PENOBSCOT RIVER MERCURY STUDY

Abstract

It has been thought that there are few available remediation options for mercury (Hg) contaminated sediments, short of capping and dredging. Recently, in-situ sorbent amendments have attracted recent attention as a low-cost, low-impact approach for remediation of organic-contaminated sediments (Ghosh et al. 2011). This study is one of the first field trials of in situ amendments for Hg and methyl Hg methyl remediation. In this study, the efficacy of in situ soil amendments was tested as a potential tool for methyl Hg risk mitigation in the Penobscot system, specifically in tidal salt marshes. Four amendments were tested, activated carbon (formulated into SediMiteTM pellets with sand and clay), a pine dust biochars, ferrous chloride (FeCl₂), and lime. The carbon treatments were chosen based on previous laboratory studies demonstrating their ability to reduced pore water Hg and methyl Hg concentrations, and uptake by benthic infauna (Gilmour et al. in review). The lime treatment was designed to alter pH, as Hg methylation rates, and % methyl Hg in soils tends to decrease with pH in the Penobscot system (see Penobscot “Mercury Methylation Studies” Report – Chapter 11). Two study sites in Mendall Marsh were chosen for the study. Field studies of methyl Hg concentrations and production rates in sediments and soils across the Penobscot system pointed to salt marshes as sites of particularly high methyl Hg production and accumulation, and therefore a key target for remediation. The study design was a fully-crossed small plot study, with five treatments at each of two sites in Mendall Marsh. At each site, four amendment treatments plus an unamended control treatment were set up in triplicate. Treatments were randomized in each of three rows at each site. Plots were roughly 1 square meter (m² ) in size. Two sites (“West” and “Central”) on the main west platform of Mendall Marsh were chosen for study. The sites were chosen to represent two different major habitats in the marsh. Both sites contain elevated Hg levels in surface soils, and produce substantial methyl Hg from that contamination. However the sediment chemistry is somewhat different at each site, with the West site more reducing than Central. Dosing rates were 1 (kg) of activated carbon (AC), biochar or Fe (as FeCl₂) per square meter, and 0.5 kg of lime. Amendments were broadcast by hand onto the plots during a tidal cycle where the marsh was not overtopped by tide. Plots were initially treated in Sept. 2010, and were sampled four times post-amendment, through Sept. 2012. The study sampling design was based on two objectives. The first was to determine if amendments reduced potential risk from methyl Hg by reducing pore water methyl Hg concentrations and sediment water partition coefficients. A secondary objective was to evaluate the impact of amendments on sediment biogeochemistry. Carbon amendments were effective in reducing Hg and especially methyl Hg concentrations in pore waters at both study sites in Mendall Marsh. Across all four sampling dates, AC amendments resulted in significant reductions in pore water Hg and methyl Hg at both test sites in the marsh. Biochar significantly reduced methyl Hg 19-2 concentrations at both sites, and total Hg concentrations at Central. AC, applied after being formulated as SediMite™ pellets, gave the highest and most consistent reduction through time. AC amendments reduced pore water methyl Hg concentrations at the both sites by >90% at the one month time point and by 60% to 70% on average across all the four time points through two years. At the one year time point, AC plots had about 75% less pore water methyl lHg than control plots at West and 50% less at Central. On average, AC reduced total Hg in pore waters at both sites by 50% to 60%. Biochar was only a little less effective than AC, providing on average 50% to 70% reductions in methyl Hg and 35% to 55% reductions in total Hg. Lime and FeCl₂ additions had inconsistent effects across the sites and through time. On average, across all time points, neither amendment had a significant impact on either total Hg of methyl Hg concentrations. The treatments were generally well-retained in the plots. In 2011, surface soils in AC plots (0-3 centimeter [cm]) contained more than 10% carbon black on average. It was visually obvious that AC and biochars penetrated deeper into surface soils over time; by Sept. 2012 the average penetration was 2-3 cm. Analysis of metals in soils showed that calcium (Ca) and iron (Fe) were retained in the plots at levels significantly above controls for the 24-month duration of the study. In general, the AC and bioachar amendments did not have any significant impact on surficial pore water chemistry, including pH, and the concentrations of nutrients, dissolved organic matter, anions, cations or redox-sensitive constituents like Fe and sulfide. Carbon amendments did not significantly impact soil bulk density or porosity. Therefore, the carbon amendments may not have significant impacts on marsh plant growth or food webs, but that would need to be explicitly tested. Vegetation cover was visually assessed in the plots in fall 2011. The major species composition and plant density was similar between control and carbon plots. In prior laboratory microcosms studies, AC amendments to Hg-contaminated sediments were highly effective in reducing methyl Hg bioaccumulation by benthic invertebrates (Gilmour et al. in review). In these studies, the effectiveness of AC and other amendments in reducing invertebrate bioaccumulation was well-correlated with their effectiveness in decreasing pore water methyl Hg concentrations, and increasing sediment:water partition coefficients. The mechanism of AC remediation appears to be mainly by reductions in methyl Hg bioavailability to worms, rather than by reductions in sediment methyl Hg production or bulk sediment methyl Hg concentration. AC as a remediation tool for Hg and methyl Hg has been tested in several laboratory studies, and in two other field studies of which the authors are aware. Across these studies, AC was most effective in sediments and soils where sediment:water Hg and methyl Hg partitioning are naturally low. Mendall Marsh soils have some of the lowest observed K* (dissociation constant) values for both Hg and methyl Hg across a broad literature on Hg in the environment (see the Penobscot “Mercury Methylation Studies” Report – Chapter 11). 19-3 In summary, activated carbon amendments could be an effective tool to reducing Hg and methyl Hg risk in contaminated Penobscot Marshes. The next steps in evaluating this tool should be larger-scale plot studies. These studies should include evaluation of animal bioaccumulation, food web structure, plant community structure, and possibly marsh productivity and Hg/methyl Hg flux. Large plot studies (perhaps an acre in size) would also provide a better estimate of the cost of treatment.

* = subscript D