2h: Mercury cycling in response to ecosystem perturbations

In this study, the effectiveness of using gas-phase sulfur as a stabilization enhancement following in-situ thermal remediation of elemental mercury is evaluated. In-situ thermal remediation involves the application of heat and vacuum to remove volatile contaminants from the subsurface. Thermal remediation is a mature technology that is commercially available for targeted cleanup of subsurface volatile organic contaminants and is being evaluated as a viable treatment technology for elemental mercury by several research groups. Laboratory experiments have demonstrated that 99.8% of elemental mercury can be removed using in-situ heating at temperatures of 244-259°C. A field-scale deployment by one group of researchers demonstrated a fourfold increase in Hg removal relative to ambient conditions, by increasing the subsurface temperature to 50-55°C. Moreover, results from numerical simulations indicate that phase transitions that occur with increasing temperature can lead to complete vaporization of elemental mercury below the boiling point of water. Although thermal remediation appears viable for substantial Hg mass removal, small residuals are likely to remain in the subsurface following deployment. These Hg residuals will likely exceed environmental thresholds, thus requiring a combined remedial solution that includes large-scale mass removal with an in-situ stabilization component. In this study the immobilization of residual elemental mercury in the subsurface with a sequestering agent (vapor-phase elemental sulfur) is examined. The novel process utilizes residual heat that is present within the system following in-situ thermal remediation to deliver gaseous sulfur, resulting in the in situ formation of recalcitrant mercury sulfide species. Conceptually the process is simple with elemental mercury and elemental sulfur reacting in the gas phase to form the recalcitrant solid phase species. Simulations indicate that the role of oxygen is significant, especially if atmospheric air is used as the delivery mechanism. These simulations indicate that a reducing environment (i.e. a transport gas with depleted oxygen) is needed to form the targeted mercury sulfides. When considering the role of other elements present in the subsurface, our simulations identified the potential formation of iron sulfide (pyrite) that will acidify the soil as moisture and oxygen enter the system. The acidification potential can be overcome by judicious selection of the transport gas used to deliver elemental sulfur. Simulations indicate that with the proper selection of the transport gas, a buffered system can be established with redox conditions that favor the targeted stabilization reaction.

Methylmercury (MeHg) production in newly flooded soils has been shown to be strongly related to soil organic carbon content. New hydroelectric developments and reservoir creation are being proposed in many ecosystems, but no data exists on whether MeHg flux from soil will vary with carbon content and composition. This information is essential to assess the environmental and human risks prior to reservoir creation. To address this knowledge gap for a hydroelectric power development under construction near Muskrat Falls on the Churchill River (Labrador, Canada), we collected 27 soil cores (15 cm long) from 3 sites for MeHg flux experiments. The core incubations included pre-flood soils, infrequently flooded soils, and wetland sites. Additional incubations with pre-flood soil cores included: 1) leaf litter layer removed, and 2) leaf litter layer and top 5 cm of soil removed. Each soil type and manipulation was replicated using three cores and flooded with Churchill River water, which was replaced every 24 hours. We sampled overlying water for MeHg analysis at intervals of 3 to 12 days for a period of 6 weeks. In previous flux core experiments using Muskrat Falls soil with leaf litter layer removed, we observed a 14-fold increase in MeHg concentration after only 5 days of incubation. We projected a 10-fold increase in riverine MeHg levels associated with flooding of the Muskrat Falls reservoir based on the relationship between soil organic carbon and soil MeHg for other systems. This resulted in a projected factor of 2.6 increase in MeHg in downstream Lake Melville estuarine surface waters, and an expected doubling of mean Inuit MeHg exposure. Our work suggests that soil organic carbon content is a useful screening criterion for hydroelectric power site selection, and that carbon removal could help reduce environmental MeHg concentrations associated with flooding. Plans for additional work at the Muskrat Falls reservoir site include soil characterization, water column MeHg measurements, MeHg flux core experiments with different levels of soil organic carbon, and probabilistic modeling to refine prior impact assessments.

The important pathway of terrestrial Hg transport into marine ecosystems is the atmosphere, however in the coastal zone, rivers represent the main source of this metal. Hg emission in the Baltic region at the beginning of the 21st century was about 20% - 30% lower than during the 1980s. The climate warming, particularly in the autumn-winter season, is another factor contributing to the decline in Hg emission from the main source, the burning of fossil fuels, into the coastal zone of the southern Baltic. On the other hand, warmer winters result in a lack of icing or its persistence for a shorter period of time, thus giving more time for remobilization of Hg and its incorporation to the marine trophic chain. In addition extreme events, such as storms, heavy rains, floods, the erosion processes along the coast are more frequent. In consequence higher load of Hg is introduced to the sea. The toxicity of mercury depends on its form, therefore the aim of this study was an estimation of input of labile mercury to the coastal zone, under the influence of intensive precipitation, catchment management strategies and erosion of the coast. Samples were collected in years 2015-2017 in the Gulf of Gdańsk watershed (southern Baltic). Mercury fractions (labile and stabile) were analyzed with DMA-80 mercury analyzer using thermodesorption method. The obtained results show that during present days, when anthropogenic emission is significantly reduced, extreme events, such as intensive precipitations, floods, coastal erosion and development of catchment area are important factors for Hg bioaccumulation in marine trophic chain.

Mercury input from rivers to the coastal zone depends on local sources of this metal, and is regulated mostly by land use and meteorological events. This study focused on four small rivers, typical for southern Baltic, with no significant mercury sources, transporting mostly Hg originating from atmospheric deposition. Samples were collected from 5 stations in Zagrska Struga and Gizdepka, 4 in Płutnica and 7 in Reda river catchment. Station represent most common land use types forest, agriculture, wetlands and urban areas. Samples were collected during average flow conditions in four seasons for two years, additionally, water was sampled during selected meteorological events floodings, downpours, snowmelt and draught. Samples included filtered and unfiltered river water, which was analysed for total mercury using CV-AFS Tekran 2600. Additional analyses included DOC, temperature, pH, dissolved oxygen and conductivity. Additionally, flow was measured by current meter. To evaluate mercury retention in the catchment, rainwater mercury concentrations were measured, and for the estimation of relative importance of small river mercury input for the Baltic, water from one of the largest rivers, Vistula, was sampled near mouth.

Results show decreased mercury retention and greater input during intense rainfalls, and floodings, showing mercury elution from catchment. This suggests that relative importance of mercury input from smaller rivers may increase due to Climate Change, which in the boreal environment may lead to more frequent extreme events.

