Response of mercury in an Adirondack (NY, USA) forest stream to watershed lime application

Geoffrey D. Millard *a, Charles T. Driscoll a, Douglas A. Burns b, Mario R. Montesdeoca a and Karen Riva-Murray b
aDepartment of Civil and Environmental Engineering, Syracuse University, Syracuse, NY 13244, USA. E-mail: gmillard@syr.edu; ctdrisco@syr.edu; mmontesd@syr.edu
bUS Geological Survey, Troy, NY 12180, USA. E-mail: daburns@usgs.gov; krmurray@usgs.gov

Received 30th October 2017 , Accepted 17th January 2018

First published on 24th January 2018


Surface waters in Europe and North America previously impacted by acid deposition are recovering in conjunction with declining precursor emissions since the 1980s. Lime has been applied to some impacted watersheds to accelerate recovery. The response to liming can be considered a proxy for future recovery from acid deposition. Increases in dissolved organic carbon concentrations have been observed in surface waters in response to increased pH associated with recovery from acid deposition. Although not previously described, recovery-related increases in dissolved organic carbon could drive increases in mercury concentrations and loads because of the affinity of mercury for dissolved organic matter. We used a before–after impact-response approach to describe the response of stream mercury cycling to the application of lime to the watershed of a small stream in the Adirondack Mountains of New York, USA. Dissolved organic carbon, total mercury and methylmercury concentrations increased significantly in streamwater within two weeks of treatment, to previously unobserved concentrations. After six months, post-treatment before–after impact-control (BACI) tests indicate that mean dissolved organic carbon concentrations and total mercury to dissolved organic carbon ratios remained significantly higher and limed site fluxes of methylmercury were lower than those at the reference stream. This pattern suggests total mercury is leaching at elevated levels from the limed watershed, but limitations in production and transport to the stream channel likely resulted in increases in methylmercury concentration that were of limited duration.



Environmental significance

Concentrations of mercury, a bioaccumulative toxin, in surface waters of northern Europe and eastern North America increase with concentrations of dissolved organic carbon as aquatic ecosystems recover from acid rain. Lime applications have been used to neutralize acid impacts and accelerate recovery of acidified aquatic ecosystems; however, impacts of liming on mercury cycling have not been described. In this study the response of an Adirondack stream to a watershed scale lime application was examined to inform management decisions and as a proxy for natural recovery. Our findings indicate that for the first water year following lime application to forested soils, streamwater has increased concentrations of dissolved organic carbon and total mercury, while methylmercury may have a more limited response.

Introduction

Acid deposition causes a variety of adverse ecological effects, as described in numerous studies conducted since environmental concerns were first raised in the late 1960s and early 1970s.1,2 Derived primarily from emissions of sulfur and nitrogen compounds (e.g., SO2, NOx), elevated inputs of strong acids have affected the structure and function of forested and aquatic ecosystems in North America, Europe and Asia.3 Leaching by strong acids removes exchangeable base cations and mobilizes dissolved inorganic aluminum from soil to water, and decreases pH and acid neutralizing capacity (ANC) in surface waters, which can result in toxic conditions for terrestrial and aquatic biota.4–6

In 1990, the United States Environmental Protection Agency (USEPA) established the Acid Rain Program (ARP) under the Clean Air Act, requiring major emissions reductions of the primary precursors of acid deposition from electric utilities.7 Through the ARP and subsequent rules, emissions have been significantly reduced, resulting in marked decreases in acidic deposition,3,8–12 and initiating recovery of acid impacted ecosystems across the northeastern United States.13 However, this recovery process has been slow as decades of elevated acidic deposition has removed available calcium (Ca) and other basic cations from soils.14

In addition to acid deposition, atmospheric mercury (Hg) inputs affect remote forested ecosystems. Direct emissions from anthropogenic activities and legacy secondary re-emissions account for the majority of global Hg emissions to the atmosphere at a rate 2–15 times greater than those from geogenic sources.15 In the northeastern United States, 60–80% of Hg deposition can be attributed to regional sources by deposition of gaseous or particulate ionic Hg.16 Mercury is deposited on trees and other plants, and is adsorbed to leaf surfaces and absorbed through stomata. Subsequent litterfall and throughfall result in elevated deposition to soils.17,18 Soil Hg is largely bound to reduced sulfur groups associated with organic matter and retained in the soil pool until the organic matter becomes water-soluble.19 Once in the aquatic environment, Hg is transported largely in association with dissolved or particulate organic matter.20

Mercury is converted to methylmercury (MeHg), a potent neurotoxin, by sulfate-reducing, iron-reducing or methanogenic bacteria and archaea in reducing21,22 and other niche environments.23 Methylmercury is readily taken up at the base of aquatic food webs, and biomagnifies along the food chain. Approximately 95% of MeHg consumed is absorbed through the gastrointestinal tract of animals.24,25 Once in the bloodstream, MeHg can pass to the brain or developing fetus, causing neurological damage in humans, fish-eating birds, and mammals.26–28

Aquatic ecosystems in northern Europe, and eastern North America are showing signs of recovery from acid deposition.3,29 Important synergistic linkages between the effects of acid and Hg deposition in remote acid-impacted watersheds suggest that decreases in acid deposition could have a co-benefit of decreasing fish Hg concentrations.30 For example, bioaccumulation of MeHg increases with decreases in pH.31 Inputs of sulfate (SO42−) from acidic deposition can drive the formation of MeHg mediated by sulfate-reducing bacteria,32,33 while inputs of nitrate (NO3) can limit this process.34 However, researchers have also reported a coincident increase in Hg concentrations in fish with increases in dissolved organic carbon (DOC)35 and ultraviolet absorbance at 254 nm (UV254)36 associated with recovery from acid deposition.37 Despite decreases in U.S. Hg emissions,38,39 Hg concentrations in recovering remote ecosystems remain elevated, an ongoing risk of bioaccumulation of this potent neurotoxin.

Previous studies40,41 have shown that calcium (Ca) additions in the form of lime (CaCO3) and wollastonite (CaSiO3) are viable options for accelerating recovery of watersheds impacted by acid deposition. These studies demonstrated that Ca base additions increase stream pH and ANC, and decrease concentrations of inorganic monomeric aluminum. Base treatment to acidic forest soils could, arguably, be an experimental analog for the ecosystem response to decreases in acid deposition. The response of Hg dynamics in ecosystems recovering from acid deposition or following base treatment is not well understood. In this study, we examined the impact of a watershed lime application on concentrations and loads of THg, MeHg, DOC, and SO42− in a small Adirondack stream.

The focus of this study was two small tributaries to Honnedaga Lake, located in the southwestern Adirondacks. Honnedaga Lake is one of only seven Adirondack lakes supporting heritage Brook trout (Salvelinus fontinalis) populations.42,43 This genetically unique population has undergone a recovery in recent years in association with increasing lake pH and ANC.44 However, this recovery is limited because of ongoing effects of acidification on tributary streams, which are critical spawning and nursery habitats. The application of lime to one of the tributary watersheds was intended to extend suitable Brook trout spawning habitat by improving tributary water quality, particularly by increasing ANC and decreasing inorganic monomeric aluminum concentrations. This lime application provided an opportunity to examine THg and MeHg concentrations in stream water draining limed and reference watersheds, to evaluate changes in Hg dynamics after liming, and to improve understanding of the potential response of Hg dynamics in an acid-impacted forest watershed to ongoing recovery from previously elevated acid deposition. We hypothesized that: (1) THg and MeHg concentrations would increase immediately following lime addition, due to increases in the release of DOC from soil (resulting from increases in soil pH); and (2) concentrations of THg and MeHg would decline within the first water year to pre-treatment/reference levels, as soil stocks of labile DOC become depleted.

Methods

Study site

Honnedaga Lake (3.1 km2 surface area) is located in the southwestern Adirondacks (43°31′06′′N and 74°48′31′′W). The 13.3 km2 watershed is completely forested, and 26 tributaries drain into the lake (Fig. 1). The two tributaries examined in this study (i.e., one limed and one reference) were both chronically acidic prior to treatment, with pH values consistently below 4.8.44 These tributary watersheds are small (limed – 0.30 km2, reference – 0.18 km2), and together, comprise less than 4% of the Honnedaga Lake watershed. Both the limed and reference watersheds are characterized by steep topography (364 m km−1) and little (i.e., limed – 4.3%) or no (i.e., reference) wetland area.
image file: c7em00520b-f1.tif
Fig. 1 Map of Honnedaga Lake and the limed and reference watersheds. Inset map of New York State shows Adirondack Park in green, and the location Honnedaga Lake within the Park. This area of New York State is at a high elevation relative to the surrounding area (667 m).

