G7 (III) Environmental Biogeochemistry: Environmental Approaches

Tuesday, 26 July, 2011

TG7-O1 — 8:30-8:45
Authors: CORBITT, Elizabeth S1, SMITH-DOWNEY, Nicole V2, JACOB, Daniel J1, CAROUGE, Claire C1, HOLMES, Chris D3, HOROWITZ, Hannah M1, AMOS, Helen M1, SUNDERLAND, Elsie M4
(1) Harvard University, corbitt@seas.harvard.edu; (2) Earth System Sciences, LLC; (3) University of California, Irvine; (4) Harvard University School of Public Health.

Terrestrial ecosystems represent the largest global biogeochemical mercury surface reservoir with 106 Mg contained in organic soils. Human activities have increased the atmospheric burden and deposition of mercury by at least threefold. Evasion of mercury from terrestrial soils exerts a major influence on atmospheric concentrations. To investigate the impacts of anthropogenic mercury on soil mercury reservoirs, terrestrial-atmospheric exchange processes, and the role of climate change and fire on terrestrial mercury evasion, we have developed a mechanistic framework for mercury cycling in soils and vegetation within the GEOS-Chem global biogeochemical model for mercury. We build on the approach of tracking mercury cycling in association with organic carbon pools in the Global Terrestrial Mercury Model (GTMM) developed by our group in previous work (Smith-Downey et al. 2010). Results from offline GTMM simulations show that the most labile soil carbon pools are most enriched in anthropogenic mercury and that soil emissions have tripled since 1840.

To enable full coupling of GEOS-Chem with GTMM we have updated the model’s dry deposition scheme for mercury including the mechanistic basis for stomatal and cuticular uptake of mercury. We have used recent studies to better constrain rates for redox reactions on surface vegetation and in soil pore waters and to parameterize throughfall of mercury and accumulation in litter. We show results of updated model simulations compared to observed mercury and carbon concentrations in different soil and ecosystem types and compare modeled fluxes to available dry deposition, throughfall, and washoff measurements. We compare results to observed seasonal patterns in the accumulation of mercury in leaves and litter and variability of atmospheric mercury near forest canopies. We analyze the interannual variability in atmosphere-terrestrial exchange by running the model with meteorological inputs from 1979 to present. Finally, we present preliminary results on long term trends including the accumulation of historical anthropogenic mercury in soils and sensitivity to future changes in emissions and climate.

TG7-O2 — 8:45-9:00
Authors: BRAATEN, Hans Fredrik Veiteberg1, RANNEKLEV, Sissel Brit1, LARSSEN, Thorjørn1
(1) Norwegian Institute for Water Research (NIVA), hbr@niva.no

Results from a recent survey show increasing levels of mercury in fresh water fish in Norway. As a result of European emission reductions of mercury, atmospheric deposition of mercury has decreased in the same time period. Hence, explanations for increased levels of mercury in fish are most likely related to processes in the lake or catchment. Sediments are important spots for microbial methylation, and the biogeochemical cycling of mercury in sediments is substantial for the understanding of how mercury transfers to water and biota.

In order to investigate the role of methylation and de-methylation in the sediments of typical Norwegian dystrophic small lakes, sediment cores and surface sediments were collected from the Lake Langtjern research site. The lake is located in southeast Norway and extensive work has previously been done to establish input and output fluxes of mercury at the site. Concentrations of total mercury (TotHg) and methylmercury (MeHg) were determined in the sediments, and redox potential and TOC content were measured. Samples were taken seasonally (autumn, winter, spring) and at different depths (1-9 metres) and distance (0-150 metres) from the typically peaty shores of the lake.

Continuous measurements of oxygen in the lake’s water column indicate a dynamic redox system. The system has highly oxic conditions throughout the water column during late autumn months and anoxic conditions in the deeper water in the winter. In addition, there is considerable variation in the redox potential in the surface sediments, with increasing oxic conditions with distance from the shore. The most anoxic conditions are found close to the peat bog shoreline. Despite the highly dynamic redox conditions in the system, concentrations of MeHg in the lake outlet stream show low seasonal variations. This indicates that local hotspots of methylation may be important for the mercury uptake in biota, but not for the overall catchment budget of MeHg.