Rising air and water temperature have been reported worldwide. Predicted increases in water temperature for temperate waters by 2100 are between 1.5 and 4.5 °C. Similarly, water temperatures in the majority of Canada are predicted to increase by 5–10 °C by 2100. We experimentally studied the impact of warming surface fresh waters on mercury cycling and accumulation in fish. Twelve tubs were filled to 8 cm depth with natural sediment collected from a lake in the Canadian Shield region, and filled with reverse osmosis filtered water to 40 cm depth. To simulate the effects of warming waters, four temperature settings with three replicates each were implemented: 12 °C (baseline), 8 °C (cooling), 16 °C and 20 °C (warming). These temperatures also cover the optimal range for white sucker, fish species used in the study. After 4 weeks, about 110 g of lumbriculus (food for fish) and sixty-five juvenile white suckers reared in the laboratory were added to each tub. The experiment was then run for 8 more weeks.Tubs were weekly topped up with reverse osmosis water to the initial volume to compensate for the evaporation loss. Water chemistry parameters, namely teperature, dissolved oxygen, pH and conductivity were measured daily, while hardness was measured weekly. Every week, three water samples from each tub were collected using a clean-hand dirty-hand procedure and acidified to 1% with ultrapure hydrochloric acid (HCL) for analyses of total and methylmercury concentrations. Five fish were randomly collected from each tub on a weekly basis and measured for length and mercury. Fish mortality was monitored daily and dead fish were removed. Observed changes in water chemistry, as well as total mercury and methylmercury concentrations in water, sediment and fish with time under the influence of varied water temperatures will be discussed.

East Fork Poplar Creek (EFPC) in east Tennessee, USA, was contaminated with large amounts of Hg between 1950 and 1963. Small amounts of Hg continue to be released from the primary source located at the headwaters of the creek. Stream restoration activities since 1983 included the initiation of a flow management program in 1996 to restore minimum baseflow in the uppermost reaches of the creek by pumping in uncontaminated water from a nearby lake. This water constituted ~13% of the baseflow measured at a point 21 km downstream but delivered substantial loads of dissolved organic carbon (DOC)(~25 kg/d) and total suspended solids (TSS)(166 kg/d).

We conducted water sampling along the length of EFPC on a monthly basis during 2011-2012 to assess patterns in Hg and monomethylmercury (MMHg) concentration along with ancillary water quality parameters (e.g., nutrients, DOC, TSS). Hg concentration decreased and MMHg concentration increased with distance downstream and MMHg concentrations were higher in spring and summer versus autumn and winter.

Upon cessation of the flow management program on 1 May 2014 monthly longitudinal sampling was repeated to determine what, if any, changes in Hg or MMHg might occur. Since cessation of flow management the Hg solid-water partitioning coefficient (Kd) has decreased by up to an order-of-magnitude throughout much of EFPC due to both lower solid phase concentrations and higher dissolved Hg (HgD) concentrations. pH values are slightly lower since flow management ended but remain in the range 7 8.3 where Hg sorption is relatively insensitive to pH changes. DOC concentrations have been more variable but are generally comparable to the earlier survey. Specific UV-absorbance at 254 nm (SUVA254), a measure of dissolved organic matter (DOM) composition, has increased.

Higher HgD concentration has potential implications for bioavailability and MMHg production. Total and dissolved MMHg concentrations have increased in EFPC and these increases are most pronounced during spring and early summer when biota may be more susceptible to exposure and uptake. Similar MMHg concentrations have not been seen in 20 years and in some locations are the highest on record. Other watershed-scale factors likely contribute to the observed patterns as these changes occurred over months rather than instantaneously after flow management stopped. Nevertheless, similar changes in MMHg have not been observed in a tributary to EFPC.

Whereas forestry provides numerous benefits for humans and wildlife, forest harvest has been hypothesized to increase mercury (Hg) mobilization to aquatic systems. Mercury tends to accumulate in the upper layer of forest soil, where it associates with organic matter. This top soil can be disturbed by clear-cutting and the exposed soil is susceptible to runoff. Harvest may also change the net balance between methylation and demethylation through addition of carbon sources from decaying logging residues. To better understand the impact of timber harvest in a mixed pine and hardwood forest in the southeast United States, we measured Hg in soil, sediment, and water from the forest floor, a forest stream, and a nearby lake before and after harvest. We also determined methylmercury (MeHg) in the stream and the lake that it flowed into. Mean concentrations of THg in the top two cm of soil from sites transecting the forest floor decreased from 52.3 ± 4.3 ng/g before harvest to 36.0 ± 3.6 ng/g after harvest (±95% CI, n=14). The post-harvest reduction in Hg in the surface soils may be due to a combination of factors including mixing of surface soil with deeper soils that have lower Hg concentrations, an increase in runoff, and increased Hg volatilization stemming from higher soil temperatures and/or direct sunlight. Soil-Hg levels varied with particle size, with organic matter content having a greater influence than particle surface area. Sequential extraction of soil showed a significant decline in the percent of Hg bound to the oxidizable fraction, which includes organic matter known to complex Hg. In the lake, concentrations of THg and MeHg increased in the month following harvest. THg increased from 0.82 to 5.0 ng/L and MeHg increased from 0.04 to 0.17 ng/L. In the stream, dissolved THg decreased after harvest from ~3 to ~1.5 ng/L. Dissolved MeHg was similar before and after harvest (0.48 and 0.40 ng/L, respectively), but the proportion of MeHg increased in the stream from 10 to 30%. A methylation rate study is needed to fully address impacts of timber harvest on in-situ production of MeHg. In any case, an unharvested stream riparian zone likely served as a buffer retaining Hg and organic matter, and minimizing impacts to the stream and lake.

Global climate change is likely to affect both temperature and resource availability in aquatic ecosystems. While higher temperatures may result in increased food consumption and increased mercury accumulation, they may also lead to increased growth and reduced mercury tissue concentration through somatic dilution. Dynamic energy budget theory provides a broad and generalizable framework based on first principles of energy metabolism that is well suited to understand these interactions, allowing joint acquisition and interpretation of chemical exposure and stressor effect information to be translated into demographic rate changes. In the current study, we conducted growth and bioaccumulation experiments to examine the interaction of temperature and resource availability on mercury accumulation and effects in the estuarine fish Fundulus heteroclitus (mummichog). In the first experiment, juvenile mummichog were fed 3.3% or 10% of their dry body weight/day with tuna naturally contaminated with mercury and held at either 15 or 27 °C for 28 days. Growth was low in most treatments, except in fish fed 10% body weight held at 27 °C (40% weight and 12% length increase). However, methylmercury bioaccumulation was similar across feeding conditions but increased with temperature (~17-fold increase in methylmercury concentration at 27 °C and ~7-fold increase at 15 °C, regardless of feeding rate). In the second experiment, mummichogs from two wild populations with differing native mercury exposures were fed either a high or low methylmercury diet. Fish were strip-spawned every two weeks during the feeding period. Adults were sampled for total mercury concentration at the start and end of the experiment, and egg methylmercury concentration was measured in unfertilized eggs from each spawning event. Danioscope software was used to assess the heart rate of developing embryos at 10 days post fertilization. A dark:light movement assay determined differences in behavior of larvae between treatments at three and 10 days post hatch using Ethovision software. Tissue analysis indicated successful maternal transfer of mercury to eggs in the high mercury feed treatment. Heart rate and movement assays indicated potential population level differences in baseline behavior. The use of these data in a dynamic energy budget model may greatly aid in understanding how temperature and resource availability affect mercury bioaccumulation. Overall, this work contributes to the ongoing development of an ecological modeling framework in a fish with an extensive toxicological and genomic background. Ultimately, we are working to connect molecular mechanistic, physiological, reproductive, and behavioral responses to population level fitness.

