2c: Sources and cycling of mercury in terrestrial ecosystems

By present time have been published large amount of data on heavy metals (but not on mercury) in environment of regions subjected to pollution of mining and nonferrous metallurgy. Mercury is constantly exist in the ores and concentrates as well as in industrial emissions and releases. Unfortunately, mercury emission from smelters not controlled officially. Research carried out in test site around the Karabash copper smelter (Southern Ural). The objective of this study was to investigate mercury cycle as result of atmospheric emission: precipitations (dry and wet) – soil – lake (water and sediments) – fish. Mercury concentrations in atmospheric dust deposition and rainwaters has been assessed. The results showed that mercury in precipitation mainly was in particulate (>2 μm) phase in the form of poorly soluble compounds, likely sulfides. The mercury content of atmospheric dust in 5-7 times higher local soils background; rainwater concentrations rarely exceed 0.5 µg/L. The area continues to suffer of acid rains (рН 3.5-4.5). Mercury concentrations in soils within 5 km zone from smelter stack (n = 20) ranged from 0.17 to 12.0 mg/kg. The maximum acceptable concentration (2.1 mg/kg) was exceeded in 30% of soils. The upper layer of bottom sediments in Lake Serebry showed extremely high concentration of mercury (up to 19 mg/kg), as well as other heavy metals from smelter emission (Pb, Zn, Cd, Cu and As). The assessment of bioaccumulation of mercury in physiological systems of fishes in comparison with background lake is made. The territory of Karabash geochemical anomaly is the natural-technogenic test area, which makes it possible to study degradation and restoration of environment at a changing technogenic load. It is shown that accumulation of mercury in fish tissue prevents the high content of Se in water and sediments.

Soil is the largest terrestrial sink for Hg and soils of Northeast US are elevated in Hg compared to US as a whole. Soil Hg concentrations are highly correlated to C and the fate of Hg in soil will be influenced by changes in climate and land use because these processes affect the C cycle. Similarly, changes in organic matter quality should affect Hg soil concentrations, because, presumably, Hg binding will change as quality changes. In this regard organic S is of particular importance because of the known high affinity of organic S groups to bind Hg. In this study, we look at the effects of OM decomposition (measured by δ13C), S, land use practice and soil type (pH) on Hg retention in forest soil in Bartlett, NH, U.S.A. The forests studied are of three ages – 15, 45 and 120 years, and contain two soil types - Inceptisol and Spodosol; Spodosols being more strongly acidic than Inceptisol soils. We collected soil from 10 replicate sites and oven dried at 30 0C. The soil samples were analyzed for C, δ13C, S, Hg, and bulk density. Soils were also extracted using DI water and analyzed for dissolved organic carbon (DOC) and Hg concentration, and partition coefficients (Kd) were calculated. The multiple regression of Log10 (Hg/C) vs. δ13C, Log10 (S/C), soil pH and forest growth was significant (p-value<0.0001; R2=0.92) and the three variables were individually significant (p-value<0.0001). The δ13C explains 33%, S/C 23%, pH 34% and forest growth 3% of the total variance in Hg/C ratio. Multiple regression of Log10 Kd vs. δ13C, Log10 S, pH and forest growth was also significant (p-value<0.0001, R2=0.56). again the first three variables were individually significant (p-value<0.0001). Most variance is explained by Log10 S (27%), δ13C (18%) and pH (10%), These results demonstrate that Hg retention (Hg/C) and binding strength (Kd) are strongly associated with the degree of OM decomposition, S concentration and soil type (pH). We do not see a significant effect of forest growth on Hg retention. However, there is slight difference in Hg retention between the three-forest growths which contributes to the total Hg retention in soil.