The limed watershed received 150 metric tons of limestone (CaCO3), distributed in a pelletized form over the 30 hectare watershed. The lime was applied just after leaf fall on October 1, 2013 by helicopter, at a dose of 1.4 Mg of Ca per ha as part of a larger study.45,46 This dosing level was comparable to application rates for other regional watershed liming studies.47,48

Field sampling of streams

USGS water quality samples were collected manually at approximately two week intervals and were supplemented by the use of automated water samplers beginning in January 2012. These samplers were programmed to collect samples that represented the rising, peak, and recession hydrograph during storm events and snowmelt. Field blank and triplicate samples were collected for additional quality control for each analyte. Water quality samples were then transported to USGS laboratories for processing. Over the pre-treatment period, water samples for Hg analysis were collected seasonally, when sample locations were physically accessible. Following lime addition, Hg samples were collected weekly during the first four weeks, and approximately monthly thereafter through September 2014. Samples were collected using a trace-metal clean technique (US EPA method 1669, U.S. EPA 1995)49 and were analyzed at the USGS Wisconsin Mercury Research Laboratory (2012 samples) or Syracuse University's Water Quality Laboratory (all subsequent samples). Simultaneously with each Hg sample, a separate water sample was also collected for analysis of DOC, SO42−, NO3, iron (Fe) and pH; these samples were analyzed by the USGS New York Water Science Center Laboratory in Troy, NY. Water samples were transported on wet ice to a field laboratory, where the Hg samples were passed through 0.4 μm polycarbonate filters, and acidified to 1% with concentrated hydrochloric acid prior to shipping to the Syracuse University and USGS laboratories. Samples for analysis of DOC, SO42−, NO3, Fe and pH at the USGS were shipped on ice to Troy, NY and processed for analysis according to procedures described at: https://www.sciencebase.gov/catalog/item/55ca2fd6e4b08400b1fdb88f.

Laboratory processing and chemical analyses

Samples sent to the USGS Wisconsin Mercury Research Laboratory were analyzed for filtered THg and MeHg. Filtered THg was determined using cold vapor atomic fluorescence spectrometry (CVAFS) according to the method described by Olson and DeWild (1999),50 a slight modification to EPA method 1631.51 Filtered MeHg was determined after distillation by ethylation, gas chromatographic separation, pyrolysis and CVAFS as described in DeWild et al. (2002),52 a modification of EPA method 1630.53

Samples sent to Syracuse University were also analyzed for filtered THg and MeHg. Total Hg was analyzed using an oxidation, purge and trap, desorption and CVAFS following U.S. EPA method 1631, revision E;51 the method detection limit is 0.2 ng L−1. Methylmercury was analyzed by distillation, ethylation, purge and trap, desorption and CVFAS according to U.S. EPA Method 1630;53 the method detection limit is 0.02 ng L−1.

Samples sent to the USGS in Troy, NY were analyzed for pH, DOC, UV254, SO42−, NO3, and Fe. Dissolved SO42− and NO3 were analyzed by ion chromatography after first passing the subsample through a 0.4 μm polycarbonate filter. The subsample for dissolved organic matter analysis was first passed through a 0.7 μm glass fiber filter (GFF) prior to analysis of DOC by UV/persulfate oxidation and infrared detection, and UV254 by measuring absorbance at 254 nm in a standard spectrophotometer54 and correcting for Fe interference, as described previously.55 Specific ultraviolet absorbance (SUVA), a broad measure of DOC quality as reflected by its relative aromatic vs. aliphatic character was calculated as (UV254 × 100)/DOC in mg L−1, and is reported in units of L per mg C per m. Dissolved Fe was analyzed by Inductively Coupled Plasma (ICP) spectrophotometry, after first passing the subsample through a 0.4 μm polycarbonate filter and then acidifying with ultrapure nitric acid. An unfiltered subsample was also analyzed for pH in the laboratory. Further USGS method details can be found at https://www.sciencebase.gov/catalog/item/55ca2fd6e4b08400b1fdb88f (accessed 01/05/2018). Laboratory QA/QC procedures are described in Lincoln et al. (2009).56 All water chemistry data are available at http://https://www.waterqualitydata.us/portal/ (site IDs limed (T16) = 0134277114, reference (T24) = 0134277112).

Statistical analysis

Three time intervals were defined for the purpose of data analysis; pre-treatment (Sept 2011–Sept 2013), transitional post-treatment (TPT, Oct 2013–Feb 2014), and longer term post-treatment (LPT, March 2014–Sept 2014). Paired Hg samples (from the limed and reference sites) were collected during each period four, six, and six times respectively. Some of these samples were below detection limits when analyzed. These samples were excluded from statistical tests, resulting in lower n values for pre-treatment MeHg, THg (n = 3), and % MeHg (MeHg × 100/THg, n = 2), as well as LPT MeHg and % MeHg (n = 5). Only one of the MeHg samples collected from the reference site during the pre-treatment period was above the detection limit. Paired samples of DOC, and SO42− were collected during each period 40, 16 and 16 times respectively. UV254 was collected 28, 16 and 15 times during each period respectively.

Statistical analysis was performed with the software package SASv9.4. Regression analysis was performed on concentration and log-transformed concentration data from each stream to describe correlations of THg and MeHg with DOC, SO42−, NO3, UV254 and SUVA. Regressions between reference and limed streams were tested for significant differences in slope using ANOVA.

Time series of THg, MeHg and DOC concentrations were used to identify changes after lime addition. A standard ANOVA was applied to all pre-treatment observations, including unpaired values, to examine statistical differences in the mean concentrations of SO42−, NO3, DOC, UV254, SUVA, THg, MeHg and % MeHg between the limed and reference streams. This was done in order to compare MeHg pre-treatment values because of the lack of paired data for this analyte. A regular ANOVA was used to test pre-treatment DOC and SO42− fluxes between sites. Using PROC GLM in SAS, a least square means (LSM) fixed model

Y = site + period + site × period
where Y is a response metric (THg, MeHg, DOC, etc.), site is sample location (limed, reference) and period is the time interval (pre-treatment, TPT, LPT), was used to test the main effect of site and period and to look for interaction. This model was applied to natural logarithm transformed, and untransformed data. For simplicity we have only reported the untransformed data as there was no effective difference in test results. Where there was probable interaction between site and period (α = 0.2, marginal α = 0.3) a before–after control-impact (BACI) paired design was used with an ANOVA test to assess changes in the difference between sites over the three periods: pre-treatment, TPT, and LPT. A p-value of less than or equal to 0.05 was considered significant and less than 0.1 marginally significant.

Determination of fluxes

Stage was measured at each site every 15 minutes with a submersible pressure transducer. Stage was converted to discharge through the use of a stage-discharge rating curve developed by making periodic discharge measurements with a current meter under a range of flow conditions. Methods for measuring stream stage and discharge are described in Sauer and Turnipseed (2010),57 and Turnipseed and Sauer (2010).58 Discharge data are available at http://https://waterdata.usgs.gov/nwis/.

Annual fluxes of SO42−, DOC, THg and MeHg were determined using the software package Flux32 v 3.03 (https://www.pca.state.mn.us/wplmn/flux32). Flux32 generated a discharge/concentration (Q/C) linear regression for each analyte and used this regression to estimate concentrations between sample points. The data were stratified based on growing and non-growing seasons to account for seasonal variation, an approach that resulted in lower variance across all sites and analytes. The concentration–discharge regression was applied to calculate an estimated mean concentration for each day, which when multiplied by mean daily discharge yielded a daily flux value. These daily fluxes were then used to estimate an annual flux for each tributary. Discharge from the tributaries is shown in Fig. S1. Insufficient Hg samples were collected to confidently stratify the Hg data into four seasons. Similarly, fluxes could not be confidently calculated for the TPT and LPT periods individually, therefore we report fluxes over the entire post-application period.