Gold mining activities in Amazonian rainforest lead to deforestation since decades and also to pollution with the use of mercury, arsenic and cyanide during the extraction process. While ecological rehabilitation is essential to restore biodiversity and decrease erosion on deforested lands such as mining sites, few studies point out the behavior of toxic traces elements such as mercury or arsenic during the rehabilitation process as well as their toxicity. Our original study focused on the potential use of microbial parameters related to C, N and P turnover, functional diversity, soil chemical proprerties and sequential extraction of Hg and As, to assess ecological impact and quality of revegetation processes of old mining sites affected by long-term acid and metal stress. We sampled two plots revegetated since 18 years in the site of Yaoni, in French Guyana. The first (R) revegetated with leguminous species (Clitoria racermosa and Acacia mangium), the second revegetated with spontaneous vegetation (Sv). We performed a microcosm experiment of soil in controlled conditions during 30 days where soil characteristics and mercury and arsenic speciation were analyzed. Along the incubation, microbial biomass, carbon, nitrogen and phosphorous mineralization processes and enzyme activities (FDA, dehydrogenase, β-glucosidase, urease, alkaline and acid phosphatase) were estimated to characterize the soil functional diversity and metal stress.

Our results showed that revegetation with leguminous (R) had a positive and significant effect on soil chemical quality, soil microbial biomass, mineralization rates of C, P, and on enzyme activities involved in C and P cycles. Revegetation process had however no effect on N mineralization and urease activity which suggested a metal stress in these conditions. Revegetation process had a significant effect on Hg speciation. In R the major proportion of Hg is bound to organic matter phase and few bound to iron oxide one. While in Sv we found more mercury mostly associated to iron oxide phase, and significantly to exchangeable and soluble Hg representing the more available carrier phases. Difference in revegetation process had less effect on As speciation. Statisticals analysis demonstrated that exchangeable Hg, total As and soil pH are key stress factors for biological parameters in rehabilitated lands.

These results also suggested that microbial activity, enzymes activities and mercury cycle are sensitive to land rehabilitation, which confirm our hypothesis that mercury speciation, microbial activity and enzymes activities must be evaluated as parameters for revegetation process quality.

Established in 1936, Kellys Slough National Wildlife Refuge (KSNWR) is a 5.1 km2 managed grassland and wetland complex located near Grand Forks, ND. In the early 1990s water control structures were constructed on an intermittent stream located within the refuge to create standing water conditions that support migratory waterfowl and shorebirds. This along with excavation of some surface soil resulted in a complex of three interconnected open-water pools: upper pool 1 [UP1], lower pool 1 [LP1], and lower pool 2 [LP2]. Immediately above the inlet to UP1, stream inflow passes through Pender pool (PP), a dense cattail marsh that was unintentionally formed during either pool or road construction at the site. A concern for the refuge managers is the observation from mercury (Hg) researchers that creating impoundments can lead to elevated levels of toxic methylmercury (MeHg), and that manipulating water levels so that sediments are alternatively dried and rewetted can further stimulate MeHg production, via sulfate re-oxidation. Starting in June 2014, the US Fish and Wildlife Service and US Geological Survey initiated a study of KSNWR to address these concerns. Surface water samples were collected from the three distinct pools during ice free periods in 2014 2016, and a sampling site was added at the inlet to Pender pool in 2015. Water samples were analyzed for total Hg, MeHg, and dissolved organic carbon (DOC) at the US Geological Survey Mercury Research Lab in Middleton, Wisconsin.

Filter passing MeHg (FMHG) concentrations increase an average 1.5X (range 0.5 5.5X) and percent of total mercury that is methylmercury increases an average of 1.3X (range 0.5 3.7X) from the inlet to the outlet of PP., suggesting that Pender pool is a source of FMHG to the downstream complex. Spatial trends of DOC show decreasing concentrations from the inlet to outlet of the study area, indicating that inflowing surface water is an important source of DOC to the system. Within the complex, average concentration of FMHG is highest at the inlet to UP1 and decreases in a downstream direction, which is likely due to photo-demethylation and reduced primary production in response to lower nutrient conditions. In addition, observed increases FMHG concentration between LP1 and LP2 during spring and early summer suggests that this part of the complex serves as a location of net MeHg production. Given that this location experiences the greatest range of water level manipulation, we surmise the MeHg increases are in response to seasonal wetting and drying of bottom sediment.

We have analyzed a database of fish muscle Hg concentrations from water bodies in northern Manitoba, Canada, in order to determine the extent of long-term recovery toward pre-impoundment levels in the decades that have elapsed since reservoir flooding for hydroelectricity generation. To date we have examined available data from 1971 2012 for walleye (Esox lucius), northern pike (Sander vitreus), and lake whitefish (Coregonus clupeaformis) species collected from 22 northern boreal water bodies.

The water bodies examined in this study include several flooded during the diversion of the Churchill River flow in the 1970s to augment the flow of the Burntwood-Nelson River System for electricity production. Prior studies have identified the extent of area flooded during impoundment as the major determinant of an initial increases in fish muscle Hg for these three species, ranging from 100 to 700% of background levels, within 2 -14 years of initial impoundment. The mechanism of the Hg increases is thought to be enhanced decomposition of organic matter in newly flooded regions, which promotes Hg methylation and entry into aquatic food webs.

Despite large variations between reservoirs, we observed increases in walleye and northern pike Hg concentrations over the past decade in the majority of examined reservoirs. In most cases, these increases occurred in the past decade after minimum fish Hg levels were observed between 1998-2005.

In contrast to the historical increases observed soon after impoundment, the recent increases are not correlated to initial flooded area nor to estimates of organic carbon content of regional soils. Furthermore, the magnitude of observed increases in walleye and northern pike muscle Hg observed in non-regulated water bodies over the last decade can equal those in regulated water bodies. Preliminary analysis suggests that regional variations, including those enhanced by climate-driven changes in the hydrologic cycle, may elevate fish Hg concentrations and in turn exacerbate fish Hg burdens in regions impacted by regulation.

