G11 Mercury monitoring and risk assessment

With the high Hg emissions in China, there is an increasing requirement to take appropriate abatement action. Better knowledge regarding Hg pollution in China is needed, related to emissions, transport, deposition, accumulation in the environment and its impacts. A large, Sino-Norwegian cooperation project (SINOMER) addresses problems related to mercury pollution in Guizhou province in southwestern China in an integrated fashion, by considering Hg releases to the environment, environmental impacts and Hg contamination in food products, as well as assessing impacts on the society and potential policy and mitigation options.

In Guizhou Province there are several considerable sources of Hg to the environment. The releases are both as direct discharge to water from mine tailings and industrial contaminated sites and to the atmosphere from coal combustion, metals smelting and other sources.

Concentrations of Hg in the environment, focusing in water, agricultural soils, rice and important vegetables have been measured in four different regions of the province. The surveys reveal that high concentrations are found in some highly contaminated areas (especially in the Wanshan Hg mining area), but show that in general Hg concentrations in water, soil and agricultural products are low. Exposure to Hg to the population from fish consumption is very low, contrary to the well known problems in Europe and North-America, due to different environmental conditions, short aquatic food chains as well as generally very low fish consumption. Rice, on the contrary is the major pathway for methylmercury exposure.

Recent researches highlight that rice cultivated in Hg contaminated areas may accumulate MeHg and the main route of human exposure to MeHg is related to frequent rice consumption. To evaluate the health risks of mercury exposure in local residents through daily rice consumption, 294 polished rice samples were collected from eight provinces (Guangdong, Hunan, Guizhou, Jiangsu, Jiangxi, Shanghai and Hubei) in South China. The eight provinces constitute 33.5% of population and 54.4% of rice production in whole China in 2009. The means of total mercury (THg) and methylmercury (MeHg) concentrations in all rice samples were 9.88 ng/g (95% CI, 2.04-30.2 ng/g) and 2.41 ng/g (95% CI, 0.53-6.56 ng/g), respectively. And 12.2% (36/294) of the rice samples exceed the tolerance limit of mercury (Hg) in human foods (20 ng/g) recommended by Chinese National Standard Agency. MeHg constitutes a large portion of THg in rice, which reaches to 35.4% on average. In addition, THg and MeHg percentages as Hg in rice showed a significant negative correlation, which indicated inorganic Hg pollution in some rice samples. Regional difference of THg concentrations were found in the rice samples, while Guizhou, Jiangsu and Jiangxi Provinces had high THg levels. The daily intakes of THg and MeHg from rice consumption were estimated for urban and rural population in the eight provinces in 2009. Urban population had significant lower intake of THg and MeHg than rural population due to less amount of rice consumption. The average of 95% upper limit of daily intake of THg and MeHg for rural population was 0.198 µg/kg/d and 0.044 µg/kg/d, which constituted 27.1% and 19.1% of PTWI recommended by JECFA. The results indicated that general population in South China had no risk of Hg exposure through rice consumption except in some special sites with Hg pollution.

Wetlands and newly flooded areas are known to be sites of enhanced production of monomethylmercury, the form of mercury that is readily biomagnified in aquatic food chains to potentially toxic levels. The Okavango Delta in Botswana, Southern Africa, is the largest inland delta in the world and a wetland ecosystem that experiences dramatic annual flooding of large tracts of seasonal floodplains. The Delta was, therefore, expected to be home to high mercury levels in fish and to be an area where local subsistence fishing communities would be at substantial risk of mercury toxicity from fish consumption. Total mercury concentrations measured in 27 species of fish from the Okavango Delta averaged (mean ± s.d., wet weight) 19 ± 19 ng g^-1 in non-piscivorous fish, and 59 ± 53 ng g^-1 in piscivorous fish. These mercury concentrations are similar to those reported for fish from lakes in other areas of tropical Africa, demonstrating that not all wetlands are sites of elevated mercury concentrations in biota. Even more intriguing is that concentrations of mercury in fish from across tropical Africa are systematically and substantially lower than those typically reported for fish from freshwater ecosystems elsewhere globally. The reasons for this apparent “African mercury anomaly” are unclear, but this finding poses a unique opportunity to improve our understanding of mercury’s biogeochemical cycling in the environment. Mercury concentrations measured in human hair collected in subsistence fishing communities in the Okavango Delta were similarly low (0.21 ± 0.22 µg g^-1 dry weight) despite high levels of fish consumption, and reflect the low mercury concentrations in the fish here.