Results

Correlations

DOC and THg concentrations exhibited a strong positive correlation for the period of study at both reference (slope = 0.514 ng-Hg per mg-C; adj-R2 = 0.627) and limed streams (slope = 0.517 ng-Hg per mg-C; adj-R2 = 0.900). The slopes of these regressions were not statistically different using all available data (Fig. 2A, p > 0.9). There was a similar positive correlation between UV254 and THg concentrations over the period of study (slope = 11.1 ng-Hg per m and 11.5 ng-Hg per m, adj-R2 = 0.528 and 0.900 respectively). The slopes of the THg-UV254 regressions were not statistically different using all available data (Fig. 2C, p > 0.9). The only other significant correlations were between SUVA and THg in the limed stream (slope = 4.29 ng-Hg per L L per mg-C per m, adj-R2 = 0.412, Fig. 2D) and nitrate and THg in the reference stream (slope = −0.27 ng-Hg per L mg-NO3 per L, adj-R2 = 0.237, Table S1). In contrast, MeHg was not significantly correlated with DOC, SO42−, NO3, UV254 or SUVA (Table S1), and log transformation of these data did not result in significant relationships.
image file: c7em00520b-f2.tif
Fig. 2 Relationship between concentrations of total mercury (THg) and dissolved organic carbon (DOC) (A), sulfate (SO42−) (B), ultraviolet absorbance at 254 nm (UV254) (C) and specific ultraviolet absorbance (SUVA) (D) in limed and reference streams prior to and following lime treatment. THg and DOC are more strongly related in the limed stream (adj-R2 = 0.900, p < 0.001) than in the reference stream (adj-R2 = 0.627, p ≤ 0.001) as is the relationship between THg and UV254 (limed: adj-R2 = 0.900, p < 0.001; reference: adj-R2 = 0.528, p = 0.001). There is no difference in slope observed at these sites for DOC (limed: m = 0.514 ng-Hg per mg-C; reference: 0.517 ng-Hg per mg-C, p > 0.9) and UV254 (limed: m = 11.12 ng-Hg per m; reference: m = 11.31 ng-Hg per m, p > 0.9). The relationship between THg and SUVA in the limed stream is the only other significant correlation (p = 0.002, adj-R2 = 0.412).

Pre-treatment period

Before lime addition, mean values of DOC, SO42−, UV254 and SUVA were significantly higher in the limed tributary than in the reference tributary (p < 0.05). However, pH was not significantly different (p = 0.111), nor were NO3, THg, MeHg, % MeHg and THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC (p > 0.38). Summary statistics for all analytes are described in Table 1 and results from the pre-treatment ANOVA test are described in Table S2.
Table 1 Summary statistics for all analytes in both study tributaries. Mean, minimum detected and maximum values during pre-treatment, transitional post-treatment (TPT) and longer-term post-treatment (LPT) periods are shown
Pre TPT LPT
Mean Min Max Mean Min Max Mean Min Max
Limed pH 4.54 4.41 4.92 6.6 5.51 7.46 5.35 5.13 5.68
DOC (mg L−1) 5.5 2.9 10.5 10.6 3.9 18.4 8.8 3.3 15.0
UV254 (cm−1) 0.227 0.113 0.401 0.442 0.147 0.781 0.375 0.130 0.647
SUVA (L per mg-C per m) 3.8 3.3 4.4 4.2 3.7 4.5 4.2 3.9 4.6
Sulfate (mg-SO42− per L) 3.1 2.0 4.1 3.6 3.0 6.4 3.1 2.4 4.0
Nitrate (mg-NO3 per L) 1.45 0.20 5.30 2.47 0.07 4.68 2.2 0.68 4.00
THg (ng L−1) 2.06 1.29 3.33 3.27 1.07 5.50 2.26 0.69 3.40
MeHg (ng L−1) 0.026 0.008 0.044 0.074 0.004 0.194 0.027 0.002 0.056
MeHg/THg (%) 1.48 0.24 2.68 3.92 0.08 18.15 2.15 0.09 8.17
THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC (μg g−1) 0.34 0.25 0.42 0.37 0.24 0.48 0.34 0.21 0.43
Reference pH 4.60 4.41 4.91 4.58 4.50 4.66 4.57 4.50 4.63
DOC (mg L−1) 4.5 2.8 7.7 5.4 3.2 8.1 4.7 3.1 7.8
UV254 (cm−1) 0.156 0.073 0.267 0.186 0.109 0.292 0.154 0.089 0.245
SUVA (L per mg-C per m) 3.3 2.4 3.9 3.4 3.2 3.7 3.3 2.9 3.6
Sulfate (mg-SO42− per L) 2.9 1.9 3.8 2.5 2.1 3.0 2.5 2.0 3.1
Nitrate (mg-NO3 per L) 1.33 0.17 4.75 1.55 0.02 3.5 1.51 0.27 3.31
THg (ng L−1) 1.54 1.15 2.15 1.37 0.53 2.13 0.73 0.33 1.17
MeHg (ng L−1) 0.007 0.006 0.008 0.019 0.008 0.046 0.022 0.002 0.040
MeHg/THg (%) 0.42 0.38 0.45 2.27 0.38 8.64 2.36 0.67 4.23
THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC (μg g−1) 0.35 0.30 0.38 0.28 0.14 0.39 0.17 0.11 0.28


The pH of pre-treatment samples collected from both the limed and reference tributaries was consistently <5.0 (Fig. 3A). Mean DOC concentration was 4.5 mg C per L (2.8–7.7 mg C per L) at the reference site, and was significantly higher 5.5 mg C per L (2.9–10.5 mg C per L, p = 0.013, Fig. 3B) at the limed site. Observations of THg revealed a similar pattern to DOC where the mean concentration was 1.54 ng L−1 (1.15–2.15 ng L−1) at the reference site, and 2.06 ng L−1 (1.29–3.33 ng L−1, Fig. 3C), at the limed site, however a statistical difference was not detected. The ratio THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC was not significantly different between the sites over this period (p = 0.84). Fluxes of THg and MeHg were not calculated during the pre-treatment period due to insufficient sampling frequency. Sulfate and DOC fluxes at the limed tributary were 4010 ± 30.9 kg-SO42− per km2 per year and 7750 ± 42.9 kg-C per km2 per year, respectively. These values were significantly (p < 0.001) higher than fluxes from the reference site which were 2830 ± 15.6 kg-SO42− per km2 per year and 5460 ± 62.1 kg-C per km2 per year (Table 2).


image file: c7em00520b-f3.tif
Fig. 3 Time-series of pH (A), dissolved organic carbon (DOC; B), total mercury (THg; C) and methylmercury (MeHg; D) of limed and reference streams before and following lime application (1 October 2013 indicated by vertical line). The blue line in panel B represents a DOC concentration of 8.5 mg L−1.
Table 2 Calculated fluxes from the reference and limed tributaries. Post-application values were calculated using data from the transitional post-treatment (TPT) and longer-term post-treatment (LPT) time periods. Insufficient Hg data were collected prior to lime addition to calculate pre-treatment MeHg and THg fluxes
Tributary Period MeHg (g-Hg per km2 per year) THg (g-Hg per km2 per year) DOC (kg-C per km2 per year) SO4 (kg-SO4 per km2 per year)
Flux Std error Flux Std error Flux Std error Flux Std error
Limed Pre-treatment NA NA 7750 42.9 4010 30.9
Post-application 0.034 0.215 5.17 1.80 12[thin space (1/6-em)]100 61.7 4380 37.1
Reference Pre-treatment NA NA 5460 62.1 2830 15.6
Post-application 0.050 0.341 1.75 1.90 5220 52.4 2850 10.9