Forestry activities have in several studies been identified to increase concentrations and loads of methylmercury (MeHg) in surface waters. There is, however, a great deal of variation in the magnitude of documented effects, varying from no effect up to manifold increases in MeHg concentrations and export. In contrast, studies of soils after forest harvest are very few. By combining molecular methods with soil geochemical analyses, we here examine how various forestry activities affect Hg methylating microbial communities and MeHg formation in soils. We investigated areas undergoing conventional stem-only-harvest and areas where both stem and stumps are removed to increase the share of bioenergy. We tested the hypothesis that environments of high MeHg net formation, so called Hg methylation hotspots, are more prone to form after stump-harvest than after stem-only-harvest. The reason would be more severe soil compaction and soil disturbance when stumps are removed. Soil concentrations of MeHg, percent MeHg of total Hg (THg) and bacterial community composition were determined at 200 sampling sites distributed across eight catchments in the north and south of Sweden. Each catchment was either stem-only-harvested (n=3), stem- and stump-harvested (n=2) or left undisturbed (n=3). A higher percentage of MeHg relative to THg was observed in one of the stump-harvested catchments. This catchment also featured the highest bacterial diversity and highest relative abundance of bacterial taxa known to include Hg methylators, such as sulfate- and iron-reducers. Environments showing high MeHg net formation and high share of microorganisms known to methylate Hg, i.e. so called Hg methylation hotspots, were most often associated with water-filled cavities created by stump removal or driving damage. We suggest limited oxygen supply in these environments in combination with access to labile carbon sources from logging residual may favor the activity of Hg methylating microorganisms and thereby enhance MeHg formation. Despite this observation, stump harvest did not cause increased concentrations of MeHg in the stream, indicating that soils with high MeHg concentrations were not directly connected to the stream. We suggest the large variability regarding forestrys influence on MeHg in runoff water reported in the literature is due to a combination of how forestry operations affect the net formation of MeHg and the hydrological connectivity between environments where MeHg is formed and surface waters.

The latest review of the current secondary National Ambient Air Quality Standards (NAAQS) for oxides of nitrogen and sulfur started in 2013. The scope of the review includes, but is not limited to, the atmospheric chemistry and ecological effects of sulfur (S) deposition. The Integrated Science Assessment (ISA) of relevant research is currently under development by the U.S. EPA and will provide the scientific foundation for review of the secondary NAAQS for oxides of sulfur and other closely related criteria pollutants. One of the subjects in the ISA is the link between atmospheric S deposition and mercury methylation within aquatic and wetland ecosystems. Relevant scientific literature was identified using machine learning and citation mapping based on the references included in the previous ISA addressing ecological effects of oxides of sulfur (published in 2008). Relevant literature was also identified by participants in an ISA Peer Input Workshop held in summer 2015. These approaches identified approximately 6500 papers published between 2008 and 2015 related to the non-acidifying effects of S deposition, and keyword searches and title screening within this set of publications identified 203 publications for detailed review. Recent research has expanded the geographic scope of inference of links between S cycling and methylmercury from the Northeastern peat bogs and lakes described in the 2008 ISA to now include streams, rivers, and freshwater marshes across the continental United States. Advances in microbial ecology have enhanced mechanistic explanations of mercury methylation. Observational studies, experimental S additions, and long-term field collections provide limited but coherent evidence of quantitative relationships between S and methylmercury production or concentration. Additional research or modeling could potentially identify dose-response relationships between S deposition and surface water sulfate concentrations, or between S deposition and methylmercury in the environment and in biota. The views expressed in this abstract are those of the authors and do not necessarily represent the views or policies of the U.S. EPA.

For several years mercury has been recognized as a one of the most dangerous world pollutants. Despite all the restrictions related to the limitation of introducing of Hg into the environment, the concentration of the metal is still at a high level. The Baltic Sea is a relatively small water body and it should be taken into account that rivers are the main source of mercury there. Due to this, the size and management of river basins have a huge impact on the level of mercury contamination in the coastal region. Soil can be considered the main sink of natural and anthropogenic mercury. Moreover, in relation to changing climate, the growth of concentration of Hg was recorded. It was caused by intense rainfall enhancing surface runoff and leaching of mercury from soil to the river water. The aim of that work was to estimate the impact of seasonal changes and intense precipitation on the leaching of labile form of Hg from soils.

The samples were collected in the Reda river (the longest river of the Pucka Bay basin (southern Baltic Sea)). These rivers where divided into 7 stations characterized by different levels of development of the basin. The material was collected seasonally in 2015 year from three depths of soil layer (0-20cm, 20-40cm, 40-60cm) and from river sediment. The fractionation of mercury was carried out by thermal desorption by DMA analyzed. Due to the combustion of the samples in different temperatures, the labile and stable form of mercury were detected.

The Reda river was characterized by high content of labile form of mercury in all studied stations. Take into account individual stations, a twofold increase of labile form of Hg in source of Reda river after flooding was detected. The leaching of labile mercury from soil and transfer to the river sediment was caused by the growth of the water level in the river. The significant increase of this form was noticed in the other stations as well. It was connected with content of fine grain and organic matter. The intense precipitation changed in the contribution share in both stable and labile form of mercury in soil and sediment of the Reda river.

Flooding has been well documented to result in increased fish mercury concentrations. Peak concentrations in adult sportfish such as northern pike or walleye can be up to 7 times greater than background levels in some new reservoirs. A mechanistic model of mercury cycling and bioaccumulation in reservoirs (RESMERC) was developed and initially calibrated using data from a series of whole-ecosystem mercury/greenhouse gas experiments at the Experimental Lakes Area (ELA) in Ontario, Canada. The ELARP experiment flooded a boreal wetland, while the FLUDEX experiment flooded three boreal forest upland sites with contrasting moisture conditions, vegetation, and stores of organic carbon. The model was then applied to two full scale reservoirs: Notigi Reservoir in Manitoba and Robert Bourassa Reservoir in Qubec. Model calibration included components for hydrology, carbon decomposition in flooded soils and sediments, and inorganic Hg(II)/methylmercury fluxes and concentrations. The model calibration reproduced several key trends observed at the ELA sites, including the rapid increase and subsequent decline of MeHg levels observed for the upland sites (FLUDEX) and sustained MeHg production in the wetland environment (ELARP). Observations and simulations suggested a pulse in decomposition after flooding that led to a pulse in MeHg production lasting a few years in flooded uplands, and longer in the flooded wetland. Large increases in porewater Hg(II) were observed for the medium carbon FLUDEX site shortly after filling. Simulations suggest that this was driven by DOC increases and that Hg bound to DOC in porewater is methylated. When the model was applied to Notigi Reservoir, the calibration based on the ELA sites resulted in peak mercury concentrationsfor northern pike and walleye that were too low, and a predicted recovery that was too fast (e.g. 10-15 years), compared to the observed 20-30 year recovery period. The model was recalibrated to fit observed fish mercury levels in Notigi and Robert Bourassa Reservoir well, but this calibration did not represent FLUDEX results adequately. There may be additional processes in full scale reservoirs with hypolimnetic waters that are not reflected in FLUDEX and ELARP, which represented shallower floodzones with seasonal water level fluctuations. Sensitivity analysis predicted that in addition to the extent of flooding and type of terrain flooded, flowrate is an important consideration when predicting the response of fish Hg levels following reservoir creation. Rapid water throughput has the potential to reduce peak MeHg levels.

