Environmental Pollution 193 (2014) 138e146

Contents lists available at ScienceDirect

Environmental Pollution journal homepage: www.elsevier.com/locate/envpol

Quantifying the effects of soil temperature, moisture and sterilization on elemental mercury formation in boreal soils Ravinder Pannu a, b, Steven D. Siciliano a, Nelson J. O'Driscoll b, * a b

Department of Soil Science, University of Saskatchewan, 51 Campus Drive, Saskatoon, SK S7N 5A8, Canada Department of Earth and Environmental Science, Acadia University, K. C. Irving Environmental Science Center, Wolfville, NS B4P 2R6, Canada

a r t i c l e i n f o

a b s t r a c t

Article history: Received 29 January 2014 Received in revised form 19 June 2014 Accepted 24 June 2014 Available online xxx

Soils are a source of elemental mercury (Hg(0)) to the atmosphere, however the effects of soil temperature and moisture on Hg(0) formation is not well defined. This research quantifies the effect of varying soil temperature (278e303 K), moisture (15e80% water filled pore space (WFPS)) and sterilization on the kinetics of Hg(0) formation in forested soils of Nova Scotia, Canada. Both, the logarithm of cumulative mass of Hg(0) formed in soils and the reduction rate constants (k values) increased with temperature and moisture respectively. Sterilizing soils significantly (p < 0.05, n ¼ 10) decreased the percent of total Hg reduced to Hg(0). We describe the fundamentals of Hg(0) formation in soils and our results highlight two key processes: (i) a fast abiotic process that peaks at 45% WFPS and depletes a small pool of Hg(0) and; (ii) a slower, rate limiting biotic process that generates a large pool of reducible Hg(II). © 2014 Elsevier Ltd. All rights reserved.

Keywords: Soil temperature Soil moisture Soil sterilization Mercury reduction Reaction rates Kinetics

1. Introduction Mercury is ubiquitous in the environment and cycles between terrestrial systems, the atmosphere, oceans, and living organisms. It is a global pollutant, and once released as Hg(0), it remains in the atmosphere for up to 1 year where it is transported globally (Lindberg et al., 2002). Soils are a reservoir in the mercury cycle (Kim and Lindberg, 1995). Research during the past decade has established the importance of natural soils in Hg cycling, showing that emission from soils may contribute substantially (700e1000 Mg Yr1) to the global atmospheric load of Hg (Engle et al., 2001; Coolbaugh et al., 2002; Engle and Gustin, 2002; Gustin et al., 2003, 2006; Zhang et al., 2003). In order to better understand the mobilization of mercury from soil reservoirs we need to know more about the processes controlling the formation of Hg(0) in soils. Elemental mercury in soil can be produced by abiotic or biotic processes. It is generally thought that most of the elemental mercury produced in soil originates from the A horizon and is produced by the high microbial activity and abundance of reductants present in this soil horizon (Carpi and Lindberg, 1997, 1998; Schluter, 2000). Experimental studies performed on natural and contaminated soils

* Corresponding author. E-mail address: [email protected] (N.J. O'Driscoll). http://dx.doi.org/10.1016/j.envpol.2014.06.023 0269-7491/© 2014 Elsevier Ltd. All rights reserved.

have demonstrated the strong dependence of Hg(0) on climatic factors (Gustin et al., 1997a; Gustin et al., 1997b; Poissant and Casimir, 1998; Gustin et al., 2002; Scholtz et al., 2003; Gustin and Stamenkovic, 2005; Lindberg et al., 2007). For example, Gillis and Miller (2000) found that Hg(0) emission rates in low-mercury, fine sandy loam soil can be largely explained by variations in surface soil temperature (r2 ¼ 0.88) and the Hg concentration gradient between the soil air and ambient air. This temperature dependence has been observed both in diurnal (Gustin et al., 2006) and seasonal studies (Sigler and Lee, 2006). Gustin and Stamenkovic (2005) have demonstrated that small additions of water significantly enhanced Hg(0) release from desert soils. Precipitation may result in Hg(0) release from natural soils (Lindberg et al., 1999; Wallschlager et al., 2000; Song and Van Heyst, 2005). Lin et al. (2010) showed that the synergistic effect from air temperature and soil moisture was 30% greater than the additive Hg flux for the two individual effects. They proposed this effect as a result of enhanced water evaporation at higher temperature, which promotes additional Hg(0) mobilization from soil; however no mechanism was suggested. Sigler and Lee (2006) suggest that Hg(0) bound to upper soil layers may be desorbed by an increase in soil temperature, thereby increasing the pool of gaseous Hg(0) in soil air spaces. However, little information is available on the controlled effect of soil temperature and moisture on Hg(0) formation in soils.