Transitional post-treatment (TPT) period

In the first six months following lime addition (TPT period), large differences were apparent between the limed and reference tributaries. Maximum concentrations for all constituents except SUVA were observed at the limed site during this period. Before–after control-impact (BACI) tests revealed significantly higher difference in pH (p < 0.001), DOC (p < 0.001; Fig. 4A), UV254 (p < 0.001; Fig. 4D), SO42− (p < 0.001; Fig. 4B), NO3 (p = 0.001; Fig. S2) and SUVA (p < 0.0001; Fig. 4C) during the TPT period relative to the pre-treatment period. During this period, the mean pH was 6.6 (5.51–7.46), DOC was 10.6 mg L−1 (3.9–18.4 mg L−1), UV254 was 0.442 cm−1 (0.147–0.781 cm−1), SUVA as 4.2 L per mg-C per m (3.7–4.5 L per mg-C per m), SO42− was 3.6 mg-SO42− per L (3.0–6.4 mg-SO42− per L) and NO3 was 2.47 mg-NO3 per L (0.07–4.68 mg-NO3 per L; Table 1) for the limed site. The difference in mean THg concentration was higher during the TPT period relative to the pre-treatment period (p = 0.043, Fig. 5B).
image file: c7em00520b-f4.tif
Fig. 4 The difference between paired observations (limed–reference) of dissolved organic carbon (DOC; A), sulfate (SO42−; B), specific ultraviolet absorbance (SUVA; C) and ultraviolet absorbance at 254 nm (UV254; D) during the pre-treatment (pre), transitional post-treatment (TPT) and longer-term post treatment (LPT) periods. The data are labelled for significant differences using BACI tests. Each box indicates the interquartile data range, the whiskers indicate the 95% central data range, and outliers are indicated by unfilled circle symbols.

image file: c7em00520b-f5.tif
Fig. 5 Differences between paired observations (limed–reference) for methylmercury (MeHg; A), total mercury (THg; B), total mercury to dissolved organic carbon ratio (THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC; C) and percent methylmercury (% MeHg; D). Only the THg during the transitional post-treatment (TPT) period and THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC during the longer-term post-treatment (LPT) period were found to be have significant differences using BACI testing. Individual datapoints are shown in red. Each box indicates the interquartile data range, the whiskers indicate the 95% central data range.

Within the first two weeks of the TPT period for the limed site, the highest observations of pH (7.46), SO42− (6.4 mg L−1), and THg (5.5 ng L−1) were observed. Maximum values of UV254 (0.781 cm−1), MeHg (0.194 ng L−1), and % MeHg (18.15%) were observed later during the TPT period. Despite these elevated values, LMS fixed model ANOVA tests did not indicate significant interaction for MeHg (Fig. 5A, Table S3), % MeHg (Fig. 5D, Table S3), MeHg[thin space (1/6-em)]:[thin space (1/6-em)]DOC (Table S3) and BACI tests indicate no significant change in THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC (Fig. 5C and Table S4), due to the sparse amount of pre-treatment data (low statistical power).

Longer-term post-treatment (LPT) period

During the LPT period, the limed–reference differences in pH, DOC, UV254, SUVA, SO42− and NO3 (BACI, p ≤ 0.001) remained significantly greater, in relation to the pre-treatment period (Fig. 4, S2 and Table S4). The limed tributary mean pH was 5.35 (5.13–5.68), DOC was 8.8 mg L−1 (3.3–15.0 mg L−1), UV254 was 0.375 cm−1 (0.130–0.647 cm−1), SUVA was 4.2 L per mg-C per m (3.9–4.6 L per mg-C per m), SO42− was 3.1 mg-SO42− per L (2.4–4.0 mg-SO42− per L), and NO3 was 2.2 mg-NO3 per L (0.7–4.0 mg-NO3 per L; Table 1). Additionally, BACI tests revealed differences in pH, DOC, UV254 and SO42− between tributaries were all significantly lower than during the TPT period (p < 0.001, p < 0.001, p = 0.003, 0.030, respectively; Table S4).

Before–after control-impact (BACI) analyses failed to detect a significant difference between the TPT and LPT periods for SUVA, and NO3, THg, and THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC ratio (Fig. 4, 5, S2 and Table S4). During the LPT period, mean THg for the limed site was 2.26 ng L−1 (0.69–3.40 ng L−1, Table 1), and THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC was 0.34 μg g−1 (0.21–0.43 μg g−1, Table 1). Similarly, BACI tests showed no significant change for THg (p = 0.363, Fig. 5B and Table S4) between pre-treatment and LPT periods, but a greater difference was detected for the THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC ratio (p = 0.015, Fig. 5C and Table S4). Although this ratio does not appear to change through time in the limed tributary, the LPT values from the reference tributary were considerably lower than during the pre-treatment period. LSM fixed model ANOVA tests indicated there was no significant interaction for MeHg, % MeHg, or MeHg[thin space (1/6-em)]:[thin space (1/6-em)]DOC ratio. These tests do indicate that MeHg between the tributaries was significantly different regardless of time period (p = 0.009, Table S3) and that % MeHg was marginally different through time (p = 0.063, Table S3).

Fluxes were examined by pooling both the TPT and LPT periods as a post-application flux. Relative to the pre-treatment period, the post-application DOC flux dramatically increased in the limed tributary to 12[thin space (1/6-em)]100 ± 61.7 kg-C per km2 per year while the reference tributary slightly decreased to 5220 ± 52.4 kg-C per km2 per year (Table 2). This dramatic increase in DOC is consistent with the changes detected using BACI. Sulfate fluxes at the limed site increased slightly to 4380 ± 37.1 kg-SO4 per km2 per year and were essentially unchanged at the reference tributary at 2850 ± 10.9 kg-SO4 per km2 per year. The post-application flux for THg had a similar pattern as the DOC flux, with the limed THg flux (5.17 ± 1.80 g-Hg per km2 per year) greater than the reference value (1.75 ± 1.90 g-Hg per km2 per year, Table 2). The post-application flux of MeHg was inconsistent with the DOC and THg fluxes, having an approximately equivalent or lower value at the limed site (0.034 ± 0.215 g-Hg per km2 per year) than that of the reference site (0.050 ± 0.341 g-Hg per km2 per year, Table 2).

Discussion

Impact of watershed liming on Hg cycling

Lime addition markedly increased the pH of streamwater, as noted in other liming studies in the Adirondacks40 and elsewhere.59–61 During the first water year after the application, pH remained above 5.0, which is protective of Brook trout health.62 Therefore the liming application accomplished the primary objective of improving Brook trout habitat. We anticipated that increases in pH associated with either decreases in acid deposition or base treatment would not only increase the mobilization of DOC but also that of THg, because of the strong correlation between these two analytes. While the drivers and mechanisms for recent increases in surface water DOC are complex,63 the mechanism in the limed watershed is likely driven by a decrease in the partitioning of DOC to soil driven by increases in soil pH.64 It remains to be observed if the liming will also affect the Hg concentrations in stream biota.

The shift towards higher SUVA values is consistent with other research65 suggesting that dissolved organic matter enriched in aromatic compounds was released from the limed watershed following application coinciding with a decrease in DOC partitioning.20 Recent studies66–70 have suggested that larger, more humic-like DOC (larger molecular mass, and more aromatic) is a stronger, less labile ligand for Hg. Our findings potentially support this hypothesis because the THg concentration (Fig. 5B) and THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC ratio (Fig. 5C) are significantly higher during the TPT and LPT periods respectively, than during the pre-treatment period with no significant change detected for MeHg with respect to time (Fig. 5A).

Observed changes in THg and the THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC ratio also suggest that the timing of Hg transport derived from watershed soils to stream water is influenced by the distance from the stream. The Hg mobilized to stream water during the TPT period, was likely derived from the near-stream environment. We suspect, due to the importance of riparian71 and hyporheic19 zones as drivers of DOC, SO42−, and MeHg, a limited near-stream MeHg pool was mobilized immediately following lime application that was largely missed by the Hg sampling frequency except for a few elevated observations of MeHg during the TPT period immediately following lime addition and during spring snowmelt. In contrast, the sustained supply of THg over the LPT period may have been derived from more upland portions of the watershed that was potentially mobilized through the riparian and hyporheic zones before methylation could occur.

Sulfate plays an integral role in MeHg formation.32,33,72,73 The limed stream had higher SO42− concentrations before and after lime application than did the reference stream. This, along with no detectable difference in % MeHg between tributaries, suggests that methylation in these watersheds is not primarily limited by SO42−, which is consistent with little anaerobic habitat in this steep upland catchment. After application, SO42− concentrations in the limed tributary did not appear to change from TPT to LPT, while SO42− concentrations in the reference tributary decreased over this period (Fig. S3). This pattern suggests a disruption in the seasonal consumption of SO42− by sulfate-reducing bacteria, which could result in a decline in the production of MeHg within the limed watershed. A possible explanation for this pattern could be the increase in NO3 following lime application (Fig. S2). As NO3 is a stronger oxidant than SO42−, the enhanced supply of NO3 could limit SO42− reduction and as a result the production of MeHg.34 Note Matthews et al. (2013)74 demonstrated addition of NO3 to a Hg-contaminated lake to limit production of MeHg from profundal sediments. A disruption in SO42− reduction by NO3, which may have limited Hg methylation, could help explain why THg and DOC showed elevated concentrations and fluxes during the post-application period, whereas MeHg fluxes were lower than those of the reference site.