The Hells Canyon Complex (HCC) consists of three hydroelectric facilities and associated reservoirs along a 145 km stretch of the Snake River on the Idaho-Oregon border (USA). Brownlee, Oxbow, and Hells Canyon Reservoirs went into service between 1958 and 1967, have surface areas of 4.6 to 59.1 square km and maximum depths from 25 to 91 m. Other human influences affecting the study area include increased regional and global atmospheric mercury (Hg) emissions since the 1800s, Hg mining in the upstream watershed that began in the mid-1800s, upstream agriculture since the 1900s, and the operation of a large cement kiln roughly 15 km from HCC, since 1979. These activities have increased Hg loading to the river system and altered conditions related to hydrology, particle dynamics, nutrient loads, system productivity, and water quality. Fish sampling in 2013-2015 clearly showed that Smallmouth Bass Hg concentrations in HCC increased progressively through the three reservoirs and declined in the Snake River downstream of the complex. Mercury concentrations in Smallmouth Bass (>250 mm) in Hells Canyon Reservoir were 0.266 µg/g, 4.4X greater than concentrations of ~0.061 µg/g in the Snake River immediately upstream of HCC. Brownlee and Hells Canyon Reservoirs are listed as impaired for Hg by the States of Idaho and Oregon. Deep water pumping from HCC has also been considered to improve downstream cold-water fish habitat, but there are concerns regarding the fate of high levels of methylmercury (MeHg) observed in hypolimnetic waters of HCC (up to 2.9 ng/L in Brownlee Reservoir and 5.5 ng/L in Hells Canyon Reservoir), if a pumping strategy were to be implemented. The U.S. Geological Survey and Idaho Power Company are collaborating on a study to better document total Hg and MeHg levels in HCC and downstream, understand factors controlling their spatial and temporal distribution, and ultimately explore remedial options. A conceptual model of Hg cycling and bioaccumulation in HCC will be presented. A leading hypothesis to explain elevated MeHg in HCC food webs is that elevated inputs of nutrients and organics enhance the supply of labile carbon, which serves as a substrate for Hg(II)-methylating microbes in deep zones of the system. Other possible contributing factors include Hg inputs from legacy Hg mining in the watershed and legacy/ongoing Hg inputs from the other nearby air and watershed sources.

With the industrialization and urbanization after the adoption of reform and opening-up policy in China, industrial and domestic wastewater has been widely produced and released into rivers, and has been used to irrigate cultivated soil due to its rich nutrient substances. Therefore, mercury (Hg) in the wastewater irrigated soil could become an important emission source of Hg to the air. In this study, we investigated and sampled surface soils from 13 traditional wastewater irrigated fields which had more than 20 years irrigation history. A continuous flow chamber was designed to simulate the emission of Hg under different temperature and solar irradiation in lab. By measuring the total gaseous mercury (TGM) concentration in the chamber, the emission rate has been calculated. We estimated that TGM emission to the air from 38 major wastewater irrigated fields in China is 2019±404 kg per year. The emission rate has a strong positive correlation with temperature and solar irradiation, respectively. Also, contents and speciation of Hg in the irrigated soils were determined and discussed. Due to wastewater irrigation, the THg concentrations in the irrigate soil range from 166±12 ng/g to 10089±72 ng/g. Most of them are much higher than the background values of Chinese soil. Our study shows that Hg emission from wastewater irrigated soil in China should not be ignored.

Wildfire has been increasing in frequency, size and intensity in recent decades due to climate change and increasing fuel loads, and these trends are expected to further rise in the future. There are many studies showing that mercury (Hg) is lost from forest ecosystems to the atmosphere due to wildfire. However, few studies have examined residual pools of Hg after wildfire, namely surface ash, which can contribute dissolved and particle-bound Hg to aquatic ecosystems upon runoff and erosion. Specifically, the Hg contents in different ash types, and more importantly the reactivity and bioavailability of Hg in ash, are largely unknown. In the present study, we quantified Hg contents and investigated the reactivity and bioavailability of Hg in different ash types generated from two wildfires in northern California (Wragg Fire and Rocky Fire in summer 2015). Black ash (200−500 °C) and white ash (>500 °C) were collected from multiple locations within the burn perimeter before any rainfall/leaching events. Two acid digestions were used to extract Hg from each ash sample: nitric acid/peroxide (reactive Hg) and aqua regia (reactive and non-reactive Hg). The difference in Hg concentrations between these two digestions was operationally defined as the non-reactive Hg pool. All ash samples contained measurable but highly variable Hg contents, ranging from 4 to 125 ng/g (n=28). Overall, the black ash had a median Hg concentration (based on aqua regia digestion) of 29 ng/g (n=14) compared to 14 ng/g in white ash (n=14). Regarding the reactivity of Hg, we found that compared to litter samples (unburned sites), the majority of black ash and white ash samples contained much larger pools of non-reactive Hg, indicating that wildfire altered Hg reactivity in the ash. Further, we performed a sealed-bottle incubation experiment in which litter, black ash, or white ash was incubated with fresh surface water for 4 and 12 weeks. Preliminary data indicated that Hg in litter was highly soluble and efficiently methylated by anaerobic microbes inside the incubation bottle. In contrast, we observed very small amounts of soluble Hg when black ash or white ash was incubated, and we detected almost no methylated Hg suggesting that Hg in ash is less reactive and bioavailable. Therefore, our work implies that Hg in ash may be less available for further biogeochemical transformation (e.g., methylation), which has important implications for the effects of wildfires on the risk of Hg to aquatic food webs and consumers of fish.

The periphyton associated with freshwater macrophyte rhizomes is the main site of Hg methylation in different wetlands in the world. The objective of this study was to test the use of water bodies connectivity metrics, in the context of landscape ecology, in a tropical wetland landscape (Guapor River, Amazonia, Brazil) as an indicator of net mercury methylation potential in periphytic communities. Fifteen lakes with different lateral connectivity patterns were sampled as well as the main channel of the river. We tested net mercury methylation potentials through incubations with local water and Eichhornia crassipes periphyton/rhizomes complexes, using 203HgCl2 as a tracer. Physical-chemical variables, landscape data (morphological characteristics, land use, types of lateral connection of water bodies) and field data were analyzed with Variation Partioning, and Generalized Additive Models (GAM). The Me203Hg net production (as% of 203Hg added) was expressive (6 - 26%). The model that best explained variations in the net production of Me203Hg (76%) was constructed by the variables: lateral connection type, total phosphorus and dissolved organic carbon (COD) in water. The type of connection was the best factor to the model fit (r2 = 0.32, p = 0.008) and the temporarily connected lakes showed higher rates of net mercury methylation. Both COD and total phosphorus showed significant covariance with mercury methylation rates (r2 = 0.26, p = 0.008 and r2 = 0.21, p = 0.012 respectively). Our study proposes a strong relationship between net MeHg production rates in tropical areas, the type of water body and its hydrological connectivity within the aquatic landscape.