Distribution of both MeHg and TotHg under temporal and spatial variation in redox potential will be shown. Correlation between MeHg and redox potential in the lake sediments may give a broader understanding of the processes involving cycling of MeHg in dystrophic lake systems. Combined with TOC levels this will also be important data in the assessment of why levels of mercury in fish in these lake systems are increasing.

TG7-O3 — 9:00-9:15
Author: RYTUBA, James1
(1) U.S. Geological Survey, jrytuba@usgs.gov

In central California saline connate ground waters within Cretaceous sandstone reach the surface along fault zones to form cold springs and provide saline waters to several watersheds. The saline waters have naturally elevated Hg concentrations, up to 234 ng/L, and elevated concentrations of Ba, W, Mg, sulfate and carbonate. During the summer dry season under low flow conditions in these watersheds, the stream waters are composed primarily of saline connate water and have only a minor component of meteoric water. High concentration of dissolved organic carbon and other nutrients in the saline stream waters results in blooms of phytoplankton and accumulation of biogenic sediment composed primarily of diatoms and lesser amounts of calcium carbonate, magnesite, and Mg clay minerals. The biogenic sediment contains micron to submicron particles of HgS, FeS, barite, and a tungsten phase dispersed within shells of diatoms that comprise the sediment. Concentrations of Hg in the biogenic sediment is high ranging from 1,000-10,000 ng/g, and monomethylmercury (MMeHg) ranges from 0.1 to 10 ng/g. Phytoplankton are known to hyper accumulate MMeHg, and at a lesser amount Hg (especially in mine impacted lakes) with bioaccumulation factors commonly over 20,000. The Hg bioaccumulation factor is 64,000 for phytoplankton in the biogenic sediment in the Harley Gulch watershed in central California and this accounts for the high concentration of Hg in the biogenic sediment. Decay of phytoplankton in the biogenic sediment releases MMeHg and Hg into the pore waters that react with sulfide derived from reduction of sulfate below the redox boundary. This results in precipitation of submicron size particles of HgS and FeS, and the decrease in sulfate concentration as sulfate is reduced to sulfide results in the precipitation of barite. In watersheds such as Harley Gulch, Hg enriched biogenic sediment is a continuing source of natural Hg that contaminates the watershed despite clean up of Hg mines at the head of watershed.

The process of Hg enrichment in biogenic sediment by phytoplankton bioaccumulation of Hg has been operative during the geologic record where small streams and ponds are dominated by saline connate waters. The class of strata bound silica-cinnabar-tungsten ore deposits are the result of biogenic and not hydrothermal processes in these specialized environments. Some diatomite deposits that form in restricted basins are enriched in Hg as a result of this process and a subclass of opalite type Hg deposits formed in this manner.

TG7-O4 — 9:15-9:30
TG7-O5 — 9:30-9:45
Authors: HO, Anthony 1, CARPI, Anthony1
(1)John Jay College, anthony.h718@gmail.com

While much progress has been made in understanding the factors that affect the reduction and emission of mercury from soil, significant gaps still exist in our understanding of the mechanisms, and thus kinetics, of this process. In an effort to better understand these mechanisms, we examined the emission of mercury under light and dark and varying temperatures, from sand surfaces doped with pure HgO, HgCl2, HgS and Hg0 as well as from undoped natural soils. Samples were monitored using a Tekran Automated Dual Sampling (TADS) system, and a Teflon dynamic flux chamber with a 1.6 chamber volume turnover rate. Surface temperature was controlled and monitored between sampling periods to ensure samples remained at 250C and 450C throughout the seven day sampling period. Full spectrum radiation caused flux spikes in both sand and soil samples. Each salt species exhibited unique behavior when subjected to light and temperature. Emissions from salt species in the dark were independent of temperature, whereas a significant effect on emissions from HgCl2 samples between different temperatures was found in the light. Thus, it appears that a minimum activation energy is required for the reduction and emission of mercury from HgCl2 containing samples. Emissions of mercury from HgO-spiked samples appear independent of temperature in both dark and light.

