An examination of the importance of air-sea exchange in mercury cycling in the Gulf of Maine

Mercury (Hg), primarily as methylmercury (MeHg), is a neurotoxin that biomagnifies up marine food chains, causing a health risk to humans and wildlife that consume fish. Coastal waters, such as the Gulf of Maine (GoM) are major fishing grounds, and understanding the cycling of Hg and MeHg in these ecosystems is important but understudied. Anthropogenic activity and climate change has increased temperature and altered atmospheric and terrestrial inputs of Hg and other constituents, such as dissolved organic carbon (DOC), which have further modified the aquatic transformations of Hg species within the GoM. Our study examined their impact on the overall fate and transport of Hg, and in particular the air-sea exchange of Hg, and the net formation of MeHg in the GoM. High resolution measurements of dissolved gaseous Hg (DGHg) were collected on two cruises in the GoM and the historically Hg-contaminated Penobscot River in April and August 2023 to help examine these fluxes. DGHg concentrations showed distinct seasonal trends and %DGHg was higher in the GoM even though unfiltered total Hg (HgT) concentrations were higher in the estuary. The role of DOC and other parameters in moderating surface DGHg and flux is discussed as well as how the levels of Hg and DGHg have changed since prior investigations more than a decade ago. Furthermore, the relative importance of gas exchange compared to other sinks (water flow offshore and sedimentation) was examined using water column measurements from four cruises, and a mass balance model developed for HgT and MeHg for the GoM. We used the additional information collected throughout the water column for HgT and MeHg, and correlations between variables to constrain the MeHg budget and discuss the importance of external versus internal sources and sinks for MeHg in the GoM, highlighting the importance of in situ methylation in this ecosystem. Overall, external inputs of MeHg are not the primary driver of water column MeHg concentrations, although further study is needed to confirm this conclusion.

1 Introduction

Mercury (Hg) is a global pollutant of international concern, primarily due to the bioaccumulation of methylmercury (MeHg), the most toxic form of Hg, in fish and seafood (Sunderland et al., 2018; Liu et al., 2018). For many populations, including the USA, consumption of estuarine or coastal seafood is as important a source of MeHg exposure as open ocean and freshwater fish (Sunderland et al., 2018). Understanding Hg cycling in estuarine and coastal waters is therefore critical to understanding ecosystem and human health. Globally, anthropogenic releases, primarily to the atmosphere and natural waters, have substantially enhanced the concentration of Hg in most reservoirs of the biosphere (Outridge et al., 2018; Driscoll et al., 2013). For the global ocean, mass balances indicate that Hg inputs are primarily from atmospheric deposition, and Hg losses are primarily due to gas evasion of elemental Hg (Hg⁰) at the sea surface. The redox chemistry of Hg in surface ocean waters is complex with the potential for both photochemical oxidation of Hg⁰ and photoreduction of ionic Hg (Hg^II) as well as microbially-driven redox transformations (Amyot et al., 1997; Soerensen et al., 2010; 2013; Mason, 2015; Mason et al., 2017; O’Driscoll et al., 2022). Overall, the rate of net reduction is a function of solar radiation and light penetration, primary productivity, Hg^II concentration, and its bioavailability while the gas exchange flux is also related to wind speed and temperature.

For coastal environments, atmospheric deposition may not be the major Hg source with inputs from rivers (Kocman et al., 2017) and from current and legacy local point sources being as, or more important, depending on location (Balcom et al., 2004; Harris et al., 2012; Sunderland et al., 2012; Liu et al., 2020; Liu et al., 2025). In many studies of coastal waters, it has been suggested that Hg sedimentation is the primary Hg sink, but many of these studies have not examined in detail the importance of the loss of Hg⁰ via gas evasion to the atmosphere compared to sedimentation, or transport of Hg species offshore (e.g., Harris et al., 2012; Sunderland et al., 2012). In Long Island Sound (LIS), for example, Balcom et al. (2004) concluded that sediment burial removed about 60% of the incoming Hg with Hg⁰ volatilization accounting for 34% of the loss, and export offshore being a minor component. In contrast, in the East China Sea, evasion was a small sink (∼10%) with sedimentation and transport offshore being equally important as the major sinks for Hg (Liu et al., 2020). For the Gulf of Maine, Sunderland et al. (2012) examined the inputs of Hg and concluded that they were dominated by inputs from offshore waters due to current flow, with atmospheric, riverine and point source inputs being more minor sources. Similarly, Harris et al. (2012) concluded that inputs from the ocean were the most important source of Hg to the Gulf of Mexico. Both these studies did not account for potential Hg loss by gas exchange.

Furthermore, sources of total Hg may not be the primary driver of methylmercury (MeHg) concentrations and fluxes in the coastal ocean, as besides external sources, MeHg can be produced in situ by methylation in the water column (Liu et al., 2023; Despins et al., 2023; Ortiz et al., 2015; Schartup et al., 2015a) and sediment (Hammerschmidt and Fitzgerald, 2006; Liu et al., 2009; Hollweg et al., 2010; Schartup et al., 2014). The potential for water column methylation has been further highlighted by recent studies suggesting a link between Hg methylation and nitrification (Despins et al., 2023; Tada et al., 2021; Starr et al., 2022) and the demonstration of the presence of Hg methylating genes in the surface ocean (Villar et al., 2020). However, water column demethylation of MeHg, primarily due to photochemical degradation, is also an important sink for MeHg (Dimento and Mason, 2017; Driscoll et al., 2013). Indeed, several recent studies have suggested that external sources of MeHg are not the major driver of the concentration of MeHg in estuarine and coastal waters, and that in situ formation is important for coastal locations not contaminated by point sources (Liu et al., 2023; Seelen et al., 2021; Schartup et al., 2015a). However, the role of gas evasion in mediating the extent of Hg methylation in coastal waters has not been clearly examined. It is likely that for ecosystems where evasion is an important sink, the buildup and bioaccumulation of MeHg will be less than in locations where this sink is hindered, as has been demonstrated for the global ocean (Amos et al., 2015). Indeed, gas evasion removes Hg from surface waters and prevents its transport either via currents or through the settling of particulate material to deeper waters where its methylation is more likely in association with the degradation of organic matter in deeper waters and sediments (Motta et al., 2022; Sunderland et al., 2009; Golombek et al., 2024).

The evasion of Hg from coastal waters globally is poorly constrained due to lack of measurements, both spatially and temporally, and because of a lack of understanding of the controlling mechanisms. Studies have focused on specific water bodies, such as LIS (Rolfhus and Fitzgerald, 2004) and the mid-Atlantic coastal waters (Mason et al., 1998; 2001) in the eastern US, the coastal waters of China (e.g., Ci et al., 2011; Ci et al., 2016) and the coastal waters of Europe (Bieser and Schrum, 2016; Kotnik et al., 2022; Floreani et al., 2023; Floreani et al., 2019), locations that range in their degree of contamination from highly polluted to relatively pristine. The extent to which the evasion Hg⁰ flux is exacerbated by human activities, and by relatively high local Hg concentrations, is not well understood. For example, the study by Floreani et al. (2023) indicated that fluxes were not substantially elevated in the contaminated reaches of the Adriatic Sea because of the impacts of other factors, such as water clarity and mixing that may influence the net rate of photochemical Hg⁰ formation and evasion. The impact of climate change, including increased inputs of Hg from the watersheds and changes in microbial community structure, are also not well understood. There is also the potential for changes in the quality and concentration of dissolved organic matter (DOM) to influence the net rate of Hg⁰ formation (Schartup et al., 2015b).

To examine the importance of air-sea exchange of Hg in coastal ecosystems more closely, and compare its effects on Hg concentrations relative to other processes (ocean exchange, river inputs), as well as evaluate its potential impact on MeHg cycling, we focused our study on an important ecosystem, the Gulf of Maine (GoM). The GoM is the largest and most important coastal fishing ground in the USA (NOAA, 2024). Recent measurements of MeHg in seafood harvested in the GoM found elevated levels of Hg in several commercial fish species (Taylor et al., 2025), and Hg concentrations in bluefin tuna, a top predator fish in the GoM, are well above EPA consumption guidelines (Lee et al., 2016). Recent modeling studies of MeHg bioaccumulation in the GoM and the Northwest Atlantic Ocean suggest there will be future increases in MeHg levels in marine fish due to climate driven ecosystem changes (Schartup et al., 2018; 2019).

As a consequence of climate change, primary production in the GoM has decreased significantly in the last 20 years (Balch et al., 2022) which could impact the transfer of Hg from surface waters to the deeper GoM through particulate flux to depth (the “biological pump”). Additionally, climate change has altered the deep water inputs to the GoM, which have shifted from being mostly from more northerly sources in previous decades to being more driven by inputs from the Gulf Stream (Balch et al., 2022; Townsend et al., 2023). Such changes have increased the temperature and decreased nutrient inputs and could impact the cycling and inputs of Hg and MeHg to the GoM from offshore. Overall, however, Brickman et al. (2021) concluded based on mode evaluations that the largescale circulation in the GoM has not changed significantly over time.

Recent data on Hg and MeHg concentrations in the water column and lower trophic levels of the GoM are scarce and the role of gas evasion has not been adequately studied so that the human and climate driven impacts on this ecosystem for Hg cannot be accurately assessed. Previous studies of the offshore Northwest Atlantic and the GoM have observed enhanced Hg concentrations relative to other regions, which were attributed to inputs from river discharge and atmospheric inputs (Soerensen et al., 2013; Zhang et al., 2015; Hammerschmidt et al., 2013; Bowman et al., 2015; Cossa et al., 2018). Earlier studies have also shown the importance of gas exchange in Hg cycling in coastal and offshore waters of the region (Soerensen et al., 2013; Mason et al., 2017).

Overall, more studies of MeHg and Hg sources, cycling and dynamics of this important ecosystem are needed, as the system is undergoing environmental change at rates faster than other regions of the ocean (Pershing et al., 2021; Balch et al., 2016; 2022; Wallace et al., 2018; Brickman et al., 2021). There is evidence that increased runoff due to changes in precipitation has increased the terrestrial dissolved organic carbon (DOC) inputs to the GoM, increasing the amount of allochthonous DOC and leading to a “yellowing” of the GoM (Balch et al., 2016; 2022; Huntington et al., 2016) which can influence Hg speciation and bioavailability (Schartup et al., 2015b; Seelen et al., 2023). In concert, the potential inputs of Hg from the Penobscot River, which was historically contaminated with Hg (Rudd et al., 2018), could exacerbate the formation of Hg⁰. Finally, the importance of terrestrial inputs to the GoM are not well constrained (Sunderland et al., 2012), and the sinks for Hg have not been previously evaluated in detail.

