2d-1: Mercury fate in aquatic and terrestrial food webs

Freshwater biofilms are natural assemblage of microorganisms, such as microalgae, bacteria and fungi, which live together in symbiosis in an exopolymeric substances matrix. Besides their implication in the ecology and biochemistry of aquatic systems, biofilms are defined to be one of the entrance of mercury (Hg) along the trophic chain. The objective of the present study was to investigate the kinetics of Hg accumulation (total and non-extractable) in biofilms of different compositions under controlled conditions during 24 hours. To that end, biofilms were grown on microscopy glass slides for 85 and 127 days in microcosms supplied with Geneva lake water and exposed for 24 hours to about 150 pM Hg concentrations (precisely measured). Both biofilms were examined for their Hg content as a function of the exposure time (10 min, 30 min, 1 h, 4 h, 6 h, 8 h and 24 h). In addition, biomass, chlorophyll content and the abundance of specific genes (16S. merA and HgcA) were also analyzed in both biofilms. Physico-chemical parameters of the exposure water were characterized such as temperature, pH, anion/cation, dissolved organic matter, dissolved Hg concentrations. For both biofilms, Hg accumulation was rapid (<10min) and total Hg concentrations in biofilms were comparable in the first hours of exposure. However, after 24 h exposure, these concentrations were 886 ± 179 pmol gdw-1 and 1528 ± 73 pmol gdw-1 for the young and old biofilm, respectively. For both biofilms, non-extractable Hg content was modelled using a non linear pseudo first order one-compartment model. The uptake rate constant of the young biofilm was 10 times higher than for the old biofilm. Furthermore, Hg accumulation in the young biofilm reached a plateau, allowing the calculation of a clearance rate constant, whereas Hg content in the old biofilm remained linear. These results might be explained by a larger thickness of the EPS matrix in the old biofilm, which was assessed with the lower biomass but similar surface coverage area in the young biofilm than in the old biofilm, indicating that the old biofilm grown in the third dimension. The thicker matrix of the old biofilm might thus limit the diffusional flux of Hg from water to the microorganisms, decreasing the constant Hg uptake rate. The 24h-Hg exposure was additionally found to induce a decrease of chlorophyll content in both biofilms, suggesting a high impact of microalgae. The present study demonstrates that biofilm composition influence Hg accumulation.

Dietary transfer of mercury (Hg) is considered as having the highest importance for its effects on higher trophic animals, but few studies have addressed the precise factors controlling this transfer as well as its mechanisms and characteristics. Recently we observed that Hg found in cell walls of macrophytes were a significant source of Hg to gammarids. Because cell wall constituents are recalcitrant by nature and are expected to remain for a long period in sediments after plant decay, we made the hypothesis that benthic organisms might as well be exposed to Hg found in tissues of macrophyte. Here, we assess how the Hg accumulated in a representative macrophyte, Elodea nuttallii was transferrable to Chironomus riparius, chosen as surrogate species for benthic organisms. Three different concentrations, comparable to those found in macrophytes from unpolluted and contaminated sites, of inorganic Hg (IHg) and methyl-Hg (MMHg), and 2 compartments (cell wall, intracellular) were used as unique food source for C. riparius. Toxicity, bioaccumulation and subcellular toxicokinetics were measured.

We observed an effect of plant intracellular Hg on chironomid growth, caused by both IHg and MMHg at the higher concentrations (11.69 ± 4.91 mg·kg-1). Bioaccumulation measured in C. riparius revealed that MMHg was more transferable than IHg in our conditions. Data also support that Hg in cell walls was more transferable than intracellular Hg, when chironomids survived the treatment. Besides, after 10-d-long exposure, MMHg was predominantly found in MMHg form in the cytosolic fraction (S), considered as the metal-sensitive fraction, while IHg showed similar concentrations in S and insoluble debris (P) fractions. After 2-day-long depuration, only MMHg showed a BAF > 1 both in S and P fractions in C. riparius. Our data support that both intracellular and cell wall accumulated Hg are dietary transferable from macrophytes to macroinvertebrates. Besides, both IHg and MMHg are dietary transferable from macrophytes to macroinvertebrates, but MMHg is more transferable to sensitive subcellular targets in chironomids than IHg, in line with field studies showing a 2- to 4-fold higher transfer of MMHg than IHg in food webs. Our data hence point out the macrophytes as a new important source of Hg transfer to benthic consumers in local food webs to be considered for enhancing aquatic environment protection.

