Mercury in Aquatic Systems of the Gulf Islands National Seashore, Southeastern USA

This study reports on levels and speciation of mercury (Hg) in different environmental compartments of selected park units in the Gulf Islands National Seashore (USA), and on potential rates of methyl-Hg (MMHg) formation and degradation in sediments. In the aqueous phase, total (THg) and MMHg concentrations ranged from 0.19 to 14.26 ng/L (n=32) and <0.5 to 1.58 ng/L (n=32), respectively. In sediments, THg concentrations varied from 6.4 to 838 ng/g wet weight (n=32), while MMHg levels ranged from 1 to 17 ng/g (n=21). Potential rates of MMHg formation (M) and degradation (D) in sediments resulted in M/D ratios that were mostly <1, suggesting a low tendency for accumulation of produced MMHg in these sediments. Finally, the detection of THg concentrations averaging 168.18 ± 48 ng/g in tissues of Ulva prolifera points to the tendency of Hg bioaccumulation, and therefore, the need for investigation of Hg levels in fish and shellfish. Overall, our findings show that coastal waters and sediments with very low Hg concentrations could support Hg-contaminated biota, which should justify the need for stringent regulations on Hg introduction to natural systems at both local and regional levels. @JASEM INTRODUCTION In 2001, the Mobile Register (Mobile, AL) reported that several popular commercial and recreational fish species caught in the Gulf of Mexico, including the restaurant delicacies amberjack and redfish, could contain very high mercury (Hg) levels that they should not be sold to the public based on fish consumption standards set by the US Food and Drug Administration (FDA). In the past few decades, biogeochemical studies of Hg in aquatic systems have shown that remote and “pristine” systems removed from direct anthropogenic impacts could contain fish with Hg burdens exceeding safe guidelines for human consumption. In the southeastern USA, water bodies in and around national parks located off the coast lines of Florida and Mississippi could receive Hg via both direct atmospheric deposition on terrestrial and aquatic landscapes and river transport of dissolved Hg to the Gulf of Mexico. However, Hg levels in different environmental compartments in these parks remain unknown, despite the fact that they do host over five million visitors each year including people who come to camp and fish in these park waters. It is likely that Hg reaching the Gulf Coast park units could result in fish Hg levels that are above safe limits for human consumption, and therefore, lead to human exposure if the park’s water bodies have high potentials for MMHg production and accumulation. Based on field and laboratory investigations, this study was initiated to (1) assess both levels and the extent of Hg contamination, if any, in different environmental compartments of selected park units within the Gulf Islands National Seashore Park Network; and (2) to determine the potential of sediments in these park units to produce and accumulate MMHg. MATERIAL AND METHODS Mercury in Aquatic Systems of the Gulf Islands National Seashore..... SEJIN YOUN; AUGUSTINE K. DONKOR, ATTIBAYEBA; JEAN-CLAUDE J. BONZONGO Study site—this study focused on selected islands that are part of the Gulf Islands National Seashore. The study area extends from Florida to Mississippi in southeastern USA, and selected sampling sites included Fort Pickens (FP) on Santa Rosa Island, Naval Live Oaks Park (NLO), and Peridido Key (PK) in Florida (Fig. 1a). In Mississippi, samples were collected from Petit Bois, Horn, East Ship, and West Ship islands (Fig. 1b). Water, sediment, and biota samples were collected during both dry and rainy seasons in Florida, and only during the wet season in park units located in Mississippi. Sample collection and analytical techniques—surface water samples were collected in acid pre-cleaned Teflon bottle with gloved hands, using “ultra-clean free-metal sampling” protocol. In the laboratory, samples were acidified with optima-HCl at a final concentration of 1% (v/v) and kept refrigerated at 4 C until analysis. Only non-filtered samples were analyzed in this study. Total-Hg (THg) concentrations were determined in water samples after cold oxidation by bromine monochloride (BrCl) prior to reduction with SnCl2. Only labile MMHg was determined on aqueous samples by direct ethylation of non-distilled water sample using sodium tetraethylborate (Bloom, 1989 and references therein). Figure 1: Map showing the location of selected national park units and sampling sites in (top figure) the state of Florida, and (bottom figure) in Mississippi, USA For MMHg analysis, the standard addition method was used to account for matrix interferences. Following the ethylation process, MMHg compounds were stripped from solution, trapped on Tenax, transferred by heating onto a GC packed column (15% OV-3 on Chromosorb), and detected after separation and breakdown of alkyl-Hg compounds at ~800 C by cold vapor atomic fluorescence spectrometry (CV-AFS). QA/QC procedures included the use of standard addition method, duplicate analyses, reagent blanks and instrument calibration using standard solutions. Details on several aspects on these analytical procedures have been described in our previous publications (e.g., Warner et al., 2003, 2005; Donkor et al., 2005). Collection and analysis of sediment samples and aquatic plant tissues—sediment samples were collected from all sampling sites by hand coring while seaweeds were grabbed from ocean floor at near shore locations with gloved hands. The latter were from the Ulvaceae family, namely Ulva prolifera. Collected samples were placed into acid-cleaned polyethylene re-sealable bags and kept in the dark on ice in coolers until return to the laboratory. THg in these solid samples (i.e. sediment and plant tissues) was determined on wet samples after hot acid digestion with concentrated 72 Mercury in Aquatic Systems of the Gulf Islands National Seashore..... SEJIN YOUN; AUGUSTINE K. DONKOR, ATTIBAYEBA; JEAN-CLAUDE J. BONZONGO HNO3/H2SO4 mixture of a pre-weighed sample (~1g) in acid-cleaned and marble capped volumetric flasks, heated overnight to a refluxing boil on a hot plate. After cooling and dilution with Nanopure water, THg was determined by the SnCl2 reduction technique and pre-concentration on gold traps prior to detection by atomic fluorescence spectrometry. MMHg in sediment and plant tissues was determined following an alkaline alcohol (KOH/CH3OH) digestion of ~1g of wet sediment in an acid-cleaned screw capped Teflon vials and heating at 75C. This step was then followed by aqueous ethylation of an aliquot of buffered digestate and MMHg determined as described earlier for water samples. For both MMHg and THg, in addition to reagent blanks and standard solutions, a certified reference material (IAEA-405, estuarine sediments containing an average THg value of 0.81 mg Kg -1 and 5.49 ng g for MMHg) was run with all digestions/analyses. The percent recovery on the IAEA-405 averaged 95 ± 11% (n=10), and 93 ± 8% (n=10) for THg and MMHg, respectively. To determine the water and organic matter content in sediment samples, a specific amount of sediment was dried in oven for 12 hours at 105°C and water content calculated from the recorded loss of weight. This first step was then followed by combustion of the dried sample in a furnace for 2 hours at 550°C to determine the loss on ignition (LOI) as proxy for organic matter content. Laboratory determination of potential rates of Hg methylation (M) and MMHg degradation (D) in surface sediments—the determination of potential rates of Hg methylation and MMHg degradation were conducted through laboratory assays following a method adapted from Warner et al. (2003). Sediments from each sampling site were mixed with the corresponding in situ water (1:1; vol/vol), homogenized, and apportioned into replicate centrifuge tubes (2-3 mL sediment in 45-mL tubes). Slurries were then bubbled for 1 hour with ultra high purity (UHP) N2 to accelerate the development of anaerobic conditions and stimulate Hg biotransformation. Individual tubes were dosed with either HgCl2 (for methylation experiments) or CH3HgCl (for MMHg demethylation experiments) to final concentrations of 1 and 0.010 ppm, respectively. Slurry containing tubes were then incubated statically for 5 days at room temperature (~22 ̊C) and in the dark. At the end of the incubation period, Hg biotransformation in the tubes was stopped by freezing and storage at -18 ̊C until analysis. The determined ratios of potential transformation rates (M/D ratios) were then used as a parameter to assess the ability of each of the studied sites to produce and ultimately accumulate MMHg. RESULTS AND DISCUSSION Mercury levels and speciation in the aqueous phase: During the dry season, water sample collection was limited to sites located in Florida and corresponding data are shown in Table 1. In August 2004 (rainy season) our sampling campaign extended from Florida to Mississippi and data obtained from samples collected during this second field trip appear in Table 2. For sample sites located in Florida, Figure 2 gives a comparative view of levels and trends of THg in water (2a) and sediments (2b) during both the dry and rainy seasons. Overall, the obtained results show that THg concentrations in these aquatic systems range from 0.19 ng/L to 14.26 ng/L (Table 1) in the dry season and from 0.71 ng/L to 4.67 ng/L in the rainy season. THg values in about 69% of the analyzed samples fell within the range of previously reported worldwide background concentrations of 0.1 to 3.5 ng/L (Lyons et al., 1999). In contrast, the few samples with measured concentration >5ng/L in this study could be an indication of potential contamination from diffuse sources. 73 Mercury in Aquatic Systems of the Gulf Islands National Seashore..... Table 1: THg and MMHg levels in water samples collected from Peridido Key (PK), Fort Picken (FP), Naval Live Oak island (NLO) during the dry season. Italiciz

