Environmental controls on the speciation and distribution of mercury in coastal sediments

Abstract

Methylmercury production by sulfate reducing bacteria in coastal sediments leads to bioaccumulation of mercury in fish, shellfish, and ultimately humans. Sulfur, organic carbon, and sediment structure and composition can all affect methylmercury production by changing the amount of bioavailable inorganic mercury and by stimulating the activity of methylating microbes. This study investigates total and methylmercury in solids and porewaters relative to total sulfide concentration, redox potential, sediment grain size, and total organic carbon in a range of sediment types from the Bay of Fundy region of Canada. Using these data, we construct a conceptual model of the biogeochemical environment surrounding methylating microbes in high sulfide, organically enriched sediments. Whereas other studies of methylmercury dynamics measured porewater sulfide concentrations in relatively low-sulfide systems (∼20–300 μM), we measured total sulfide levels using a method developed to indicate organic enrichment across a much wider range of sulfidic sediments (10–4000 μM). We observed that higher sulfide concentrations correspond to an elevated fraction of mercury in methylated form suggesting higher net methylation rates in these sediments. This relationship is strongest in sediments that are moderately impacted by organic enrichment, but weak in less impacted, aerobic sediments. Higher sulfide concentrations in porewaters containing dissolved organic matter appear to yield a geochemical environment that is conducive to uptake of Hg(II) by methylating bacteria. Data collected in this study imply that moderate levels of organic enrichment through fish farming may enhance methylmercury production in the Bay of Fundy.

Introduction

Despite the fact that the majority of human exposure to mercury in the United States is the result of consumption of estuarine and marine fish and shellfish (Carrington and Bolger, 2002, Tran et al., 2004), information on mercury fate and transport in coastal ecosystems is limited. Only the organic form of mercury, methylmercury (MeHg), biomagnifies to significant concentrations inside living cells and tissues of aquatic organisms (Lawrence and Mason, 2001, Lawson and Mason, 1998). Methylmercury produced in near-shore sediments accounts for the majority of mercury in primary producers that is bioaccumulated at higher trophic levels (Hammerschmidt et al., 2004, Sager, 2002). Coastal ecosystems that produce large amounts of MeHg may therefore lead to localized “hot spots” in human exposure. Since direct measurements of MeHg in coastal systems are often limited, environmental managers would benefit from screening level models that identify geographic regions likely to produce significant amounts of MeHg. This study analyses several factors that may assist in developing such models for MeHg production in coastal sediments.

Many coastal systems contain large reservoirs of “legacy” mercury in the sediment compartment (Hammerschmidt and Fitzgerald, 2004, Sager, 2002, Sunderland et al., 2004). Microbes convert a small fraction of this pool of inorganic mercury (Hg(II)) to MeHg over time. The rate of Hg(II) conversion to MeHg in coastal sediments is therefore crucial for anticipating accumulation of mercury in coastal organisms.

Sulfate reducing bacteria are thought to be the principal agents responsible for MeHg production in coastal sediments (Compeau and Bartha, 1984, Gilmour et al., 1992, King et al., 1999). These microbes thrive at the geochemical interface between oxic and anoxic conditions (Hintelmann et al., 2000). A number of environmental factors may affect the rate of MeHg formation by influencing the supply of bioavailable Hg(II) and/or activity of methylating microbes. In addition to Hg(II) concentrations, effective proxy indicators for MeHg production and accumulation identified by previous research include sulfide concentrations, total organic carbon (TOC), and redox potential (Eh) (Baeyens et al., 1998, Benoit et al., 1999a, Benoit et al., 2001, Compeau and Bartha, 1984, Mason and Lawrence, 1999, Stoichev et al., 2004).

Past studies show organic matter can influence MeHg production by reducing the amount of bioavailable Hg(II) in the dissolved phase and stimulating the activity methylating bacteria by providing a substrate for mineralization (Hammerschmidt et al., 2004, Hammerschmidt and Fitzgerald, 2004, Mason and Lawrence, 1999, Stoichev et al., 2004). In addition, a number of studies show that in certain systems elevated sulfide concentrations inhibit mercury methylation by reducing the bioavailable pool of Hg(II) (Benoit et al., 1999a, Benoit et al., 1999b, Benoit et al., 2001). Overall, the ultimate influence of changes in environmental factors on methylation appears to depend largely on the initial characteristics of a specific ecosystem that limit the production of MeHg by methylating microbes, whether this is the pool of bioavailable Hg(II) or other factors that affect microbial activity.

This study investigates some of the main geochemical factors known to affect the speciation and distribution of mercury in coastal sediments from the Bay of Fundy, Canada. Although there are no large point sources of mercury in this region, high levels of mercury in fish and wildlife continue to be a problem (NESCAUM, 1998). Organic enrichment of sediments from salmon mariculture is another ongoing management concern in this region. To help identify impacted regions, Wildish et al., 1990, Wildish et al., 1993, Wildish et al., 1999, Wildish et al., 2001 developed an empirically derived organic enrichment gradient based on benthic macrofaunal characteristics. Geochemical indicators like Eh and S²⁻ have been widely applied in the region to monitor the level of organic enrichment in the sediments. A recent interlaboratory calibration exercise found that total sulfide measurements in interfacial marine sediments are the most reliable indicator of enrichment in the Bay of Fundy; i.e. sulfides increase after organic materials are added to an ecosystem (Wildish et al., 2004).

We hypothesized that the Bay's predominantly clay mineral sediments (Loring, 1979, Loring et al., 1998) support higher levels of mercury methylation when the amount of organic matter is enhanced by activities such as fish farming. To test this hypothesis, we measured concentrations of total mercury (THg) and MeHg in the solid and dissolved phases as a function of TOC, Eh, sediment grain size and total sulfides in surface sediments from multiple sites across the mouth of the Bay of Fundy. In this paper, we calculate the covariance between sulfide, Eh and TOC throughout the Bay, and assess the utility of these measurements as indicators of an ecosystem's ability to methylate mercury. We use the empirical data on mercury distributions as a function of different environmental characteristics to make inferences about mechanistic processes controlling MeHg production in coastal sediments. By isolating these relationships in the context of empirical data collected in this study, we gain useful insights into the mechanism of MeHg formation in coastal sediments that can be tested with future research.

Other studies that have investigated MeHg dynamics have focused on coastal systems that are moderately to highly contaminated by urbanization and industry (Bloom and Fitzgerald, 1988, Hammerschmidt and Fitzgerald, 2004, Mason and Lawrence, 1999, Sager, 2002, Stoichev et al., 2004). The majority of data presented in this study were collected in a region where the enrichment of mercury in surface sediments can be attributed mainly to enhanced atmospheric deposition from a combination of regional and global mercury sources (Sunderland et al., 2004). Therefore, results from this study present new information on the biogeochemical cycling of atmospherically derived mercury in a macrotidal estuary that has not been significantly impacted by local mercury sources. Finally, we use the empirical data to develop simple statistical models of MeHg distribution across the Bay of Fundy. Combined with our previous work (Sunderland et al., 2004), this information contributes to the development of a mercury cycling model for this region (Sunderland, 2003).