R. Pannu et al. / Environmental Pollution 193 (2014) 138e146

The mechanism by which Hg(0) is formed in soil is not well understood, however some research suggests that abiotic processes such as the desorption of Hg(0) adsorbed onto soil particles or alternatively, abiotic reactions in the soil can produce Hg(0) from available Hg(II) (Pehkonen and Lin, 1998; Zhang and Lindberg, 1999; Scholtz et al., 2003; Gu et al., 2011). Hg(0) can readily adsorb onto the surface of particles and remain there; for example Bouffard and Amyot (2009) found that 200 pg of Hg(0) adsorbed via Van der Waals type forces onto 1 g of sediment in less than 1 h with maximum adsorption (approximately 85%) taking place in the first 5 min. Similarly, in soil there is a pool of Hg(0) that can readily desorb into the soil airspace. In addition, there are a wide range of aqueous abiotic processes such as reduction of Hg(II) mediated by humic acids, fulvic acids, free radical electrons, and sunlight mediated photoreduction that transform Hg(II) in the soil solution to Hg(0) (Schluter, 2000; Gabriel and Williamson, 2004). These abiotic processes can be enhanced in the presence of mixed valence (Fe(II)/Fe(III)) iron oxide minerals and elevated pH (Wiatrowski et al., 2009). In addition to physical and chemical processes, microbial activity may contribute to Hg(0) production in soils (Fritsche et al., 2008; Choi and Holsen, 2009). This bacterial production of Hg(0) can occur at ambient Hg(II) concentrations via mercury-specific detoxification pathways (Barkay et al., 1991) or non-specific microbial reduction of the Hg(II) linked to the microbial detoxification of reactive oxygen species (Siciliano et al., 2002). The production of Hg(0) is not linked to total Hg(II) in soil but to the bioavailable fraction of Hg(II) in soil. For example, Rasmussen (1994) found that the differences in response of the two mer-lux derivatives of Escherichia coli in agricultural and beech forest soil dosed with equal amounts of total Hg were likely due to differences in the bioavailability of Hg(II). Soil properties not only influence the bioavailability of Hg(II) but also the rate of microbial transformation and affinity of the soil for Hg(0). Organic matter content not only alters the affinity between Hg(0) and sediments under anoxic conditions but also accelerates biotic reduction of Hg(II) to Hg(0) (Bouffard and Amyot, 2009). Thus, it is plausible that microbial reduction may contribute to Hg(0) production in soil, however the relative importance of abiotic to biotic microbial processes has not been quantified. Soil is a large reservoir for mercury in which temperature and moisture changes affect mercury speciation processes; however the effects of these variables on mercury reduction processes in soils have not been quantified in a controlled manner. As such, the objectives of this study were to (a) characterize the effects of temperature and moisture on rate of abiotic and combined abiotic/ biotic Hg(0) formation in soil under controlled conditions and (b) to estimate the proportion of Hg(0) production arising due to microbial activity. 2. Methods and materials 2.1. Soil sampling Soil samples were collected from Kejimkujik National Park (KNP) and Antigonish County, Nova Scotia, Canada. The mercury content of bedrock geology (0.2 ng Hg g1e38.9 ng Hg g1 for surface outcrops) in KNP is similar to other areas within Canada (O'Driscoll et al., 2005). Four surface soil samples (0e15 cm depth) were collected from KNP on April 16e17, 2010, and six from Antigonish County on June 28e29, 2010 (Table 1). The organic litter was removed and surface soil (0e10 cm depth) homogenized with a stainless steel spade. Each sample was a pooled composite of four samples (~0.5 kg each) collected within 100 m2. The 2 kg pooled samples were further homogenized and stored in polypropylene bags (253 K in dark) until analysis. Thawed soil samples were dried in the dark in a growth chamber (Conviron E15) at 298 K and 0% relative humidity for 72-h each. The dried samples were sieved (2 mm stainless steel) and stored in polypropylene containers. Soil pH was measured in 0.01 M CaCl2 solution with 1:2.5 soil:solution ratio (Mehlich, 1976) and electrical conductivity (EC) was measured in Milli-Q water with 1:2 soil:water ratio. Total Hg concentrations were quantified using aqua regia (37% HNO3 þ 63% HCl, 1:3) digestion and cold vapor e atomic absorption spectroscopy.