The Athabasca oil sands, situated in Alberta’s Regional Municipality of Wood Buffalo, continues to be the subject of considerable environmental debate, both on a regional and global scale. Concerns have been raised in regards to water use, changes in the quality of the “receiving” environment (air, water, soil and biota), effects on the landscape, and impacts on indigenous ways of life. Most recently, it has been suggested that the deterioration in water quality downstream from the industrial oil sands development is leading to the gradual poisoning of communities along the Athabasca River, the largest watercourse in the oil sands, and in the Peace-Athabasca Delta.

Ongoing monitoring programs in the oil sands indicate that measured concentrations of mercury in fish exceed health-based subsistence guidelines in the region. However, mean mercury concentrations in fish downstream of the industrial oil sands development generally fall within the range of regional mercury concentrations in fish from waterbodies outside the influence of the oil sands. As well, mean mercury concentrations in fish from the oil sands region typically fall inside the range of mercury concentrations reported in freshwater fish from other North American waterbodies. Concentrations of mercury in fish from waterbodies in the Athabasca oil sands do warrant certain limitations on consumption for higher fish intake groups and certain “at-risk” groups (e.g., pregnant women and children).

This paper will provide a community based and probabilistic analysis of mercury exposures to people consuming fish from the oil sands region. The analysis will be based on mercury concentrations in various species of fish collected from the region over the past 10 years. The risks to human health will be quantified based on the various health endpoints associated with mercury exposure. Analysis will highlight the variability in exposures and health risks using stochastic exposures and modelled dose-response relationships from literature. In addition, particular attention will be paid to the model parameter uncertainties and their influence on population and individual health risks. Overall, the goal is to develop a tool that aids individuals and health authorities in determining the frequency and magnitude of adverse health outcomes in communities that consume fish from the Athabasca oil sands.

Recent monitoring of total Hg in fillets from sport fish collected throughout the Ohio River by regulatory agencies has suggested potential human health fish consumption concerns. To compliment these existing data, skin-off fillet concentrations of total Hg, MeHg, Se, and As were determined in samples from four species (channel catfish, hybrid striped bass, sauger, freshwater drum) collected near twelve coal-fired power plants. Fillet samples (mostly composite) were analyzed by ICP-MS using EPA Method 1638 DRC (As, Se), Method 1631 (total Hg), and Method 1630 (MeHg). A total of 47 samples were collected along a 556 river mile distance. Three samples (one channel catfish, two freshwater drum) had MeHg concentrations that exceeded U.S. EPA’s MeHg human health criterion value of 0.3 mg/kg wet wt. The drum samples were from older fish (ages 19+ to 21+), suggesting that consumption of very large drum may be of regulatory concern. The percent of total Hg as MeHg ranged from 66 to 100% for hybrid striped bass and from 76 to 100% for sauger. There were no evident longitudinal (spatial) trends in fillet trace metal concentrations, suggesting relatively similar bioaccumulation rates along the Ohio River. For all species combined, there was no significant correlation between total Hg and Se levels (P > 0.47). There was a significant positive correlation (P < 0.001) between fillet Se concentration and selenium/mercury molar ratio. These ratios exceeded 1.0 for all species evaluated. For sauger and hyrbrid stiped bass, weight-standardized MeHg content and Se content were significantly correlated (P < 0.001 for sauger, P < 0.004 for hybrid striped bass). Several water chemistry variables were evaluated as possible co-variants with MeHg tissue concentrations in sauger and channel catfish; results of Principal Components Analysis, however, indicated very little segregation of MeHg levels among the water quality variables. We conclude that, with the exception of very large, old freshwater drum, the levels of total Hg and MeHg in fish collected near 12 power plant sites do not constitute a fish consumption human health risk.

