3c-1: Risk assessment of mercury exposure to fish, birds, and wildlife

Assessing risk to a wildlife population from a toxicant is generally done in one of two ways. Researchers may compare the tissue concentrations, health, or reproductive success of populations at contaminated and reference sites. Alternately, dosing may be carried out on a model species in captivity to measure effects on reproductive success or physiological condition. Both methods have generated a wealth of knowledge on the mechanisms through which mercury harms wildlife, and the scale of this global problem. However, both research strategies are likely to underestimate the effects of the contaminant and to suggest toxicity reference values that are not protective of typical populations. This is because when studying natural populations exposed to environmental contaminants, researchers are studying only the surviving lineages, which likely have experienced strong natural selection for contaminant resistance. This problem can be eliminated by experimentally exposing naïve subjects in the laboratory. However, the lives of laboratory animals lack many of the stressors facing their wild counterparts, such as predation, competition, weather, and challenges that require cognition to overcome (e.g., the need to remember spatial locations). Thus, contaminant exposure in captivity may fail to cause harm at levels that would be deleterious in the wild. Using data from recent studies on mercury-exposed songbirds illustrates these points. Avian mercury toxicity reference values, recently argued by some to be too low, are likely too high if they are derived from resistant wild populations or captive animals with unchallenging lives. For more accurate results, dosing studies on captive animals should provide housing conditions with multiple challenges such as temperature fluctuations, competition for mates and hidden food. Dosing could also be attempted on free-living animals at uncontaminated sites, which circumvents the problem of studying resistant populations. At contaminated sites, new populations of non-resistant individuals can be recruited, for example by using nestboxes to attract tree swallows to habitats that did not support them beforehand. All of these techniques are more difficult and expensive than current methods, but if we are interested in identifying meaningful effects thresholds, continuing with the current strategy is imprudent.

High levels of mercury in the San Francisco Bay region pose a threat to local wildlife. Recent studies have documented unexpectedly high concentrations of methylmercury in terrestrial invertivores indicating possible risk from exposure via the food web. Such discoveries prompted interest in developing a biosentinel for methylmercury in stream riparian food webs of the Bay Area. The Song Sparrow (Melospiza melodia) was determined to be the best riparian biosentinel candidate on the basis of its natural history, sampling feasibility, and sensitivity to mercury. The ability of Song Sparrows to reflect a wide range of mercury concentration in their blood, and thereby reflect differences in methylmercury risk between areas, was assessed by sampling individuals from riparian sites across the region. The sampling design was based on a conceptual model in which the key drivers of biosentinel exposure were Total Mercury Contamination of sediment (Total Mercury Contamination) and physicochemical environmental conditions associated with production of methylmercury (Net Methylation Environment). Net Methylation Environment was assumed based on landscape-level indicators. Both Total Mercury Contamination and Net Methylation Environment were found to influence biosentinel mercury concentrations. Results of this study confirmed the appropriateness of the Song Sparrow as a riparian mercury biosentinel species.

Mercury (Hg) contamination in aquatic and terrestrial ecosystems is a widespread issue that poses considerable reproductive, behavioral and physiological risks to wildlife populations. Songbirds are now recognized as indicators of mercury in terrestrial ecosystems, where invertivore foodwebs biomagnify methylmercury (MeHg) to levels that can adversely affect reproductive success. With upcoming changes to Hg emissions regulations, understanding how MeHg bioavailability currently varies across the landscape is important for quantifying the effects of these changes. Building upon 13 years of mercury research in New York State, a five-year project was initiated in 2013, to identify at-risk songbird species, classify sensitive habitat types, and to assess spatial and temporal trends of mercury across a variety of ecosystems. To date, approximately 1,900 blood and feather samples have been collected and analyzed from songbirds across the state. This project focuses on: (1) annual sampling and monitoring at established study sites to evaluate temporal patterns in songbird mercury exposure; (2) sampling selected sites statewide to identify mercury hotspots for inclusion into a predictive map documenting spatial gradients of methylmercury availability; and (3) analysis of museum specimens to quantify trends in Hg exposure over the 20th Century.