139

Table 1 Locations and physical and chemical characteristics of the soils used in this study. Soil IDa

Longitude

Latitude

pH

EC dS/m

WHC ml kg1

OC g kg1

Total Hg ng g1

K1 K2 K5 K7 A11 A12 A13 A14 A15 A18

44 260 4500 N 44 270 3000 N 44 190 4900 N 44 170 5500 N 45 450 0600 N 45 450 0600 N 45 420 1400 N 45 400 3100 N 45 390 2200 N 45 390 2700 N

65 150 1900 W 64 590 0100 W 65 140 0600 W 65 140 5400 W 61 560 4900 W 61 560 4600 W 61 590 2500 W 61 430 2900 W 61 500 3200 W 61 510 2500 W

4.6 4.4 4.8 4.7 4.2 4.3 4.4 5.4 5.1 4.3

0.03 0.02 0.02 0.03 0.04 0.05 0.08 0.19 0.18 0.05

46.2 49.6 31.1 37.3 43.7 69.8 59.3 59.1 56.3 30.7

36 26 15 20 16 65 35 24 37 05

105 104 66 66 28 106 69 50 96 13

a

K: Kejimkujik National Park; A: Antigonish County.

Soils were analyzed for organic matter (OM) (Walkley and Black, 1934) and water holding capacity (WHC) (Franzluebbers, 1999). To eliminate microbial activity without significantly altering physical and chemical properties (see Supporting information), homogenized sub samples were gamma (ɤ) irradiated at the Canadian Irradiation Center (CIC) (Thuerig et al., 2009). As such, 500 g each of air dried soil in a Ziploc bag within an air-sealed polypropylene plastic container was placed inside a carrier irradiator (JS-8900, s/n: IR-147) and irradiated in continuous mode. Each soil sample was uniformly exposed to a maximum dose of 30 kGy (Harwell Red 4034 dosimeter at 640 nm). Sterility was tested according to Berns et al. (2008) and the soil samples were considered to be sterile when no microbial growth occurred after a 3 week incubation (20 ± 0.2  C). 2.2. Quartz reaction chamber system A quartz beaker system (O'Driscoll et al., 2003) was used to quantify the effects of soil temperature, WFPS and soil sterilization on Hg(0) formation in soils. This system consists of a fused silica quartz beaker (6 cm diameter, 9.6 cm height, 0.3 L) with Teflon inlet and outlet through a platinum-cured silicone stopper (O'Driscoll et al., 2006). All components were acid washed in 20% HCl acid for 24 h followed by deionized water rinse and drying in a laminar flow hood (Microzone VPFX-4). A Tekran model 1100 mercury zero-air generator supplied mercury-free air (1 L min1) to the chamber. To achieve the temperature ranges (278e303 K), the quartz beaker was immersed in a temperature-controlled water bath (ThermoHakke model K20) and the soil surface temperature was monitored with a Teflon insulated thermocouple (digital thermometer, Omega Eng.). Since no significant difference (n ¼ 10, p < 0.05) was observed for cumulative Hg formed over a 24-h analysis period under both dry and humidified zero air conditions, dry mercury-free zero air was used throughout this experiment to prevent the condensation of water vapors at low temperatures. The analysis consists of soil placed in a quartz beaker (temperature and humidity controlled) and Hg(0) stripped from the chamber using air flow (small turnover time). Hg(0) formation within a shallow soil sample is calculated by continuously removing Hg(0). Soil Hg flux readings were taken every 5 min for 24 h and repeated in triplicate. Soil analysis under dark conditions began with initial blanking of the chamber by passing mercury-free zero air through the chamber without soil in the chamber until no mercury was detectable. Soil (20 g) was then brought to 45% water filled pore space (WFPS) and uniformly placed at the bottom of the quartz beaker in a thin layer with bulk density of 1.60 g cm3. This was done by gently tapping the soil sample in the quartz beaker before analysis until a soil depth of approximately 0.44 cm was achieved. Temperature effects on Hg(0) emissions were evaluated at environmentally relevant surface soil temperatures of 278, 283, 288, 293, and 303 K (Carpi and Lindberg, 1998). To determine the effect of soil moisture, soil (20 g) maintained at 15, 30, 45, 60 and 80 WFPS was uniformly placed at the bottom of the quartz glass beaker in a thin layer with a bulk density of 1.60 g cm3. Humid air was circulated over the soil sample by passing dry zero air through a Milli-Q water-containing bubbler prior to the quartz chamber to maintain moisture levels. Analysis proceeded over a 24-h period while keeping environmental parameters constant (soil temperature 20  C, radiation, air flow rate 1 L min1). Filter packs (0.2 mm Teflon) were placed in-line after the soil chamber to avoid particle contamination. No UV radiation and minimal visible radiation (