Other studies have provided evidence suggesting that Hg methylation can be limited by high DOC concentrations. Using a series of Arctic lakes with a wide DOC gradient, French et al. (2014)67 showed that MeHg concentrations are diminished in biota at DOC concentrations greater than approximately 8.5 mg L−1. They suggest this effect is due to increased binding of THg with larger molecular weight humic acids, which are not easily degraded, limiting the release of ionic Hg and the formation of MeHg. While a potentially similar pattern of reduced MeHg concentrations in biota with elevated DOC has been reported for Adirondack lakes,75 studies of Adirondack streams have not observed a similar decrease in Hg bioaccumulation with elevated DOC.76,77

Before lime addition, DOC concentrations in both the limed and reference tributaries were generally below the 8.5 mg L−1 threshold value reported by French et al. (2014).67 Base treatment increased pH and consequently increased mobilization of DOC to streamwater. This elevated DOC would likely include more humic-based material because the solubility of humic acids increases with increased pH,64 as evidenced by the increase in SUVA following lime treatment. This pattern suggests a mobilization of humic organic matter when lime is applied to forest soils. A higher humic acid DOC fraction could limit bioavailability of Hg in downstream lakes.

Liming as a management strategy

Potential effects of liming on Hg cycling and bioaccumulation complicates an assessment of liming as a management strategy.78 Lime can be applied in forested areas to protect against or reverse the impacts of acidic deposition and preserve or restore sensitive or impacted ecosystems. At Honnedaga Lake, CaCO3 was applied to tributary watersheds to rapidly improve spawning habitat for heritage Brook trout by reducing streamwater acidity and decreasing concentrations of inorganic monomeric aluminum. Our results suggest that the increase in pH associated with watershed lime treatment can result in increases in stream water concentrations of DOC and THg. Higher concentrations and fluxes of THg in aquatic ecosystems may, thus, be unintended consequences of watershed liming. The effects of this management strategy on Hg bioaccumulation in aquatic ecosystems requires further investigation given the variety of interacting factors that can affect the Hg cycle.79,80

Increased mobilization of THg is a particular problem in forested areas because Hg accumulates in forest soils31 and much of the northeastern United States is forested.81 Applying lime to acid sensitive watersheds in the Adirondack region and elsewhere could drive improvements in pH and ANC, but simultaneously mobilize THg. Elevated transport of THg from watersheds to downstream lakes would likely lead to deposition of THg to anaerobic lake sediments, which might later promote the formation of MeHg.72,73 Given that the limed watershed is only 2.7% of the drainage area of Honnedaga Lake, any change in flux would be expected to have only a minor impact on Hg fluxes to the lake. However, if a complete liming of the Honnedaga Lake watershed were to be conducted a 50% increase in load could be expected based on the response of the limed tributary. This observed increase in DOC and Hg[thin space (1/6-em)]:[thin space (1/6-em)]DOC ratio may potentially persist beyond the first year after treatment with Hg dynamics different from those associated with the ongoing recovery from acidic deposition.

The natural recovery of Honnedaga Lake from ongoing decreases in acid deposition further complicates assessment of this management strategy. Increases in DOC of approximately 1% per year have been reported in the northeastern United States following decreases in SO42− deposition.82 The effect of observed DOC increases and other changes in lake chemistry related to decreases in acid deposition at Honnedaga Lake,44 on Hg concentrations of lake-resident Brook trout, and the effect of liming on Hg bioaccumulation in tributary-resident macroinvertebrates and young-of-the-year Brook trout are subjects of ongoing research.

Implications for acid rain recovery

The results of this experiment have important implications for understanding the natural recovery of lake watersheds from acid deposition. This multi-decadal process of recovery83 may lead to continued increases in DOC and THg concentrations in recovering ecosystems. The increasing fish-Hg reported in aquatic ecosystems following decreases in SO42− deposition may be indicative of this DOC-driven change.37,44 Some studies have suggested that changes in the quality of DOC may be coincident with increases in concentrations of DOC,84–87 but it is unclear whether or not changes in DOC quality would impact formation and subsequent transport of MeHg from soils to aquatic ecosystems.

In the first water year after liming, THg and DOC concentrations increased significantly in the limed tributary relative to those observed in the reference tributary. The strength of the THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC correlation, the increase in DOC flux post-application, and the elevated THg flux from the limed watershed, indicate increased mobilization of Hg following CaCO3 application. This pattern suggests the potential for future release of Hg associated with soil organic matter88,89 within the reference watershed as part of ongoing pH recovery from decreases in acid deposition.

Conclusions

Stream Hg concentrations and fluxes at a calcium carbonate treated and reference watersheds at Honnedaga Lake were examined for the first water year after lime addition. The response of Hg and DOC to base treatment can help inform our understanding of the natural recovery of watersheds impacted by acid deposition. These results also have implications for the downstream lake ecosystem where THg could be deposited to sediments and eventually converted to MeHg.

The lime application resulted in a large increase in pH of the limed tributary. This increase in pH mobilized organic carbon in treated soils, causing DOC and THg concentrations and fluxes to be elevated over reference and pre-treatment observations. In contrast, elevated MeHg concentrations did not persist in the limed stream and the calculated annual flux was slightly lower than in the reference tributary during the post-application period. We speculated that MeHg was mobilized from near-stream pools, while an increase in the THg[thin space (1/6-em)]:[thin space (1/6-em)]DOC ratio and SO42− flux suggests transport from upland sources, but did not result in enhanced MeHg production. The presumably small pools in the near stream region resulted in the limited MeHg treatment response. Further monitoring of these tributaries is required to evaluate the longer term impacts of using lime to accelerate recovery from acidification on Hg dynamics and Hg concentration in biota.

Author contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript.

Conflicts of interest

There are no conflicts to declare.

Abbreviations

HgMercury
THgTotal mercury
MeHgMethylmercury
DOCDissolved organic carbon
UV254Ultraviolet absorbance at 254 nm
SUVASpecific ultraviolet absorbance of dissolved organic carbon at 254 nm
ANCAcid neutralizing capacity
CaCO3Lime or calcium carbonate
CaSiO3Wollastonite
% MeHgPercent of methylmercury from total mercury
USEPAUnited States Environmental Protection Agency
ARPAcid rain program
TPTTransitional post-treatment
LPTLong-term post-treatment

Acknowledgements

This project was supported by funds from the New York State Energy Research and Development Authority (NYSERDA). Additional support was provided by NSF through the Syracuse University EMPOWER NRT program. Special thanks to the Adirondack League Club for access to Honnedaga Lake, the surrounding tributaries and the field lab. The authors would also like to thank Greg Lawrence, Jason Siemion, Tia-Marie Scott, Dan Josephson, Selam Lemma, Mariah Taylor and John Byrnes for their contributions in the field and the laboratory. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government.