Suspended matter (SM) plays an important role in the transport of mercury (Hg) in aquatic systems. Information about Hg-species in this material is crucial to understand risk potential especially for Hg methylation and bioavailability. In the Idrija Hg mine (Slovenia) cinnabar (α-HgS) was mined and processed for centuries. These activities caused contamination of the Idrijca river system by dumping of ore roasting residues and atmospheric Hg deposition to soils attributed to excessive Hg emissions from the roasting plant. Cinnabar is the dominant Hg-species in the coarse-grained sediments of the Idrijca river where Hg methylation is general low, whereas organically-bound Hg has caused intense Hg methylation in the Gulf of Trieste (GT), the final sink for Hg released from the Idrijca catchment. Hydrology of the Idrijca river is characterized by high discharge events during heavy rains and snowmelt, which transports large amounts of SM towards the GT. However, the dominant Hg-species transported in SM and their specific source under varying hydrological conditions is largely unknown, but crucial to predict future transport of bioavailable Hg forms from the mining area to the GT.

We analyzed Hg concentrations and Hg-species (Hg-thermo-desorption) in SM and different grain size fractions of soils from the Idrijca catchment to evaluate changes in solid phase Hg-species under low and high flow conditions. Concentrations of dissolved Hg did not change significantly during low and high flow (median: 21.3 to 28.1 ng L^-1, respectively). Hg concentrations in SM in tributaries decreased with distance (~ 30 km) from the mine from 32.7 to 0.47 mg kg^-1 related to Hg concentrations in fine particulate soil fractions (0.45-20 μm) which ranged from 56.2 to 0.53 mg kg^-1. Hg speciation in SM was dominated by organically-bound Hg forms at low flow (median: 95.5%) which decrease during high flow conditions (median: 53.8%) attributed to increased mobilization of cinnabar from riverbanks and mine residues. Our results show that aqueous Hg transport in the Idrija mining area is dominated by heavy rain events and mobilization of organically-bound Hg from soils indicating that the frequency of heavy rain events will control transport of bioavailable Hg to the marine environment.

The influence of temperature, redox conditions and organic matter on the release and biomethylation of mercury (Hg) is well documented but mechanistic processes in the terrestrial environment are ill-characterized although they might have a substantial impact on Hg behavior. This is particularly true for agricultural floodplain soils, where manure is commonly applied and also because climate predictions postulate an increase in temperatures and extreme weather events such as flooding. To study the effects of temperature, redox conditions and manure addition on the biogeochemical cycling of Hg, we chose a Hg-contaminated agricultural floodplain situated near Visp, Switzerland. There, between 1930 and 1976, an acetaldehyde-producing company released an estimated 50 t of Hg into a canal. During this period, the canal was cleaned multiple times and the Hg-rich sediments were distributed throughout the agricultural fields and private gardens, contaminating the whole valley. We studied three agricultural soils situated near the canal. The soils were incubated in triplicate microcosms under flooded conditions both with and without the addition of 2 wt.% organic matter (OM) for 11 days. Bulk soils were sampled before and after the incubation to test soil biomethylation while the soil pore water was sampled throughout the experiment to assess the presence of dissolved MeHg and the release of Hg. MeHg was extracted from soils with an optimized three-step selective extraction using 35% HCl, dichloromethane and L-cysteine. Hg species were measured with HPLC-ICP-MS, while total Hg and a range of relevant elements (e.g. Fe, Mn) were measured by ICP-MS. Further soil properties potentially affecting biomethylation and release of Hg were also measured (e.g. DOC, Eh and pH). Under these conditions, Hg concentrations in the pore water increased to over 40 µg/L within a few days, with a higher release in the manure-amended microcosms. Increasing Hg concentrations were correlated with those of Mn in pore water, suggesting a reductive dissolution of Mn (oxy-)hydroxides. The Hg release was followed by a subsequent decrease within two days. Interestingly, an increase in MeHg concentration was measured in the soils during the incubation (up to 28 µg/kg) while no MeHg could be detected in the pore water. Ongoing investigations include incubation experiments to determine more precisely the biomethylation mechanism and the decrease of dissolved Hg in pore water. This work is important in order to understand the influence of climate change and agricultural practices on the biomethylation and the release of Hg in the terrestrial environment.