Comparison of Hg emissions from samples doped with elemental mercury to those containing mercury salts suggests that reduction is the rate limiting step in the soil flux process. Proposed mechanisms of the reduction of Hg+2 have previously involved a third-party electron donor. We found evidence for the reduction and emission of mercury from HgCl2 and HgO spiked samples in the absence of a third-party electron donor. Molecular energetics modeling of HgCl2shows that the molecule enters a transition state in which the simultaneous release of both chloride ligands is possible at energies that can be delivered by incident radiation. This is typical of charge transfer absorption processes, and evidence exists that HgCl2 undergoes reduction through this pathway. HgO does not appear to undergo reduction via the same process. In addition to further elucidating the mechanisms of the soil-mercury reduction and emission process, these results may provide a mechanism for characterizing different mercury salts in soil.

TG7-O6 — 9:45-10:00
Authors: BRIGGS, Christian W.1, GUSTIN, Mae S.1
(1) University of Nevada, Reno, briggsc9@gmail.com

Parameters that have been shown to influence Hg release from soils include substrate and air Hg concentration, light, atmospheric oxidants, temperature and soil moisture. However, for low Hg containing soils the influence of specific parameters can vary across space and time. We hypothesized that soil moisture loss, because it integrates the influence of multiple parameters, may be a useful tool for predicting Hg flux from background soils. For this project data was collected from low mercury containing soils from Indiana, Alabama and Ohio, and those amended with flue gas desulfurization derived gypsum as an agricultural conditioner. Hg flux was measured for these materials on a 20 minute time step over at least 24 hours using a dynamic flux chamber linked to a Tekran 2537A. Because integrated 24 h Hg fluxes from FGD amended soils were not significantly different from the corresponding unamended soil (two sample t-test, p>0.05) data was combined for the treatments and controls to give a more robust data set. The following environmental parameters were also quantified over the flux measurement period: solar radiation, local atmospheric ozone concentration, relative humidity, soil and air temperature, mass wetness and loss of soil moisture by evaporation. Data were divided into two sets: one that was collected from materials that were constantly wet due to weekly watering to 15% soil moisture and that derived during water addition experiments when soils had been allowed to dry for at least a month after which soil moisture was then increased by at 5-15%. Multilinear regression analyses of data collected using the soil maintained at 15% moisture using all parameters measured showed that 50% of the flux variability could be explained and water lost over the flux period was the most significant independent variable correlated with flux (p<0.001, n=132). Using only soil moisture lost accounted for 43% of the variation in flux. For the dry material watering experiments the soil moisture lost over the measurement period had the highest correlation with Hg flux (r2=0.31, p<0.001, n=36) and using all other parameters in a multilinear regression analyses increased the r2 to 0.47. As such soil moisture evaporation appears to be a parameter that may be used to model flux, however the equations developed need to be tested using field data.

TG7-O7 — 10:00-10:15
Author: ZHANG, Hua1
(1)Institute of Geochemistry, Chinese Academy of Sciences, zhanghua0851@gmail.com

Long-range atmospheric transport of Hg from pollution sources has led to increased Hg contamination even in remote areas that are free from direct anthropogenic influences. This research was designed to evaluate Hg impact on ecosystem in a remote high-altitude mountain area (range: 867–2185m asl.) in southwestern China. A mean of 0.21±0.06 mg/kg for THg and a geometric mean of 1.66±1.45µg/kg for MeHg in soil were observed. Altitude was observed positively correlated with concentrations of both THg and MeHg in soil (R2=0.75 for THg and 0.45 for MeHg, respectively, P<0.001). Multiple linear regression analyses shown that variable of altitude could explain 66% and 63% (i.e. adjusted R2) of the concentration of THg and MeHg in the soil, respectively. Higher concentrations of THg and MeHg were also observed in local rice compared to other regions in China, with a geometric mean of 2.67± 2.15µg/kg for THg and a mean of 2.14± 1.69µg/kg for MeHg, respectively. These results confirmed increased contamination in local ecologic environments. At the first time, our study supplies a new knowledge of Hg contamination in remote high-altitude areas in China.

TG7-O8 — 10:15-10:30
Tuesday, 26 July, 2011