in Florida (Fig. 1a). In Mississippi, samples were collected from Petit Bois, Horn, East Ship, and West Ship islands (Fig. 1b). Water, sediment, and biota samples were collected during both dry and rainy seasons in Florida, and only during the wet season in park units located in Mississippi.

Sample collection and analytical techniques-surface
water samples were collected in acid pre-cleaned Teflon ® bottle with gloved hands, using "ultra-clean free-metal sampling" protocol. In the laboratory, samples were acidified with optima ® -HCl at a final concentration of 1% (v/v) and kept refrigerated at 4 0 C until analysis. Only non-filtered samples were analyzed in this study. Total-Hg (THg) concentrations were determined in water samples after cold oxidation by bromine monochloride (BrCl) prior to reduction with SnCl 2 . Only labile MMHg was determined on aqueous samples by direct ethylation of non-distilled water sample using sodium tetraethylborate (Bloom, 1989 and references therein). in Mississippi, USA For MMHg analysis, the standard addition method was used to account for matrix interferences. Following the ethylation process, MMHg compounds were stripped from solution, trapped on Tenax ® , transferred by heating onto a GC packed column (15% OV-3 on Chromosorb), and detected after separation and breakdown of alkyl-Hg compounds at ~800 0 C by cold vapor atomic fluorescence spectrometry (CV-AFS). QA/QC procedures included the use of standard addition method, duplicate analyses, reagent blanks and instrument calibration using standard solutions.
Details on several aspects on these analytical procedures have been described in our previous publications (e.g., Warner et al., 2003Warner et al., , 2005Donkor et al., 2005). To determine the water and organic matter content in sediment samples, a specific amount of sediment was dried in oven for 12 hours at 105°C and water content calculated from the recorded loss of weight. This first step was then followed by combustion of the dried sample in a furnace for 2 hours at 550°C to determine the loss on ignition (LOI) as proxy for organic matter content.

Laboratory determination of potential rates of Hg methylation (M) and MMHg degradation (D) in surface
sediments-the determination of potential rates of Hg methylation and MMHg degradation were conducted through laboratory assays following a method adapted from Warner et al. (2003). Sediments from each sampling site were mixed with the corresponding in situ water (1:1; vol/vol), homogenized, and apportioned into replicate centrifuge tubes (2-3 mL sediment in 45-mL tubes). Slurries were then bubbled for 1 hour with ultra high purity (UHP) N 2 to accelerate the development of anaerobic conditions and stimulate Hg biotransformation. Individual tubes were dosed with either HgCl 2 (for methylation experiments) or CH 3 HgCl (for MMHg demethylation experiments) to final concentrations of 1 and 0.010 ppm, respectively. Slurry containing tubes were then incubated statically for 5 days at room temperature (~22˚C) and in the dark. At the end of the incubation period, Hg biotransformation in the tubes was stopped by freezing and storage at -18˚C until analysis. The determined ratios of potential transformation rates (M/D ratios) were then used as a parameter to assess the ability of each of the studied sites to produce and ultimately accumulate MMHg.

