Net Degradation of Methyl Mercury in Alder Swamps - ACS Publications

Article pubs.acs.org/est

Net Degradation of Methyl Mercury in Alder Swamps Rose-Marie Kronberg,†,§ Ida Tjerngren,†,§ Andreas Drott,† Erik Björn,‡ and Ulf Skyllberg*,† †

Department of Forest Ecology and Management, Swedish University of Agricultural Sciences, SE-901 83 Umeå, Sweden Department of Chemistry, Umeå University, SE-901 87 Umeå, Sweden

‡

S Supporting Information *

ABSTRACT: Wetlands are generally considered to be sources of methyl mercury (MeHg) in northern temperate landscapes. However, a recent input-output mass balance study during 2007−2010 revealed a black alder (Alnus glutinosa) swamp in southern Sweden to be a consistent and significant MeHg sink, with a 30−60% loss of MeHg. The soil pool of MeHg varied substantially between years, but it always decreased with distance from the stream inlet to the swamp. The soil MeHg pool was significantly lower in the downstream as compared to the upstream half of the swamp (0.66 and 1.34 ng MeHg g−1 SOC−1 annual average−1, respectively, one-way ANOVA, p = 0.0006). In 2008 a significant decrease of %MeHg in soil was paralleled by a significant increase in potential demethylation rate constant (kd, p < 0.02 and p < 0.004, respectively). In contrast, the potential methylation rate constant (km) was unrelated to distance (p = 0.3). Our results suggest that MeHg was net degraded in the Alnus swamp, and that it had a rapid and dynamic internal turnover of MeHg. Snapshot stream input-output measurements at eight additional Alnus glutinosa swamps in southern Sweden indicate that Alnus swamps in general are sinks for MeHg. Our findings have implications for forestry practices and landscape planning, and suggest that restored or preserved Alnus swamps may be used to mitigate MeHg produced in northern temperate landscapes.

■

INTRODUCTION Formation and bioaccumulation of the potent neurotoxin methyl mercury (MeHg) in soils and waters is an environmental issue of global concern. Apart from atmospheric deposition of inorganic Hg (Hginorg), this situation is linked to the formation of MeHg in soils, sediments and waters under reducing conditions. Artificial and natural wetlands have been identified as principal sources of MeHg in northern temperate landscapes.1−3 In Sweden alone about 50 000 lakes have Hg concentrations in fish exceeding health guidelines (0.5 mg kg−1).4 Approximately half of the world’s wetland area has been drained for agricultural and silvicultural purposes.5 However, due to an increased awareness of the positive effects wetlands have on biodiversity, hydrology, and nutrient filtration, significant efforts are currently undertaken to restore previously drained areas.6 Society thus has to balance the positive effects of wetland restoration against the potentially negative effects of increased MeHg production and subsequent accumulation in food webs. In a recent input−output mass balance study we reported a considerable variation in MeHg yield (the difference in mass between stream output and the sum of all inputs) in boreal wetlands with different nutrient status.7 Across all four years of the study seven out of eight wetlands, including various types of fens and a bog, were sources of MeHg with yields corresponding to 6−670% of inputs. Fens with an intermediate nutrient status (pH ∼5 and C/N ratio ∼20) were the largest MeHg sources, confirming and extending previous findings.8−11 In sharp contrast, a black alder (Alnus glutinosa) swamp, © 2012 American Chemical Society

Edshult, was shown to be a consistent MeHg sink, with 29− 58% of the annual MeHg input retained or degraded in the swamp during the four years of study.7 To our knowledge, this is the first report of a northern temperate wetland acting as a significant MeHg sink. Black and gray alders (Alnus incana) are widely distributed across the boreal zone.12 In Sweden, Alnus glutinosa and Alnus incana cover roughly 107 000 ha and 50 000 ha, respectively.13 Alnus glutinosa is found in the south, whereas Alnus incana is more common in the north.14 Due to N-fixation, Alnus swamps are considered as productive environments. They have therefore been liable to extensive drainage for agricultural and silvicultural purposes. In Sweden roughly 50% and 16% of the areas previously covered by Alnus glutinosa and Alnus incana, respectively, have been drained.15 Since alders grow in discharge areas they are usually located downstream MeHg sources such as bogs, fens, and forest clear-cuts,16 forming buffer zones along lakes, streams, and rivers.17 In this study we report on the soil processes important for making the Alnus swamp Edshult a MeHg sink. The Edshult site has all the important characteristics of a well-developed Alnus glutinosa swamp,12 most importantly permanently flooded soils with overland flowing water. A key question is whether MeHg is net degraded or accumulated in the soil. To Received: Revised: Accepted: Published: 13144

December 28, 2011 November 8, 2012 November 19, 2012 November 19, 2012 dx.doi.org/10.1021/es303543k | Environ. Sci. Technol. 2012, 46, 13144−13151

Environmental Science & Technology

Article

Figure 1. Total catchment area of the Alnus swamp Edshult (0.58 km2). The size of the upstream watershed drained by the stream forming the inlet to the swamp is 0.46 km2 and the Alnus swamp area is 0.042 km2. Lantmäteriet, I2011/0032.