The agroforestry system (SAF) is a land use category, that integrates in the same geographic space: perennial woody plants with crops and / or animal breeding. This farming system has been identified as strategic in the mitigation of greenhouse gas emissions (GHG) is an alternative use of the land capable of reducing mercury remobilization adsorbed on leaf tissue of native or cultivated vegetation. In this context, the objective of the present study was to assess the influence of macro invertebrate soil on mercury concentrations in deployed agroforestry system under red ultisol in the Umari rural settlement, South of Amazonas state. Four methods were used: i) digital image processing and production of thematic maps; ii) demarcation of fixed plots and construction of soil monoliths; iii) determination of Hg concentrations in soil and individuals of the macro invertebrate soil; and iv) statistical analysis. A total of, 11 groups of macro soil invertebrates, especially in the layer of 0 -10cm depth of soil. The dominant groups of soil invertebrates were: Oligochaetas (72.32%), Diplpodes (miripodes) adults and young (13.20%), Isoptera anaplotermes (5.66%) e Arachnida (2.51%). Among these, in descending order, high [Hg] were recorded in Oligochaetas (19.337 ng/g), Larvas de Diplpodes (16.939 ng /g), Arachnida (781 ng / g) e Chilopoda (445 ng / g). The average concentrations of mercury in the soil compartment (0-10cm depth) varied of 38 ng / g and 292 ng / g In the rainy season and 22 ng / g an 219 ng /g In the dry season. Among land use systems assessed (shifting cultivation, livestock and agroforestry system), the agroforestry systems with (SQF) ten years or more of deployment presented greater diversity of macrofauna taxonomic groups of the soil and, the same time high levels of Hg (Med. 286.5 ng / g) in soil and macrofauna of soil. On the other hand, the conventional farming systems with low diversity of soil invertebrate groups, had lower [Hg] in the soil and in soil macrofauna individuals. These results demonstrate the potential of the agroforestry system to retain Hg and potentially other heavy metals, from natural and / or anthropogenic sources. furthermore, the presence of macrofauna of the soil tends to favor the fixation of Hg among other substances the natural ecosystem and cultivated southern Amazon.

Northern peatlands can be sites of net methylmercury (MeHg) production and sources of MeHg to downstream aquatic ecosystems. Northern peatlands are ‘hot spots’ of MeHg production and transport because of their anoxic organic peat soils that support sulphate-reduction and hence, mercury methylation. Northern peatlands are not a single ecosystem type, but are a continuum of hydrological connectivity and nutrient status (nutrient-rich, groundwater-fed fens to nutrient-poor, rain-fed bogs). Although they are all peat accumulating systems, fens tend to be dominated by vascular plants (e.g. sedge spp.) while bogs are moss-dominated (typically sphagnum spp.). Although a considerable number of intensive studies on MeHg have been undertaken on nutrient-limited, moss dominated peatlands, no intensive field studies have been undertaken to evaluate patterns and controls on MeHg concentrations in circumneutral pH, sedge-dominated fen peatlands. We investigated pore water and stream total mercury (THg) and MeHg along with other hydrochemical parameters in a boreal intermediate fen in northern Ontario, Canada to evaluate the overall distribution of MeHg and to assess whether or not within-peatland near-stream zones of potential groundwater-surface water interaction had higher MeHg concentrations than more hydrologically decoupled areas of the peatland. Pore water sampling was conducted monthly during the summer and fall 2016 at both higher resolution transects (8 m length) near incised fen streams and lower resolution transects (190 ± 65 m length) across the broader fen area. A clear pattern of greater net MeHg production near incised fen streams than the broader area of the peatland emerges as a result local groundwater discharge as indicated by other water chemistry and hydraulic head measurements. The proximity and high degree of hydrologic connectivity of near-stream pore waters to the streams represents an important linkage between net MeHg production and transport to downstream aquatic ecosystems in these understudied peatland types. It has been suggested that climate change has the potential to shift northern peatlands toward more vascular-plant dominated ecosystems, so greater insight into their Hg biogeochemistry will contribute to the broader understanding of Hg in northern aquatic ecosystems in the future.

To understand the influence of cage aquaculture activity on the distribution and methylation in river-reservoir system, Wujiangdu (WJD) and Yinzidu (YZD) reservoirs which located in Wujiang River Basin, Southwest China, were chosen in this study. The WJD and the YZD are characterized by absent and popular cage aquaculture activities, respectively. Concentration and distribution of Hg species in different sectors of the river-reservoir system (e.g. water column, sediment, inflow-outflow rivers of reservoirs) were systematically investigated over a one-year period. On the basis of the measured data, a detailed Hg mass balance in selected reservoirs were developed as well.

Our results showed that the concentrations of THg in the WJD and the YZD present the same level. However, concentrations of MeHg in water column and sediment of WJD were significantly elevated compared to those of YZD, suggesting the active Hg methylation in the WJD. In contrast, no discernable Hg methylation was observed in the YZD during the sampling periods. The results suggest that high primary productivity resulting from cage aquaculture activities in the WJD is an important control on Hg methylation in the reservoir, increasing the concentrations of MeHg in water in the Wujiang River basin, southwestern China. Our Hg mass balance calculations showed that YZD represented a net sink of MeHg, WJD acted as a source for MeHg annually.

The YZD as a newly formed reservoir were not active sites of net Hg methylation because of the low organic matter in submerged soil. On the opposite, organic matter in surface sediment originating from cage aquaculture activities in WJD principally lead to a much more active net Hg methylation. Current study confirmed that MeHg production in reservoirs located in southwestern China cannot be predicted using previous observations in North America and Europe because the Hg methylation process in the Wujiang River Basin is driven by a completely different biogeochemical dynamic.

Coastal marine ecosystems have ecologic, economic, and social value, but they are threatened by human induced climate change and increased mercury (Hg) loadings. These changes may influence the bioaccumulation of methylmercury (MeHg) into aquatic food webs where it becomes an exposure risk to humans. The most important step for MeHg incorporation into the coastal food web is via phytoplankton uptake and assimilation. Uptake depends on the chemical form (speciation) of MeHg, and has been correlated to the concentration and composition of organic matter (OM) in the system. Specifically, terrestrially derived OM in coastal waters decreases MeHg bioavailability compared to that of autochthonous origin (here forward referred to as marine). Thiol ligands associated with dissolved OM (DOM) and particulate OM (POM) primarily control the chemical speciation and bioavailability of MeHg in oxidizing coastal waters, so we have hypothesized that differences between thiol composition or concentration may be controlling uptake via thermodynamic and kinetic constraints. To test this, DOM was extracted from water samples collected at discrete points along the NE coast of the United States that represent variable organic matter and Hg loadings. The overall binding strength of the DOM to MeHg was then quantified through competing ligand experiments using a low molecular mass thiol ligand as the competing ligand. DOM binding capacity was evaluated through S K-edge XANES measurements and speciation modeling. Our results provide a refined molecular level understanding of how MeHg speciation is influenced by DOM from various sources, and give insight towards the observed higher bioavailability of MeHg bound to marine compared to terrestrial DOM. Understanding these processes is critical for reliably predicting MeHg levels in marine biota, and human exposure, now and following expected human and climate driven ecosystem changes.