References

  1. S. Oden, The acidification of air precipitation and its consequences in the natural environment, in Ecological Research Communications Bulletin of NFR Arlington (VA), 1968, p. Translation Consultants Parker Search PubMed.
  2. G. E. Likens, F. H. Bormann and N. M. Johnson, Acid Rain. Environ. Sci. Policy, 1972, 14(2), 33–40,  DOI:10.1080/00139157.1972.9933001.
  3. C. T. Driscoll, G. B. Lawrence, A. J. Bulger, T. J. Butler, C. S. Cronan and C. Eagar, et al. Acidic deposition in the northeastern United States: sources and inputs, ecosystem effects, and management strategies, Bioscience, 2001, 51(3), 180 CrossRef.
  4. B. P. Baldigo, G. B. Lawrence, R. W. Bode, H. A. Simonin, K. M. Roy and A. J. Smith, Impacts of acidification on macroinvertebrate communities in streams of the western Adirondack Mountains, New York, USA, Ecol. Indic., 2009, 9(2), 226–239,  DOI:10.1016/j.ecolind.2008.04.004.
  5. B. P. Baldigo, G. Lawrence and H. Simonin, Persistent mortality of Brook trout in episodically acidified streams of the southwestern Adirondack mountains, New York, Trans. Am. Fish. Soc., 2007, 136(1), 121–134,  DOI:10.1577/T06-043.1.
  6. G. B. Lawrence, K. M. Roy, B. P. Baldigo, H. A. Simonin, S. B. Capone and J. W. Sutherland, et al. Chronic and episodic acidification of Adirondack streams from acid rain in 2003–2005, J. Environ. Qual., 2008, 37(6), 2264–2274,  DOI:10.2134/jeq2008.0061.
  7. U.S. EPA, Acid Rain and Related Programs: 2009 Highlights [Internet], EPA reports. 2009 [cited 2016 May 3], available from: http://nepis.epa.gov/Exe/ZyPDF.cgi/P1009I5K.PDF?Dockey=P1009I5K.PDF.
  8. C. T. Driscoll, K. M. Driscoll, M. J. Mitchell and D. J. Raynal, Effects of acidic deposition on forest and aquatic ecosystems in New York State, Environ. Pollut., 2003, 123(3), 327–336,  DOI:10.1016/S0269-7491(03)00019-8.
  9. C. T. Driscoll, K. M. Driscoll, K. M. Roy and J. Dukett, Changes in the chemistry of lakes in the Adirondack region of New York following declines in acidic deposition, Appl. Geochem., 2007, 22(6), 1181–1188,  DOI:10.1016/j.apgeochem.2007.03.009.
  10. L. Chen and C. T. Driscoll, Regional assessment of the response of the acid–base status of lake watersheds in the Adirondack region of New York to changes in atmospheric deposition using PnET-BGC, Environ. Sci. Technol., 2005, 39(3), 787–794,  DOI:10.1021/es049583t.
  11. D. A. Burns, J. A. Lynch, B. J. Cosby, M. E. Fenn and J. S. Baron, National Acid Precipitation Assessment Program report to congress 2011: An integrated assessment, National Science and Technology Council, Washington, DC, 2011 Search PubMed.
  12. T. L. Greaver, T. J. Sullivan, J. D. Herrick, M. C. Barber, J. S. Baron and B. J. Cosby, et al. Ecological effects of nitrogen and sulfur air pollution in the US: What do we know?, Front. Ecol. Environ., 2012, 10(7), 365–372,  DOI:10.1890/110049.
  13. C. T. Driscoll, K. M. Driscoll, H. Fakhraei and K. Civerolo, Long-term temporal trends and spatial patterns in the acid–base chemistry of lakes in the Adirondack region of New York in response to decreases in acidic deposition, Atmos. Environ., 2016, 146, 5–14,  DOI:10.1016/j.atmosenv.2016.08.034.
  14. R. A. F. Warby, C. E. Johnson and C. T. Driscoll, Chemical recovery of surface waters across the northeastern United States from reduced inputs of acidic deposition: 1984–2001, Environ. Sci. Technol., 2005, 39(17), 6548–6554,  DOI:10.1021/es048553n.
  15. C. T. Driscoll, R. P. Mason, H. M. Chan, D. J. Jacob and N. Pirrone, Mercury as a global pollutant: sources, pathways, and effects, Environ. Sci. Technol., 2013, 47(10), 4967–4983,  DOI:10.1021/es305071v.
  16. N. E. Selin, Global biogeochemical cycling of mercury: a review, Annu. Rev. Environ. Resour., 2009, 34(1), 43–63,  DOI:10.1146/annurev.environ.051308.084314.
  17. H. Hintelmann, R. Harris, A. Heyes, J. P. Hurley, C. A. Kelly and D. P. Krabbenhoft, et al. Reactivity and mobility of new and old mercury deposition in a boreal forest ecosystem during the first year of the METAALICUS study, Environ. Sci. Technol., 2002, 36(23), 5034–5040,  DOI:10.1021/es025572t.
  18. B. D. Blackwell and C. T. Driscoll, Using foliar and forest floor mercury concentrations to assess spatial patterns of mercury deposition, Environ. Pollut., 2015, 202, 126–134,  DOI:10.1016/j.envpol.2015.02.036.
  19. U. Skyllberg, J. Qian, W. Frech, K. Xia and W. F. Bleam, Distribution of mercury, methyl mercury and organic sulphur species in soil, soil solution and stream of a boreal forest catchment, Biogeochemistry, 2003, 64(1), 53–76,  DOI:10.1023/A:1024904502633.
  20. J. A. Dittman, J. B. Shanley, C. T. Driscoll, G. R. Aiken, A. T. Chalmers and J. E. Towse, et al. Mercury dynamics in relation to dissolved organic carbon concentration and quality during high flow events in three northeastern U.S. streams, Water Resour. Res., 2010, 46(7), W07522,  DOI:10.1029/2009wr008351.
  21. C. C. Gilmour, G. S. Riedel, M. C. Ederington, J. T. Bell, J. M. Benoit and G. A. Gill, et al. Methylmercury concentrations and production rates across a trophic gradient in the northern Everglades, Biogeochemistry, 1998, 40, 327–345,  DOI:10.1023/A:1005972708616.
  22. E. J. Kerin, C. C. Gilmour, E. Roden, M. T. Suzuki, J. D. Coates and R. P. Mason, Mercury methylation by dissimilatory iron-reducing bacteria, Appl. Environ. Microbiol., 2006, 72(12), 7919–7921,  DOI:10.1128/AEM.01602-06.
  23. C. C. Gilmour, M. Podar, A. L. Bullock, A. M. Graham, S. D. Brown and A. C. Somenahally, et al. Mercury methylation by novel microorganisms from new environments, Environ. Sci. Technol., 2013, 47(20), 11810–11820,  DOI:10.1021/es403075t.
  24. B. Aberg, L. Ekman, R. Falk, U. Greitz, G. Persson and J. Snihs, Metabolism of methyl mercury (203 Hg) compounds in man, Arch. Environ. Health, 1969, 19(4), 478–484,  DOI:10.1080/00039896.1969.10666872.
  25. Agency for Toxic Substances and Disease Registry (ATSDR), Toxicology profile for mercury. U.S. Department of Health and Human Services, Public Health Service, Atlanta, GA, 1999 Search PubMed.
  26. N. M. Burgess and M. W. Meyer, Methylmercury exposure associated with reduced productivity in common loons, Ecotoxicology, 2008, 17(2), 83–91,  DOI:10.1007/s10646-007-0167-8.
  27. T. Syversen and P. Kaur, The toxicology of mercury and its compounds, J. Trace Elem. Med. Biol., 2012, 26(4), 215–226,  DOI:10.1016/j.jtemb.2012.02.004.
  28. N. Schoch, M. J. Glennon, D. C. Evers, M. Duron, A. K. Jackson and C. T. Driscoll, et al. The impact of mercury exposure on the Common Loon (Gavia immer) population in the Adirondack Park, New York, USA, Waterbirds, 2014, 37(sp1), 133–146,  DOI:10.1675/063.037.sp116.
  29. D. S. Jeffries, T. G. Brydges, P. J. Dillon and W. Keller, Monitoring the results of Canada/U.S.A. acid rain control programs: Some Lake responses, Environ. Monit. Assess., 2003, 88(1–3), 3–19,  DOI:10.1023/A:1025563400336.
  30. U.S. EPA, Risk and exposure assessment for review of the secondary National Ambient Air Quality Standards for oxides of nitrogen and oxides of sulfur [Internet], 2011 [cited 2016 May 3], available from: https://www3.epa.gov/ttn/naaqs/standards/no2so2sec/data/20110114pamain.pdf.