Regional geochemical survey was conducted in whole territory of Slovenia (20273 km2). High, medium and low sample density surveys were compared. High sample density represented the regional geochemical data set supplemented by local high-density sampling data (irregular grid, n=2835). Medium-density soil sampling was performed in a 5 x 5 km grid (n=817) and low-density geochemical survey was conducted in a sampling grid 25 x 25 km (n=54). Mercury distribution in Slovenian soils was determined with models of mercury distribution in soil using all three data sets. A distinct Hg anomaly in western part of Slovenia is evident on all three models. It is a consequence of 500-years of mining and ore processing in the second largest mercury mine in the world, the Idrija mine. The determined mercury concentrations revealed an important difference between the western and the eastern parts of the country. For the medium scale geochemical mapping is the median value (0.151 mg /kg) for western Slovenia almost 2-fold higher than the median value (0.083 mg/kg) in eastern Slovenia. Besides the Hg median for the western part of Slovenia exceeds the Hg median for European soil by a factor of 4.

Comparing these sample density surveys, it was shown that high sampling density allows the identification and characterization of anthropogenic influences on a local scale, while medium- and low-density sampling reveal general trends in the mercury spatial distribution, but are not appropriate for identifying local contamination in industrial regions and urban areas. The resolution of the pattern generated is the best when the high-density survey on a regional scale is supplemented with the geochemical data of the high-density surveys on a local scale.

References:
Gosar, M, Šajn, R, Teršič, T. Distribution pattern of mercury in the Slovenian soil: geochemical mapping based on multiple geochemical datasets. Journal of geochemical exploration, 2016, 167/38-48.

The Cedar River watershed (20,800 km2 at study location) comprises much of eastern Iowa where row crop agriculture is the dominant land use resulting in disturbances of hydrology and nutrient cycling. Climatic change has likely driven increases in baseflow and the frequency of flood events in the region. Within this context, we report on hydrogeochemical controls on methylmercury (MeHg) production within a savanna oak floodplain and terrace fen complex along the Cedar River, with emphasis on microtopography and groundwater-surface water interaction in controlling MeHg production.

Total Hg (THg) in the Cedar River averaged 8.4±1.2 ng/L, with a mean filterable THg of 1.1±0.55 ng/L and a mean filterable MeHg of 0.085±0.020 ng/L. Within the floodplain where groundwater-surface water exchange occurs, filterable MeHg was elevated compared to the Cedar River, averaging 0.14±0.16 ng/L in shallow alluvial groundwaters and 0.34±0.17 ng/L in wetland ponds. Wells in sandy ridges had lower TDS, were more oxic (higher dissolved oxygen, lower Fe(II) and dissolved Mn) and had lower filterable MeHg (median = 0.049 ng/L) than silty swales (median = 0.098 ng/L). Our findings point to the importance of microtopography in controlling groundwater flow paths, dominant terminal electron acceptor processes and Hg methylation in alluvial groundwaters. Across all wells and sampling events, dissolved Mn was the only variable significantly correlated with filterable MeHg (r = 0.46, p = 0.01). Wetland surface water MeHg was also significantly correlated with dissolved Mn (r = 0.70, p = 0.01), suggesting involvement of Mn-reducing bacteria in Hg methylation at our study location.

Within the terrace fen complex, a strong redox gradient across the terrace (oxic) to fen (anoxic) boundary was accompanied by increased filterable MeHg in groundwater (0.054±0.035 ng/L in upland terrace vs. 0.11±0.12 ng/L in fen). MeHg concentrations in fen peat reached 6 ng/g dry weight or nearly 5% of total Hg; low MeHg concentrations in peat groundwater are attributable to strong MeHg sorption to peat organic matter and/or iron sulfide minerals. Hg concentrations in the macroinvertebrate Gammarus were nearly twice as high in fen organisms compared to those dwelling near the terrace discharge (52 vs. 28 ng/g dry weight median Hg), and indicate potential for trophic transfer of Hg in this high value ecosystem. Overall, our results suggest that alluvial groundwaters may be important zones of Hg methylation in terrestrial ecosystems. Ongoing work aims to quantify the importance of flood events to Hg cycling within this system.