Many research indicate that in the catchments atmospheric Hg retention takes place. Deposited Hg is accumulated in the organic-rich soil where it effectively binds to reduced sulphur sites as well as with oxygen/nitrogen-groups in the organic molecules. However there are some factors which stimulate the outflow of Hg from soil to the river. The aim of this research was to assess the impact of the river engineering on the outflow of mercury from soil to the river. For this purpose in years 2015-2016 mercury concentrations were measured in precipitation, soil and also in water, sediment and SPM from rivers in different parts of the Southern Baltic coastal zone. Total Hg in liquid samples were analyzed by means of Tekran 2600 analyzer. The direct mercury analyzer DMA-80 was used for the detection of Hg in solid samples. Apart from total Hg concentration in solid samples, also four fraction of Hg were separated. Hg fraction associated with labile compounds such as: mercury halides, mercury perchlorate, mercury nitrate, mercury cyanide, mercury thiocyanate,mercury fulminate,mercury acetate,methyl mercury and humus-like substance (labile-1); mercury sulfate, mercury oxide (red) andmercury fluoride (labile-2) as well as with mercury sulphide (stabile-1) and mineral matrix (stabile-2) was analyzed using the thermo-desorption method.

The obtained results show that activities related to river engineering such as dredging, desludging and rebuilding of the riverbed had a significant effect on mercury outflow to the river. As a result of these actions significant increase of mercury concentration in water and sediment river was observed. Also the percentage contribution of mercury fraction in soil and sediment was changed. Elevated levels of mercury in the sediment remained during few next weeks. This indicate that catchment management affects the load and the form of Hg reaching the sea with surface runoff. While less important on a global scale, it is crucial in mediterranean seas and coastal zones, where rivers can be the main source of mercury.

Uncertainties in emissions and key kinetic processes such as reduction-oxidation (redox) reactions, mass transfer coefficients and partition coefficients limit our quantitative understanding of the global mercury mass balance. Such an understanding helps us comprehend how confident can we be in the results from global models. Also, research can then be directed towards most uncertain parameters or processes. At present, a substantial effort is made towards constraining anthropogenic emissions of mercury, particularly to atmosphere. Uncertainty around global atmospheric mercury emissions is generally accepted to be +/- 50%. However, it is also known that the uncertainty surrounding rate constants for reduction and oxidation span orders of magnitude, factor 10 to 100! This brings out the question if more research, and money, should be directed towards updating emissions or towards, for example, better understanding rate constants for reactions.

Recent GEOS-Chem publications attempt to constrain mercury emissions to the atmosphere using sensitivity analysis type updating of model results by using field observations of mercury. While this is important, a true uncertainty analysis should involve probabilistic Monte Carlo sampling, and a Bayesian updating of key input parameters using all available mercury observations. Previously, Qureshi et al. (2011, Global Biogeochem. Cycles, doi: 10.1029/2011GB004068) had conducted a complete but unbounded Monte Carlo uncertainty analysis on the global mercury multimedia mass balance, including fluxes, transfer processes and compartmental inventories. Here, we use all the available observations of mercury and mercury species in air and water to bound the uncertainties in the global mass balance of mercury, and determine important influential parameters. A Markov Chain Monte Carlo technique called Metropolis Hastings is adopted. A priori uncertainties in model inputs, such as emissions, rate constants, partition coefficients, are determined from literature. Then 10,000 model runs are made, and uncertainties in inputs are refined based on the bounds forced by actual observations. Finally, a posterior distribution of inputs is obtained. This is used to calculate the refined and more constrained global mass balance of mercury and to determine influential parameters.

It is found that the uncertainties in key fluxes, such as evasion of mercury from ocean to atmosphere are reduced from factor 650 to factor 50, and in net reduction of Hg(II) to Hg(0) in atmosphere from factor 36 to 10. An important finding is that uncertainties in oceanic parameters such as redox rate constants, and the partition coefficients of mercury in surface oceans are much more influential than anthropogenic mercury emissions. Therefore, more efforts should be made on understanding oceans rather than emissions. This finding also questions the inferences from some global models that Asian emissions shoud be higher just because results do not match observations it could be because of rate constants!

Although mercury dynamics have been relatively well-studied in freshwater ecosystems, they are understudied in estuarine environments. It is suggested that nutrient loading into coastal waters may have an inverse relationship with methylmercury availability due to growth dilution from eutrophication, decreases in methylation due to reducing conditions and the binding and sequestration of mercury with organic matter at higher nutrient concentrations. In order to fully understand the relationship between nutrient concentrations and mercury availability, water and sediment samples were collected in 2015 and 2016 in two estuaries and analyzed for mercury and nutrients, with a focus on nitrogen and carbon species. The two study estuaries were chosen to represent a gradient of residential and industrial development. Jamaica Bay is highly urbanized with dense residential and industrial zoning and receives 89% of its nutrient load from wastewater effluent, whereas Great South Bay is characterized largely by residential properties and has a more diverse input of nutrients. Results from 2015 show a stronger relationship between particulate nutrient and mercury concentrations than for dissolved concentrations. Mercury bioavailability appears to decrease with increases in sediment nutrient concentration after an initial increase in mercury bioavailability at very low concentrations. Patterns of coastal mercury will be investigated further with the addition of the 2016 analysis and the relationships will be presented.

Acidithiobacillus ferrooxidans is an acidophilic chemolithoautotrophic iron-oxidizing bacterium. The bacterium is highly resistant to many kinds of heavy metal ions, but sensitive to Hg2+. Mercury is toxic to almost all organisms because it has a strong affinity for the thiol groups in protein. It is expected that A. ferrooxidans resistant to mercury can be used as a powerful agent for bioremediation of mercury under acidic conditions. A number of heterotrophs have a mercury reductase that reduces Hg2+ with NADPH and generates elemental mercury (Hg0) under neutrophilic pH. In this work, we tried to isolate a mercury resistant A. ferrooxidans.

Of 100 strains of iron-oxidizing bacteria isolated, A. ferrooxidans SUG2-2 was the most resistant to HgCl2 and could grow in an iron medium (pH 2.5) with 6 μM HgCl2. In contrast, a mercury sensitive A. ferrooxidans AP19-3 could not grow on the medium with 0.7 μM Hg2+. When incubated for 3 h in 20 ml of salt solution (pH 2.5) with 0.7μM Hg2+ and SUG 2-2 or AP19-3 cells (1.0 mg protein), 20% and 1.7% of the total mercury added were volatilized, respectively. When SUG 2-2 cells (0.01 mg protein) was incubated for 10 d at 30℃ in 20 ml of sulfuric acid (pH 2.5) with ferrous sulfate (2.2 mmol) and 1 g of mercury-polluted soil (7.5 nmol Hg), 4.1 nmol of Hg0 was volatilized. Further addition of ferrous sulfate (2.2 mmol) and SUG 2-2 cells (0.01 mg) to the reaction mixture, with an additional 30 d incubation, increased the amount of Hg0 volatilized to 6.9 nmol. When 50 μl of an acidic-mercury wastewater containing 17.5 nmol of Hg, which is usually produced in the course of everyday laboratory work in Okayama University (pH, 0.98 and COD, 600 mg/L), was incubated in 20 ml sulfuric acid (pH 2.5) with ferrous sulfate (100 μmol) and SUG cells (0.05 mg protein) for 10 d at 30℃, 47% of the total mercury in the reaction mixture (8.23 nmol Hg0) was volatilized. However, when organic compounds in the mercury wastewater were decomposed by Fenton’s method and then treated with SUG cells, 17.5 nmol of Hg (100% of the total mercury in the wastewater) was volatilized. It was found that cytochrome oxidase on the plasma membrane of A. ferrooxidans SUG 2-2 plays a crucial role in the rapid reduction of Hg2+ with ferrous iron to produce Hg0.