As a contrasting reference to Edshult, a peatland (Långedalen) shown to be a consistent source of MeHg,7 was included in the study. Långedalen is located in southern Sweden (N 58° 19.870′, E 12° 30.181′) and the total wetland area measures 1.03 km2. The upstream part of the peatland consists of a peat moss (Sphagnum spp.) dominated bog, transitioning into a more nutrient rich fen dominated by broadleaved grasses. In that respect, the site represents the two most common boreal wetland types (bog and fen). Ten additional sites were selected for the snapshot study (see description below). Nine of the sites were characterized as forest swamps, eight of which with a tree layer dominated by Alnus glutinosa and one dominated by downy birch (Betula pubescens). The swamps are all located in discharge areas, have flooded soils with overland flow, and are nonpeat forming. The last site was a semiopen bog with a Betula pubescens tree cover. A map showing the location of all sites and a table reporting the tree composition and upstream land use is provided in the Supporting Information (SI) (Figure S1 and Table S1, respectively). Soil Sampling and Sample Preparation. Soil samples were taken in the Alnus swamp Edshult in November 2006, September 2007, September 2008, May 2009, and May 2010. The upper 10 cm of the soil is highly organic (350−450 mg organic carbon g−1, Table S2, SI). Stones and boulders become highly abundant beneath 20 cm depth, restricting both sampling and water penetration. Soil samples (0−10 cm) were taken in spots with at least 10 cm deep organic soil, using a steel peat corer (10.5 cm inner diameter) with a cutting edge. The soil surface was defined as 0 cm depth, after removal of living mosses or plants. On all sampling occasions the soil had at least one cm of standing water at the surface, and was thus water saturated. During 2008−2010, soil samples were taken and put into double zip-lock plastic bags. Excess air was forced out manually before the bags were sealed. Any air that was mixed into the soil during sampling was likely rapidly consumed by microbial processes, well before samples were processed in the laboratory. In 2006 and 2007, 1 L plastic buckets with rubber lids were used, which were filled with soil

investigate this, we quantified soil pools of MeHg and total Hg (HgTOT) and made soil incubation experiments to assess potential methylation and demethylation rates. In addition, stream input-output budgets were determined during one single rain event at nine other forest swamps, eight of which dominated by Alnus glutinosa. Our results suggest that MeHg is net degraded in the soil along the water flow path at Edshult and that Alnus glutinosa swamps may be important sinks for MeHg in general. These findings have implications for the restoration of Alnus swamps, and their potential role in forestry and landscape planning to counteract MeHg production from upstream located environments.

■

MATERIALS AND METHODS Site Descriptions. The main site of this study is the Alnus swamp Edshult located in southern Sweden (N 57° 34.278′, E 15° 12.052′, Figure 1). The site is described in more detail by Tjerngren et al,11 but a short description follows. The Alnus swamp measures 0.042 km2 and is a nonpeat forming wetland where the incoming water have good contact with the soil due to branching of the main stream and overland flow. Except at very dry periods, the swamp has more than half of its area covered by standing water. The tree density in the Alnus swamp is highest in the upstream third, becomes lower going into the middle third, and the downstream third is an open area with few trees where the understory vegetation predominates. The tree height in the swamp is ∼20 m. The understory vegetation changes along the flowpath of water, which is roughly parallel to the south to north axis, with the upstream third dominated by acidophilic feather mosses, common woodsorrel (Oxalis acetosella) and low growing ferns. The middle third is dominated by the high lady-fern (Athyrium f ilix-femina) and wood club rush (Scirpus sylvaticus), and the downstream third is dominated by Scirpus sylvaticus. The swamp receives runoff from 0.37 km2 of formerly drained peaty soils forested by ∼50year old Picea abies, a Pinus silvestris vegetated partly drained ombrotrophic bog of 0.084 km2, and a 0.034 km2 area previously forested with Picea abies that was clear-cut in 2008. 13145

dx.doi.org/10.1021/es303543k | Environ. Sci. Technol. 2012, 46, 13144−13151


Article

swamp. In order to remove variability due to differences in bulk density and time of sampling, MeHg concentrations were normalized to soil organic carbon (MeHg SOC−1) and to the average soil MeHg SOC−1 concentration obtained at the specific sampling occasion. Because the bulk density of soil samples was only determined in 2010, the concentration of MeHg normalized to SOC was used as a proxy for the soil pool mass of MeHg per volume of soil (or per wetland area, if the soil depth is considered reasonably stable throughout the Alnus swamp). The strong relationship observed between MeHg normalized to SOC and MeHg expressed per dm3 (Figure S4, SI), verifies that MeHg normalized to SOC is a suitable proxy for the MeHg soil pool. This is explained by the fact that the organic matter content largely controls the spatial variability in the soil bulk density. Linear regression analysis was used to evaluate the relationships between distance from stream inlet of Edshult and %MeHg of HgTOT in soil, km, kd, and km/kd. Prior to linear regression analysis, data were tested for normal distribution and homogeneous variance. The software Minitab 16 (Minitab Inc., Saltsjöbaden, Sweden) was used for these tests.