  31. X. Yu, C. T. Driscoll, M. Montesdeoca, D. Evers, M. Duron and K. Williams, et al. Spatial patterns of mercury in biota of Adirondack, New York lakes, Ecotoxicology, 2011, 20(7), 1543–1554,  DOI:10.1007/s10646-011-0717-y.
  32. P. E. Drevnick, D. E. Canfield, P. R. Gorski, A. L. C. Shinneman, D. R. Engstrom and D. C. G. Muir, et al. Deposition and cycling of sulfur controls mercury accumulation in Isle Royale fish, Environ. Sci. Technol., 2007, 41(21), 7266–7272,  DOI:10.1021/es0712322.
  33. J. K. Coleman Wasik, D. R. Engstrom, C. P. J. Mitchell, E. B. Swain, B. A. Monson and S. J. Balogh, et al. The effects of hydrologic fluctuation and sulfate regeneration on mercury cycling in an experimental peatland, J. Geophys. Res.: Biogeosci., 2015, 120(9), 1697–1715,  DOI:10.1002/2015JG002993.
  34. S. G. Todorova, C. T. Driscoll, D. A. Matthews, S. W. Effler, M. E. Hines and E. A. Henry, Evidence for regulation of monomethyl mercury by nitrate in a seasonally stratified, eutrophic lake, Environ. Sci. Technol., 2009, 43(17), 6572–6578,  DOI:10.1021/es900887b.
  35. I. F. Dennis, T. A. Clair, C. T. Driscoll, N. Kamman, A. Chalmers and J. Shanley, et al. Distribution patterns of mercury in lakes and rivers of northeastern North America, Ecotoxicology, 2005, 14(1–2), 113–123,  DOI:10.1007/s10646-004-6263-0.
  36. J. A. Dittman, J. B. Shanley, C. T. Driscoll, G. R. Aiken, A. T. Chalmers and J. E. Towse, Ultraviolet absorbance as a proxy for total dissolved mercury in streams, Environ. Pollut., 2009, 157(6), 1953–1956,  DOI:10.1016/j.envpol.2009.01.031.
  37. D. Hongve, S. Haaland, G. Riise, I. Blakar and S. Norton, Decline of acid rain enhances mercury concentrations in fish, Environ. Sci. Technol., 2012, 46(5), 2490–2491,  DOI:10.1021/es3002629.
  38. P. E. Drevnick, D. R. Engstrom, C. T. Driscoll, E. B. Swain, S. J. Balogh and N. C. Kamman, et al. Spatial and temporal patterns of mercury accumulation in lacustrine sediments across the Laurentian Great Lakes region, Environ. Pollut., 2012, 161, 252–260,  DOI:10.1016/j.envpol.2011.05.025.
  39. Y. Zhang, D. J. Jacob, H. M. Horowitz, L. Chen, H. M. Amos and D. P. Krabbenhoft, et al. Observed decrease in atmospheric mercury explained by global decline in anthropogenic emissions, Proc. Natl. Acad. Sci. U. S. A., 2016, 113(3), 526–531,  DOI:10.1073/pnas.1516312113.
  40. C. P. Cirmo and C. T. Driscoll, The impacts of a watershed CaCO3 treatment on stream and wetland biogeochemistry in the Adirondack Mountains, Biogeochemistry, 1996, 32(3), 265–297,  DOI:10.1007/BF02187142.
  41. Y. Cho, C. T. Driscoll and J. D. Blum, The effects of a whole-watershed calcium addition on the chemistry of stream storm events at the Hubbard Brook Experimental Forest in NH, USA, Sci. Total Environ., 2009, 407(20), 5392–5401,  DOI:10.1016/j.scitotenv.2009.06.030.
  42. D. A. Webster, An unusual lake of the Adirondack Mountains, New York, Limnol. Oceanogr., 1961, 6(1), 88–90 CrossRef.
  43. C. L. Schofield, Water quality in relation to survival of Brook trout, Salvelinus fontinalis (Mitchill), Trans. Am. Fish. Soc., 1965, 94(3), 227–235 CrossRef.
  44. D. C. Josephson, J. M. Robinson, J. Chiotti, K. J. Jirka and C. E. Kraft, Chemical and biological recovery from acid deposition within the Honnedaga Lake watershed, New York, USA, Environ. Monit. Assess., 2014, 186(7), 4391–4409,  DOI:10.1007/s10661-014-3706-9.
  45. S. D. George, B. P. Baldigo, G. B. Lawrence and R. L. Fuller, Effects of watershed and in-stream liming on macroinvertebrate communities in acidified tributaries to an Adirondack lake, Ecol. Indic., 2018, 85, 1058–1067,  DOI:10.1016/j.ecolind.2017.11.048.
  46. C. Homan, C. Beier, T. McCay and G. Lawrence, Application of lime (CaCO3) to promote forest recovery from severe acidification increases potential for earthworm invasion, For. Ecol. Manage., 2016, 368, 39–44,  DOI:10.1016/j.foreco.2016.03.002.
  47. S. C. Peters, J. D. Blum, C. T. Driscoll and G. E. Likens, Dissolution of wollastonite during the experimental manipulation of Hubbard Brook Watershed, Biogeochemistry, 2004, 67(3), 309–329,  DOI:10.1023/B: BIOG.0000015787.44175.3f.
  48. C. T. Driscoll, C. P. Cirmo, T. J. Fahey, V. L. Blette, P. A. Bukaveckas and D. A. Burns, et al. The experimental watershed liming study: comparison of lake and watershed neutralization strategies, Biogeochemistry, 1996, 32(3), 143–174,  DOI:10.1007/BF02187137.
  49. U.S. EPA, Method 1669: Sampling Ambient Water for Trace Metals at EPA Water Quality Criteria Levels, United States Environ Prot Agency, 1995, pp. 1–42 Search PubMed.
  50. M. L. Olson and J. F. DeWild, Techniques for the collection and species-specific analysis of low levels of mercury in water, sediment, and biota, U.S. Geological Survey Water Resource Investigations Rep, 1999, p. 19 Search PubMed.
  51. U.S. EPA, Method 1631, revision E: mercury in water by oxidation, purge and trap, and cold vapor atomic fluorescence spectrometry, United States Environ Prot Agency, 2002, (August), pp. 1–46 Search PubMed.
  52. J. F. DeWild, M. L. Olson and S. D. Olund, Determination of methyl mercury by aqueous phase ethylation, followed by gas chromatographic separation with cold vapor atomic fluorescence detection, U.S Geological Survey, Open-File Report, 2002, p. 19 Search PubMed.
  53. U.S. EPA, Method 1630 Methyl Mercury in Water by Distillation, Aqueous Ethylation, Purge and Trap, and Cold Vapor Atomic Fluorescence Spectrometry, United States Environ Prot Agency, 2007, pp. 1–55 Search PubMed.
  54. J. Weishaar, G. Aiken, B. Bergamaschi, M. Fram, R. Fujii and K. Mopper, Evaluation of specific ultra-violet absorbance as an indicator of the chemical content of dissolved organic carbon, Environ. Sci. Technol., 2003, 37(20), 4702–4708,  DOI:10.1021/es030360x.
  55. B. A. Poulin, J. N. Ryan and G. R. Aiken, Effects of Iron on Optical Properties of Dissolved Organic Matter, Environ. Sci. Technol., 2014, 48(17), 10098–10106,  DOI:10.1021/es502670r.
  56. Lincoln T. A., Horan-Ross D. A., McHale M. R., Lawrence G. B., Quality-assurance data for routine water analyses by the U.S. Geological Survey laboratory in Troy, New York—July 2001 through June 2003, U.S. Geological Survey Open-File Report 2009–1232 [Internet], 2009, p. 32, available from: http://pubs.usgs.gov/of/2009/1232/.
  57. V. B. Sauer and D. P. Turnipseed, Stage measurement at gaging stations: U.S. Geological Survey Techniques and Methods book 3, chap. A7, 2010 Search PubMed.
  58. D. P. Turnipseed and V. B. Sauer, Discharge measurement at gaging stations: U.S. Geological Survey Techniques and Methods book 3, chap. A8, 2010 Search PubMed.
  59. Y. Cho, C. T. Driscoll and J. D. Blum, The effects of a whole-watershed calcium addition on the chemistry of stream storm events at the Hubbard Brook Experimental Forest in NH, USA, Sci. Total Environ., 2009, 407(20), 5392–5401,  DOI:10.1016/j.scitotenv.2009.06.030.