RESULTS AND DISCUSSION
Mercury levels and speciation in the aqueous phase: During the dry season, water sample collection was limited to sites located in Florida and corresponding data are shown in  concentrations of 0.1 to 3.5 ng/L (Lyons et al., 1999). In contrast, the few samples with measured concentration >5ng/L in this study could be an indication of potential contamination from diffuse sources.

Mercury levels and speciation in sediments:
concentrations of THg and MMHg determined on non-sieved wet sediment samples ranged from 29 to 838 ng/g, and 1 to 17 ng/g, respectively.
Corresponding data are presented in tables 4 and 5.  Both sites had very shallow water with surface sediments under less than 1 foot of water. Therefore high summer temperatures in addition to sulfate reducing bacteria (SRB) have probably played an important role in the production and accumulation of MMHg in these sites (Gilmour et al., 1992;Bloom, 1989 Warner et al., 2005). These MMHg data will be discussed further in relation with the ability of these sediments to produce and accumulate MMHg based on our laboratory methylation and demethylation assays.
Mercury levels in plant tissues-Some of the sites sampled during this study provide a variety of habitats suitable for the growth of benthic marine algae. The most common species to these sites, Ulva prolifera, was collected from 7 sites during the wet season only, and the determined THg concentrations ranged from 105.9 to 237.1ng/g on wet weight basis (Table 6).  (Greger et al., 2005). Windham et al. (2003) found that 70 to100% of the metal concentration of a whole plant was contained in the roots. Accordingly, this preliminary investigation needs to be followed by a well-designed study to look at both Hg speciation in aquatic plants and Hg levels in higher trophic level organisms. Finally, the lack of relationship between THg levels in analyzed plant tissues and THg in either sediments or water could simply be due to the fact THg is not necessarily equal to the bioavailable Hg fraction in water or sediments.

Potential rates of Hg methylation and MMHg
demethylation-determined potential rates of Hg methylation (M), and MMHg demethylation (D) as well as calculated M/D ratios are given in Table 7.
Potential rates of MMHg demethylation ranged from zero to 9.07 ng g -1 day -1 and potential rates of Hg methylation ranged from 0 to 42.90 ng g -1 day -1 . M/D ratios have been widely used to compare the relative rates of Hg biotransformation from different environment types (Warner et al., 2003). M/D ratios determined in this study ranged from zero when no Hg methylation was observed to a value of 6.765.
However, the majority of sites had M/D ratios <1 and orders of magnitude lower than the above mentioned maximum value. The assumption when using the M/D approach has been that the determined ratios would correlate with the observed ambient MMHg concentrations as they represent the net balance of the co-occurring processes of Hg methylation and MMHg demethylation.  (Ullrich et al., 2001). In addition, the sediment organic matter content which tends to enhance methylation rates was quite low in most of these sediment samples (Table 5).  Warner et al. 2005).
This lack of direct relationship finds its explanation in the complexity of the bioaccumulation process and to the poorly known number of steps and mechanisms that lead to the transfer of Hg from sediment following sedimentary biotransformation of inorganic Hg to MMHg to biota. This is because the bioaccumulation of Hg depends not only on Hg levels, but also on several site specific conditions. Based on data obtained in this study, one could speculate that the ability of sampled sediments to produce and accumulate MMHg should be low. However, and as stated above, the complexity of Hg bioaccumulation and biomagnification in aquatic food chains may not support the above conclusion, and an investigation of Hg levels in fish and shellfish is probably the only effective way to address the issue of fish contamination in these systems.