The radioisotopes beryllium-7 (7Be, half-life 54 days) and lead-210 (210Pb, half-life 22 years) are produced in the atmosphere, are globally distributed, and by virtue of their radioactive decay provide time information for a variety of earth system processes. Because the occurrence of 7Be, 210Pb and Hg in terrestrial systems derives from a shared atmospheric source, an understanding of their relative behaviors during deposition and interaction with vegetation may provide insights into Hg post-depositional fate. To investigate the deposition and interception of atmospherically derived elements we measured 7Be, 210Pb, total Hg (HgT) and other metals in tree leaves (Quercus rubra, Eastern red oak) over four years in annual time series. Depositional flux of 7Be, 210Pb and major and trace elements was also measured weekly for five years at a site co-located with leaf collections. In oak leaves the concentrations of 7Be, 210Pb, HgT, Al, stable Pb and certain transition metals all increased linearly with time through the growing season, and this trend continued through autumn and winter senescence; red oak retains some dead leaves in the canopy through the dormant season. After 1 year of atmospheric exposure HgT concentrations averaged 92 ±33 ng g-1 leaf dry weight (mean ±1SD, n=7). The 7Be:210Pb ratio of the leaves maintained a dynamic equilibrium with measured atmospheric deposition, but due to the short 7Be half-life this ratio is sensitive to seasonal changes in leaf interception efficiency. A monotonic increase in 210Pb, 7Be, total Pb and HgT leaf concentrations through the annual seasonal cycle indicates that a portion of the depositional flux of each metal becomes irreversibly bound to leaf matter. Litterfall is the primary source of Pb and Hg to underlying soils in vegetated terrestrial systems, and our observations thus suggests that 210Pb may provide a valuable tracer and chronometer of Hg fate during its deposition from the atmosphere and transfer to soil systems.

Mercury is a neurotoxin that has significant impact on human health. Mercury is commonly found in its divalent form in environmental soils; however it can be converted to elemental mercury (Hg0), which can move from the soil to the atmosphere. It has been found that the addition of water to a soil can result in an increase in Hg0 emissions. The mechanism by which water influences mercury flux from soil, however, remains unclear. In 2005, Gustin Stamenkovic proposed that water could lead to a redistribution of reducible mercury within the soil matrix thus facilitating subsequent reduction and emission. However, water appears to increase mercury reduction in soils in the absence of a third party mechanism of said reduction. We conducted a number of controlled laboratory experiments in which dry, sifted soil samples were placed under Teflon flux chambers, and purified water was then added in distinct events at timed intervals. Consistent with reported studies, an increase in Hg0 flux was seen with the addition of water, with larger volumes of water led to larger Hg0 fluxes up to a point. However, in our work when a soil sample was wet with a given volume of water, allowed to dry and then wet with the same volume of water and left in the dark in the dark, the second wetting led to a larger increase in Hg0 emissions than the first water addition this was consistent across multiple samples and multiple subsequent wettings. These preliminary results suggest that additional factors may be at play when soil is wet in addition to the redistribution of reducible mercury.

Mercury is a global health concern because of its toxicity to the human brain. Coal fired plants and sewage sludge incinerators are anthropogenic sources of mercury emission to the environment. Ambient air monitoring of such sources is expensive; however, moss has been used as a passive bio-monitor for airborne mercury due to its ability to absorb ions from the atmosphere. The purpose of this research was to study potential sources of mercury in New Haven, Connecticut, and determine whether moss is sensitive enough to detect a relatively small point source of atmospheric mercury pollution in a non-pristine environment. We hypothesized that there would be a direct relationship between the distance of the moss to identified sources of mercury, and the concentration of mercury in the moss samples. Local sources and patterns of distribution were studied by placing 21 samples of moss in three transects around a sewage sludge incinerator in New Haven and exposing them to ambient conditions for three weeks in July of 2015 and 2016. After collection, samples were analyzed and the patterns of mercury accumulation in the samples were correlated with distance from local sources of the metal. In 2015, moss samples closest to the incinerator showed mercury concentrations from 0.03 to 0.04 ppm, while moss placed further away had mercury concentrations from 0.02 to 0.03 ppm. Although samples from 2016 are still undergoing analysis, preliminary data corroborates the 2015 findings. Our data suggests New Haven (small source, non-pristine environment), and that the sewage sludge incinerator is acting a s a mercury source in the area.