and water leaving no headspace. Samples were kept dark and cool until sample preparation, and in the laboratory handled in a glovebox filled with N2 (g). The samples were homogenized by hand and pH was measured. Subsamples were taken for incubation studies to assess potential rates of Hg inorg methylation and MeHg demethylation, and for analyzes of HgTOT (Hginorg+MeHg), MeHg, N, C, and S concentrations. Snapshot Budgets and Streamwater Sampling. To test whether Alnus swamps in general are sinks for MeHg, streamwater was sampled from ten sites during one single rain event. From this single sampling, stream mass input-output budgets were determined, hereafter referred to as snapshot budgets. The relevance of the snapshot budget results was evaluated by a comparison of snapshot results with results from complete annual input-output budgets at Edshult and Långedalen in 2009 and 2010. The main criteria used when selecting wetlands for the snapshot budgets were that each site should have one welldefined inlet and outlet, high concentrations of DOC in the inlet, and a high degree of contact between the incoming streamwater and the wetland soil. Sites with draining ditches were disregarded. We choose end of May as sampling date for the snapshot budgets because bacterial activity, and thus processes affecting MeHg net production, have been shown to be significant at that time of year in our previous studies of wetlands in southern Sweden.7,11 In May 2009 and 2010, one streamwater sample was collected from the inlet and one sample from the outlet at each site within approximately one hour. These two samples taken at each site lay the basis for the snapshot budgets. At the same time, water flow rates were measured at in- and outlets by a salt dilution method.18 The aim was to sample after a medium-sized rain event, when water flow rates were about the same at the inlet and outlet. The reported flow rate is an average of three measurements taken within approximately 15 min. Fluxes of MeHg, Hginorg and DOC at in- and outlets were calculated by multiplying concentrations (mass L−1) in streamwater with the water flow rate (Q = L s−1). The snapshot budgets are reported as masses and mass-% net output, as calculated by eq 1, where ∑output and ∑input denote masses of MeHg, Hginorg or DOC.

■

RESULTS AND DISCUSSION Input of Hginorg, MeHg and DOC to the Alnus Swamp, Edshult. As we have previously reported, the Alnus swamp Edshult was a consistent sink for MeHg during the time period 2007−2010 with 1.7−6.2 g km−2, corresponding to 29−58% of the MeHg input.7 For a sink to exist, import is needed. This import was provided by the inlet stream, having average annual, nonvolume weighted MeHg concentrations of 2.4−5.9 ng L−1 (n = 28) as compared with much lower concentrations in the outlet (0.8−1.6 ng L−1 (n = 28). The annual mass import of MeHg was 0.15−0.44 g and the export 0.08−0.22 g (Table S3, SI). The swamp was a sink for Hginorg as well, even though concentrations in inlet and outlet differed less (16−18 and 10− 15 ng L−1 respectively) and the sink accounted for 6.2−21% of the total input. Note that the inlet water was high in acidic and aromatic humic substances, as reflected by an average (±SD) pH of 4.6 ± 0.47, DOC concentration of 99 ± 33 mg L−1, and SUVA254 nm of 4.6 ± 0.75 L mg−1 C1− m−1, during the period 2007−2010. In contrast to Hginorg, for which concentrations did not show any obvious seasonality, stream concentrations of MeHg showed marked peaks during May to September in the inlet but not in the outlet (Figure S2, SI). These peaks reflect periods of net methylation in upstream environments, often occurring during warm periods with low to medium precipitation. Our previously reported mass balance calculations show that the annual average export of Hginorg and MeHg from the catchment area located upstream the swamp was 4.8 ± 1.7 g km−2 and 0.56 ± 0.23 g km−2 respectively. These values are in the high end of the range of exports from that study, which for Hginorg ranged from 0.51 ± 0.011 to 6.1 ± 0.089 g km−2, and for MeHg from 0.04 ± 0.003 to 0.65 ± 0.03 g km−2.7 The average annual export of DOC during 2007− 2010 was 27.0 ± 15.5 mg km−2, which is three times higher than the DOC areal production before, and on the same order as after, flooding of a boreal peatland.20 These observations all show that the environments located upstream supported the Alnus swamp Edshult with a significant input of MeHg in association with Hginorg and DOC. Spatial Patterns in Soil Pools of HgTOT and MeHg at Edshult. We used HgTOT and MeHg normalized to SOC as

mass − %net output = 100 × (∑ output − ∑ input)/∑ input

(1)

Sampling protocols were followed to avoid trace metal contamination, with blanks included. Samples for HgTOT and MeHg analyses were collected in acid-washed FEP-Teflon bottles, and HDPE-bottles were used for ancillary chemistry samples. All samples were kept dark and cool until sample preparation. Chemical Analyses and Soil Incubation Studies. The methods for chemical analyses of concentrations of total Hg and MeHg in soil and water, ancillary chemistry and methods for soil incubations and subsequent calculations of potential rates of Hg methylation and MeHg demethylation are described in detail elsewhere,11 with a brief description in SI. Statistics and Data Normalization. MeHg concentration data were tested for normal distribution and homogeneous variance using Shapiro-Wilkison test and Levene statistics, respectively.19 The software PASW Statistics 18 (Chicago, IL) was used for these tests. One-way ANOVA was used to test for differences in soil MeHg concentrations between the upstream (0−250 m) and downstream (270−440 m) part of the Alnus 13146



Article

Figure 2. Soil data from the Alnus swamp Edshult September 2008. Linear regressions of (a) %MeHg, (b) kd, (c) km/kd, and (d) km with distance from inlet, and (e) correlation between %MeHg and km/kd. Error bars for kd and MeHg represent SD for duplicate samples.