Exposure to methylmercury through fish consumption is a public health concern. American Indians and other ethnic populations are potentially exposed to higher levels of methylmercury than the general population through reliance on subsistence fishing practices. The goal of our project is to determine whether ethnic populations (American Indians, Hispanics, and Micronesians) and recreational anglers in northeastern Oklahoma are exposed to elevated mercury levels resulting from their fishing practices. Our study site is the Grand Lake watershed, which is downwind and downstream of several mercury pollution sources. Our project addresses community concerns about mercury exposure through consumption of locally-caught fish and the lack of information about mercury contamination and bioaccumulation in this watershed.

The project has three major components: a survey of mercury concentrations in commonly-consumed types of fish throughout the watershed; a study of mercury exposure among high-end fish consumers, including quarterly collection of food frequency questionnaires (FFQs) and hair samples as biomarkers of mercury exposure; and the establishment of ongoing community education and outreach programs and partnerships with state and regional agencies.

Community participation is an integral part of our study design and implementation. Throughout the development of our FFQ, we incorporated feedback from a community advisory board, a council of fishing experts, and multiple focus groups. In this way, we have included many perspectives; however, inclusion of members of the area’s small, insular Micronesian population has posed a greater challenge. Working with high school age Micronesian youth has been a powerful tool for incorporating their unique perspective into our study design and identifying potential participants from their community.

In addition, we are relying on community members and local government agencies as resources for fish collection. We have welcomed individuals in the community, including participants in the exposure study, to share samples of fish caught throughout the watershed, and provided the option for them to receive the results of mercury analyses of their samples. We also have collected samples of bycatch from fish population surveys by the local wildlife conservation department. These approaches, which reduce sampling size requirements for our own fish collection, increase community buy-in and raise community awareness of our study and broader issues relating to mercury and public health.

For over a decade, Hg has been listed as one of the most hazardous substances in the U.S. due to its potential threat to human health. Despite its importance, our understanding of current human exposure to Hg through fish consumption remains limited. One of the largest sources of uncertainty in estimating exposure risk in the U.S. is the lack of comprehensive data on Hg concentrations in commercial fish. We conducted a large-scale study to assess general trends and variability in Hg content of commonly consumed commercial seafood items in the U.S. We evaluated over 1000 peer-reviewed scientific publications and government agency reports for inclusion in the study based on 10 selection criteria. We extracted data from approximately 350 sources. For each seafood item, we calculated the range and grand mean Hg concentration, weighted by sample size, using mean Hg concentrations reported from each study. We found that swordfish, halibut species, Atlantic cod, sardine species and shark species are among the most studied seafood items. In contrast, tilefish, monkfish and farmed fish (relative to their wild counterparts) are among those relatively understudied. Seafood items with the most variable Hg concentrations include shark species, tilapia species, shrimp species, pollock species and blue marlin. In contrast, seafood items including Atlantic cod, Pacific cod and Spanish mackerel, have the least variable Hg concentrations. Preliminary analyses revealed that the majority of seafood items varied in Hg content by at least 10-fold. Grand mean Hg values calculated from our study were largely consistent with mean Hg values from the FDA Monitoring Program. However, we found that the FDA values overestimate Hg content in several seafood items, including blue crab and catfish. Additionally, FDA values underestimate Hg content in several other seafood items, including striped bass, shrimp and blue marlin. Overall, our findings reveal that data-rich estimates of Hg concentrations of commercially important seafood items, based on a synthesis of the literature, warrant a re-assessment of consumption advisories and Hg exposure risk.