Atmospheric mercury in litterfall is an important form of deposition to forests, which comprise 50 percent of the land cover in the eastern USA. Mercury (Hg) was measured in autumn litterfall samples during 2007 to 2014 at a total of 27 National Atmospheric Deposition Program (NADP) precipitation-Hg monitoring sites in deciduous and mixed deciduous-coniferous forests in 16 states in the eastern USA. These data comprise one of the broadest and longest studies to date and captured the maximum litterfall-Hg concentrations and litterfall mass at the study sites each year, along with annual precipitation-Hg data. Rates of litterfall-Hg deposition were higher than or equal to precipitation-Hg deposition rates in 70 percent of the annual data, which indicates a substantial contribution from litterfall to total atmospheric-Hg deposition. Annual litterfall-Hg deposition rates from this study matched modeled dry-Hg deposition rates at NADP Hg-monitoring sites and may indicate the low end of the range of dry-Hg deposition to forests. Atmospheric-Hg deposition rates were statistically different among five forest-cover types in this study, and rates were highest in oak-hickory and maple-beech-birch types. Forest canopies apparently recorded changes in atmospheric-Hg concentrations over time because litterfall-Hg concentrations were significantly higher in 2007-2009 than in 2012-2014 and litterfall-Hg concentrations decreased year to year for all sites in the study periodfindings consistent with reported decreases in Hg emissions and atmospheric elemental Hg concentrations during this same time period. Methylmercury, the organic form of Hg which accumulates and concentrates in food webs, was detected in all litterfall samples at the study sites, documenting part of a terrestrial Hg cycle reported to affect wildlife such as songbirds and raptors and their prey. Findings from this study indicate that long-term measurements of Hg in litterfall and precipitation help to quantify most of the atmospheric-Hg deposition to deciduous and mixed deciduous-coniferous forests in the eastern USA.

The relative importance of root exposure versus air exposure for plant uptake of oxidized mercury compounds has not been investigated extensively in controlled experiments. This project was a component of a larger one investigating the accuracy of tree rings as proxies for air exposures (i.e. atmospheric composition of Hg). This project involves the use of data collected from pines that were moved to different air exposures at 4, and 6-to-7 years old. In addition, tree cores were collected in multiple locations from trees >80 years old.

Here we discuss data collected using 4-year old Austrian pine trees grown from common genetic stock in Oregon that were exposed to three air treatments differing in GOM chemistry and GEM concentrations, and retained in the same soil. The trees were placed in: a greenhouse with GOM emitted from mining material (Hg-sulfur compounds), a greenhouse with HgBr2 permeated into ambient air, and a greenhouse with ambient air. In addition, a subset of trees was spiked with HgBr2 concentrations in solution of 100 ng/L applied to the roots 2-times a week for 5 weeks prior to being placed in different settings to investigate the impact of a root exposure versus air exposure. Foliar Hg concentrations, after growing in the different settings, were greater than the control trees (trees sacrificed when initially received), and increased over the growing season (June-September) for all exposures. There was no significant effect of the spike on tree ring concentrations. In addition, all tree rings were consistently higher relative to the controls. These results indicate permeation throughout the whole young tree.

Mines in remote northern regions have limited options for handling their wastewater produced through industrial and processing activities due to geographic isolation, cost, and energy requirements. Wetlands have a recognized capacity to assimilate nutrients and particulate matter, and since they are often geographically ubiquitous in these regions, are thus are an appealing water polishing option. However, some wastewaters contain elevated concentrations of sulphate, which is known to stimulate methylmercury (MeHg) production in wetlands, and in particular peatlands. To evaluate the potential impact of wastewater discharge containing sulphate on net MeHg production in peatlands, we conducted a large-scale field experiment that added simulated wastewater (containing nutrients and ~35 ppm sulphate) to a 0.98 hectare fen in the James Bay Lowland. In the summers of 2014 and 2015, simulated wastewater was loaded to the peatland at ~40,000 L/day for ~45 days each year. Samples were taken throughout the experimental site from pools and porewater via wells and piezometers. In 2016, the site water was sampled as in previous years, but no additions were made to evaluate the potential recovery from nutrient loading. Two reference peatlands were also sampled. In both 2014 and 2015, the fen showed a dramatic increase in MeHg concentrations (site average 1.53 ng/L vs. 0.075 ng/L for reference sites) concurrent with the addition of sulphate, and an accompanying site average increase in %MeHg from ~4.21% before loading to ~28.9% during additions and ~18.0% post additions. In 2016, MeHg, concentrations at the experimental site dropped significantly from the years with additions, but remained above the reference site (background) concentrations. In all years, there was no measurable increase in MeHg concentrations at the outflow of the peatland, despite the dramatic increases in MeHg concentrations seen in the porewaters. These results reveal the sensitivity of pristine northern peatlands to sulphate addition with respect to MeHg, however we also found that increases in production do not necessarily translate into increases in export, and that recovery after additions may be relatively rapid.