to be 1998 ng MeHg dm−3. Integrated over the whole swamp of 0.042 km2 and 10 cm depth, the total mass of MeHg in the wetland was 8.4 g. This value can be compared with an annual input of MeHg varying between 0.15 g (2009) and 0.44 g (2007) from the inlet stream (Table S3, SI). Thus, the annual input roughly corresponded to 2−5% of the total mass of MeHg in the wetland. Spatial and Seasonal Patterns of Methylation and Demethylation Processes at Edshult. In September 2008, %MeHg of HgTOT, km, and kd were determined in seven soil samples from the Alnus swamp Edshult. A spatial analysis revealed significant negative relationships between the distance from the inlet and the two parameters %MeHg and km/kd, (p = 0.020 and p = 0.001, respectively, Figure 2). The latter relationship builds on both the significant positive relationship between distance and kd (p < 0.004) and the nonsignificant relationship between distance and km (p = 0.254). The observed patterns indicate that MeHg demethylation processes become increasingly important with distance from the stream inlet. The km/kd ratio determined during 48 h of laboratory incubations can be considered a short-term proxy for MeHg net production or degradation, whereas %MeHg in soil or sediment is suggested to be a good proxy for net MeHg production during a time span of weeks to months.23,24 However, in environmental settings with substantial input of MeHg, such as in the swamp at Edshult, a decrease in %MeHg along the hydrological pathway strongly suggest a net degradation of MeHg. The observed pattern indicates that MeHg demethylation reactions become successively more important further into the Alnus swamp. Lastly, the significant relationship between %MeHg and km/kd (p = 0.001, Figure 2e) illustrates a linkage between longer and shorter term net MeHg production.

proxies for the HgTOT and MeHg pools per unit soil volume, as justified in the Materials and Methods section. The soil of the Alnus swamp is shallow and underlain by a dense layer of stones and boulders, and the top 10 cm contained the majority of plant fine roots. Given that root exudates is expected to drive the microbial activity responsible for methylation and demethylation processes,21,22 it may be assumed that most of the soil MeHg pool and its dynamics was confined to the top 10 cm of soil. The soil pool of HgTOT (Hg SOC−1) did not follow any clear spatial pattern along the S−N axis from stream inlet to stream outlet in the Alnus swamp (Table S2, SI). In contrast, MeHg SOC−1 in general showed lower concentrations past 250 m from the stream inlet during the three years with most extensive sampling (2007, 2008, and 2010, Table S2, Figure S3, SI). In most years MeHg SOC−1 also decreased from central to peripheral parts along the west to east (W-E) axis of the swamp at 250 m. This decrease follows the fan-like flowpath of water from the center to the outer parts of the swamp. In agreement with these observations there was a significant difference in soil MeHg SOC−1 pools between the upstream sampling points (0−250 m) and the downstream sampling points (270−400 m). The average upstream and downstream values were 1.34 and 0.66 ng MeHg g−1 SOC−1, respectively, (one-way ANOVA, p = 0.00056, Table S4 and S5, SI). In this calculation, MeHg SOC−1 values were normalized to the annual average in order to remove the large variability in absolute numbers among years. The average MeHg concentration in the soil samples taken 2007, 2008, and 2010 was 39 ng MeHg SOC−1. Using the regression equation in Figure S4 (SI), this concentration can be recalculated to a volume-weighted soil pool, which is estimated 13147



Article

wetland and uplands within two years after artificial flooding.3,20 Even if absolute numbers of km and kd cannot be compared, a qualitative comparison of their relative changes over time shows that on occasions with a relatively high kd and low km, as in November 2006, %MeHg was also very low (Figure 3). In 2010 and 2007 isotope tracer additions were too low for determination of kd, but relatively high km and %MeHg values indicate a buildup of MeHg due to high net methylation. The sampling occasions in May 2010, September 2008, and September 2007 were preceded by relatively warm 15-day periods (11 ± 3.1 °C, 11 ± 3.1 °C, and 10 ± 2.1 °C, respectively), as compared to a colder 15 day period preceding the sampling occasion in May 2009 (2.3 ± 0.42 °C). Higher temperatures might explain the markedly higher km in 2010, 2008, and 2007 as compared to 2009. The higher temperatures probably stimulated microbial activity through higher plant root exudation, litter decomposition, and an increased availability of electron donors. The low temperatures in November (2.1 ± 4.1 °C for the preceding 15 day period) may partly explain the relatively higher %MeHg and km in May and September. Additionally, the supply of low molecular mass organic substances is probably higher during the growing season (including May-September), as compared to November. The importance of organic substances as electron donors and thus their importance for MeHg production is well-known.21,26 The results show a significantly smaller soil MeHg pool in the downstream compared to the upstream part of the swamp, a concurrent increase in kd and decrease in %MeHg and km/kd with distance from the stream inlet, as well as fast dynamics in the MeHg soil pool from one year (and season) to another. These results all point in the same direction, that the 30−60% loss of imported MeHg in the stream is due to a net degradation of MeHg along the water flow path in the Alnus swamp Edshult.

