G15 (I) Mercury in Fish

At the La Grande hydroelectric complex, mercury concentrations were measured in over 25 000 fish as part of an ongoing environmental effects monitoring program. Five fish species were monitored over a 30-year period: the non-piscivorous lake whitefish (Coregonus clupeaformis) and longnose sucker (Catostomus catostomus) and the piscivorous northern pike (Esox lucius), walleye (Sander vitreus) and lake trout (Salvelinus namaycush). Total mercury concentrations were measured by standard cold vapour atomic absorption spectrophotometry. Fish population characteristics, such as recruitment, fishing yields, growth rates and condition factors were also monitored over a 20-year period, as part of the same program. In reservoirs, mercury concentrations in all species increased rapidly after impoundment, peaking after 5 to 9 years in non-piscivorous fishes, and after 10 to 13 years in piscivorous species, at levels 3 to 6 times those measured in surrounding natural lakes, then significantly and gradually declined. Concentrations in most species have returned to levels typical of natural lakes, 10 to 30 years after flooding. Despite such increases, the monitored fish showed no mercury-related ill effects at the population level. Indeed, for most species, increases were observed in fishing yields (by factors ranging from 2 to over 20), in growth rates and in condition factors (for more than a decade). The post-impoundment increase in growth rate was particularly spectacular for lake whitefish of the Robert-Bourassa Reservoir, as the 400-mm length was reached after 5 years compared to 12 years in natural surrounding lakes. The percentage of small specimen of the main species was generally maintained or increased after flooding, indicating a good recruitment. This was particularly obvious for northern pike of the Robert-Bourassa reservoir for which the highest average mercury concentrations were measured (3.28 µg g^-1 for 700-mm fish), 11 years after flooding. The biological boom usually observed after reservoir impoundment may have prevented potential mercury-related ill effects.

Mercury is a global pollutant whose potential toxicity varies according to its different chemical forms. It is known that due to bioaccumulation and biomagnification, elevated Hg values can be detected in high trophic levels. Consequently, a fish based diet is considered the primary pathway of human exposure to methyl-Hg. This is of special concern in the Amazon region. The scenario of human mercury exposure in the Amazon has changed as gold extraction has diminished and it has been observed a decrease in human hair mercury concentration in the last decade. However it is important to understand that the Amazon basin represents a natural laboratory for studying the effects of low mercury exposure. The interaction with diet, clinical history, and susceptibility to endemic disease mechanisms still remain unclear. The complexity of the Amazonian ecosystem and the limited knowledge of mercury’s biogeochemical cycle in tropical rain forests explain the difficulties faced by the scientific community in assessing the impact of mercury contamination. These issues affect the potential for exposure of the local populations and consequently the risk involved. High mercury levels in fish have been reported in sites close to gold mining (where mercury was extensively used) as well as in areas with no history of such activity. The present work is part of a program of Health Risk Assessment of Hg in a riparian extractives reserve at the Cuniã lake, located alongside the left shore of the Madeira River (a tributary of the Amazon River). This area will be under indirect impact of the hydroelectric power construction as a consequence of the new economic profile of this region. Fishing is the main activity of this community, and there is no history of gold extraction. Fish is their main protein source, eventually complemented with caiman meat and water birds. The studied population age ranged from 2 to 90 years old and hair was used as biomarker. A total of 152 hair samples were collected from the fishermen families. Hg concentrations averaged 6.0 (IC95% 5.58-6.58) ± 3.9 µg.g^-1 (SD). The fish consumption for adults in this community range from 40 to 600 grams per day. The results showed that a significant gender differences for Hg levels (p-valor < 0,05), for man the average is 7,4 (IC 95% 5.56- 8.28)±4,5 µg.g^-1 µg/g and for woman 5.0 (IC 95% 4.49-5.54) ±3.3 µg.g^-1.Studies on neurological symptoms and diet of this community are under way along with other clinical trials and biochemical measurements.

China is the world’s largest emitter of mercury (Hg). As Hg hotspots often occur in the vicinity of local sources, it is necessary to investigate whether important Chinese inland fisheries reflect these large Hg emissions. Fish culture in China occurs mainly in reservoirs, ecosystems that are predisposed to high Hg accumulation in the food web if seasonal fluctuations in water levels are frequent.

