1c-2: Stable isotope studies of global mercury cycling and bioaccumulation

River and lake subsistence fisheries nourish hundreds of millions of people globally, particularly in nations with high poverty rates. Food security in freshwater fisheries could be undermined by accumulation of mercury (Hg) and other contaminants in these food webs. We find that the mean Hg concentrations in food fishes from freshwater locations in Central Africa, Thailand, Venezuela, and Hawaii exceed limits for safe consumption up to ten-fold, yet have no obvious local source for the elevated levels. The aim of this study is to utilize Hg stable isotopes to interpret source similarities and food chain dynamics that have led to these unusually high Hg concentrations. We compare 12 distinct freshwater communities that support food fishes on four continents. Mercury isotope results demonstrate striking similarities between ecosystems with elevated Hg concentrations in Central Africa and Hawaii, including highly negative δ202Hg and Δ199Hg in fish tissue. These signatures likely arise from terrestrial Hg sources originating from volcanic activity and a lack of photochemical transformations. The lowest mean Hg concentrations in the study were observed in Africa’s Lake Tanganyika, one of the world’s largest and oldest lakes. Isotopically, Lake Tanganyika also had the highest Δ199Hg and δ202Hg values, which may indicate that photochemical demethylation is an important control on Hg levels in these fish. Although exact source tracking is challenging, significant Δ200Hg indicates that Hg is derived from atmospheric origin for all sites examined in this study. Additionally, preliminary results of different feeding guilds within in each food web show similar δ202Hg signatures after photochemical corrections, signifying that different trophic level organisms are still receiving Hg from a singular source. This study marks the first global use of stable isotopes to examine Hg sources in vastly different food webs, and clarifies patterns of Hg bioaccumulation in freshwater systems that may affect human health and food security.

The extent of Hg contamination in the Southern Ocean ecosystems remains largely unknown and determining its fate and impact in these remote areas involves a major challenge. Seabirds are exposed to large quantities of Hg via their marine food chain and have been identified as effective biomonitors of Hg marine contamination. Penguins, as non-flying seabirds, feed on approximately constant prey and exploit similar foraging habitats during their annual cycle, therefore, they are representative of local Hg contamination. Since they are ubiquist marine birds over the entire Southern Ocean, analyses of their tissues provide information about Hg contamination of the different latitudinal areas. In this work, Hg speciation (GC-ICPMS) and isotopic signatures (CVG-MC-ICPMS) were investigated following non-lethal sampling of feathers and blood in 7 penguin species from the marine environments around the French Southern Ocean territories. These lands cover a wide latitudinal gradient from the Adélie Land (66°39′S, Antarctic) to Crozet Islands (46°25′S, subantarctic) and Amsterdam Island (37°47'S, subtropical). Carbon (δ13C) and nitrogen (δ15N) isotopes were also determined to better understand the food web structure of the penguin species and Hg isotopic information. A good correlation of isotopic signatures between feathers and blood demonstrates that both tissues are valuable bioindicators of Hg local contamination. Mass dependent (MDF) and mass independent (MIF) fractionation signatures obtained in both tissues clearly separated populations geographically, permitting the identification of the distinct ecosystems studied. Feathers of Antarctic penguins displayed lower MDF (δ202Hg) values (0.4-1.0‰) than subantarctic (1.3-2.5‰) and subtropical (2.3-2.7‰) penguins along with increasing southern latitude. For MIF (∆199 Hg), less variation is observed and most values ranged between 1.3 and 2.3 ‰. These specific isotopic signatures for each penguins population are attributed to different foraging habitats (spatial or depth variations) and different extent of photochemical versus dark demethylation or reduction processes. Overall, this work demonstrates that Hg isotopic variations recorded in penguins’ samples provide new information on the major sources of methylmercury (MeHg) in the Southern Ocean ecosystems. For instance, MeHg accumulated in population from subtropical to subantarctic latitude seems to be mainly of marine pelagic origin, while in the Antarctic zone, bioaccumulated MeHg could originate from both coastal and marine waters.