It should be noted that it is not possible to quantitatively compare absolute numbers of km and kd, partly because they are measured using different amounts of tracers.25 Figure 3 illustrates temporal variations in soil %MeHg, km, and kd, at Edshult. There was a considerable change in %MeHg

Figure 3. Annual and seasonal variation of km (shaded bar), kd (white bar) and %MeHg in soil (circles with dashed line) at the Alnus swamp Edshult. Error are ± SE for replicate samples (n = 7, 3, 7, 6, 4 for % MeHg, n = 7, 3, 7, 6, 4 for km and n = 0, 2, 14, 0, 4 for kd, for 2010, 2009, 2008, 2007, and 2006, respectively). In 2007 and 2010 kd measurements were unsuccessful. Sample locations for the different years are illustrated in Figure S3 (SI).

from one year to another, for example from September 2008 to May 2010, which was mainly due to changes in MeHg concentrations (as reflected by variations in average MeHg SOC−1 in Table S2, SI). There are few reports on seasonal variations in MeHg concentrations in soils and sediments unaffected by artificial treatments such as flooding, but changes on the order of 100% was observed on the time-scale of months in sediments of a freshwater lake in northern Sweden (Skyllberg et al., in preparation.). Also, concentrations of MeHg was shown to increase by 8 to 37 times in a riverine

Table 1. Snapshot Budget Results for 10 Forest Swamps and Two Peatlands during One Day in May 2009 and/or in May 2010a annual budgetsd MeHg site and year

% net output

Forest Swamps Edshult 2009b −44 ± 6.7 Edshult 2010b −29 ± 3.9 Speltorpet 2009b Trestena 2009b Nybygget 2010b Klasentorp 2010b Steglehylte 2010b Löneberg 2010b Kolsboda 2010b Kvillehult 2010b Åryd 2010c Peatlands Långedalen 2009 198 ± 30 Långedalen 2010 676 ± 25 Ystebo 2010c

mass MeHg in

out

mass DOC in

−1

μg day 247 362 176 95 2623 623 2397 1365 3485 764 1991

151 1074 124 42 766 431 1383 1411 3595 802 2419

15 26 101

47 420 231

mass Hginorg

out g day

10 10 4.1 1.9 60 36 83 88 150 15 76 0.24 0.63 8.7

−1

in mg day

7.1 34 4.3 1.0 64 26 48 104 174 17 98 0.50 6.9 11

water flow rate

snapshot budget results

out

MeHg

−1

MeHg/DOC

MeHg/Hgin

mass-% net output

in

out Ls

−1

2.2 2.8 3.2 0.21 11.8 7.3 17.8 29.0 33.6 4.7 23.8

1.2 5.6 2.6 0.15 12.0 5.2 9.4 34.5 38.1 4.6 31.4

−39 197 −30 −56 −71 −31 −42 3 3 5 21

−14 −13 −33 −16 −73 −4 0 −13 −11 −7 −6

12 48 −13 −38 −71 −3 9 −13 −9 7 −8

1.4 1.3 4.1 0.6 10 11 25 29 38 4.3 24

1.1 3.8 4.1 0.4 11 10 13 34 46 5.1 32

0.10 0.17 1.6

0.16 1.8 1.9

213 1515 129

50 47 80

96 53 93

0.1 0.4 2.3

0.2 1.8 2.8

a

Masses of MeHg, DOC and Hginorg in inlet (in) and outlet (out) streams were used to calculate a mass-% net output of MeHg, MeHg/DOC and MeHg/Hginorg, by the equation: 100 × (∑output − ∑input)/∑input. As a comparison to the snapshot budgets, annual budgets for 2009 and 2010 reported are for masses of MeHg at sites Edshult and Långedalen.7 Negative values indicate that the wetland is a net sink. bTree layer dominated by Alnus glutinosa. cTree layer dominated by Betula pubescens. dAnnual budgets, reported in Tjerngren et al. 2012.7 13148



