical transformations of inorganic Hg and MeHg occurring downstream. It has recently been demonstrated that photo-reduction of inorganic Hg and photo-decomposition of MeHg produce Hg with mass independent fractionation (MIF) of odd Hg isotopes (i.e., 199Hg and 201Hg), but this does not occur through dark abiotic or microbial Hg transformations. Therefore, to test whether photochemical reactions are becoming increasingly important along this stream size gradient in the stream network we are measuring the stable Hg isotopic composition of the stream biota along this gradient. Biota are good monitors of Hg isotopic composition because they integrate MeHg from the ambient environments and can provide adequate amounts of Hg for isotopic analysis. Preliminary analyses of archived invertebrate samples demonstrated significantly higher MIF signals of Hg isotopes in two functional feeding groups of insect larvae (?201Hg = 0.29 ‰ higher for filter-feeding hydropsychid caddisflies and ?201Hg = 0.38 ‰ higher for predatory perlid stoneflies) in downstream channels (drainage area > 100 km2) compared to upstream tributaries (drainage area < 17 km2). We will present all results in the context of Hg cycling in the watershed as well as the importance of stream network position (or stream size) in determining relative rates of in-stream Hg transformations. Overall, our studies in this stream network provide new insight into how spatial variability and in-stream processes in a forested watershed may mediate Hg transformation and bioaccumulation. These experimental observations complement other stream studies that most often focus on upstream sources of MeHg such as connected wetlands in the watershed.

Monday, 25 July, 2011