Sulphate additions to peatlands are known to increase methylmercury (MeHg) concentrations over time but little information exists about the effects of different levels of sulphate loading on the magnitude in intact natural peats. In remote, northern higher latitude peatlands, atmospheric loading of sulphate is extremely low areas where sulphate is higher are some sites of groundwater-surface water interaction, or where there has been land use change such as extractive mining. Here we present results from a replicated laboratory column experiment designed to examine a) the impact of different sulphate loads on net MeHg production in pristine peats naturally low in sulphate, and b) the degree to which sulphate derived from crushed mine waste rock stimulated net MeHg production in these same peats. Glass chromatography columns (30cm long, 4.8cm wide) with Teflon fittings were packed with peat, under anaerobic conditions. For the first experiment, continuous additions of three different sulphate concentrations (1, 5, and 30 ppm, plus controls) into the columns to simulate/represent the addition of sulphate-rich groundwater or wastewater containing sulphate to peat. For the second experiment, sulphate was delivered to the peat column by passing pH adjusted distilled water through a column packed with milled waste rock to simulate the pulse of sulphate that might be delivered from mine tailings. All sulphate additions stimulated net MeHg production, however the highest MeHg concentrations were seen in the 5 and 30 ppm treatments (1.82 +/- 0.18 ng/L and 1.23 +/- 0.11 ng/L, respectively vs 0.242 +/- 0.03 ng/L for the control columns) with the 5 and 30 ppm treatments also showing a more lagged response (48-72 hours) than 1 ppm, which was <24 hours. Results suggest that even very modest additions of sulphate to these pristine peats would result in a significant increase in MeHg in pore waters and potential downstream effects requiring careful consideration of both water and waste rock management for sulphate.

Forests are considered as a pool of mercury (Hg) in the global mercury cycle, Hg pollutions have also attracted many attentions for decades in suburb of southwestern China with a rapid economic growth. Studies have pointed out that the forest Hg is from atmosphere, dynamics and budget of mercury in some remote forests are not comprehensive and this risk of Hg releasing probably occur with the increasing of Hg accumulating in forest. However, limited data were published about dynamics and budget of mercury in a suburb area. In this study, a subtropical forest (nearby an industrial city, Chongqing) in southwestern China was chosen to study dynamics and budget of mercury in the forest, aiming to: 1) investigate the Hg concentration in various forest layers and analyze the dynamics of Hg in this forest; 2) understand the characteristics of Hg migration; 3) calculate the budget of Hg in this system. The results showed that the volume-weighted mean concentrations (VWM) and Hg fluxes of total Hg (THg) in precipitation, throughfall and runoff were 14.68 ± 6.13 ng·L-1, 31.64 ± 15.04 ng·L-1and 4.86 ± 0.31 ng·L-1, respectively and 16.51 µg·m-2·yr-1, 29.92 µg·m-2·yr-1, 1.29 µg·m-2·yr-1, respectively, which were higher than remote forests. Seasonal variations were also observed with peak THg value in winter and highest MeHg concentration in summer. Hg evaporation from soil/air interface (62.49 µg·m-2·yr-1) was the dominant pathway for Hg release from forest. The main pathway for Hg input into forest was litterfall (42 µg·m-2·yr-1), followed by throughfall and precipitation. The storage of Hg in forest field (including forest floor and soil profiles) was 130.37 mg·m-2·yr-1, which was much larger than input, this implied that the forest field was a big receiver for Hg. In this study, the Hg source was mainly from atmosphere in this forest system, and the Hg VWM in runoff has no risk for drinking water, but with the increasing accumulation of Hg in soil, this risk may be emerged.