Article

Are Alnus glutinosa Swamps Generally Sinks for MeHg? In order to test if Alnus swamps generally are sinks for MeHg, snapshot budgets were determined at Edshult and at nine other swamps during one rain event in a period of the year when net methylation is expected to be high.7 In addition, two peatlands (Långedalen and Ystebo), the former shown to be a consistent sink in 2007−2010,7 were sampled as contrasting references to the swamps. In order to evaluate the representativeness of the snapshot budgets, we compared the snapshot budgets for the Alnus swamp Edshult and the peatland Långedalen with their previously reported annual budgets, Table 1. In agreement with the annual budgets, the peatland Långedalen was a source for MeHg according to the snapshot budgets both in May 2009 and 2010 (Table 1). As for Edshult the snapshot budget from May 2009 was in agreement with the annual budget, showing the swamp to be a sink. In May 2010 however, the snapshot budget indicated the Alnus swamp to be a source of MeHg (196% net output), in disagreement with the annual budget. This discrepancy could be explained by a three times higher water flow at the outlet as compared to the inlet on the snapshot budget sampling occasion, the result of a heavy rain event starting four days prior to sampling. At the time of sampling the inlet water flow had decreased while the outlet flow remained high, causing a high MeHg export. To correct for water flow differences between stream inlet and outlet MeHg was normalized to DOC (Table 1). This normalization relies on that changes in the MeHg/DOC mass ratio between inlet and outlet is due to processes in the wetland such as degradation, production, immobilization and mobilization. This is reasonable with the assumptions that (1) MeHg is bound to functional groups in DOC,27 transported as a MeHg−DOC complex in the inlet stream, through the wetland, and in the outlet stream, and (2) the MeHg/DOC ratio in the inlet stream is reasonably constant during the course of the rain event (rising and falling limb of the hydrograph). That is, the MeHg/DOC ratio in the inlet is assumed to be constant during the time period when MeHg in the outlet was generated. Some of the imported DOC may be degraded and new DOC may be released from the wetland, but given the high DOC concentrations during the rain event and its in average high aromatic character, these processes should have a minor impact during the time period giving rise to the snapshot budgets. Support for this is given by similar values of SUVA254 nm and pH in inlet and outlet of the wetlands (Table S5, SI). Based on the DOC normalized data, the Alnus swamp Edshult was a sink for MeHg in both 2009 and 2010, in agreement with annual budgets (Table 1). Based on the MeHg snapshot budgets in Table 1, five out of eight Alnus swamps were indicated to be large sinks for MeHg, with net output ranging between −31 and −71% of input (Speltorpet, Trestena, Nybygget, Klasentorp and Steglehylte). Except for Steglehylte, the flow rates at the inlet and outlet were similar at these five sites. The three remaining Alnus swamps: Lö neberg, Kolsboda and Kvillehult, had MeHg snapshot budgets roughly in balance (3−5%). The Betula swamp Åryd was the only source (21%) of MeHg. It may be noted that these four last mentioned swamps had far lower sulfate concentrations in the outlet steam as compared to the inlet (Table S6, SI) indicating net sulfate reduction. When differences in water inflow and outflow rates were corrected for by normalization to DOC (MeHg/DOC), all swamps were either major sinks or had budgets roughly in balance. Notably,

the Alnus swamps at Löneberg, Kolsboda and Kvillehult, and the Betula swamp Åryd all turned from sources to sinks. Also note that the Alnus swamps Klasentorp and Steglehylte turned from significant sinks to having budgets roughly in balance. In contrast to the swamps, the peatlands Långedalen and Ystebo were substantial sources for both MeHg and MeHg/DOC with a respective net export of 129−1515% and 47−80%. Despite the many limitations of the snapshot budgets, we conclude that there seems to be a difference in net MeHg output between more productive swamps dominated by Alnus glutinosa (C/N-ratios below 20), and more nutrient poor peatlands dominated by Sphagnum spp. (C/N-ratios above 22, Table S6, SI). The difference between the peatland Långedalen and the Alnus swamp Edshult has been reported earlier both in terms of %MeHg, km, and kd in soil, and by annual MeHg mass balance budgets.7,11 In the same study we concluded that methylation rates appear to peak at intermediate nutrient status of wetlands. Furthermore, demethylation seems to play a relatively greater role in the nutrient poor and nutrient rich ends of the wetland spectra, where Alnus swamps are found.7,11 Since bacterial processes are important for degradation of MeHg, factors likely to have an effect are soil productivity, pH, and the site’s position in the landscape. These factors are reflected in the vegetation to some extent, and though this study focuses on Alnus glutinosa our findings may also apply to other types of productive wetlands. Processes in Control of net MeHg Degradation at Edshult. With the present state of knowledge, processes responsible for the MeHg degradation in the Alnus swamp Edshult, both biotic and abiotic, and factors in control of those cannot be elucidated in detail. However, a few important points can be made. The net loss of MeHg in the swamp (0.1−0.3 g/ year) as calculated by the annual input-output budgets,7 is small relative to the significant changes in the MeHg soil pool between seasons and years (on the order of several g MeHg). This suggests that the Alnus swamp behaves similarly to some lakes shown to be sinks for MeHg. Boreal drainage lakes with significant input of MeHg from wetlands have also shown a large internal production and degradation of MeHg.28,29 Of the total import and production the internal production of MeHg was 73−90%. Photolysis is the by far dominant process for MeHg degradation in lakes. This process may be of importance in Alnus swamps because of their standing water, although during the growing season low light penetration through the tree canopy and understory vegetation may limit this. The process may dominate after the trees have shed their leaves, but when DOC concentrations are high photodemethylation is reduced, as for example in Edshult. Note that photon driven processes were excluded in the incubation experiments of this study. It is also tempting to speculate about the possible role of the nitrogen regime since Alnus swamps are N-fixing and nitrate is a potential electron acceptor for bacteria. However, nitrate concentrations were not particularly high at any of our sites (Table S6, SI), and demethylation studies in the Florida Everglades, U.S., have indicated that nitrate reducers do not demethylate MeHg.30 The electron acceptor showing the greatest differences between the swamps and the peatlands was sulfate with the highest concentrations in the swamps (Table S6, SI). This difference is intriguing since molybdate incubation studies indicated SRBs to be responsible for both methylation and demethylation in the Alnus swamp Edshult in May 2009.11 Similarly, SRB have been shown capable of both MeHg 13149