There is ample evidence that methylmercury (MeHg) exposure can cause reductions in avian fitness, particularly in aquatic species during breeding. However, recent work indicates that terrestrial songbirds that live in association with MeHg contaminated environments can bioaccumulate MeHg to high levels in body tissues . How elevated MeHg levels in migratory songbirds affect flight performance and migration is still unknown. We conducted a MeHg dosing experiment (target concentration: 0.5 ppm and 1 ppm (w.w.) actual concentration: 0.38 ± 0.04 ppm and 0.73 ± 0.08 ppm, w.w.)with yellow-rumped warblers (Setophaga coronata) while they were in a hyperphagic, migratory state. Warblers rapidly bioaccumulated dietary MeHg by 20-fold in blood (w.w.), 40-fold in brain and muscle (d.w.), 60 fold in liver (d.w.), and over 100 fold in kidneys (d.w.) after 2 weeks. These increases in MeHg concentrations did not affect vertical takeoff, but in two-hour wind tunnel flights, MeHg treated warblers had a greater median of numbers of strikes (landing or losing control) in the first 30 minutes, longer strike duration, and shorter flight duration than control birds. In MeHg treated warblers, the number of strikes in first 30 minutes was related to blood Hg concentration in a sigmoid, dose-dependent fashion with threshold near 11 ppm. From our results, we conclude that hyperphagic migratory birds rapidly accumulate MeHg from the diet, which can subsequently lead to decreased flight performance and migratory endurance.

The National Institute of Standards and Technology (NIST) Seabird Tissue Archival and Monitoring Project (STAMP) expanded to the Pacific Islands Region in 2010 and began collecting Laysan albatross (Phoebastria immutabilis) and black-footed albatross (Phoebastria nigripes) eggs from the Northwestern Hawaiian Islands (NWHI) on Midway Atoll in 2011. Laysan albatross and black-footed albatross are sympatric, synchronous breeders that nest in dense colonies on Midway Atoll. Abandoned, non-viable eggs were collected by USFWS staff and volunteers. Definitive species identification of abandoned albatross eggs can be difficult on NWHI due to close nesting colonies and visually indistinguishable eggs between species resulting in some uncertainty of the egg species by field collectors. The original intent of this study was to determine mercury concentrations in eggs of albatross species for Midway Atoll. Total mercury was measured by atomic absorption spectrometry (AAS) in an aliquot of the homogenized egg content from each albatross egg collected (n = 57) from 2011 to 2013. Mercury analysis showed significantly (p <0.001 ANOVA) greater concentrations in black-footed albatross eggs compared with Laysan albatross eggs, with some notable exceptions, raising questions about the accuracy of species identification in the field. Since species identification of some eggs was questionable and the species for several eggs were not identified, genetic analysis was performed on eggshell membranes where DNA was recoverable (n = 50). Genetic species identification matched suspected species based on mercury concentrations with 100 % accuracy. Trace element concentrations have been used as an alternate for genetics in other species such as fish for population structure delineation. Mercury may therefore serve as an alternate method of species identification for sympatric breeding albatross species that have significantly different concentrations due to prey and foraging grounds differences when genetic material is not available. Since mercury is routinely measured in STAMP eggs, this would be an attractive alternative to expensive genotyping. To test whether mercury can serve as an alternative to genotyping for NWHI albatross eggs, albatross eggs collected (n = 41) on Midway Atoll in 2014 and 2015 are currently being measured for mercury.