We examined the temporal Hg concentrations and food web structure, using stable isotopes of carbon and nitrogen, of 25 fish species from a large, oligo-mesotrophic reservoir to determine Hg biomagnification in a food web of stocked and wild fish. Data showed that 1) overall, fish have very low Hg concentrations, especially in stocked species (e.g. Aristichthys nobilis and Carassius auratus). Wild fish generally have double the Hg concentrations compared to stocked fish, and two carnivorous wild species are regularly above the Chinese consumption guidelines of 0.3 mg/kg wet weight (Siniperca chuatsi and Silurus asotus); 2) overall fish Hg concentrations were significantly related to trophic position (i.e. delta 15-N) but not to organic matter pathway (i.e. delta 13-C); 3) for almost all species, season did not have an effect on food web position and Hg concentrations, except for two species with higher Hg concentrations in the season that corresponded with their higher delta 15-N value (Ictalarus punctatus – Fall; Pelteobagrus fulvidraco – Spring); 3) no effect of an indicator of lipid to protein ratio (C:N) was found on Hg among species, or within a species between seasons; and 4) biomagnification of Hg, derived using the log10[Hg]-d¹⁵N slope, was about 0.1. Though we observed similar Hg concentrations for many reservoir species from other regions of China, we believe species feeding preferences do lead to important differences in Hg exposure for some fish. Differences in biomagnification of Hg in reservoirs from other subtropical regions will be evaluated and Hg concentrations compared to the same species outside of China. This study finds Hg content in wild Chinese fish can exceed consumption guidelines, despite very low concentrations in most reservoir species.

Fish mercury concentrations can vary considerably among species within a waterbody. For example, piscivores usually have much higher mercury concentrations than benthivores or herbivores. However, even within the piscivore trophic guild, mercury concentrations can still vary among species at a standard body size. We examined the relative roles of food web position and growth rate in accounting for interspecific variation in mercury concentration of four co-habiting, native piscivore species (walleye, northern pike, lake trout, burbot) in boreal shield lakes of northern Ontario, Canada. Linear models relating total mercury concentration (THg), to carbon source (inferred from d¹³C), trophic position (inferred from d¹⁵N) and growth rate in a standard size of fish (1 kg) were compared using Akaike’s Information Criterion (AIC). In general, variation among lakes was higher than variation among species. Following adjustment for lake effects, the piscivores divided into two distinct groups – those with higher growth rate and lower d¹⁵N (walleye, northern pike), and those with lower growth rate and higher d¹⁵N (lake trout, burbot). Models containing d¹³C or d¹⁵N ranked higher than those containing growth rate. But, models containing either food web variables or growth rate all ranked higher than a model where species was defined as a categorical variable. Much of the interspecific variation in mercury concentrations among boreal piscivores can be accounted for by subtle differences in food web position.

This presentation will provide an overview of our recent research on littoral–pelagic differences in mercury biomagnification in lake food webs. We compared methylmercury (MeHg) concentrations in pelagic zooplankton to those in littoral macroinvertebrates from 52 mid-latitude lakes in North America. Invertebrate MeHg concentrations were primarily correlated with water pH, and after controlling for this influence, pelagic zooplankton had significantly higher MeHg concentrations than littoral primary consumers but lower MeHg than littoral secondary consumers. Intensive sampling of eight lakes indicated that habitat-specific bioaccumulation in invertebrates (of similar trophic level) may result from spatial variation in aqueous MeHg concentration or from more efficient uptake of aqueous MeHg into the pelagic food web. These findings showed that littoral–pelagic differences in MeHg bioaccumulation are wide spread in small mid-latitude lakes. We also investigated the consequences of MeHg variation in prey for its transfer to four fish species in one of the study lakes in Quebec, Canada. The dietary pathways in the food web were characterized using fish stomach contents and a Bayesian stable isotope mixing model. The diet analysis showed MeHg concentrations in fish increased with piscivory as well as greater consumption of pelagic invertebrates. While trophic position is widely recognized as a key driver of fish MeHg concentrations, habitat use is likely also an important element of food web structure for MeHg bioaccumulation in many small, temperate lakes of North America.

There is considerable interest in the Hg dynamics of the peatland-dominated watersheds of Canada’s Hudson Bay Lowlands in light of current and future climate and land-use changes. First Nations residents of Northern Ontario harvest fish as an important food source. Determining the factors governing Hg exposure in this ecosystem is important to the health and well-being of these residents. Little data exist on mercury and methylmercury in the various abiotic and biotic components of the freshwater ecosystems of the region, making assessments of future change difficult. Young-of-year (YOY) fish are useful for monitoring changes in mercury (Hg) exposure in aquatic ecosystems because their tissue Hg concentrations reflect their juvenile (short-term) exposure that is not masked by cumulative bioaccumulation as would be the case in larger adult fish. This research focuses on the use of YOY fish as sentinels of spatiotemporal variability in aquatic ecosystem Hg exposure.