We report here on the air-sea exchange of Hg and the sources and sinks for Hg and MeHg in the GoM, including the potential role of the historically-contaminated Penobscot River (Rudd et al., 2018; Geyer and Ralston, 2018; Turner et al., 2018), and compare these imports and exports to those of an earlier study (Sunderland et al., 2012). A companion paper examines in more detail the distributions and drivers of Hg and MeHg concentrations in the GoM and the historically-contaminated Penobscot River (Smith et al., 2025). To examine more closely the controlling factors on net Hg⁰ formation and evasion in the GoM, and its importance as a sink of Hg within the GoM relative to other sinks, we collected high resolution samples of surface water dissolved gaseous Hg (DGHg) on two cruises in 2023, in conjunction with extensive water column sampling on four cruises in both 2023 and 2024 (Figure 1; Smith et al., 2025). We contrasted the coastal waters of GoM with those of the lower Penobscot estuary, a region contaminated by historical manufacturing (Rudd et al., 2018). Additionally, we compared our findings in terms of gas exchange to those of Soerensen et al. (2013) from more than a decade ago.

Figure 1

Figure 1. Map of the Gulf of Maine and surrounding areas with the sampling locations indicated. The insert in the map shows the transit from Narragansett, RI to Station one and back for the two cruises. Underway collection of dissolved gaseous mercury occurred during transit to and from Station 1. An inset shows the transect to/from Station 1 starting from Narragansett, Rhode Island.

2 Methods

2.1 Dissolved gaseous mercury

Sampling for DGHg occurred during two cruises on the RV Endeavor in April and August 2023. The ship departed from Narragansett, Rhode Island and transited through the Cape Cod Canal to the Gulf of Maine (Figure 1). Two transects were completed during each cruise with water column sampling using a trace metal clean rosette system. The Gulf of Maine transect was from Portland, Maine to Canada, a transect that has been sampled many times by others (the Gulf of Maine North Atlantic Time Series (GNATS) transect; Balch et al., 2016). Additional samples were collected at stations within the Penobscot estuary, with the location of the furthest upstream station (St. 11) being determined by the ship’s capabilities, and St. 14 being the link between the two transects (Smith et al., 2025). For DGHg, underway sampling began before the first station and ended after the vertical profile sampling was complete. In April 2023, the ship did not return to port through the canal but transited offshore of Cape Cod, so during that cruise offshore waters were sampled for DGHg as well. In addition to the DGHg collections, vertical profile Hg speciation and ancillary measurement samples were collected from the ship at the stations indicated. Vertical profile sampling occurred during two additional cruises in November 2023 and April 2024. Additionally, samples were also collected using a small boat further upstream on the river in August 2023 and April 2024 for Hg and ancillary measurements (Smith et al., 2025).

For the quantification of DGHg, seawater from the ship’s intake at 5–10 m depth (which is used to measure surface temperature, salinity, and fluorescence) was subsampled with an automatic continuous equilibrium system (Andersson et al., 2008; Soerensen et al., 2013; Mason et al., 2017) during the cruises for the quantification of DGHg (Figure 1; Supplementary Figure S1). As there was no significant difference in concentration between when the ship was moving and on station, which never exceeded 6 h, the data for the time on station was not removed, as has been done for other studies where the time on station was longer (e.g., Mason et al., 2017). Quantification was done with 10 min resolution and relied on a Tekran 2537 Hg vapor analyzer for determining the air concentration which is then related to the water concentration using Henry’s Law. Presented data is hourly averages of the collections during each hour. The instrument was calibrated daily using the internal calibration source (every 25 h) and had a detection limit of <2 fM for seawater DGHg. Moisture from the equilibrator was prevented from reaching the Tekran as air passed through both an in-line Teflon filter and a soda lime trap, which were routinely changed. High-resolution ancillary measurements of wind speed, salinity, temperature, precipitation, and other parameters were aggregated into 1-h averages for statistical analyses comparable with the DGHg data.

For the calculations of concentration and flux, DGHg in the sampled seawater was assumed to be entirely Hg⁰. This assumption was confirmed with concurrent onboard measurements of dissolved gaseous Hg species and no dimethylmercury was found in surface waters. The gas exchange flux was estimated using a similar parameterization to that of Soerensen et al. (2010) which uses the Nightingale et al. (2000) relationship between wind speed and gas exchange coefficient, the Henry’s law coefficient for Hg⁰ from Andersson et al. (2008) and the Wilke-Chang method for estimating a temperature and salinity-corrected Hg⁰ diffusivity (Wilke and Chang, 1955). Additionally, the air concentration was estimated based on recent measurements for the region (Mao and Talbot, 2012; AMNET, 2025) and was assumed to be constant over time (1.2 ng m^-3) given the low seasonal variability found by these studies. The assumed concentration is lower than that found by Soerensen et al. (2013) during their cruises in 2008/2009 (average 1.4 ng m^-3) and this is consistent with the overall decreasing trend in concentration over time for the Atmospheric Mercury Network (AMNet) sites in the region (NH06 and NS01; AMNET. 2025).

2.2 Mass balance calculations

Concentrations of Hg used in the mass balance and flux calculations were derived from measurements made on four cruises between April 2023 and 2024 (Smith et al., 2025) as well as data in the literature. For the 2023/24 cruises, vertical profile samples collected were of unfiltered Hg (HgT) and MeHg, as well as particulate Hg and MeHg, and the mass balance estimates used the average values for all cruises for each water mass in the flux calculations so values presented in Supplementary Table S1 may differ from those in Table 1. Estimates of the rate of input of Hg and MeHg from offshore waters, via the surface inflow of Scotian Shelf waters into the northern part of the GoM (Figure 1; Supplementary Figure S2; Supplementary Table S1) and from the deeper inputs of slope water, are estimated based on current flows in the literature (Flores-Cervantes et al., 2009; Sunderland et al., 2012; Smith et al., 2014; Townsend et al., 2023) and available Hg data from this region (Bowman et al., 2015; Hammerschmidt et al., 2013; Soerensen et al., 2013; Cossa et al., 2018). The GOM data from 2023 to 2024 (Smith et al., 2025), integrated for the water mass (mixed layer or deeper waters), were used to estimate the export of Hg and MeHg in surface waters from the southwest GOM and export of intermediate waters offshore (Supplementary Figure S2). For river inputs, average flows are given in Supplementary Table S1. River inputs relied on flow data in the literature (Flores-Cervantes et al., 2009; Sunderland et al., 2012) for the rivers not measured in this study and/or Hg data from the literature (Peckenham et al., 2003; Sunderland et al., 2012; Reinhart, 2018; Seelen et al., 2021). For the Penobscot River, which contributes 12% of the freshwater inflow to the GoM, the flux was based on measurements during 2023–24 and Seelen et al. (2021), and the flow data taken from the USGS monitoring location at West Enfield near Bangor, ME (Smith et al., 2025; Supplementary Figure S3). Our estimate of Hg river export for the Penobscot River is lower than that of Rudd et al. (2018).

Table 1

Table 1. Concentrations of surface mercury and ancillary parameters measured in the Gulf of Maine and in the Penobscot estuary during two cruises in 2023 (April and August) as well as data collected during two cruises in August 2008 and October 2009 (Soerensen et al., 2013).

Atmospheric deposition to the GoM, both wet and dry, was estimated from the measurements at coastal sites in Maine and Nova Scotia that have been sampled as part of the Mercury Deposition Network (MDN, 2025). This data is reproduced in Supplementary Figure S4 (Sites ME96 Casco Bay (44.83 ^oN, 70.06 ^oW); ME98 Acadia National Park (44.38 ^oN, 68.26 ^oW); and NS01 (44.43 ^oN, 65.21 ^oW)). Overall, the Hg wet deposition flux decreased from 1997 to 2016 but has been higher in more recent years. The wet flux was based on the average deposition data since 2020. For dry deposition estimates, literature data from studies in coastal Maine were used (Feddersen et al., 2012; Mao and Talbot, 2012) along with estimates of the relative magnitude of wet versus dry deposition for this region (10% dry deposition; Mason et al., 2017).

The gas exchange flux was determined from the data in this study which used the parametrization of Soerensen et al. (2010). As the April cruise was early in the season and the DGHg concentrations were close to equilibrium with the atmosphere, this was assumed to be the case for the colder months of the year (November-April). The flux found during the cruise was used to scale the overall annual flux for these months using the average values for the GoM and the variation in wind speed over time. As the average flux for the cruise was based on a significant number of measurements it was assumed to be a representative value for the cruise and the time period. Similarly, the August data was scaled for the warmer months (May-October) to obtain an annual estimate. Other studies have also found highest concentrations in summer (e.g., Ci et al., 2016). As discussed below, for the GoM, the gas exchange flux was strongly correlated with wind speed rather than DGHg concentration so the seasonal differences in wind speed were taken into account in deriving the overall annual gas exchange flux.

Sedimentation was estimated using various datasets to obtain a more robust removal flux for this parameter. In Figure 4, sedimentation flux is determined from mass balance, and the comparison with the other calculations are detailed in the Results and Discussion. Smith et al. (2014) used radioisotope data and concentration ratios for metals to thorium (²³⁴Th) to estimate the water column flux and burial of metals in GoM sediments. Using these calculations for Pb (sediment flux 40 nmol m^-2 d^-1) and the ratio of Hg/Pb in deep GoM waters (7 × 10⁻³) through comparison of our recent data (Smith et al., 2025) and that of Smith et al. (2014), the sedimentation of Hg was estimated as 17.5 kmol yr^-1. Our unpublished particulate ²³⁴Th values from the 2023/24 study are comparable to the previously measured values (Smith et al., 2014; Charette et al., 2001). Flores-Cervantes et al. (2009) measured black carbon (BC) and particulate organic carbon (POC) concentrations and used radioisotope measurements to estimate fluxes in the GoM, giving a sedimentation rate of 200 Gg BC yr^-1. From their measurements coupled with our measurements of POC and particulate Hg during this study (i.e. 2 × 10⁻⁹ mol Hg/g POC on average; Smith et al., 2025), we estimated the sedimentation flux for Hg as 8 kmol yr^-1. In comparison, Liu et al. (2025) estimated the burial fluxes for various coastal ocean regions based on available data and modeling. Their estimate of the burial flux for the GoM region is 10–20 ug m^-2 yr^-1, which converts into a Hg flux of 8–16 kmol yr^-1. The flux range from Liu et al. (2025) is at the low end of the coastal data compiled and reported by Sun et al. (2023) but comparable to other uncontaminated northerly locations (Bering and Chukchi Seas, Hudson Bay, Baltic Sea). Overall, the coastal ocean Hg sediment burial estimates range from 8–17.5 kmol yr^-1. Finally, as the mass balance is based on a two-box model with inputs into and export from the mixed layer and the deeper waters, a flux calculation for the transport between the two boxes is needed to balance each model box and was calculated to balance the sources and sinks. This flux is from the mixed layer to depth for HgT and represents the removal of Hg via particle settling.