Anthropogenic environmental mercury contamination in Michigan dates back over one hundred years to the mining activities in Michigans Upper Peninsula and in the 1970s, massive mercury discharges to the waters of Michigans Lower Peninsula were identified. More recently, atmospheric sources of mercury were characterized as the leading source of loadings to state waters. Significant progress has been made in the state towards reducing all uses and environmental releases of mercury including water discharges, air emissions, biosolids and product use. However, the environmental media and bioindicator data do not necessarily demonstrate consistent temporal trend reductions. The mercury use and release trend data will be presented along with temporal environmental trends and an assessment for successful phase out of use within the state of Michigan. A Michigan state-wide goal is being established that limits atmospheric emissions and deposition, thereby reducing the mercury available for methylation and bioaccumulation in fish. However, further national and international efforts are needed to reduce these impacts because most of the atmospheric mercury deposition in Michigan originates from beyond the states borders.

Coal ash is the left over material after coal is burned and contains various kinds of toxic metals including mercury (Hg). On February 2, 2014, massive amounts of coal ash and contaminated water from a decommissioned coal-fired power plant spilled into the Dan River near Eden, North Carolina, representing the third worst coal ash spill in the history of the United States. Although analyses on water samples immediately after the incident indicated very low levels of contaminants, including Hg, it is widely speculated that these contaminants would deposit in the bottom sediment where biogeochemical transformations, such as Hg methylation, can occur. Therefore, it is necessary to assess whether there is Hg contamination in the Dan River sediment and biota at downstream sites following the coal ash spill. In July 2015, about 17 months after the spill, we performed longitudinal sampling in 10 sites along Dan River, with 3 sites in the upstream section (up to 3.6 km) and 7 sites in the downstream section (up to 65 km) relative to the spill site. At each sampling site, we collected river channel coarse sand and finer depositional sediments along the shore. We also collected Asian clams (Corbicula fluminea) when available. We analyzed total Hg (THg) for the sediment samples, and THg and methyl-Hg (MeHg) for the clam tissue. Overall, THg in sediment samples was generally considered low but variable, from 2 to 76 ng/g. Sediment THg correlated well with organic matter (OM) content, and thus we normalized sediment THg to OM; OM-normalized sediment THg ranged from 293 to 741 ng/g OM. OM-normalized sediment THg in the upstream section ranged from 319 to 661 ng/g OM. If we assumed that the sediment sample with the highest THg in the upstream section was the upper limit of Hg in Dan River sediment without coal ash, the majority of downstream sediment samples (90%) would have OM-normalized THg lower than this threshold. In fact, the few samples exceeding this upper limit had OM-normalized THg only slightly higher than this upper limit. Clam tissue THg and MeHg levels were variable among sites but did not appear to be different between upstream and downstream sections. Thus, our findings indicated essentially no Hg contamination in Dan River up to 65 km of downstream of Eden, North Carolina due to the coal ash spill after almost 1.5 years of the incident. However, it may be possible that Hg contamination hotspots were either not sampled (e.g., spatially too variable, or buried in deeper layers) or occurred further downstream of our most downstream sampling site (65 km).

Due to global climate warming, the frequency and intensity of wildfires are expected to increase in the next few decades. This is especially important in regions with prolonged dry summers, such as in the Mediterranean climate of California where forests have accumulated a large fuel load over the past century. Wildfires are known to release substantial amounts of previously sequestered mercury (Hg) in natural forests but residual pools of Hg may be left behind in the ash layer and burned soils. These residual Hg pools can be mobilized by intensive rainfall events leading to Hg transport dissolved in surface runoff and attached to eroded soil particles. This study examined the fluvial transport of total Hg (THg) and methyl Hg (MeHg) in two streams in northern California whose watersheds were previously burned in summer 2015: Cold Creek burned by the Wragg Fire (33 km2) and Cache Creek partly burned by the Rocky Fire (281 km2). Event and baseflow water samples were collected in these two streams during the first wet season (January to April, 2016) after the wildfires, and samples were analyzed for THg and MeHg in both unfiltered and filtered samples, as well as for total suspended solids (TSS) and dissolved organic carbon (DOC). Remarkably, stream water from the burned watersheds had extremely high TSS levels, approaching 20 g/L in Cold Creek compared to a maximum of ~50 mg/L in an adjacent unburned reference stream. We observed highly variable unfiltered THg in streams from both burned watersheds, ranging from 0.9 ng/L (baseflow) to 1,379 ng/L (event) for Cold Creek and from 3.8 ng/L (baseflow) to 688 ng/L (event) for Cache Creek. We observed much higher particulate THg during the first or second flush due to the extensive mobilization of surface ash and soil layers by erosion. For filtered samples, we observed only small variations of filtered THg among event and baseflow samples (i.e., 0.81 to 5.6 ng/L for Cold Creek and 1.6 to 15.0 ng/L for Cache Creek), and most of these variations can be explained by differences in DOC concentrations. MeHg levels were low across all water samples but we observed higher MeHg levels and % MeHg in water and particulates at the beginning of the rainy season.

It is widely predicted that global climate change would cause extensive changes in weather patterns, one of which would result in higher frequency of extreme weather events such as hurricanes and droughts. In two consecutive years, two strong hurricanes (Hurricane Joaquin in October, 2015, and Hurricane Matthew in October, 2016) affected much of the Southeastern United States. These extreme weather events caused extensive flooding in the low-lying coastal plain of the southeastern region where wetlands are widespread and representing substantial sources of mercury (Hg) to the freshwater and coastal environment. In this work, we assessed filtered total Hg (THg) and methyl-Hg (MeHg) concentrations in a blackwater river (Waccamaw River) in South Carolina, in which the river drains much of forested wetlands and swamps in the region and represents a model system to examine the impacts of these extreme flooding on fluvial export of THg and MeHg over time. Overall, river discharge increased by about 5-10 times during and right after the surging hydrograph in both years and the flooding lasted for about 2-3 weeks. In fall 2015, we found that dissolved organic carbon (DOC) and THg were both lower than under normal conditions (up to 50% lower). Different from THg, MeHg was much lower during the high flow period and the aqueous concentrations of MeHg were about 5-10 times lower than low flow period, suggesting different biogeochemical controls on the fluvial export of inorganic Hg vs. MeHg. Percentage of total Hg as MeHg (%MeHg) increased from 4% during high flow period to about 25% during low flow period, implying exports of MeHg from those wetlands and possibly some new production of MeHg in the system. Overall, during the high flow period THg flux would be about 2-3 times higher than normal conditions but MeHg flux would be somewhat similar between high and low flow period. Sample analyses are still on-going for 2016 samples and the data will be presented, compared with 2015 data, and discussed in the context of extreme flooding events and biogeochemical Hg cycling in the coastal plain.