Mercury (Hg) is a toxin shown to have serious health effects for humans and animals. While Hg is a global problem due to the world-wide distribution of anthropogenic Hg emissions, and local pollution from industrial processes, Asia represents the majority of current and future emissions and hence requires particular attention in Hg research. In this study in China, we concentrated on spiders, can play an important role in the bioaccumulation of Hg in the food web, as they themselves are carnivorous and then are consumed by birds, in rice paddies ecosystems, known to be vulnerable ecosystems for Hg bioaccumulation. In 2014 we sampled soil and invertebrates over a gradient of distances from the abandoned mines and at a control site also in Guizhou Province, but without a history of mercury mining, at the end of the rice harvest. In addition, to test the hypothesis that Hg accumulation in the ecosystem is seasonal, we collected separate samples at different seasons in 2015. We measured THg and MeHg levels in range of invertebrates. In the 2014 data, we uncovered a pattern of THg and MeHg concentrations in spiders and grasshoppers that decreased with increasing distance from the abandoned mines. For all species, the distance to the mines significantly influenced THg. Patterns in the percentage of THg that was MeHg were complicated. The spider species varied in their THg and MeHg levels, with orb weavers generally have higher concentrations, and this pattern was correlated to their trophic levels, as measured by C and N stable isotopes. In the 2015 data, THg concentrations in soil were stable during two seasons, but soil MeHg concentrations were significantly higher when rice was nearly mature. For all invertebrate species, and all sites, there was a strong seasonal increase in MeHg. While there was also an increase in THg concentrations, that increase was less than the increase in MeHg (MeHg was on average 2.15 times higher at harvest, compared to THg, that was on average 1.52 times higher). These results indicate that there is a pulse of MeHg in the ecosystem towards the end of the rice growing season, which through spiders may flow into the vertebrate component of the food web. We argue that multinational studies need to be conducted in rice-based ecosystems throughout Asia to further investigate whether there may be an undocumented loss of biodiversity related to Hg local pollution and deposition in Asia.

Methylmercury (MeHg) is the toxic and bioavailable mercury species of greatest interest to monitor in aquatic biota, but laboratory analysis of MeHg in biological tissue is much more expensive than is analysis for total Hg (THg). However, when the percentage of MeHg relative to THg (hereafter MeHg%) is high and well-established, as in fish tissue (generally > 90%), THg analysis results can adequately represent MeHg concentrations. Aquatic macroinvertebrates have great potential as MeHg sentinels in monitoring programs. They are important in MeHg trophic transfer to fish, and are often abundant and relatively easy to collect. However, macroinvertebrates are not widely used in MeHg monitoring because the suitability of THg as a MeHg surrogate in macroinvertebrate tissues has not been empirically established. The U.S. Geological Survey (USGS), with support from the New York State Energy Research and Development Authority, conducted a study to evaluate the suitability of THg as a MeHg surrogate for macroinvertebrates from streams in undeveloped settings by documenting taxonomic, spatial, and temporal variation in MeHg%. MeHg and THg concentration data from more than 300 samples were compiled from published USGS studies conducted in the Adirondack Mountain region of New York (nine streams) and, for geographic comparison, from Coastal Plain streams in South Carolina (four streams). The macroinvertebrate taxa from these samples were categorized by scraper, shredder, collector-gatherer, filterer, omnivore, and predator functional feeding groups (FFGs). Median MeHg% in Adirondack samples ranged from 35% in scrapers to 92% in predators, and varied significantly among feeding groups (Kruskall-Wallis test, F=103, p<0.0001). MeHg% in predators, mainly dragonflies (Odonata: Aeshnidae and Libellulidae) from Adirondack and Coastal Plain streams was typically high (>90%) and exhibited minimal taxonomic, spatial and temporal variation. In contrast, MeHg% within most of the lower FFGs considered was highly variable. Median MeHg% in the northern case-maker caddisfly (Trichoptera: Limnephillidae), a shredder, collected from six Adirondack sites ranged from 18% to 72%, and differences were highly significant (F=18.0, p<0.0001). Median MeHg% in the flat-head mayfly (Ephemeroptera: Heptageniidae), a scraper, collected from five Adirondack sites ranged from 28% to 55%, and was significantly higher in summer than in spring from the same site (Wilcoxon rank-sum test, p<0.005). These results indicate that THg has good potential to function as a surrogate for MeHg in dragonflies and other predatory aquatic macroinvertebrates in streams of largely undeveloped watersheds, but MeHg analyses are warranted for taxa that feed at lower trophic levels.