For MeHg, external input fluxes are estimated based on the HgT flux and the average %MeHg in the source waters and a similar approach was used for export based on %MeHg in water flowing offshore, and in surface sediment. Gas evasion of MeHg is considered ta minor component. Given the strong relationship found between MeHg concentration and apparent oxygen utilization (AOU) and nutrient concentrations (Smith et al., 2025), and the negative correlation with chlorophyll concentration, we modeled the potential in situ methylation using the data from Despins et al. (2023) who measured methylation rates in water samples collected in the waters around Alaska. Additionally, methylation estimates were calculated using the MeHg relationship with AOU derived by Sunderland et al. (2009) for the North Pacific Ocean. For photochemical demethylation, the model developed by Dimento and Mason (2017) was used to estimate this for the upper waters of the GoM. In their model, estimates are derived for typical river, estuarine and coastal waters, based on DOC and suspended solids concentrations, and for the photon flux at three latitudes (the equator, mid-latitude, and polar). For the GoM, the calculation assumed that the GoM waters are equivalent in terms of their rate of demethylation to a model mix of 75% estuarine and 25% ocean waters.

The particle vertical transport flux for MeHg was calculated from the value for HgT and the average %MeHg in particulate from Smith et al. (2025). As the mass balance box model has two layers (mixed layer and deeper waters), the overall flux between the two boxes needed to balance the overall budget was calculated to ensure balance.

The uncertainty in the flux estimates is smaller for those based on recently measured concentrations which typically varied by <30% within a cruise for a particular parameter in a particular water mass and varied by a similar amount between cruises. This variability is greater than the analytical errors. The HgT and MeHg concentrations used for the export fluxes were the overall averages for each parameter for the four cruises for each box (2-3 depths per station per box and 10 stations for the GoM; Smith et al., 2025). The extrapolation of the DGHg data from the two cruises to estimate the annual gas exchange flux has higher uncertainty than the estimate for wet deposition which is based on a more extensive dataset. The MDN data typically showed similar annual wet fluxes between the sampling sites but more inter-annual variability (Supplementary Figure S3) so the difference in the rainfall amount in a year significantly contributes to the uncertainty in this flux estimate. The precipitation in 2024 was near the long-term average while it was above average in 2022 and 2023 and lower in 2021 (NOAA, 2024) and so using the average flux over the period is reasonable. Similarly, the riverine flux estimate is likely affected as much by the water flow as the estimated Hg and MeHg concentrations, which, for example, were relatively constant across the sampling periods for the Penobscot estuary. Estimates for the inputs from offshore have a larger uncertainty given the lack of recent data and the fact that the associated studies (primarily Bowman et al. (2015); Cossa et al. (2018)) were not specifically focused on the GoM. The exchange flow estimates are likely not a major source of error in these estimates (Brickman et al., 2021). Uncertainty in sediment flux based on the various estimates is a factor of 2, as discussed in Section 2.2. The overall lack of Hg data for the rivers flowing into the GoM makes this estimate also more uncertain than the estimates that are based on the measured data from Smith et al. (2025). Overall, we conclude that the uncertainty range in a specific flux estimate is a factor of two for riverine inputs, dry deposition inputs, gas exchange flux, and sediment flux, and less than a factor of two for the other fluxes given the uncertainties discussed above, which is a similar range in error to that of other Hg box model studies (Sunderland et al., 2012; Driscoll et al., 2013; Outridge et al., 2018).

3 Results and discussion

3.1 Elemental mercury concentrations and flux

The underway DGHg flux data are compiled in Figures 2, 3, along with the underway data for wind speed and temperature, with mean Hg and ancillary measurement concentrations tabulated by date and transect in Table 1, for both the two recent cruises and for the earlier cruises (Soerensen et al., 2013). The unfiltered HgT and ancillary data in Table 1 are those for the shallowest sample collected from the rosette sampler (typically 10 m) from the stations during each transect as these are most representative of the water sampled with the underway system. Additional ancillary underway water chemistry parameters are shown in Supplementary Figure S5. The temperature was low and relatively constant during the spring cruise compared to the summer (Table 1; t-test; p < 0.05). Underway fluorescence (Chl a) (2.4 ± 1.2 μg L^-1) was not elevated in the Penobscot River compared to the GoM and was lowest when the ship was further offshore and on the eastern edge of the GoM near Station 6 (Supplementary Figure S5). Fluorescence was slightly lower in August 2023 (1.9 ± 0.9 μg L^-1) and again there were no obvious trends with location. However, there was an oscillating trend in the fluorescence concentration with peaks in concentrations alternating with low concentrations (Supplementary Figure S5). However, the lower concentrations did not always coincide with nighttime, or with other parameters such as temperature and salinity. Periods of higher DGHg, and substantial changes in concentration did not coincide with major changes in fluorescence. In April, the salinity was much lower in the estuary than offshore due to the higher river discharge at this time compared to August (Supplementary Figures S4 & S5).

Figure 2

Figure 2. Data from the April 2023 cruise. (A) Plot of the hourly averaged measured dissolved gaseous mercury (DGHg) concentration (fM), the estimated flux to the atmosphere (pmol m^-2 d^-1), and the calculated concentration at equilibrium with the atmosphere (fM; dotted line). (B) the hourly averaged wind speed (m s^-1) and water temperature (^oC) recorded by the ship. The x-axis represents the time since departure from Narragansett, RI and the station labels on the figure indicate the position of the ship over time.

Figure 3

Figure 3. The data plotted are comparable to that for Figure 2 but for the August 2023 cruise.

Fluxes of DGHg were low in April, when water DGHg concentrations were at or near equilibrium with the atmosphere (Figure 2). The low fluxes are especially apparent in the GoM. Fluxes were highest in the offshore waters east of Cape Cod in April (Figure 1), which was sampled during the latter part of the cruise. It should be noted that concurrent batch measurements of DGHg species (Hg⁰ and dimethylmercury (DMHg)) on board demonstrated that levels of DMHg in surface waters were at the detection limit, so DGHg is entirely Hg⁰ for these surface waters.

In August, HgT concentrations were slightly higher than April in the GoM but more so in the Penobscot estuary and this difference likely accounts, to some degree, for the higher DGHg concentrations found in August (∼30% higher on average; Table 1). However, the %DGHg was overall higher in both regions in August (two to three X’s), especially in the Penobscot estuary (Figure 3), so the higher DGHg concentration is not entirely attributable to higher HgT (Table 1). The %DGHg was higher in the GoM compared with the estuary in August. Wind speeds were more variable in April as were fluorescence and DOC concentrations (comparing those measured only on station in the mixed layer). The high DOC in the GoM in April coincided with an unusual dinoflagellate bloom that occurred in the GoM in Spring 2023 (NOAA, 2024). The temperatures were significantly higher in August and this likely enhanced the evasion flux given the effect of temperature on Hg⁰ solubility.

Overall, the fluxes measured in this study are within the range of values reported for other coastal regions. Rolfhus and Fitzgerald (2004) measured concentrations of DGHg between 37 and 530 fM in Long Island Sound over several cruises and found that concentrations were higher in summer, were not strongly related to dissolved Hg, and increased with salinity. More recent data collected in LIS ranged from 50 to 250 fM during a spring cruise in 2018, and similar concentrations were found with longer term measurements at Avery Point in Groton, CT (He, 2018). There was evidence for a diurnal signal in the shallow water samples at Avery Point, as has been found by others for coastal waters (Osterwalder et al., 2021). The diurnal variation was driven primarily by differences in radiation levels rather than changes in water chemistry or air concentrations. No diurnal variability is apparent in the current dataset.

Mason et al. (2001) found higher concentrations than those in this study in the mid-Atlantic Bight (up to 1400 fM). They found that DGHg concentrations on the shelf were higher, as was the %DGHg, than the Chesapeake Bay, and to those measured further offshore. Our data show a similar trend of higher %DGHg offshore. In the coastal waters off China, concentrations of DGHg were in a similar range to the GoM with concentrations higher in the summer than the fall, and lowest in the spring (Ci et al., 2011; Ci et al., 2016; Tseng and Lamborg, 2013; Wang et al., 2020). In the Mediterranean and other European waters, studies have also found a similar range in concentrations and the controlling factors (Floreani et al., 2023; Kotnik et al., 2022; Osterwalder et al., 2021).

The DGHg concentrations in August for the GoM are within the range of those measured by Soerensen et al. (2013) in their studies in August 2008 and October 2009 (Table 1). The locations sampled in the prior study included sites further offshore and south of Cape Cod. The earlier study did not collect samples in the Penobscot estuary. The HgT concentrations were higher in the earlier studies for the GoM compared to those measured in the 2023/24 cruises. Fluxes were in a similar range for the two studies, with the exception of August 2023, which were higher than in the earlier studies. The higher HgT measured in 2008/09 suggests that there has been a decrease in the concentrations in the GoM in the last 15 years, as discussed further below. The earlier study found a relationship between DGHg and ancillary variables (temperature–salinity + fluorescence) in October but no relationship with DGHg in August as the variables were strongly colinearly related.

3.2 Factors influencing dissolved concentration and evasion flux in the Gulf of Maine

In April, DGHg concentrations were comparable for the GNATS transect to those found in the Penobscot. The %DGHg was slightly lower in April in the Penobscot estuary than in the GOM, driven by HgT concentrations being ∼2 times higher on average (Table 1). Correlations between DGHg concentrations and physicochemical parameters were examined across locations and time points. There was no correlation between DGHg concentration and underway Chl a fluorescence during either cruise in 2023. The higher temperature in August could have enhanced the rate of reduction and be the factor driving the overall higher %DGHg in August. Using data collected at each station, the influence of nutrients, Chl a, and DOC on DGHg was examined, but many of these relationships were not significant. For both cruises, the %DGHg decreased with increasing total Hg, and in August, with nitrate (NO₃⁻) as well (no nitrate data was collected in April; Smith et al., 2025):

%DGHg (April) = 16.34–8.905HgT; r² = 0.404; p = 0.006

%DGHg (Aug.) = 26.31–12.02HgT +6.54NO₃⁻; r² = 0.699; p < 0.001.

Overall, there were higher HgT concentrations in the Penobscot estuary than in the GoM (2–3 times higher on average), but less difference in the DGHg, especially in April. This similarity in DGHg concentration suggests that the estuarine Hg is less available for conversion into Hg⁰, which is corroborated by the lower %DGHg. Furthermore, while the HgT and DOC were strongly correlated in the Penobscot River and estuary for both cruises, there was no significant correlation in the GoM (Smith et al., 2025). These correlations suggest that different factors are controlling the rate of formation of Hg⁰ in the estuary compared to the more offshore waters.