  60. C. Sjöstedt, C. Andrén, J. Fölster and J. P. Gustafsson, Modelling of pH and inorganic aluminium after termination of liming in 3000 Swedish lakes, Appl. Geochem., 2013, 35, 221–229,  DOI:10.1016/j.apgeochem.2013.04.014.
  61. S. Sandoy and A. J. Romundstad, Liming of acidified lakes and rivers in Norway, Water, Air, Soil Pollut., 1995, 85(2), 997–1002,  DOI:10.1007/BF00476960.
  62. B. A. Fost and C. P. Ferreri, pH preference and avoidance responses of adult Brook trout Salvelinus fontinalis and brown trout Salmo trutta, J. Fish Biol., 2015, 86(3), 952–966,  DOI:10.1111/jfb.12610.
  63. J. M. Clark, S. H. Bottrell, C. D. Evans, D. T. Monteith, R. Bartlett and R. Rose, et al. The importance of the relationship between scale and process in understanding long-term DOC dynamics, Sci. Total Environ., 2010, 408(13), 2768–2775,  DOI:10.1016/j.scitotenv.2010.02.046.
  64. D. a. N. Ussiri and C. E. Johnson, Sorption of organic carbon fractions by spodosol mineral horizons, Soil Sci. Soc. Am. J., 2004, 68(1), 253,  DOI:10.2136/sssaj2004.0253.
  65. D. A. Burns, G. R. Aiken, P. M. Bradley, C. A. Journey and J. Schelker, Specific ultra-violet absorbance as an indicator of mercury sources in an Adirondack river basin, Biogeochemistry, 2013, 113(1–3), 451–466,  DOI:10.1007/s10533-012-9773-5.
  66. S. A. Chiasson-Gould, J. M. Blais and A. J. Poulain, Dissolved organic matter kinetically controls mercury bioavailability to bacteria, Environ. Sci. Technol., 2014, 48(6), 3153–3161,  DOI:10.1021/es4038484.
  67. T. D. French, A. J. Houben, J.-P. W. Desforges, L. E. Kimpe, S. V. Kokelj and A. J. Poulain, et al. Dissolved organic carbon thresholds affect mercury bioaccumulation in Arctic lakes, Environ. Sci. Technol., 2014, 48, 3162–3168,  DOI:10.1021/es403849d.
  68. H. Chen, R. C. Johnston, B. F. Mann, R. K. Chu, N. Tolic and J. M. Parks, et al. Identification of mercury and dissolved organic matter complexes using ultrahigh resolution mass spectrometry, Environ. Sci. Technol. Lett., 2017, 4(2), 59–65,  DOI:10.1021/acs.estlett.6b00460.
  69. M. Haitzer, G. R. Aiken and J. N. Ryan, Binding of mercury(II) to aquatic humic substances: influence of pH and source of humic substances, Environ. Sci. Technol., 2003, 37(11), 2436–2441,  DOI:10.1021/es026291o.
  70. P. R. Gorski, D. E. Armstrong, J. P. Hurley and D. P. Krabbenhoft, Influence of natural dissolved organic carbon on the bioavailability of mercury to a freshwater alga, Environ. Pollut., 2008, 154(1), 116–123,  DOI:10.1016/j.envpol.2007.12.004.
  71. J. L. J. Ledesma, M. N. Futter, H. Laudon, C. D. Evans and S. J. Köhler, Boreal forest riparian zones regulate stream sulfate and dissolved organic carbon, Sci. Total Environ., 2016, 560–561, 110–122,  DOI:10.1016/j.scitotenv.2016.03.230.
  72. M. Podar, C. C. Gilmour, C. C. Brandt, A. Soren, S. D. Brown and B. R. Crable, et al. Global prevalence and distribution of genes and microorganisms involved in mercury methylation, Sci. Adv., 2015, 1(9), 1–13,  DOI:10.1126/sciadv.1500675.
  73. J. M. M. Benoit, C. C. C. Gilmour, A. Heyes, R. P. Mason and C. L. Miller, Geochemical and biological controls over methylmercury production and degradation in aquatic ecosystems, in ACS symposium, 2002, pp. 262–297,  DOI:10.1021/bk-2003-0835.ch019.
  74. D. A. Matthews, D. B. Babcock, J. G. Nolan, A. R. Prestigiacomo, S. W. Effler and C. T. Driscoll, et al. Whole-lake nitrate addition for control of methylmercury in mercury-contaminated Onondaga Lake, NY, Environ. Res., 2013, 125, 52–60,  DOI:10.1016/j.envres.2013.03.011.
  75. C. T. Driscoll, C. Yan, C. L. Schofield, R. K. Munson and J. Holsapple, The Mercury cycle and fish in the Adirondacks, Environ. Sci. Technol., 1994, 28(3), 136–143 CrossRef.
  76. K. Riva-Murray, L. C. Chasar, P. M. Bradley, D. A. Burns, M. E. Brigham and M. J. Smith, et al. Spatial patterns of mercury in macroinvertebrates and fishes from streams of two contrasting forested landscapes in the eastern United States, Ecotoxicology, 2011, 20(7), 1530–1542,  DOI:10.1007/s10646-011-0719-9.
  77. D. A. Burns and K. Riva-Murray, Variation in fish mercury concentrations in streams of the Adirondack region, New York, A simplified screening approach using chemical metrics, Ecol. Indic., 2018, 84, 648–661,  DOI:10.1016/j.ecolind.2017.09.031.
  78. G. B. Lawrence, D. A. Burns and K. Riva-Murray, A new look at liming as an approach to accelerate recovery from acidic deposition effects, Sci. Total Environ., 2016, 562, 35–46,  DOI:10.1016/j.scitotenv.2016.03.176.
  79. H. A. Simonin, J. J. Loukmas, L. C. Skinner and K. M. Roy, Lake variability: key factors controlling mercury concentrations in New York State fish, Environ. Pollut., 2008, 154, 107–115,  DOI:10.1016/j.envpol.2007.12.032.
  80. D. M. Ward, K. H. Nislow and C. L. Folt, Bioaccumulation syndrome: identifying factors that make some stream food webs prone to elevated mercury bioaccumulation, Ann. N. Y. Acad. Sci., 2010, 1195, 62–83,  DOI:10.1111/j.1749-6632.2010.05456.x.
  81. J. R. Thompson, D. N. Carpenter, C. V. Cogbill and D. R. Foster, Four centuries of change in northeastern United States forests, PLoS One, 2013, 8(9), e72540,  DOI:10.1371/journal.pone.0072540.
  82. D. T. Monteith, J. L. Stoddard, C. D. Evans, H. a. de Wit, M. Forsius and T. Høgåsen, et al. Dissolved organic carbon trends resulting from changes in atmospheric deposition chemistry, Nature, 2007, 450(7169), 537–540,  DOI:10.1038/nature06316.
  83. C. T. Driscoll, K. M. Driscoll, K. M. Roy and M. J. Mitchell, Chemical response of lakes in the Adirondack region of New York to declines in acidic deposition, Environ. Sci. Technol., 2003, 37(10), 2036–2042,  DOI:10.1021/es020924h.
  84. P. Vidon, W. Carleton and M. J. Mitchell, Spatial and temporal variability in stream dissolved organic carbon quantity and quality in an Adirondack forested catchment, Appl. Geochem., 2014, 46, 10–18,  DOI:10.1016/j.apgeochem.2014.04.008.
  85. H. Laudon, M. Berggren, A. Ågren, I. Buffam, K. Bishop and T. Grabs, et al. Patterns and dynamics of dissolved organic carbon (DOC) in boreal streams: the role of processes, connectivity, and scaling, Ecosystems, 2011, 14(6), 880–893,  DOI:10.1007/s10021-011-9452-8.
  86. M. B. David and G. F. Vance, Chemical character and origin of organic acids in streams and seepage lakes of central Maine, Biogeochemistry, 1991, 12(1), 17–41 CrossRef CAS.
  87. H. Fakhraei and C. T. Driscoll, Proton and aluminum binding properties of organic acids in surface waters of the northeastern U.S., Environ. Sci. Technol., 2015, 49(5), 2939–2947,  DOI:10.1021/es504024u.
  88. D. A. Burns, L. G. Woodruff, P. M. Bradley and W. F. Cannon, Mercury in the soil of two contrasting watersheds in the eastern United States, PLoS One, 2014, 9(2), e86855,  DOI:10.1371/journal.pone.0086855.
  89. J. D. Demers, C. T. Driscoll, T. J. Fahey and J. B. Yavitt, Mercury cycling in litter and soil in different forest types in the Adirondack region, New York, USA, Ecol. Appl., 2007, 17(5), 1341–1351,  DOI:10.1890/06-1697.1.

Footnote

Electronic supplementary information (ESI) available. See DOI: 10.1039/c7em00520b

This journal is © The Royal Society of Chemistry 2018