Article

formation and degradation in the Florida Everglades.30 Iron has been shown to decrease the net MeHg production in wetlands. The explanation is a decrease in methylation rates presumably caused by a lowering in concentrations of dissolved, biovailable Hg-sulfides.31 Even though Edshult was the site with the highest concentrations of iron in the outlet stream among the eight boreal wetlands studied by Tjerngren et al.,7 the methylation rates were also among the highest.11 So another explanation could be that iron stimulates demethylation rates rather than inhibits methylation rates. Controlled microbial experiments are in progress to investigate the influence of various electron acceptors and donors on methylation and demethylation rates at Edshult. Implications for Forestry and Landscape Planning. Even if the Alnus swamp Edshult receives a higher input of MeHg than most other wetland sites, the snapshot budgets indicate that the net degradation of MeHg in soils may be a common feature of Alnus glutinosa swamps. This finding can be used in forestry and landscape planning to decrease MeHg loads from, for example, clear-cuts17,32 and peatlands,7,10 both identified as MeHg net exporting sites. There is a policy in Sweden and other countries with the aim to restore large areas of historically drained wetlands,33 which is likely to increase the export of MeHg to downstream located aquatic ecosystems. Thus, maintaining existing and restoring previously drained Alnus glutinosa swamps may be an efficient measure to reduce the negative effect of upstream net MeHg producing sites.

■

the Centre for Environmental Research (CMF, project nr 0822333) and the Kempe foundations.

■

(1) St.Louis, V. L.; Rudd, J. W. M.; Kelly, C. A.; Beaty, K. G.; Bloom, N. S.; Flett, R. J. Importance of wetlands as sources of methyl mercury to boreal forest ecosystems. Can. J. Fish. Aquat. Sci. 1994, 51 (5), 1065−1076. (2) Rudd, J. W. M. Sources of methyl mercury to fresh-water ecosystems - a review. Ecosystems 1995, 80, 697−713. (3) Hall, B.; Louis, V.; Rolfhus, K.; Bodaly, R. Impacts of reservoir creation on the biogeochemical cycling of methyl mercury and total mercury in boreal upland forests. Ecosystems 2005, 8, 248−266. (4) Lindquist, O.; Johansson, K.; Aastrup, M.; Andersson, A.; Bringmark, L.; Hovsenius, G.; Håkanson, L.; Iverfeldt, Å.; Meili, M.; Timm, B. Mercury in the Swedish environment: Recent research on causes, consequences and corrective methods. Water, Air, Soil Pollut. 1991, 55, 221−251. (5) Zedler, J. B.; Kercher, S. Wetland resources: Status, trends, ecosystem services, and restorability. Annu. Rev. Environ. Resour. 2005, 30, 39−74. (6) Tracking the Ramsar convention, About Ramsar. http://www. ramsar.org/cda/en/ramsar-about-about-ramsar/main/ramsar/136%5E7687_4000_0_ (accessed September 16, 2011). (7) Tjerngren, I.; Meili, M.; Björn, E.; Skyllberg, U. Eight boreal wetlands as sources and sinks for methyl mercury in relation to soil acidity, c/n ratio, and small-scale flooding. Environ. Sci. Technol. 2012, 46, 8052−60. (8) Branfireun, B.; Roulet, N. Controls on the fate and transport of methylmercury in a boreal headwater catchment, northwestern Ontario, Canada. Hydrol. Earth Syst. Sci. 2002, 6, 783−794. (9) Mitchell, C. P. J.; Branfireun, B. a; Kolka, R. K. Spatial characteristics of net methylmercury production hot spots in peatlands. Environ. Sci. Technol. 2008, 42, 1010−6. (10) Louis, V.; Rudd, J.; Kelly, C. Production and loss of methylmercury and loss of total mercury from boreal forest catchments containing different types of wetlands. Environ. Sci. Technol. 1996, 30, 2719−2729. (11) Tjerngren, I.; Karlsson, T.; Björn, E.; Skyllberg, U. Potential Hg methylation and MeHg demethylation rates related to the nutrient status of different boreal wetlands. Biogeochemistry 2011, 108, 335− 350. (12) Claessens, H.; Oosterbaan, a.; Savill, P.; Rondeux, J. A review of the characteristics of black alder (Alnus glutinosa (L.) Gaertn.) and their implications for silvicultural practices. Forestry 2010, 83, 163− 175. (13) Axelsson, A.-L.; Ståhl, G.; Söderberg, U.; Peterson, H.; Fridman, J.; Lundström, A. National Forest Inventories reports: Sweden. In National Forest InventoriesPathways for Common Reporting; Tomppo, E.; Gschwantner, T.; Lawrence, M.; McRoberts, R. E., Eds. Springer: 2010; pp 541−553. (14) Hultén, E. Atlas of the Distribution of Vascular Plants in NW. Europé; Generalstabens litografiska anstalts förlag: Stockholm, 1971; p 512. (15) Rudqvist, L. Sveriges Sumpskogar - Resultat Av Sumpskogsinventeringen 1990−1998; Skogsstyrelsen: Jönköping, Sweden, 1999. (16) Skyllberg, U.; Westin, M. B.; Meili, M.; Björn, E. Elevated concentrations of methyl mercury in streams after forest clear-cut: A consequence of mobilization from soil or new methylation? Environ. Sci. Technol. 2009, 43, 8535−41. (17) Evans, J. Silviculture of Broadleaved Woodland; Her Majesty’s Stationery Office: London, 1984; Vol. 62, p 232. (18) Moore, R. D. D. Introduction to Salt Dilution Gauging for Streamflow Measurement Part III: Slug injection using salt in solution. Streamline Watershed Manage. Bull. 2005, 8, 1−6. (19) Zar, J. H. Biostatistical Analysis; Prentice-Hall International (UK) Limited: London, 1996. (20) St Louis, V. L.; Rudd, J. W. M.; Kelly, C. a; Bodaly, R. a D.; Paterson, M. J.; Beaty, K. G.; Hesslein, R. H.; Heyes, A.; Majewski, A.