The Dragonfly Mercury Project spans more than 200 unique locations across over 60 national parks throughout the US. Since 2014, parks have sampled dragonfly larvae for total mercury (THg) and methylmercury (MeHg) as well as surface-water (filtered THg and MeHg, DOC, sulfate, SUVA, and pH), and sediment (THg and MeHg) to determine: (1) baseline Hg concentrations in national parks prior to implementation of national and international mercury use and emission controls; and, (2) which habitats and catchment characteristics influence vulnerability to Hg accumulation in foodwebs. Dragonfly larvae are useful bio-sentinels for Hg spatial patterns because they are widespread in freshwaters, exhibit high site fidelity, and as predators, contain almost all of their Hg as MeHg (83% in this research). In an initial subset of the data including 127 freshwater lakes, ponds, wetlands, streams, and river sites in 34 national parks, THg in dragonfly larvae was 141.1±2.5 (mean±SE) ppb, dry weight (dw). We observed 76-fold variation between the sites with the greatest (>1000 ppb, dw) and least (~20 ppb, dw) mean dragonfly larval THg concentration across parks, and up to 44-fold variation among sites within a single park, highlighting the importance of spatial variability at the scale of individual water bodies. Preliminary analyses in Olympic and Acadia National Parks, where >10 water bodies of varying size, type, and hydroperiod were sampled, indicated that stream sites had the greatest dragonfly THg concentrations as compared to wetlands, ponds, and lakes. We observed taxonomic variation in dragonfly MeHg:THg only for one of the six families present in these parks. Larvae of the family Aeshnidae, which tend to be aggressive predators that can consume larval frogs and small fish, generally had greater MeHg:THg than larvae of other families. Linking water chemistry parameters with dragonfly and sediment THg allows for comparison of modeled vulnerability to methylation with actual geographic patterns that include these point-specific, layered MeHg production drivers. Although use of dragonfly larvae as bio-sentinels for Hg continues to be a promising approach and allows for Hg assessment at sites that are fishless, it is important to statistically account for habitat, taxonomic, and ontogenetic differences in interpreting and comparing data among sites.

Since the 1980s, predatory Indo-Pacific lionfish (Pterois volitans/miles) have successfully invaded and rapidly spread throughout the western Atlantic Ocean and Caribbean Sea. Invasive lionfish can be detrimental to coral reefs and fisheries, and efforts to control lionfish population size include targeted removals and spearfishing derbies. Spearfishing and consumption of lionfish are increasingly popular and effective tools for maintaining lionfish populations at reduced levels. However, few studies have examined whether consumption of lionfish poses any danger to human health through exposure to contaminants, such as mercury. Here, we analyzed lionfish collected along the southern coast of Curacao, relative to the island’s oil refinery (Isla Refinery), for total mercury (THg), stable isotopes of nitrogen (d¹⁵N) and carbon (d¹³C), and stomach contents. Fish ranged in size from 86 and 331 mm standard length and total mercury concentrations ranged from 0.008 to 0.106 ppm. THg significantly increased with fish length and fish body condition, suggesting bioaccumulation of mercury as lionfish age. THg also increased linearly with d¹⁵N, but not d¹³C, suggesting lionfish feeding on higher trophic level prey resulted in greater mercury bioaccumulation, but different carbon sources did not differ in mercury concentration. Stomach content analyses complement these findings, as larger lionfish with higher THg consumed more vertebrate prey items than smaller, younger lionfish. THg did not significantly differ across sampling sites, indicating that Isla Refinery has no direct impact on lionfish mercury levels in Curacao. However, sampling site was related to d¹⁵N; specifically, Punda, a sampling site located directly off the coast from Isla Refinery, had significantly higher d¹⁵N than all other sampling sites. d¹³C was significantly different across sampling depths and, in particular, was enriched in lionfish captured at the shallowest (5-15m) depth range, depths approximately shallower than Curacao’s fringing reef. Overall, these results indicate that lionfish around the island of Curacao contain low concentrations of total mercury, even in the presence of a local oil refinery.