Overall, the DOC concentrations were not substantially different between the GoM and the Penobscot (Table 1), although both were higher in April, suggesting that Hg^II complexation by DOC may not be the main driver of the differences in %DGHg between the GoM and the estuary. But the DOC within the estuary has a higher terrestrial component (Smith et al., 2025) compared to the GoM, which would likely bind to Hg more strongly and hinder Hg⁰ formation (Schartup et al., 2015b), indicating a potential role of terrestrial DOC in hindering Hg⁰ formation (Seelen et al., 2023). To explore this, we examined the relationship between the total fraction of terrestrial DOC (DOC_T), obtained from PARAFAC analysis of the fluorescence spectra (total of four identified terrestrial fractions in Raman units; Smith et al., 2025), as well as the humification index (HIX), and %DGHg (Supplementary Figure S6). For the April cruise, there was a significant linear correlation (p < 0.001) between the %DGHg and both DOC_T (r² = 0.469) and HIX (r² = 0.475). Overall, the range in the DOC_T across all stations in the GoM and Penobscot estuary was smaller in April than in August where there were some high values for the estuarine locations (Supplementary Figure S6). In August, there was a significant exponential relationship (p < 0.001) for DOC_T (r² = 0.342) suggesting that the impact of DOC_T is likely related to its ability to complex with Hg^II and hinder its reduction and that at the higher fractions of DOC_T essentially all the Hg^II is complexed so adding more terrestrial DOC has little impact. The relationship with HIX was linear (r² = 0.408; p < 0.001). Schartup et al. (2015b) also found a relationship between %DGHg and the terrestrial DOC fraction and O’Driscoll et al. (2022) reached a similar conclusion about the role of DOC in hindering reduction from their studies in lakes. Overall, the rate of formation of Hg⁰ was lower in the estuary compared to the more offshore waters, reflecting the higher fraction of DOC_T in the estuary (Smith et al., 2025).

For April, the flux was not related to the wind speed for the entire dataset. However, for the GNATS transect, higher fluxes coincided with higher wind speed (r² = 0.317; p < 0.001) but fluxes were not related to DGHg concentration (r² = 0.127; p > 0.001). In contrast, the flux in August was related to wind speed for the overall dataset (r² = 0.41, p < 0.001) as well as to DGHg concentration (r² = 0.284; p < 0.001). These relationships are consistent with the findings in many other studies that have concluded that wind speed is the major driver of Hg⁰ flux for coastal and open ocean waters.

The hourly averaged DGHg concentrations and fluxes were higher and statistically different for the August 2008 sampling (Soerensen et al., 2013) compared to the current study (t-test; n > 100; p < 0.001). Similarly, the HgT and %DGHg were overall higher in August 2008 than in 2023 although the data is more limited (n < 10) given that HgT was only measured on station. The higher concentrations in 2008 were previously attributed to enhanced rainfall in the region at the time (Soerensen et al., 2013). In 2023, sampling occurred during a period of low river output, based on USGS data from West Enfield, ME on the Penobscot River (Supplementary Figure S4). Additional data collected within the Penobscot River in August 2023 and April 2024 (Smith et al., 2025) suggest that there was little difference in the river endmember HgT concentration with season and that the river flux is primarily driven by the flowrate (Smith et al., 2025; Seelen et al., 2021). A comparison of unfiltered HgT in the Penobscot River at lower salinity sites (Smith et al., 2025) to those measured in summer 2016 (Seelen et al., 2021) suggest a decrease in HgT in the last 8 years. This decrease is also consistent with our estimate of the Penobscot River flux which is lower than the value estimated by Rudd et al. (2018). The wet deposition Hg data for two sites on the Maine coast and one in Nova Scotia indicate an overall decrease in the wet deposition HgT flux from 1995 until 2016, with substantial interannual variability, but an increasing trend in the last few years (Supplementary Figure S3). However, the Hg wet deposition flux was above the long-term average in 2008, supporting the contention of Soerensen et al. (2013). Overall, variability in external inputs to the GoM and seasonal differences in wind speed, temperature, and salinity at the sites sampled in the Penobscot estuary are likely the main drivers of the Hg⁰ flux and its seasonal variation.

It is apparent from the discussions above that the external Hg inputs to the GoM have changed over time, but it is not clear which sources are the major drivers of the HgT, DGHg and MeHg concentrations in the GoM. In particular, the importance of gas exchange versus other sinks for Hg has not been previously investigated. Furthermore, the importance of external sources relative to internal cycling (i.e., net in situ production of MeHg) has also not been previously evaluated. To examine these factors further and compare fluxes to previous estimates, we discuss below the sources and sinks for Hg in the GoM.

3.3 Sources and sinks for mercury in the Gulf of Maine

Sunderland et al. (2012) previously evaluated the external inputs of Hg to the GoM and concluded that river and point source inputs were a small fraction of the external sources (∼15%) with the dominant sources being atmospheric deposition (∼25%) and inflow of seawater into the GoM from offshore (∼60%). Other studies suggest that the sources for the offshore water input have changed in the last decades, and this shift could be driving the changes in HgT, MeHg and DGHg in the GoM over time, even given the potentially larger changes in the other external Hg inputs (e.g., rivers and atmospheric deposition). Seidov et al. (2021) noted a warming of the deeper inflowing water over the last decade, which was greater than that for shallower waters, and a resultant shift in the thermal regime. Townsend et al. (2023) also concluded that the deeper inflowing water had become warmer, saltier and nutrient deficient compared to earlier studies and that the change over time was the result of more inflow of Gulf Stream waters rather than waters from more temperate sources, especially in the summer. The potential impacts of different deepwater sources on Hg and MeHg inputs to the GoM can be assessed using Hg data in the literature (Hammerschmidt et al., 2013; Bowman et al., 2015; Cossa et al., 2018; Supplementary Table S1). Overall, currently there is more incoming surface water from the Scotian Shelf than is exported in the Maine Surface Water southward and this difference is compensated for by a higher offshore flow in the intermediate waters compared to the inflow of slope water at depth (Supplementary Table S1; Supplementary Figure S2).

While Sunderland et al. (2012) estimated the external inputs, they did not evaluate the sinks for Hg from the GoM, which are critical to understanding the dynamics of Hg in the GoM. The details of the mass balance calculations are given in the Methods section and the overall mass balance for Hg in the GoM is shown in Figure 4, with the calculations detailed in Supplementary Tables S1–S3. As found by Sunderland et al. (2012), external inputs of Hg from offshore, including both surface (Scotian Shelf Water; 9.45 kmol yr^-1) and deeper water (Slope Water; 17.4 kmol yr^-1) are the dominant sources to the GoM (total 26.9 kmol yr^-1; 75% of all inputs). This trend is due to the large inflow from offshore, which is ∼200 times greater than the river flow (Supplementary Table S1), rather than a higher Hg concentration in these waters. The outflow flux (surface outflow 3.75 kmol yr^-1; subsurface outflow12.5 kmol yr^-1) is somewhat lower primarily because of the lower concentrations in surface waters in the western GoM compared to the deeper waters (Smith et al., 2025; Supplementary Table S1). The other external inputs to the GoM include river and point source inputs (2.0 kmol yr^-1; 6% of inputs) and atmospheric deposition (7.1 kmol yr^-1; 20% of the inputs), which are based on measured concentrations of HgT in rivers and in wet deposition and atmospheric Hg concentrations (Supplementary Tables S1–S3; Supplementary Figures S3, S4).

Figure 4

Figure 4. The mass balance for total mercury inputs and outputs from the Gulf of Maine. Fluxes are given in kmol yr^-1.

Overall, the current mass balance differs from the earlier estimates in that all the previous atmospheric, river and point source Hg inputs were calculated to be higher than the current estimates (Sunderland et al., 2012; Rudd et al., 2018). The current estimate of atmospheric Hg inputs is 28% lower, and rivers and point sources are estimated as being 58% lower than in the previous model (Sunderland, et al., 2012). The estimate of Hg input from offshore is, however, 29% higher based on the more recent data available (Bowman et al., 2015; Cossa et al., 2018; Soerensen et al., 2013). Overall, the changes in Hg fluxes are greater for the river inputs than for atmospheric inputs, reflecting the paucity of data that was available when the estimates were made by Sunderland et al. (2012), or resulting from the greater decrease in anthropogenic inputs to the rivers surrounding the GoM compared to atmosphere deposition, which is influenced by more regional inputs. In support of decreases in river inputs, our estimate of the flux from the Penobscot River is 40% of the value estimated by Rudd et al. (2018). Overall, there is still a paucity of information on the Hg concentrations in the majority of the rivers flowing into the GoM, with the contaminated Penobscot River being the primary exception as it has been well studied (Rudd et al., 2018; Gilmour et al., 2018; Seelen et al., 2021; Smith et al., 2025). River Hg inputs to the Bay of Fundy have also been studied more than the other major tributaries (Harding et al., 2018; Reinhart, 2018).

The previous study did not evaluate the export of Hg from the GoM and, as shown in Figure 4, the major sink is the transport of Hg offshore, which accounts for 45% of the total export fluxes, with Hg export being predominantly at depth. Gas evasion is estimated based on the data reported above and is slightly smaller than the atmospheric input flux (6.1 kmol yr^-1; 17% of the total sink). The estimated removal by sediment burial is given in the figure as the value required for mass balance (13.6 kmol yr^-1). As detailed in Section 2.2, the sediment burial flux was estimated in three different ways based on previous estimates for Pb sedimentation (Smith et al., 2014), black carbon removal (Flores-Cervantes et al., 2009) and a modeling study of global Hg burial in coastal environments (Liu et al., 2025). The estimates ranged from 8–17.5 kmol yr^-1. Thus, the mass balance estimated value falls within this range, and given the uncertainty associated with the various mass balance estimates, is consistent with the values estimated by the alternative means. Thus, the gas evasion flux is smaller than the sedimentation flux.

For MeHg, the mass balance is also dominated by inflow of offshore waters and outflow from the GoM (Figure 5; Supplementary Table S3) when not considering the internal cycling (i.e., in situ methylation and demethylation). There is some uncertainty in the %MeHg concentration in the inflowing waters (Bowman et al., 2015; Cossa et al., 2018), making it difficult to estimate a flux. Using 3% MeHg for surface inflow, and 10% for deeper waters, the flux is estimated at 1.6 kmol yr^-1 (1.3 kmol yr^-1 for the deep inputs). The inputs of MeHg from the atmosphere, rivers, and point source inputs are estimated as 0.17 kmol yr^-1 based on the low %MeHg in precipitation (1%) and rivers (5%; Smith et al., 2025; Seelen et al., 2021; Sunderland et al., 2012).

Figure 5

Figure 5. The mass balance for methylmercury inputs and outputs from the Gulf of Maine. Fluxes are given in kmol yr^-1.

The %MeHg of the GoM waters are more constrained by our measurements (Smith et al., 2025). The outflow flux (1.5 kmol yr^-1) is estimated based on an average surface water %MeHg of 5% and a mid-depth value of 10%. The sediment flux is estimated by assuming 1% MeHg in surface sediments (0.14 kmol yr^-1), based on Shi et al. (2025). The mass balance so far does not consider in situ formation and degradation of methylated Hg. The main internal sink is the net demethylation of MeHg, primarily from photochemical degradation in surface waters. Predictions of the potential for photochemical demethylation of MeHg in the GoM can be made based on the measurements of Dimento and Mason (2017) and their integrated estimates for water column demethylation, assuming the GoM is intermediate in its integrated demethylation between estuarine and offshore waters, and accounting for the photon flux at this latitude, as described in the Methods section. Using this information, an estimate for demethylation of 4.0 kmol yr^-1 is derived, which is much larger than the external inputs (Figure 5; Supplementary Table S3). Removal of MeHg through particulate settling and water mixing is estimated at 0.9 kmol yr^-1. Thus, there is removal of 5.2 kmol yr^-1 MeHg on a yearly basis from the mixed layer.