Mercury contamination is observed in Arctic food webs despite the absence of significant local Hg point sources, making understanding Hg sources and dynamics in the Arctic an area of active research. It is understood that much of the Hg contamination in the Arctic is a result of long-range transport of atmospheric Hg. Thus, identifying and quantifying different sources is needed to predict how changing sources may affect Hg cycling in the Arctic along with making targeted strategies for mitigation. There is also unique chemistry that occurs in the Arctic that contributes to elevated Hg deposition. In polar spring, atmospheric mercury depletion events occur where gaseous elemental Hg is nearly completely oxidized and deposited to snow. Some fraction is photoreduced back to the atmosphere, which affects the extent these events impact the overall net accumulation of Hg. Stable Hg isotope fractionation can be used to trace, distinguish and potentially quantify various Hg processes. Natural Hg isotopes exhibit a large range in mass-dependent fractionation (MDF) and several types of mass-independent fractionation (MIF). MDF is ubiquitous in nature and occurs during many environmental transformations. In contrast, Hg MIF only occurs during a subset of reactions and is largely produced during photochemical transformations with different reactions displaying different MIF extent and signatures. Odd-mass MIF occurs during aqueous and possibly surficial photochemistry, whereas atmospheric gas-phase MIF occurs for all the isotopes and is reported as even-mass MIF. As part of an exploratory study, stable Hg isotopes in aerosols and snow were measured at Alert in the Arctic spring for years 2011, 2013 2015 to assess the potential of Hg isotopes to help understand and distinguish the various Hg sources and transformations. Preliminary Hg isotope data for aerosols and snow exhibit a large variability in MDF and MIF signatures. The aerosols are characterized by negative odd-mass MIF, which is in contrast to the published data that show mostly positive odd-mass MIF for atmospheric oxidized Hg species. Only three of the aerosol samples display significant even-mass MIF with most aerosols having negligible even-mass MIF. These results also differ from the current published data on atmospheric oxidized Hg species. All snow samples exhibit negative odd mass MIF, consistent with photochemical reduction in surface snow, and show no preservation of even-mass MIF. In addition, aerosol metal and lead isotope data were measured and passive samplers for gaseous Hg were deployed in 2016-2017. Results will be discussed.

Mass independent fractionation (MIF) of Hg isotopes in marine organisms has been used to estimate the relative proportion of monomethylmercury (MeHg) formed in the open ocean that is photochemically degraded prior to entry into the foodweb. Flying fish normally feed in the upper ~10 m of the ocean and a previous study has shown that they contain Hg with the highest level of MIF (Δ199Hg ) of all fish in the Northern Pacific open ocean food web. In this study we utilized flying fish as a monitor of the relative photochemical degradation of (MeHg) over large areas of the western Atlantic Ocean (12N to 26S) and the North Central Pacific Ocean (22N to 5N). The Δ199Hg values of the fish range widely between 2.7 and 5.5‰ indicating highly variable proportions of MeHg photodegradation. Based on experimental fractionation experiments and the flying fish Δ199Hg values we estimate that in the surface ocean between 55 and 80% of the MeHg formed is photochemically degraded. The Δ199Hg values do not display a simple relationship with latitude, but do display some spatial patterns. Three processes that may contribute to the observed spatial variability with respect to the amount of photo-demethylation were investigated, including: 1) the average amount of incoming solar radiation (e.g., anglet of solar incidence and cloud cover), 2) light penetration depth (depends on water clarity and is related to chlorophyll content) and 3) the role of dissolved organic carbon (DOC) (acts as complexation agent and photosensitizer). There was no correlation observed between Δ199Hg with solar irradiance satellite data (R2=0.12, p=0.25). In contrast, there is a strong negative correlation between Δ199Hg and satellite chlorophyll concentration (R2=0.67, p<0.001). DOC content was estimated from the MITgcm model and there was a moderate correlation between Δ199Hg (R2=0.43, p=0.015) and DOC. Lastly, we compare variability in Δ199Hg values to methylation and photochemical demethylation rates simulated within a global 3-D ocean circulation model (MITgcm). We find a strong negative correlation between Δ199Hg and each of these rates (R2=0.75, p<0.001 and R2=0.80, p<0.001 respectively). We suggest that the Hg isotopic signature of flying fish is dependent on the environmental conditions and biological processes at specific locations in the oceans and can be used to monitor the extent and mechanism of photochemical degradation of MeHg prior to its entry into marine food webs.