We evaluated dietary Hg exposure in Bald Eagle (Haliaeetus leucocephalus) nestlings throughout Maine’s Penobscot River watershed in 2004 – 2015, focusing particularly on exposure relative to the location of a significant Hg point source (HoltraChem) located in a brackish portion of the Penobscot River. The Penobscot River watershed encompasses more than one-quarter of the state, stretching over 150 km from northern lakes to an estuary and Penobscot Bay. Geometric mean Hg concentrations in nestling blood (range 0.06 – 1.51 µg/g) and feathers (range 2.9 – 46.8 µg/g) varied significantly across the four major habitat types in the study area (p < 0.05). Means generally followed the pattern: lakes > freshwater rivers > brackish rivers > marine. The overarching influence of habitat type on Hg exposure in Bald Eagle nestlings masked detection of the possible influence of Hg exposure attributable to HoltraChem. We also compared reproductive measures (productivity, young fledged / occupied nest; nest success, proportion of nests successfully fledging ≥1 young) to eagle tissue Hg concentrations and will discuss those relationships and influential factors. Bald Eagle tissue Hg concentrations found in lakes and freshwater rivers in the Penobscot River watershed are similar or higher than virtually all comparably sampled Bald Eagle populations elsewhere. Eagles at lakes in the Penobscot River Watershed were most similar to a site in BC Canada associated with a Hg mine. In contrast, eagle sampling sites below HoltraChem, one of the most substantial Hg pollution point sources in the country, exhibited relatively low concentrations of Hg due to the predominant influence of factors associated with habitat type. We attribute the notable exposure of Hg in lake- and freshwater river-dwelling Bald Eagles in the Penobscot River watershed upstream from HoltraChem to the combined influences of atmospheric deposition, other point sources, and site-specific biogeochemical factors. Bald Eagles have been repeatedly proven to be valuable contaminant bioindicators at site-specific and watershed scales in both freshwater and marine ecosystems; however, eagles’ large home range and dietary plasticity limits their use in smaller scale, traditional ‘upstream vs. downstream’ contaminant risk evaluations, particularly those spanning multiple habitat types.

Mercury (THg) is a pollutant of global concern. Sentinel species, such as river otter and mink, are often used to monitor environmental concentrations. Tissue THg concentrations are often used as biomarkers of exposure. However, there is no comprehensive model relating mercury tissue concentrations in different tissues, making inter-study comparisons challenging. Our objective is to establish universal conversion factors relating fur, brain, liver, kidney, and muscle THg tissue concentrations using mean concentrations and standard errors reported in literature. Data from over 6000 samples, pooled across 16 studies, and 96 sampling sites in North America and Europe were used. A total of 16 regressions were derived for the river otter and mink models which were statistically significant at a 95% confidence interval and yielded high explained variance. The models were validated using an external data set of individually measured THg tissue concentrations. The validated conversions were used to evaluate the current 20 μg/g and 30 μg/g fur mercury screening guideline. At both of these fur concentrations, brain concentrations are of concern for altering brain neurochemistry. We suggest a more conservative 15 μg/g fur mercury screening guideline to protect more sensitive furbearers. The model conversion factors can be used to predict internal organs from fur, thus eliminating the need to collect invasive tissue samples (e.g. brain or liver) for future monitoring programs.

We reviewed effects of chronic methylmercury exposure on non-primate mammals, with the goal of identifying toxicity reference values for the ecological risk assessment of mammalian wildlife populations. The review focused on dietary exposures in which effects on reproductive success or survival were investigated. Issues related to applying toxicity study results in ecological risk assessments were also reviewed. Relevant toxicity data were available primarily for mink and rats. Dose-response relationships and effect thresholds will be summarized. In contrast to birds, mink exhibited little difference in sensitivity between survival and reproductive endpoints, suggesting that methylmercury effects on reproductive success may be secondary to more general systemic toxicity. Important considerations in applying the available mammalian toxicity data include interspecies differences in mercury tolerance adaptations, mercury-selenium interactions, methylmercury bioaccessibility, and effects of body weight differences on dose extrapolation. Marine mammals are well adapted to demethylate, sequester, and eliminate mercury; data to quantify methylmercury exposure levels that would overwhelm these tolerance mechanisms remain elusive. Susceptibility of bats to methylmercury toxicity is a matter of increasing interest, and uncertainties in dose extrapolation are particularly significant for bats due to their small body sizes and high food ingestion rates.