We further note the strong correlations that were found between MeHg and ancillary variables in deeper waters of the GoM, suggesting that there is substantial net methylation in these waters (Smith et al., 2025). MeHg concentrations correlated with apparent oxygen utilization (AOU) and nitrate concentration, and negatively correlated with chlorophyll concentration and temperature, suggesting the potential importance of water column methylation of Hg as an internal source (Smith et al., 2025). Other studies have also suggested the potential for Hg methylation in concert with nitrification and organic matter degradation (Sunderland et al., 2009; Despins et al., 2023; Tada et al., 2021; Starr et al., 2022) and this is likely occurring in the deeper waters of the GoM. Overall, for the cruises in August 2023 November 2023 and April 2024, Smith et al. (2025) found that the MeHg concentration positively correlated with AOU (p < 0.001) and total oxidized N concentration (NO_X; p < 0.01) and negatively correlated with dissolved oxygen concentration, salinity and temperature (all p < 0.01). These results are shown in Variable Influence on Projection (VIP) correlation plots in Supplementary Figure S6.

These observations suggest that net methylation of MeHg in the deeper waters of the GoM is the additional source needed to balance the loss terms for the mixed layer. For net methylation within the GoM deeper waters, an estimate can be determined by mass balance. This value is 3.7 kmol yr^-1. There are no experimental rate data to confirm this flux magnitude but assuming that deep methylation in the GoM is coupled with nitrification, and using the rates of methylation measured in the Gulf of Alaska by Despins et al. (2023) for the GoM deeper waters (>100 m), produces an estimate that is larger (6.6 kmol yr^-1). This estimate is consistent with the flux estimated by difference, given that there will also be demethylation in the deeper waters. While sediment MeHg flux could also contribute, based on the sediment-water mass balance for MeHg determined by Shi et al. (2025) for sediments around Iceland, and the estimated sedimentary flux, this value is small.

Concentrations of MeHg in the deeper waters of the GoM are also consistent with the model results of Sunderland et al. (2009) for the subsurface North Pacific given the overall low AOU (<100 μM; Smith et al., 2025). Both the methylation and demethylation estimated fluxes are substantially greater than the external input and export fluxes for MeHg and suggest that in situ processes are the main drivers of the MeHg concentration and distribution in the water column of the GoM. This conclusion is consistent with the analysis of Smith et al. (2025) who concluded, based on a multiple regression analysis, that external fluxes were not the dominant control over MeHg distributions in the GoM. The mass balance estimates support this contention that there is substantial Hg methylation in the deeper waters and demethylation in the mixed layer, but future measurements of methylation and demethylation rates in the GoM are needed to further constrain the MeHg cycle.

Overall, the mass balance indicates that there is net transport of HgT from the surface waters to depth (Figure 4), likely primarily due to particulate settling with a small contribution from dissolved Hg transport. The studies of Smith et al. (2014) and others indicate that the water column particulate flux for other metals is greater than the sedimentation flux but this does not appear to be the case for HgT (Figure 4). One reason is that the MeHg mass balance suggests there must be substantial transport of MeHg from deeper waters to the surface (Figure 5), given the gradient in concentration with higher MeHg at depth. This transport upward suggests that there is actually a higher overall flux of inorganic Hg (Hg^II) from the surface to depth (13.2 kmol yr^-1, Figures 3, 4) which is mitigated by the conversion of Hg^II to MeHg and the flux of MeHg back to the mixed layer. More detailed modeling of the system would be needed to confirm this notion.

For HgT, the mass balance suggests that the GoM is a net sink for Hg with around 38% of the HgT coming into the Gulf being deposited into the sediment. The mass balance for Hg in the GoM can be compared with that for Pb (Smith et al., 2014). For Pb, 48% of the inputs are from offshore water inflow, with atmospheric deposition and terrestrial inputs (river inputs and direct discharge) being equally important, respectively, 25% and 27% of the inputs. Thus, rivers are a relatively more important source of Pb to the GoM than they are for Hg, but this may also reflect the time differences between the studies as the Pb samples were collected more than 20 years ago. In terms of sinks, there is no gas evasion for Pb, so the two sinks are sediment burial and export offshore. Sediment burial at 56% of the total Pb sink is somewhat more important than offshore export. For Hg, gas evasion at the sea surface removes Hg from the water column decreasing the amount that can be scavenged by particles and transported to the sediment, and this evasion likely accounts for the lower relative burial rate for Hg (38% of inputs). Additionally, as noted above, methylation of Hg in deeper waters reduces the overall burial of HgT given the lower particle reactivity of MeHg.

However, the differences in the relative importance of the sedimentation and burial flux for the two metals could also be accounted for by the decreasing primary productivity in the GoM over time. Balch et al. (2022) have demonstrated that there has been a ∼50% decline in primary productivity in the GoM in the last 20 years and also a change in the phytoplankton community structure (less diatoms) which could further reduce the vertical flux. Given the earlier sampling of the Pb study, this could account for the differences between the relative magnitude of the fluxes for Pb and Hg. Indeed, the Hg burial estimate based on the Pb data is at the high end of the range of values discussed in Section 2.2 (Supplementary Table S2), and 30% greater than the value in Figure 4. Clearly, the impact of changes in primary productivity on Hg, and Pb, cycling in the GoM needs further investigation.

Overall, there is a net import of Pb from offshore into the GoM, as found for Hg, and the GoM is a substantial net sink for both metals. The importance of sedimentation as a sink likely reflects the high particle reactivity of both metals. Fitzgerald et al. (2018) measured both Hg and Pb sediment fluxes in the Pettaquamscutt estuary in Rhode Island and found the ratio of Pb/Hg sedimentation flux to be around 400, which is of the same order as the ratio found for the GoM (165), derived by comparing the Hg mass balance with that of Smith et al. (2014), further suggesting the magnitude of the Hg sedimentation flux estimated here is reasonable.

That sedimentation is the major sink for Hg in the GoM is consistent with other studies that have examined the mass balances of Hg in coastal ecosystems. For the Gulf of Mexico, Harris et al. (2012) concluded that gas evasion and sedimentation flux were of similar magnitude. For this ecosystem, outflowing waters to the Atlantic Ocean was the major sink and was much larger than the other sinks. The sedimentation rate estimated by Harris et al. (2012) was 10 μg m^-2 yr^-1, comparable to the value estimated here for the GoM (16 μg m^-2 yr^-1). Thus, overall, the behavior of Hg in the GoM and the Gulf of Mexico are similar. The sedimentation flux estimated for the GoM is comparable to that for other temperate and polar coastal waters such as the Baltic Sea, Hudson Bay and the Bering and Chukchi Seas (as summarized in Sun et al., 2023).

However, for the coastal waters of China (Bohai, Yellow, and East China Seas) overall Hg sedimentation fluxes are much higher, ranging from an average of 67 μg m^-2 yr^-1 for the Yellow Sea to 452 μg m^-2 yr^-1 for the Bohai Sea and 467 μg m^-2 yr^-1 for the East China Sea (Sun et al., 2023). The overall average flux is 245 μg m^-2 yr^-1 for these coastal waters. Liu et al. (2016) modeled the sources and sinks for Hg and found that the export of Hg from a sea to other seas or offshore was not always the dominant sink. It was not the major sink in the Bohai Sea but was for the South China Sea. Evasion was the dominant sink for the South China Sea and sedimentation was the major sink in the Bohai Sea. Fluxes were orders of magnitude higher in these locations given the high levels of anthropogenic Hg inputs to these coastal waters. Similarly, Hg accumulation in sediments of the Penobscot River and estuary are higher than in the GoM (Yeager et al., 2018). Overall, besides variations in the magnitude of anthropogenic inputs, several factors impact the relative importance of sedimentation and gas evasion as sinks for Hg, such as the surface area of the sea/gulf, the average depth of the water, and the extent of its isolation from offshore waters. Other factors, such as the number and size of the rivers, primary productivity and suspended solids concentrations, also play a role.

Contrasting the GoM to other regional locations provides further evidence of the controlling factors. For a smaller, shallower and more contained ecosystem such as Long Island Sound (LIS), the net flow offshore is small (<10% of the sinks; Balcom et al., 2004) and the major sink is burial in the sediment (60%; 110 μg m^-2 yr^-1), somewhat higher than the gas evasion flux (34%). Additionally, the river and point source inputs are the dominant input to this highly urbanized estuary (66% of the inputs). The burial flux in LIS is comparable to that estimated for the Penobscot River. In contrast, for the Chesapeake Bay, Mason et al. (1998) concluded that terrestrial inputs were the dominant source (49%) with inputs from offshore (21%) and atmospheric deposition (30%) being the lesser sources. In terms of outputs, sedimentation and transport out of the Bay were of equal importance (respectively, 43% and 42% of the total) with gas exchange being a minor component (13%). Both LIS and Chesapeake Bay have large riverine inputs and relatively restricted exchange between the estuarine system and the coastal waters, and these features likely drive the differences in the importance of exchange with offshore waters and air-sea exchange. The GoM is less constrained by land mass. Also, while the surface area of the GoM and the Chesapeake Bay are of similar magnitude, the depth is very different, such that the surface area/volume ratio for the GoM is several orders of magnitude smaller than the Chesapeake Bay, which is overall shallow. This difference clearly impacts the relative importance of sedimentation versus gas exchange and offshore exchange of water and its constituents. LIS is intermediate between the other two east coast systems in terms of its surface area/volume ratio.

The discussion above, the discussion above supports the unique characteristics of the GoM which impacts the overall cycling of Hg, and likely other contaminants and chemicals, and the degree to which in situ processes such as primary productivity, particle flux, methylation/demethylation and gas exchange are controlling Hg fate compared to transport via ocean currents. The average residence time for water in the GoM is around 6 months, longer than LIS, and similar to the Chesapeake Bay. However, the relative importance of offshore versus freshwater inputs is much greater for the GoM than for the other systems where the exchange is more tidal in nature, and the freshwater input is predominantly from one large river. As an example, the freshwater input to the Chesapeake Bay is of the same order as the export offshore while for the GoM, as noted above, the transport offshore is ∼200 times the freshwater input. Thus, the dominant sources and sinks for these different ecosystems are controlled to a large degree by their relative depth and the importance of their exchange with offshore waters.