Relative to aquatic ecosystems, we have a very limited understanding of the biogeochemical processes of methylmercury (MeHg) in forests. In this study, we examined total mercury (THg) and MeHg and their isotopic compositions in food webs (from basal resources to predatory invertebrates) in four temperate forest ecosystem reserves across the United States, including Coweeta Hydrologic Laboratory in North Carolina (NC), Hubbard Brook Experimental Forest in New Hampshire (NH), University of Michigan Biological Station in Michigan (MI), and Angelo Coast Range Reserve in California (CA). We calculated the trophic biomagnification slope (TMS) of MeHg in forest food webs at each site, and found that the mean TMS values were 0.18 (NC), 0.20 (NH), 0.16 (MI), and 0.24 (CA), which are within the range of TMS values found for freshwater ecosystems (mean=0.16, range from -0.19 to 0.48), implying that biomagnification potential of MeHg is in general similar between freshwater and forest food webs. Stable Hg isotope measurements are still in progress for part of the study, but the isotopic data analyzed so far (n=58) among sites showed a wide range of mass-dependent fractionation (MDF, as d202Hg) ranging from -3.04 to +1.32 ‰ and a smaller range of mass-independent fractionation (MIF, as D199Hg) ranging from -0.44 to +1.72 ‰. When we examined the (incomplete) isotopic data at three forest sites (i.e., NC, NH and CA), we regressed d202Hg or D199Hg against %MeHg in each system and extrapolated to estimate endmember d202Hg and D199Hg of “pure” MeHg. If we assume the starting substrate for MeHg in forest ecosystems is inorganic Hg in the forest floor, the isotopic compositions (as estimated in surface litter samples) in these three sites are quite similar with d202Hg ranging from -2.53 to -1.98 ‰ and D199Hg ranging from -0.41 to -0.31 ‰ (n=7). However, it is intriguing that the estimated “pure” MeHg (based on the available data so far) in these three forests are very different: NC has d202Hg = -0.35 ‰ and D199Hg = +0.32 ‰, NH has d202Hg = +2.50 ‰ and D199Hg = +2.05‰, while CA has d202Hg = +1.15 ‰ and D199Hg = +1.55 ‰. Our preliminary interpretation on this isotopic data is that MeHg, before entering these forest food webs, underwent variable degrees of photodemethylation but it is not yet completely clear what differences in biogeochemical processes among sites leads to the variable d202Hg values of MeHg.

The mechanism of mercury (Hg) re-emission from vegetation foliage is currently poorly understood. Here, we systematically applied stable isotope technique to various compartments of a pristine subtropical evergreen forest ecosystem to gain insight into this important process. We observed that the Hg isotope signature of bulk leaf sprout samples from the dominant tree species (10-15 days old; δ202Hg = 0.08±0.74‰, Δ199Hg = −0.20±0.14‰, n=3, ±2σ) was similar in composition to surrounding gaseous Hg (Hg0) in ambient air (δ202Hg= 0.37±0.44‰, Δ199Hg = −0.18±0.04‰, n=9) suggesting that atmospheric Hg0 is a major source of foliar Hg. In relation to sprouts, mature foliage samples (0.5-1.5 years old, n=16) from the dominant tree species were substantially enriched in light Hg isotopes (negative MDF, δ202Hg = −2.82±0.76‰) and slightly depleted in odd mass isotopes (negative MIF, Δ199Hg = −0.32±0.12‰). The large negative δ202Hg-shift (on an average exceeding 3‰) between new and mature leaves is likely caused by kinetic MDF introduced during foliar uptake and air-leaf exchange of Hg0 over time. Furthermore, sunlight-mediated Hg0 gas efflux from intact mature foliage enclosed in a dynamic chamber system exposed to zero Hg air displays δ202Hg in a broad range (−2.47±1.36‰, n=18) that is statistically similar to mature foliage. However, interestingly, most Hg0 efflux samples have a profoundly positive MIF signature (Δ199Hg = 0.17±0.40‰, n=18) in contrast to ambient air and leaf samples. On an average, the Δ199Hg shift between Hg0 efflux and mature foliage Hg runs up to 0.48‰. In-turn, our data support for a temporal evolution towards slightly more negative MIF in mature foliage. The ∆199Hg/∆201Hg ratio of all foliar samples is 1.06±0.11, which is diagnostically divergent from that of Hg0 efflux data at 0.81±0.06. Unlike leaf uptake of Hg0 with subsequent in vivo oxidation and sorption, processes that presumably trigger diminutive MIF, reduction pathways of Hg2+ bound to reduced sulfur groups in the leaf interior potentially offer an explanation for departing observed (+)MIF in the Hg0 re-emission flux and (−)MIF in the foliage HgII pool after reductive loss. To explain the ∆199Hg/∆201Hg fractionation trajectory for Hg0 re-emission using existing data in the literature, a mass-balance suggests that maximum 76% of daily Hg re-emission was caused by photo-reduction, and a minimum of 24% by dark/thermal pathways.