A field campaign was conducted from 2 November 2014 to 22 March 2015 at Nam Co (Tibet, China) using a set of automatic atmospheric mercury speciation analyzers (Tekran 2537B with 1130 and 1135 units) for the measurements of Gaseous Elemental Mercury (GEM), Reactive Gaseous Mercury (RGM) and Particle-bound mercury (HgP). This study is one of attempts to present the valuable data sets of continuous measurements of atmospheric mercury species in the inland of the TP (Tibetan Plateau). Average concentrations of GEM, HgP and RGM were 0.94±0.19 ng m-3, 50.61±26.74 pg m-3 and 3.15±2.79 pg m-3 (mean±SD), respectively. The range of GEM, HgP and RGM as: GEM: 0.39-1.66 ng m-3; HgP: 4.88-157.84 pg m-3 and RGM: 0.18-29.73 pg m-3. For daily variation, GEM and RGM were relatively stable during monitoring, and HgP concentrations were much higher in wintertime than in later autumn and early spring. For details in diurnal variations of the mercury species, maximum GEM concentration was measured normally after sunrise whereas HgP reached its maximum before sunrise and RGM was maximum in the afternoon. Both convention and photochemical production were responsible for diurnal cycle of atmospheric mercury. Modeling results were relatively similar from potential source contribution function (PSCF) and concentration weighted trajectory (CWT) and results indicated that Nam Co Station was mainly affected by the air mass from South Asia. Western and central Nepal, central and east Pakistan and northern India were considered as potential source of Nam Co.

It is widely accepted that the amount of nature mercury emission is much larger than that of anthropogenic mercury emission. Multiple soil and meteorological parameters have been suggested to influence mercury release from soils. However, the relationships between the parameters and mercury release are still not very clear. In this study, we conducted a series of simulation experiments in lab to explore the effects of soil temperature and solar radiation on emission of mercury from soil. Field measurements were also conducted to verify the results. Three types of soil were collected in China including brown earth, black soil and red earth. The mercury concentrations in the three soils were 25.1 ± 0.4, 73.8 ± 2.1, 75.7 ± 2.4 ng/g, respectively, which are close to the background values of Chinese soils. Different gradients of solar radiation and soil temperature were designed in the orthogonal experiment. In addition, the concentrations of soil organic matter and other soil properties were examined, in order to investigate their influences. There are significantly logarithmic linear positive correlation between mercury emission and solar radiation, and significantly positive correlation between mercury emission and soil temperature. The minimum mercury emission flux (0.2, 0.6, 0.8 ng/m3/h) was observed under the dark condition and the maximum (2.1, 4.8, 6.3 ng/m3/h) was observed under the highest radiation intensity. The results showed that solar radiation is the primary driver for mercury emission from soil, which is the same as previous studies. Mercury emission flux in dark condition is about 5% - 15% of the flux in light condition and the percentage increases significantly as the soil temperature increasing. The flux in dark condition could be explained by thermal reduction of reactive mercury. Soil samples with similar mercury concentrations could have a difference of ~ 30% in mercury emission flux under the same experiment condition. The difference might be caused by the variation of soil organic matter and other soil properties.

Based on the measurement of domestic sewage samples from 59 sewage treatment plants in China, we built the mass flows of total mercury (THg) and methylmercury (MeHg) release from this sector to the environment. As an important anthropogenic source which was frequently neglected before, THg and MeHg release from domestic sewage increased rapidly due to the population increase and economic development in China. There were 160 (-17% to 24% of overall uncertainties) tons of THg and 260 (-17% to 25%) kg of MeHg released to the environment in 2014. Both THg and MeHg released to the aquatic environment decreased from 2000 to 2014, since the improvement of treatment rate in this period. THg and MeHg release to the pedosphere, such as land application and dumping of sludge produced by the sewage treatment plants, increased rapidly and threatened the local residents. THg emission from the disposition of sludge might be a neglected atmospheric source in China. The study on the release of THg and MeHg from urban areas to the environment by domestic sewage can fill the vacancy of anthropogenic source inventory, and could help identify feasible management measures in China.