The Pinchi Lake Mercury Mine in northern British Columbia, Canada, produced elemental mercury from 1940 to 1944 (historic operation) and from 1968 to 1975 (modern operation). During the historic operation, mercury-contaminated calcines (primarily in the form of cinnabar [HgS]) were deposited directly into Pinchi Lake (55 km²). Mercury concentrations in surface sediments throughout most of the lake were between 2 to 20 mg/kg, but increased to 50 to 850 mg/kg in calcine-dominated sediments immediately adjacent to the mine. A range of monitoring and risk assessment studies have been conducted since 1974 to understand mercury dynamics in the aquatic and terrestrial environments at Pinchi Lake. This presentation focuses on the work conducted to date on fish.

Pinchi Lake fish mercury concentrations are elevated relative to the same species in nearby lakes and elsewhere in BC. Based on stable isotope analyses, these differences are mine related and not due to differences in trophic position. Although fish mercury concentrations in lake trout (0.81 ppm in 2006 for a standardized 55 cm fish) have declined significantly since the 1970s (5 ppm), concentrations have not diminished much between the 1980s and 2006 (~1 ppm). Stable isotope work has shown that food chain length is relatively short and similar to food web structure in adjacent lakes. Despite reductions in tissue mercury concentrations, levels are sufficiently elevated to potentially cause sublethal effects to individual fish based on laboratory studies. From a risk-based perspective, there is a high degree of uncertainty as to whether laboratory-based results can be extrapolated to the field and how any effects to individual, if occurring, ultimately affect fish populations. Lake trout should have the highest potential for sublethal effects to individuals given their elevated tissue concentrations and we compare literature-derived thresholds with site-specific tissue concentrations. However, available information on meristics, catch effort and growth rates suggests that population-level effects are unlikely. The long-term fishing advisory for the lake has limited fishing, which confounds comparisons to other lakes that sustain fishing pressure. Planned long-term monitoring will investigate the potential for effects to individual fish in Pinchi Lake relative to fish from reference lakes.

This study was conducted in August and October 2010, in Poboya gold mining area in Palu, Central Sulawesi, Indonesia, to reveal the magnitude of mercury pollution both in water and sediment and its potential risk to human health due to the consumption of fish contaminated with mercury. Gold mining activities have been started with more than thousands of gold mining processing plants since 2008 and cause mercury used in gold separation process may enter the environment.

The concentrations of mercury in the river water from the Poboya River ranged from 34 ng/L to 333 ng/L. While in Palu River, the concentrations ranged from 44 ng/L to 264 ng/L. Among all sampling points, the highest mercury concentration was at downstream of the gold mining activities. In sediment, mercury concentrations reached a value of 0.56 mg/kg DW. In August, 80 % of mercury collected from the Poboya River sediments exceeded the mercury concentration guideline which is set up by the US EPA (0.2 mg/kg). Whereas, the mercury concentrations in Palu River were lower than the guideline because the wastewater derived from the gold processing plants might not enter to the river directly. Wastewater from gold processing plants in Poboya area contained mercury up to 17400 ng/L. This concentration exceeded the wastewater discharge limit permitted by the US EPA (10000 ng/L). Mercury concentrations in the three sources of water consumption (well water, spring water and water supply) were 50 ng/L, 14 ng/L and 254 ng/L, respectively. The water supply had the highest mercury concentration among those three sources of water consumption. However, these concentrations were still lower than the Indonesian drinking water standard (1000 ng/L).

In water samples from the Palu Bay, mercury concentrations varied from 31.5 ng/L to 82.8 ng/L; while, in the sediment ranged from 0.002 mg/kg DW to 0.06 mg/kg DW. The highest concentration of mercury in the water samples was found at the Poboya River mouth in which mercury was transported from the river to the bay. Total of 21 fish specimens of 3 fish species had mercury concentrations ranged from 0.03 mg/kg to 0.22 mg/kg. These values were still below the mercury concentrations allowed for fish consumption (0.3 mg/kg) by the US EPA.