Native fish populations, including four federally endangered fish species (Colorado Pikeminnow, Bonytail, Humpback chub, Razorback sucker), in the Upper Colorado River Basin (UCRB), have endured many threats, including altered natural flow and thermal regimes, fragmentation of habitat by dams, invasions by nonnative species, and climate change. Exposure to contaminants has been understudied in the UCRB, largely because of its remoteness and high proportion of federally managed lands. However, global transport of mercury (Hg) and deposition within the basin has become a concern. In addition, the UCRB has naturally high background concentrations of selenium (Se) in the surrounding soils and bedrock. Within the basin, mining and combustion of coal may be a local source of Hg while an expansive conversion of arid land to irrigated land has exacerbated transport of Se to the riverscape. Both Hg and Se can be toxic to wildlife, but in combination Se may abate Hg toxicity. In response to these concerns, the U.S. Geological Survey has compiled and analyzed a historical database of fish tissue Hg and Se concentrations for the UCRB. Over 2000 fish tissue samples have been collected basin-wide since 1962; however, 73% of these samples were collected in only 8 of the last 26 years, and only 12% were from federally-endangered fishes. Historically, average basin-wide Hg concentrations in two fishes of conservation concern (i.e., Roundtail Chub and Colorado Pikeminnow) were >2.4X the US EPA human health standard of 0.3 ppm; whereas only 1 native fish species, speckled dace, averaged above the 2016 US EPA tissue standards for Se. Considerable geographic variability in Hg and Se concentrations in fish was observed among the eight UCRB subbasins. The White-Yampa River and the Lower Gunnison River basins had the highest average (pooling across species) Hg and Se concentrations (respectively). There was also considerable variability in Se:Hg ratios among subbasins, however typically ratios exceeded 5. These relatively high Se:Hg ratios suggest that Hg toxicity is abated by Se, however, locally high Se concentration may pose risks to native fish. Preliminary findings suggest that some imperiled fishes of the UCRB face potential risks associated with high Hg and Se exposures. In 2016, we augmented the historical Hg and Se data by collecting more than 1000 samples, including >300 from native endangered species. These samples and those collected in the coming years will help identify if Hg and Se exposures pose risks to native fishes of the UCRB.

Mercury (Hg) contamination in freshwater fish is a severe environmental problem throughout the world, including the Scandinavian region. Health advisory limits are often exceeded in perch (Perca fluviatilis) in Sweden and Norway, even in remote boreal lakes. Historically deposited and catchment stored long-range transported atmospheric Hg is the main pollution source to the lakes. The last 20-30 years, fish Hg concentrations show patterns – including long-term increases – that currently lack an explanation. Generally, absence of data from the same lakes over time leads to confusing temporal patterns drawn from spatial variation.

We designed a sampling campaign to revisit lakes with historical data on Hg in fish along a north-south gradient in Norway and Sweden. The lakes have been subject to varying degree of browning and climate warming, enabling us to test hypotheses on factors driving Hg levels in fish. In freshwaters, elevated concentrations of Hg appear to be connected to humus-rich waters, which makes a connection between the recent rise in surface water colour and increased Hg in fish plausible. Similarly, a warmer climate may potentially affect Hg concentrations in top predators through a range of direct and indirect processes.

We selected 27 lakes that 1) had Hg in fish data available from the 1990s; 2) were located in three different climatic regions, i.e. South-boreal, Mid-boreal and Subarctic (from 59.0° N to 69.5° N); 3) had gradients in dissolved organic carbon (DOC) concentrations, pH and productivity within each region. Sampling was conducted during the autumn of 2016. We tested hypotheses on the main drivers of temporal trends of Hg in fish in Sweden and Norway by collecting data on catchment characteristics, water chemistry, fish, gridded downscaled weather data, and reconstructed DOC trends based on monitoring data and process-understanding of DOC drivers.

All lakes were sampled in the early 1990s, the 2000s and 2016. Temporal trends were identified using a covariance analysis, creating general linear models. Explanatory variables included season, year, fish morphology and stable isotopes (δ15N and δ13C), in addition to fatty acids where available, to test for dietary preferences. We will show robust tests of hypotheses on effects of climate and browning on Hg in fish based on these datasets. In addition, we will be able to test effects of short-term climatic variation on annual and seasonal variation in Hg in fish from a sub-set of lakes, using growth and trophic status as additional variables.