However, current flow is not the major driver of the sinks for Hg in the GoM and does not control its residence time. Overall, the residence time of water in the GoM mixed layer is short (2.3 months) compared to the deeper waters (9 months) and the residence time of HgT is smaller for both (0.5 months for the mixed layer and 3.5 months for the deeper waters). These residence time differences further demonstrate the importance of other transport processes besides water transport, such as gas evasion and removal from the mixed layer by particle settling, and sediment removal at depth.

In conclusion, atmospheric deposition and terrestrial sources do not exert major controls over the concentrations of HgT and MeHg in the GoM. For total HgT, external inputs are the major source while for MeHg its major source is in situ production. Given the impacts of changing climate on the offshore circulation and the connectivity between the GoM and the North Atlantic Ocean surface and deeper currents, it is likely that the inputs of Hg to the GoM will change in the future. For MeHg, climate change and the rapid increase in temperature will impact the net rate of methylation in the GoM water column and clearly this should be further investigated given that this may be the main driver of fish MeHg concentrations. A recent study of commercial fish species from the GoM found higher Hg in pelagic species from offshore relative to inshore locations (Taylor et al., 2025), consistent with patterns of MeHg in the water column from this study. The modeling of Schartup et al. (2019) and the measurements of Taylor et al. (2025) of MeHg in several fish species also suggest that other climate-driven factors such as changes in fish diet seasonally and over time, as well as changes due to fish migration north, can have a substantial impact on fish MeHg. In conclusion, the GoM should receive further scientific scrutiny in terms of the sources and cycling of MeHg given its importance to commercial and recreational fishing.

Our findings highlight differences in Hg⁰ formation and evasion between the coastal and estuarine waters and document the lack of substantial change in concentrations and gas evasion of DGHg over the last decade even given the potential decline in atmospheric Hg inputs and other Hg sources to the GoM over that time period. Finally, our examination of the current sources and sinks for Hg and MeHg to the GoM show that the major inputs for HgT are from inflowing surface and deep waters and atmospheric deposition, with the sinks being primarily outflow and sediment burial, and gas evasion being a smaller flux. For MeHg, we conclude that in situ net formation of MeHg is the most important source with external inputs being a small contribution. Overall, the mass balance suggests that the GoM is a net sink for Hg being transported into the region from offshore waters and other sources.

Data availability statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: (URL: https://www.bco-dmo.org/project/896019). The data is being archived as required by the funding organization at BCO-DMO under the project link shown above.

Author contributions

RM: Conceptualization, Data curation, Formal Analysis, Funding acquisition, Investigation, Methodology, Supervision, Writing – original draft, Writing – review and editing. HI: Data curation, Investigation, Methodology, Writing – review and editing. SS: Data curation, Investigation, Methodology, Writing – review and editing. VT: Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Project administration, Supervision, Writing – review and editing.

Funding

The authors declare that financial support was received for the research and/or publication of this article. National Science Foundation Chemical Oceanography Program (Grant # 2148407 UConn; # 2148683 Dartmouth).

Acknowledgements

We thank members of the RM Lab and Dartmouth colleagues, and the crew of the R/V Endeavor, for their assistance with sample collection, analysis and other project related activities. Additional thanks to the NSF Chemical Oceanography program for providing funding for this project (Grant # 2148407 UConn; # 2148683 Dartmouth). The data collected during this study formed the basis of the MS theses of SS and HI.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

The author(s) declared that they were an editorial board member of Frontiers, at the time of submission. This had no impact on the peer review process and the final decision.

Generative AI statement

The authors declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fenvc.2025.1701684/full#supplementary-material

References

AMNET (2025). The atmospheric Mercury network. Available online at: https://nadp.slh.wisc.edu/networks/atmospheric-mercury-network/(Accessed October, 2025).

Google Scholar

Amos, H. M., Sonke, J. E., Obrist, D., Robins, N., Hagan, N., Horowitz, H. M., et al. (2015). Observational and modeling constraints on global anthropogenic enrichment of mercury. Environ. Sci. Technol. 49, 4036–4047. doi:10.1021/es5058665

PubMed Abstract | CrossRef Full Text | Google Scholar

Amyot, M., Gill, G. A., and Morel, F. M. M. (1997). Production and loss of dissolved gaseous mercury in coastal seawater. Environ. Sci. Technol. 31 (12), 3606–3611. doi:10.1021/es9703685

CrossRef Full Text | Google Scholar

Andersson, M. E., Gardfeldt, K., and Wangberg, I. (2008). A description of an automatic continuous equilibrium system for the measurement of dissolved gaseous mercury. Anal. Bioanal. Chem. 391, 2277–2282. doi:10.1007/s00216-008-2127-4

PubMed Abstract | CrossRef Full Text | Google Scholar

Balch, W., Huntington, T., Aiken, G., Drapeau, D., Bowler, B., Lubelczyk, L., et al. (2016). Toward a quantitative and empirical dissolved organic carbon budget for the Gulf of Maine, a semi-enclosed shelf sea. Glob. Biogeochem. Cycles 30 (2), 268–292. doi:10.1002/2015GB005332

CrossRef Full Text | Google Scholar

Balch, W., Drapeau, D. T., Bowler, B. C., Record, N. R., Bates, N. R., Pinkham, S., et al. (2022). Changing hydrographic, biogeochemical, and acidification properties in the Gulf of Maine as measured by the Gulf of Maine north Atlantic time series, GNATS, between 1998 and 2018. JGR-Biogeosci 127 (6), e2022JG006790. doi:10.1029/2022JG006790

PubMed Abstract | CrossRef Full Text | Google Scholar

Balcom, P. H., Fitzgerald, W. F., Vandal, G. M., Lamborg, C. H., Rolfhus, K. R., Langer, C. S., et al. (2004). Mercury sources and cycling in the connecticut river and Long Island Sound. Mar. Chem. 90, 53–74. doi:10.1016/j.marchem.2004.02.020

CrossRef Full Text | Google Scholar

Bieser, J., and Schrum, C. (2016). Impact of marine Mercury cycling on coastal atmospheric mercury concentrations in the North- and Baltic Sea region Elementa 4, 000111, doi:10.12952/journal.elementa.000111

CrossRef Full Text | Google Scholar

Bowman, K. L., Hammerschmidt, C. R., Lamborg, C. H., and Swarr, G. (2015). Mercury in the North Atlantic Ocean: the U.S. GEOTRACES zonal and meridional sections. Deep Sea Res. 116, 251–261. doi:10.1016/j.dsr2.2014.07.004

CrossRef Full Text | Google Scholar

Brickman, D., Alexander, M. A., Pershing, A., Scott, J. D., and Wang, Z. (2021). Projections of physical conditions in the Gulf of Maine in 2050. Elementa 9 (1), 00055. doi:10.1525/elementa.2020.20.00055

CrossRef Full Text | Google Scholar

Charette, M. A., Moran, S. B., Pike, S. P., and Smith, J. N. (2001). Investigating the carbon cycle in the Gulf of Maine using the natural tracer thorium 234. J. Geophys. Res. 106 (6), 11553–11579. doi:10.1029/1999jc000277

CrossRef Full Text | Google Scholar

Ci, Z. J., Zhang, X. S., Wang, Z. W., Niu, Z. C., Diao, X. Y., and Wang, S. W. (2011). Distribution and air-sea exchange of mercury (Hg) in the Yellow Sea. Atmos. Chem. Phys. 11 (6), 2881–2892. doi:10.5194/acp-11-2881-2011

CrossRef Full Text | Google Scholar

Ci, Z., Zhang, X., and Wang, Z. (2016). Air–sea exchange of gaseous mercury in the tropical coast (Luhuitou fringing reef) of the South China Sea, the Hainan Island, China. Environ. Sci. Poll. Res. 23 (11), 11323–11329. doi:10.1007/s11356-016-6346-5

PubMed Abstract | CrossRef Full Text | Google Scholar

Cossa, D., Heimburger, L.-E., Perez, F. F., Garcia-Ibanez, M. I., Sonke, J. E., et al. (2018). Sources, cycling and transfer of mercury in the Labrador Sea (GEOTRACES GEOVIDE cruise). Biogeosci 15, 2301–2323. doi:10.5194/bg-15-2309-2018

CrossRef Full Text | Google Scholar

Despins, M. C., Mason, R. P., Aguilar-Islas, A. M., Lamborg, C. H., Hammerschmidt, C. R., and Newell, S. E. (2023). Linked mercury methylation and nitrification across oxic subpolar regions. Front. Environ. Chem. 4, 1109537. doi:10.3389/fenvc.2023.1109537

CrossRef Full Text | Google Scholar

Dimento, B. P., and Mason, R. P. (2017). Factors controlling the photochemical degradation of methylmercury in coastal and oceanic waters. Mar. Chem. 196, 116–125. doi:10.1016/j.marchem.2017.08.006

PubMed Abstract | CrossRef Full Text | Google Scholar

Driscoll, C. T., Mason, R. P., Chan, H. M., Jacob, D. J., and Pirrone, N. (2013). Mercury as a global pollutant: sources, pathways, and effects. Environ. Sci. Technol. 47 (10), 4967–4983. doi:10.1021/es305071v

PubMed Abstract | CrossRef Full Text | Google Scholar

Feddersen, D. M., Talbot, R., Mao, H., and Sive, B. C. (2012). Size distribution of particulate mercury in marine and coastal atmospheres. Atmos. Chem. Phys. 12 (22), 10899–10909. doi:10.5194/acp-12-10899-2012

CrossRef Full Text | Google Scholar

Fitzgerald, W. F., Engstrom, D. R., Hammerschmidt, C. R., Lamborg, C. H., Balcom, P. H., Lima-Braun, A. L., et al. (2018). Global and local sources of mercury deposition in coastal New England reconstructed from a multiproxy, high-resolution, estuarine sediment record. Environ. Sci. Technol. 52 (14), 7614–7620. doi:10.1021/acs.est.7b06122

PubMed Abstract | CrossRef Full Text | Google Scholar

Floreani, F., Acquavita, A., Petranich, E., and Covelli, S. (2019). Diurnal fluxes of gaseous elemental mercury from the water-air interface in coastal environments of the northern Adriatic Sea. STOTEN 668, 925–935. doi:10.1016/j.scitotenv.2019.03.012

PubMed Abstract | CrossRef Full Text | Google Scholar

Floreani, F., Barago, N., Klun, K., Faganeli, J., and Covelli, S. (2023). Dissolved gaseous mercury production and sea-air gaseous exchange in impacted coastal environments of the northern Adriatic Sea. Environ. Poll. 332, 121926. doi:10.1016/j.envpol.2023.121926

PubMed Abstract | CrossRef Full Text | Google Scholar

Flores-Cervantes, D. X., Plata, D. L., MacFarlane, J. K., Reddy, C. M., and Gschwend, P. M. (2009). Black carbon in marine particulate organic carbon: inputs and cycling of highly recalcitrant organic carbon in the Gulf of Maine. Mar. Chem. 113 (3), 172–181. doi:10.1016/j.marchem.2009.01.012