ASSOCIATED CONTENT

S Supporting Information *

Description of methods for chemical analyses and the determination of potential methylation and demethylation rates, site map, tables on soil data 2007−2010, summary of input-output results, ANOVA calculations for differences in MeHg soil pools, figures on variations in Hginorg and MeHg concentrations in the in- and outlets, maps with soil MeHg concentrations normalized to soil organic carbon, km and kd for 2007, 2008 and 2010, a figure justifying the use of MeHg SOC−1 and Hg SOC−1 as proxies for soil pools at Edshult, and streamwater data for sites used in the snapshot budgets. This material is available free of charge via the Internet at http:// pubs.acs.org.

■

REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Phone: +46 (0)90-786 84 60; e-mail: [email protected]. Author Contributions §

These authors contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We are grateful for help with site information provided by Johan Nitare, the Swedish Forestry Board, and Per Peterson, Sveaskog AB. For assistance in the laboratory, we thank Bengt Andersson and Helen Genberg, SLU. We thank David Krabbenhoft, and four anonymous reviewers for comments on the manuscript. The project was supported financially by The Swedish Research Council for Environment, Agricultural Sciences and Spatial Planning (Formas) (project no. 214-20051462 & 229-2009-1207), the Oscar & Lili Lamm foundation, 13150



Article

R. The rise and fall of mercury methylation in an experimental reservoir. Environ. Sci. Technol. 2004, 38, 1348−58. (21) Windham-Myers, L.; Marvin-Dipasquale, M.; Krabbenhoft, D. P.; Agee, J. L.; Cox, M. H.; Heredia-Middleton, P.; Coates, C.; Kakouros, E. Experimental removal of wetland emergent vegetation leads to decreased methylmercury production in surface sediment. J. Geophys. Res. 2009, 114, 1−14. (22) Mitchell, C. P. J.; Gilmour, C. C. Methylmercury production in a Chesapeake Bay salt marsh. J. Geophys. Res. 2008, 113, 1−14. (23) Drott, A.; Lambertsson, L.; Björn, E.; Skyllberg, U. Do potential methylation rates reflect accumulated methyl mercury in contaminated sediments? Environ. Sci. Technol. 2008, 42, 153−8. (24) Fitzgerald, W. F.; Lamborg, C. H.; Hammerschmidt, C. R. Marine biogeochemical cycling of mercury. Chem. Rev. 2007, 107, 641−62. (25) Hintelmann, H. Constants of mercury methylation and demethylation rates in sediments and comparison of tracer and ambient mercury availability. Environ. Toxicol. Chem. 2009, 19, 2204− 2211. (26) Lambertsson, L.; Nilsson, M. Organic material: The primary control on mercury methylation and ambient methyl mercury concentrations in estuarine sediments. Environ. Sci. Technol. 2006, 40, 1822−9. (27) Qian, J.; Skyllberg, U.; Frech, W. Bonding of methyl mercury to reduced sulfur groups in soil and stream organic matter as determined by X-ray absorption spectroscopy and binding affinity studies. Geochim. Cosmochim. Acta 2002, 66, 3873−3885. (28) Sellers, P.; Kelly, C. A.; Rudd, J. W. M. Fluxes of methylmercury to the water column of a drainage lake: The relative of internal sources importance and external. Limnol.. Oceanogr. 2001, 46, 623−631. (29) Watras, C. J.; Morrison, K. a; Kent, A.; Price, N.; Regnell, O.; Eckley, C.; Hintelmann, H.; Hubacher, T. Sources of methylmercury to a wetland-dominated lake in northern Wisconsin. Environ. Sci. Technol. 2005, 39, 4747−58. (30) Marvin-DiPasquale, M. C.; Oremland, R. S. Bacterial methylmercury degradation in Florida Everglades peat sediment. Environ. Sci. Technol. 1998, 32, 2556−2563. (31) Mehrotra, A. S.; Sedlak, D. L. Decrease in net mercury methylation rates following iron amendment to anoxic wetland sediment slurries. Environ. Sci. Technol. 2005, 39, 2564−70. (32) Bishop, K.; Allan, C.; Bringmark, L.; Garcia, E.; Hellsten, S.; Högbom, L.; Johansson, K.; Lomander, A.; Meili, M.; Munthe, J.; Nilsson, M.; Porvari, P.; Skyllberg, U.; Sorensen, R.; Zetterberg, T.; Akerblom, S. The effects of forestry on Hg bioaccumulation in nemoral/boreal waters and recommendations for good silvicultural practice. Ambio 2009, 38, 373−80. (33) National Strategy for Thriving Wetlands; Swedish Environmental Protection Agency: Stockholm, 2005; www.naturvardsverket.se/ Documents/publikationer/620-1254-1.pdf.

13151


Net Degradation of Methyl Mercury in Alder Swamps - ACS Publications

Recommend Documents