As part of an ongoing monitoring program at the De Beers Victor Mine near Attawapiskat, Ontario, YOY fish have been collected annually since 2008 for Hg analysis. To date, over 1,500 YOY fish have been collected from first- and second-order tributaries and rivers (Attawapiskat and Nayshkatooyaow Rivers and their smaller tributaries). Muscle Hg concentrations of these individual fish, water quality measures (dissolved organic carbon (DOC), total mercury (THg) and MeHg), and (sub)watershed hydrogeomorphic information will be presented. Analysis of the data shows evidence of with-species variability over a relatively small geographic region, confounding the concept of ‘reference’, or ‘unimpacted’ sampling locations and populations. The research presented will address the relative importance of biogeochemical, hydrological and ecosystem controls on the spatiotemporal variability of Hg in these YOY fish. This unprecedented dataset will provide the framework for determining whether climate and land-use changes influence fish mercury dynamics and generate a baseline for the ongoing monitoring of Hg in Ontario’s far north.

Environmental and health advisory limits of Hg concentrations ([Hg]) in freshwater fish (0.2-0.5 mg kg^-1) are exceeded in thousands or even tens of thousands lakes. Temporal trends and deviations from background in fish [Hg] were examined using data from Swedish environmental monitoring programs. Temporal trends were detected using lake-specific, multiple linear regression models to estimate fish [Hg]. Significant temporal trends were detected in a minority of the analysed lakes with an average increase in fish [Hg] of 4% per year. The relative increase in fish [Hg] was positively correlated with dissolved total organic carbon. The performance of statistical models in detecting temporal trends depended on the number of fish sampled from each lake thus indicaing the importance of sampling efforts over time for the ability to reveal temporal trends. Deviations in contemporary fish [Hg] from pre-industrial fish [Hg] were determined using sediment cores. A sub-selection of the lake sediment cores was dated using lead-210. Hg concentrations were higher (0.09-0.72 mg kg^-1) in surficial sediments, representing an average contemporary Hg deposition rate, compared to that in deeper layers, (0.03-0.29 to mg kg^–1) representing pre-industrial (≈1850 A.D.) Hg deposition. Fish [Hg] appears to have increased in lakes in southern Sweden while lakes in northern Sweden had a lower degree of increment above the modelled background level. The patterns revealed in this study contribute to the understanding of human-induced changes in water quality and ecosystems affect Hg exposure and accumulation within freshwater ecosystems and if measures to decrease Hg exposure have been successful.

California has more than 1,200 miles of ocean shoreline that supports an active recreational fishery. An estimated 16 million fish were landed by recreational anglers in 2009. Despite a number of localized surveys of tissue contamination, there has never been a statewide survey of mercury levels in sportfish tissues. The goal of this study was to assess the proportion of popular fishing zones statewide with safe levels of seafood concentrations and to determine the spatial distribution of contaminant concentrations. More than 500 samples were collected, comprising over 2 dozen species, and edible tissues analyzed for total mercury and chlorinated hydrocarbons such as polychlorinated biphenyls (PCBs) and dichlorodiphenyltrichloroethane (DDTs). Total mercury concentrations ranged from <0.012 to 1.3 parts-per-million wet weight (ppm). The highest concentrations were found in sharks, followed by pelagivores (i.e., rockfishes, basses), with the lowest concentrations observed in planktivores (i.e., surfperches). Nearly all of the fishing zones sampled exceeded the lowest risk guidelines recommending limited consumption for kids and women of child bearing age for at least one species. Approximately 20% of the fishing zones exceeded the highest risk guidelines that recommend no consumtion of sportfish for at least one species. Risk guideline exceedences occurred more frequently, and at greater magnitude, for mercury compared to PCBs or DDTs even though large hot spots for these chlorinated hydrocarbons occur in California. Spatial distributions identified a propensity for greater concentrations and extent of risk guideline exceedences along coastline near large urban centers such as San Francisco and Los Angeles; however, risk guideline exceedences occurred even at offshore islands distant from known point sources.