CrossRef Full Text | Google Scholar

Geyer, R. W., and Ralston, D. K. (2018). A mobile pool of contaminated sediment in the Penobscot Estuary, Maine, USA. STOTEN 612, 694–707. doi:10.1016/j.scitotenv.2017.07.195

CrossRef Full Text | Google Scholar

Gilmour, C., Bell, J. T., Soren, A. B., Riedel, G., Riedel, G., Kopec, A. D., et al. (2018). Distribution and biogeochemical controls on net methylmercury production in Penobscot River marshes and sediment. STOTEN 640–641, 555–569. doi:10.1016/j.scitotenv.2018.05.276

PubMed Abstract | CrossRef Full Text | Google Scholar

Golombek, N. Y., Kienast, M., Pilskaln, C. H., Algar, C., and Sherwood, O. (2024). Origin and alteration of sinking and resuspended organic matter on a benthic nepheloid layer influenced continental shelf. Geochim. Cosmochim. Acta 366, 31–47. doi:10.1016/j.gca.2023.12.008

CrossRef Full Text | Google Scholar

Hammerschmidt, C. R., and Fitzgerald, W. F. (2006). Methylmercury cycling in sediments on the continental shelf of New England. Geochim. Cosmochim. Acta 70, 918–930. doi:10.1016/j.gca.2023.12.008

CrossRef Full Text | Google Scholar

Hammerschmidt, C. R., Finiguerra, M. B., Weller, R. L., and Fitzgerald, W. F. (2013). Methylmercury accumulation in plankton on the continental margin of the Northwest Atlantic Ocean. Environ. Sci. Technol. 47 (8), 3671–3677. doi:10.1021/es3048619

PubMed Abstract | CrossRef Full Text | Google Scholar

Harding, G., Dalziel, J., and Vass, P. (2018). Bioaccumulation of methylmercury within the food chain of the outer Bay of Fundy, Gulf of Maine. PLoS ONE 13 (7). doi:10.1371/journal.pone.0197220

PubMed Abstract | CrossRef Full Text | Google Scholar

Harris, R., Pollman, C., Hutchinson, D., Landing, W., Axelrad, D., et al. (2012). Mercury in the Gulf of Mexico: sources to receptors. Environ. Res. 119, 42–52. doi:10.1016/j.envres.2012.08.001

PubMed Abstract | CrossRef Full Text | Google Scholar

He, Y. (2018). Air-sea exchange of mercury species in the coastal and open ocean. University of Connecticut, 65.

Google Scholar

Hollweg, T. A., Gilmour, C. C., and Mason, R. P. (2010). Mercury and methylmercury cycling in sediments of the mid-atlantic continental shelf and slope. Limnol. Oceanogr. 55, 2703–2722. doi:10.4319/lo.2010.55.6.2703

CrossRef Full Text | Google Scholar

Huntington, T. G., Balch, W. M., Aiken, G. R., Sheffield, J., Luo, L., Roesler, C. S., et al. (2016). Climate change and dissolved organic carbon export to the Gulf of Maine. JGR-Biogeosci. 121, 2700–2716. doi:10.1002/2015jg003314

CrossRef Full Text | Google Scholar

Kocman, D., Wilson, S. J., Amos, H. M., Telmer, K. H., Steenhuisen, F., Sunderland, E. M., et al. (2017). Toward an assessment of the global inventory of present-day mercury releases to freshwater environments. Int. J. Environ. Res. Publ. Health 14, 138. doi:10.3390/ijerph14020138

PubMed Abstract | CrossRef Full Text | Google Scholar

Kotnik, J., Žagar, D., Novak, G., Ličer, M., and Horvat, M. (2022). Dissolved gaseous mercury (DGM) in the Gulf of Trieste, Northern adriatic Sea. J. Mar. Sci. Eng. 10 (5), 587. doi:10.3390/jmse10050587

CrossRef Full Text | Google Scholar

Lee, C.-S., Lutcavage, M. E., Chandler, E., Madigan, D. J., Cerrato, R. M., and Fisher, N. S. (2016). Declining Mercury concentrations in Bluefin Tuna reflect reduced emissions to the North Atlantic Ocean. Ocean. Environ. Sci. Technol. 50 (23), 12825–12830. doi:10.1021/acs.est.6b04328

PubMed Abstract | CrossRef Full Text | Google Scholar

Liu, M., Chen, L., He, Y., Baumann, Z., Mason, R. P., Shen, H., et al. (2018). Impacts of farmed fish consumption and food trade on methylmercury exposure in China. Environ. Int. 120: 333–344.

PubMed Abstract | CrossRef Full Text | Google Scholar

Liu, B., Schaider, L. A., Mason, R. P., Bank, M. S., Rabalais, N. N., Swarzenski, P. W., et al. (2009). Disturbance impacts on mercury dynamics in northern Gulf of Mexico sediments. JGR-Biogeosci 114 (2). doi:10.1029/2008JG000752

CrossRef Full Text | Google Scholar

Liu, M., Chen, L., Wang, X., Zhang, W., Tong, Y., Ou, L., et al. (2016). Mercury export from Mainland China to adjacent seas and its influence on the marine mercury balance. Environ. Sci. Technol. 50 (12), 6224–6232. doi:10.1021/acs.est.5b04999

PubMed Abstract | CrossRef Full Text | Google Scholar

Liu, C., Chen, L., Liang, S., and Li, Y. (2020). Distribution of total mercury and methylmercury and their controlling factors in the East China Sea. Environ. Poll. 258, 113667. doi:10.1016/j.envpol.2019.113667

PubMed Abstract | CrossRef Full Text | Google Scholar

Liu, M., Mason, R. P., Vlahos, P., Whitney, M. M., Zhang, Q., Warren, J. K., et al. (2023). Riverine discharge fuels the production of methylmercury in a large temperate estuary. Environ. Sci. Technol. 57 (35), 13056–13066. doi:10.1021/acs.est.3c00473

PubMed Abstract | CrossRef Full Text | Google Scholar

Liu, M., Zhou, C., Zhang, Q., Raymond, P. A., Li, X., Mason, R. P., et al. (2025). Continental shelves as the largest marine mercury sinks threatened by fish trawling and climate-related perturbations. Nat. Sustain. doi:10.1038/s41893-025-01642-5

CrossRef Full Text | Google Scholar

Mao, H., and Talbot, R. (2012). Speciated mercury at marine, coastal, and inland sites in new england-part 1: temporal variability. Atmos. Chem. Phys. 12 (11), 5099–5112. doi:10.5194/acp-12-5099-2012

CrossRef Full Text | Google Scholar

Mason, R. P. (2015). “Geomicrobiology of mercury,” in Ehrlich’s geomicrobiology. 16 (Boca Raton, FL: CRC Press), 323–342.

Google Scholar

Mason, R. P., Lawson, N. M., Lawrence, A. L., Leaner, J. J., Lee, J. G., and Sheu, G.-R. (1998). Mercury in the Chesapeake Bay. Mar. Chem. 65, 77–96. doi:10.1016/s0304-4203(99)00012-2

CrossRef Full Text | Google Scholar

Mason, R. P., Lawson, N. M., and Sheu, G. R. (2001). Mercury in the Atlantic Ocean: factors controlling air-sea exchange of mercury and its distribution in the upper waters. Deep Sea Res. II 48 (13), 2829–2853. doi:10.1016/s0967-0645(01)00020-0

CrossRef Full Text | Google Scholar

Mason, R. P., Hammerschmidt, C. H., Lamborg, C. H., Bowman, K. L., Swarr, G. J., and Shelley, R. U. (2017). The air-sea exchange of mercury in the low latitude Pacific and Atlantic oceans. Deep Sea Res. 122, 17–28. doi:10.1016/j.dsr.2017.01.015

CrossRef Full Text | Google Scholar

MDN (2025). Mercury deposition network. Available online at: https://nadp.slh.wisc.edu/;networks/mercury-deposition-network/&/atmospheric-mercury-network/.

Google Scholar

Motta, L. C., Blum, J. D., Popp, B. N., Umhau, B. P., Benitez-Nelson, C. R., Close, H. G., et al. (2022). Mercury isotopic evidence for the importance of particles as a source of mercury to marine organisms. PNAS 119 (44), e2208183119. doi:10.1073/pnas.2208183119

PubMed Abstract | CrossRef Full Text | Google Scholar

Nightingale, P. D., Malin, G., Law, C. S., Watson, A. J., Liss, P. S., Liddicoat, M. I., et al. (2000). In situ evaluation of air-sea gas exchange parameterizations using novel conservative and volatile tracers. Glob. Biogeochem. Cycles 14 (1), 373–387. doi:10.1029/1999gb900091

CrossRef Full Text | Google Scholar

NOAA (2024). National oceanic and atmospheric administration fisheries. Available online at: https://www.fisheries.noaa.gov/species-directory/.

Google Scholar

Ortiz, V. L., Mason, R. P., and Evan Ward, J. (2015). An examination of the factors influencing mercury and methylmercury particulate distributions, methylation and demethylation rates in laboratory-generated marine snow. Mar. Chem. 177, 753–762. doi:10.1016/j.marchem.2015.07.006

PubMed Abstract | CrossRef Full Text | Google Scholar

Osterwalder, S., Nerentorp, M., Zhu, W., Jiskra, M., Nilsson, E., Nilsson, M. B., et al. (2021). Critical observations of gaseous elemental mercury air-sea exchange. Glob. Biogeochem. Cycles 35 (8), e2020GB006742. doi:10.1029/2020gb006742

CrossRef Full Text | Google Scholar

Outridge, P. M., Mason, R. P., Wang, F., Guerrero, S., and Heimburger-Boavida, L.-E. (2018). Updated global and oceanic mercury budgets for the United Nations Global Mercury Assessment 2018. Environ. Sci. Technol. 52, 11466–11477. doi:10.1021/acs.est.8b01246

PubMed Abstract | CrossRef Full Text | Google Scholar

O’Driscoll, N. J., Christensen, T. M., Mann, T. A., Keenan, R., and Klapstein, S. J. (2022). Temporal changes in photoreducible Mercury, photoreduction rates, and the role of dissolved organic matter in freshwater lakes. Bull. Environ. Contam. Toxicol. 108, 635–640. doi:10.1007/s00128-021-03422-1

PubMed Abstract | CrossRef Full Text | Google Scholar

Peckenham, J. M., Kahl, J. S., and Mower, B. (2003). Background mercury concentrations in river water in Maine, U.S.A. Environ. Monit. Assess. 89 (2), 129–152. doi:10.1023/A:1026077824228

PubMed Abstract | CrossRef Full Text | Google Scholar

Pershing, A. J., Alexander, M. A., Brady, D. C., Brickman, D., Curchitser, E. N., Diamond, A. W., et al. (2021). Climate impacts on the Gulf of Maine ecosystem: a review of observed and expected changes in 2050 from rising temperatures Elementa, 9, (1):doi:10.1525/elementa.2020.00076