Mercury (Hg0) emission from soil is an important source of atmospheric Hg on a global scale. Soil temperature and solar radiation are the major factors inducing Hg0 emission from soil. We investigated Hg isotope fractionation during thermal- and photo-induced emission of Hg0 from naturally Hg-enriched soils (agricultural and forest soils). Both mass dependent fractionation (MDF) and mass independent fractionation (MIF) were observed. The isotopic composition of emitted Hg0 from the agricultural soil displayed negative δ202Hg (-2.87 to -1.10‰) and positive ∆199Hg values (0.08 to 1.34‰) during both thermal- and photo-induced experiments. The negative shifts in δ202Hg values of emitted Hg0 from agricultural soil relative to reactant soil (δ202Hg = -0.41±0.12‰, ∆199Hg = -0.06‰) were generally larger during thermal-induced experiments (mean = -1.96 ‰) than those during photo-induced experiments (mean = -1.39‰), whereas photo-induced experiments induced larger shifts (mean = 0.68 ‰) in ∆199Hg values than the thermal-induced experiments (mean = 0.21%). The isotopic composition of emitted Hg0 from the forest soil (δ202Hg = -1.17±0.12‰, ∆199Hg = 0.07‰) displayed more negative δ202Hg values (-4.64 to -1.42‰) during both thermal- and photo-induced experiments. Similarly, thermal-induced experiments induced larger positively shifts (mean = -3.32‰) compared to photo-induced experiments (mean = -1.70‰). In contrast to agricultural soil, photo-induced experiments induced smaller shifts (mean = 0.07 ‰) than the thermal-induced experiments (mean = 0.22%). This implies that the organic compounds that affect the MIF of the odd-mass-number Hg isotopes in the presence of light in forest soil were more or less different from that in agricultural soil.

Wet deposition is an important source of mercury (Hg) input to remote montane forest ecosystems. Although Hg deposition flux through rainfall is typically lower than that from litterfall by a factor of two or more, the contribution of precipitation to biomass production and the associated enhancement on Hg uptake did not receive much attention. In this study, we characterized Hg, C, N concentrations and their isotope signatures in litter and soil horizons along two slopes of a subtropical montane site (Mt. Ailao) in Southwest China, where the precipitation intensity on the west slope (1200 2650 m) is significantly higher than on the east slope (800-2650 m), to understand the influence of precipitation on Hg accumulation on forest floor. Hg concentration in litter shows little influence by altitude in both slope. However, the concentration in soil exhibits a strong trend with altitude under 2550 m, increasing from 18-216 ng g-1 in west slope and 13-213 ng g-1. At the mountain top, soil Hg concentrations are 60 ng g-1 and 75 ng g-1 on the west and east slope, respectively. Hg input from litter is more important than wet deposition to Hg accumulation on the forest floor, as evidenced by the negative D199Hg found in the surface soil samples (-0.28±0.7‰ in litter vs. -0.37±0.12‰ in surface soil). Hg accumulation in soil can be explained by the increasing rainfall that enhances litter biomass production. Positive d202Hg gradient from 0.5 to 1 ‰ was observed from the litter layer to Oe, Oa and soil horizons at 2100-2650 m altitude on both slopes, while 0 to 0.2 ‰ negative D199Hg-shift was dependent on the density of forest canopy. Principal component analysis of Hg isotopes, C and N data suggests that 88% variations of d202Hg in soil profiles can be explained by the Hg loss during the processes of carbon and nitrogen mineralization, and 70% of D199Hg is caused by the re-emission induced by photoreduction. Significantly higher Hg concentrations and more negative D199Hg in the surface soil on the west slope suggest that precipitation imposes an indirect effect on Hg accumulation by influencing litter biomass production. This study provides new insights in understanding the role of precipitation in Hg accumulation in montane forested areas.