CrossRef Full Text | Google Scholar

Reinhart, B., Kidd, K. A., Curry, R. A., O'Driscoll, N. J., and Pavey, S. A. (2018). Mercury bioaccumulation in aquatic biota along a salinity gradient in the Saint John River estuary. J. Environ. Sci. (China) 68, 41–54. doi:10.1016/j.jes.2018.02.024

PubMed Abstract | CrossRef Full Text | Google Scholar

Rolfhus, K. R., and Fitzgerald, W. F. (2004). Mechanisms and temporal variability of dissolved gaseous mercury production in coastal seawater. Mar. Chem. 90, 125–136. doi:10.1016/j.marchem.2004.03.012

CrossRef Full Text | Google Scholar

Rudd, J. W. M., Bodaly, R. A., Fisher, N. S., Kelly, C. A., Koper, D., and Whipple, C. (2018). Fifty years after its discharge, methylation of legacy mercury trapped in the Penobscot Estuary sustains high mercury in biota. STOTEN 642, 1340–1352. doi:10.1016/j.scitotenv.2018.06.060

PubMed Abstract | CrossRef Full Text | Google Scholar

Schartup, A. T., Balcom, P. H., and Mason, R. P. (2014). Sediment-porewater partitioning, total sulfur and methylmercury production in estuaries. Environ. Sci. Technol. 48, 954–960. doi:10.1021/es403030d

PubMed Abstract | CrossRef Full Text | Google Scholar

Schartup, A. T., Balcom, P. H., Soerensen, A. L., Gosnell, K. J., Calder, R. S. D., Mason, R. P., et al. (2015a). Freshwater discharges drive high levels of methylmercury in Arctic marine biota. PNAS 112 (38), 11789–11794. doi:10.1073/pnas.1505541112

PubMed Abstract | CrossRef Full Text | Google Scholar

Schartup, A. T., Ndu, U., Balcom, P. H., Mason, R. P., and Sunderland, E. M. (2015b). Contrasting effects of marine and terrestrially derived dissolved organic matter on mercury speciation and bioavailability in seawater. Environ. Sci. Technol. 49 (10), 5965–5972. doi:10.1021/es506274x

PubMed Abstract | CrossRef Full Text | Google Scholar

Schartup, A. T., Qureshi, A., Dassuncao, C., Thackray, C. P., Harding, G., and Sunderland, E. M. (2018). A model for methylmercury uptake and trophic transfer by marine plankton. Environ. Sci. Technol. 52 (2), 654–662. doi:10.1021/acs.est.7b03821

PubMed Abstract | CrossRef Full Text | Google Scholar

Schartup, A. T., Thackray, C. P., Qureshi, A., Dassuncao, C., Gillespie, K., Hanke, A., et al. (2019). Climate change and overfishing increase neurotoxicant in marine predators. Nature 572 (7771), 648–650. doi:10.1038/s41586-019-1468-9

PubMed Abstract | CrossRef Full Text | Google Scholar

Seelen, E., Chen, C. Y., Balcom, P. H., Buckman, K. L., Taylor, V. F., and Mason, R. P. (2021). Historic contamination alters mercury sources and cycling in temperate estuaries relative to uncontaminated sites. Water Res. 190, 116684. doi:10.1016/j.watres.2020.116684

PubMed Abstract | CrossRef Full Text | Google Scholar

Seelen, E., Liem-Nguyen, V., Wünsch, U., Baumann, Z., Mason, R., Skyllberg, U., et al. (2023). Dissolved organic matter thiol concentrations determine methylmercury bioavailability across the terrestrial-marine aquatic continuum. Nat. Commun. 14 (1), 6728. doi:10.1038/s41467-023-42463-4

PubMed Abstract | CrossRef Full Text | Google Scholar

Seidov, D., Mishonov, A., and Parsons, R. (2021). Recent warming and decadal variability of Gulf of Maine and Slope water. Limnol. Ocean. 66 (9), 3472–3488. doi:10.1002/lno.11892

CrossRef Full Text | Google Scholar

Shi, X., Annett, A., Jones, R., Middag, R., and Mason, R. P. (2025). Benthic deposition and burial of total mercury and methylmercury estimated using thorium isotopes in the high-latitude North Atlantic. Geochim. Cosmochim. Acta 399, 191–204. doi:10.1016/j.gca.2025.04.029

CrossRef Full Text | Google Scholar

Smith, J. N., Yeats, P. A., Knowlton, S. E., and Moran, S. B. (2014). Comparison of 234Th/238U and mass balance models for estimating metal removal fluxes in the Gulf of Maine and Scotian Shelf. Cont. Shelf Res. 77, 107–117. doi:10.1016/j.csr.2014.01.008

CrossRef Full Text | Google Scholar

Smith, S., Inman, H., Wunsch, U., Mason, R. P., and Taylor, V. F. (2025). Seasonal characterization of mercury and organic matter from the contaminated Penobscot River to the Gulf of Maine. JGR Biogeosciences.

Google Scholar

Soerensen, A., Sunderland, E. M., Holmes, C. D., Jacob, D. J., Yantosca, R. M., Skov, H., et al. (2010). An improved global model for air-sea exchange of mercury: high concentrations over the North Atlantic. Environ. Sci. Technol. 44 (22), 8574–8580. doi:10.1021/es102032g

PubMed Abstract | CrossRef Full Text | Google Scholar

Soerensen, A. L., Mason, R. P., Balcom, P. H., and Sunderland, E. M. (2013). Drivers of surface ocean mercury concentrations and air–sea exchange in the West Atlantic Ocean. Environ. Sci. Technol. 47 (14), 7757–7765. doi:10.1021/es401354q

PubMed Abstract | CrossRef Full Text | Google Scholar

Starr, L. D., McCarthy, M. J., Hammerschmidt, C. R., Subramaniam, A., Despins, M. C., Montoya, J. P., et al. (2022). Mercury methylation linked to nitrification in the tropical North Atlantic Ocean. Mar. Chem. 247, 104174. doi:10.1016/j.marchem.2022.104174

CrossRef Full Text | Google Scholar

Sun, X., Hu, L., Sun, X., Fan, D., Liu, M., Wang, H., et al. (2023). Mercury burial in modern sedimentary systems of the east China Marginal Seas: the role of coastal oceans in global mercury cycling. Glob. Biogeochem. Cycles 37 (9), e2023GB007760. doi:10.1029/2023GB007760

CrossRef Full Text | Google Scholar

Sunderland, E. M., Krabbenhoft, D. P., Moreau, J. W., Strode, S. A., and Landing, W. M. (2009). Mercury sources, distribution, and bioavailability in the North Pacific Ocean: insights from data and models. GBC 23. doi:10.1029/2008gb003425

CrossRef Full Text | Google Scholar

Sunderland, E. M., Amirbahman, A., Burgess, N. M., Dalziel, J., Harding, G., Jones, S. H., et al. (2012). Mercury sources and fate in the Gulf of Maine. Environ. Res. 119, 27–41. doi:10.1016/j.envres.2012.03.011

PubMed Abstract | CrossRef Full Text | Google Scholar

Sunderland, E. M., Li, M., and Bullard, K. (2018). Decadal changes in the edible supply of seafood and methylmercury exposure in the United States. Environ. Health Persp. 126 (1), 017006. doi:10.1289/EHP2644

PubMed Abstract | CrossRef Full Text | Google Scholar

Tada, Y., Marumoto, K., and Takeuchi, A. (2021). Nitrospina-like bacteria are dominant potential mercury methylators in both the Oyashio and Kuroshio regions of the western North Pacific. Microbiol. Spectr. 9 (2), e00833-21. doi:10.1128/Spectrum.00833-21

PubMed Abstract | CrossRef Full Text | Google Scholar

Taylor, V. F., Buckman, K. B., Chen, C. Y., and Kraft, J. (2025). Geographic and ecological drivers of contaminants and nutrients in commercial fish species. STOTEN 986, 179786. doi:10.1016/j.scitotenv.2025.179786

PubMed Abstract | CrossRef Full Text | Google Scholar

Townsend, D. W., Pettigrew, N. R., Thomas, M. A., and Moore, S. (2023). Warming waters of the Gulf of Maine: the role of Shelf, Slope and Gulf Stream Water masses. Prog. Ocean. 215, 103030. doi:10.1016/j.pocean.2023.103030

CrossRef Full Text | Google Scholar

Tseng, C. M., Lamborg, C. H., and Hsu, S. C. (2013). A unique seasonal pattern in dissolved elemental mercury in the South China Sea, a tropical and monsoon-dominated marginal sea. Geophys. Res. Lett. 40, 167–172. doi:10.1029/2012gl054457

CrossRef Full Text | Google Scholar

Turner, R. R., Kopec, A. D., Charette, M. A., and Henderson, P. B. (2018). Current and historical rates of input of mercury to the penobscot river, Maine, from a chlor-alkali plant. STOTEN 637–638, 1175–1186. doi:10.1016/j.scitotenv.2018.05.090

PubMed Abstract | CrossRef Full Text | Google Scholar

Villar, E., Cabrol, L., and Heimbürger-Boavida, L. (2020). Widespread microbial mercury methylation genes in the global ocean. Environ. Microbiol. Rep. 12, 277–287. doi:10.1111/1758-2229.12829

PubMed Abstract | CrossRef Full Text | Google Scholar

Wallace, E. J., Looney, L. B., and Gong, D. (2018). Multi-decadal trends and variability in temperature and salinity in the Mid-Atlantic Bight, Georges Bank, and Gulf of Maine. J. Mar. Res. 76 (5), 163–215. doi:10.1357/002224018826473281

CrossRef Full Text | Google Scholar

Wang, C., Wang, Z., and Zhang, X. (2020). Characteristics of mercury speciation in seawater and emission flux of gaseous mercury in the Bohai Sea and Yellow Sea. Environ. Res. 182, 109092. doi:10.1016/j.envres.2019.109092

PubMed Abstract | CrossRef Full Text | Google Scholar

Wilke, C. R., and Chang, P. (1955). Correlation of diffusion coefficients in dilute solutions. AIChE 1 (2), 264–270. doi:10.1002/aic.690010222

CrossRef Full Text | Google Scholar

Yeager, K. M., Schwehr, K. A., Louchouarn, P., Feagin, R. A., Schindler, K. J., and Santschi, P. H. (2018). Mercury inputs and redistribution in the penobscot river and Estuary, Maine. STOTEN 622–623, 172–183. doi:10.1016/j.scitotenv.2017.11.334

PubMed Abstract | CrossRef Full Text | Google Scholar

Zhang, Y., Jacob, D. J., Dutkiewicz, S., Amos, H. M., Long, M. S., and Sunderland, E. M. (2015). Biogeochemical drivers of the fate of riverine mercury discharged to the global and arctic oceans. Glob. Biogeochem. Cycles 29 (6), 854–864. doi:10.1002/2015GB005124

CrossRef Full Text | Google Scholar