- Email: [email protected]
Effects of road salt deicers on an urban groundwater-fed kettle lake
Applied Geochemistry 89 (2018) 265–272
Contents lists available at ScienceDirect
Applied Geochemistry journal homepage: www.elsevier.com/locate/apgeochem
Effects of road salt deicers on an urban groundwater-fed kettle lake 1
D. Allie Wyman , Carla M. Koretsky
T
∗
Department of Geosciences, Western Michigan University, Kalamazoo, MI, 49008, USA
A R T I C L E I N F O
A B S T R A C T
Handling Editor: Prof. M. Kersten
Road salt deicers significantly influence the chemistry and physical mixing of urban lakes, even causing transition from dimixis to monomixis or meromixis. In this study, the water column geochemistry of Asylum Lake, a primarily groundwater-fed, eutrophic kettle lake in urban Kalamazoo, MI, was monitored for over a year to determine the extent of road salt deicer influence on the lake chemistry and the physical mixing of the lake. Water column samples from the surface to the deepest part of Asylum Lake were analyzed monthly for nineteen months for a suite of parameters including dissolved oxygen, pH, conductivity, temperature, alkalinity, Fe2+, Mn2+, orthophosphate, total NH4+, alkalinity, total sulfide, Cl−, SO42−, Ca2+, Mg2+, K+, and Na+. During the study period, spring mixing was never observed and a nearly complete fall turnover was observed only in November 2013. The hypolimnion of Asylum Lake was always hypoxic or anoxic and redox-stratified, with seasonal development of suboxic and sulfidic zones, which were disrupted in fall and winter following partial fall turnover and subsequent ice cover. This study suggests that road salt deicers have caused Asylum Lake to transition from dimixis to meromixis or periodic monomixis with significant consequences for biogeochemical cycles in the lake waters.
Keywords: Eutrophication Road salt Anoxia Redox stratification Mixing
1. Introduction The widespread application of road salt deicers has the potential to significantly influence the salinity of freshwater systems. In the United States, an estimated 8 Mt of salt was used for deicing in 1975, increasing to 55 Mt in 2003 (U.S. Geological Survey, 2014). Rising chloride levels related to the use of road salts have been documented in groundwater and surface waters in the United States, Canada and elsewhere (e.g. Environment Canada, 2001; Chapra et al., 2009, 2012; Mullaney et al., 2009; Kaushal et al., 2014; Rogora et al., 2015; Dugan et al., 2017). For example, Novotny et al. (2008) showed that urban lakes in Minnesota have 10 to 25 times greater sodium and chloride levels as compared to rural Minnesota lakes, with chloride concentrations in the urban lakes increasing at an annual rate of 1.8%. Other studies have demonstrated increasing chloride levels in streams and rivers (e.g. Kaushal et al., 2005; Kelly et al., 2012; Dailey et al., 2014; Corsi et al., 2015), the Great Lakes (Chapra et al., 2009, 2012), as well as in lakes of New Hampshire (Likens and Buso, 2010), Michigan (Koretsky et al., 2012; Sibert et al., 2015), Italy (Rogora et al., 2015), Canada (Environment Canada, 2001), and Sweden (Thunqvist, 2004) due to influx of road salt deicers. With continued use of road salt deicers, chloride concentrations in urban freshwater systems are likely to continue increasing.
∗
1
Levels of chloride in many surface water systems now periodically or even permanently exceed the chronic toxicity threshold for freshwater species of 250 mg/L (e.g., Environment Canada, 2001; Kaushal et al., 2005; Mullaney et al., 2009; Corsi et al., 2015; Dugan et al., 2017). Increasing chloride levels can also affect the density structure, and thus potentially change the mixing dynamics, of freshwater lakes. For example, Judd (1970) found that wind action was insufficient to cause seasonal turnover of First Sister Lake in Michigan because of the increased salinity, and thus increased density of the lake bottom waters, created by the influx of road salts. Similarly, Novotny et al. (2008) observed that road salt contamination caused two urban lakes in Minnesota to transition from dimixis (twice per year mixing) to monomixis (once per year mixing), and Sibert et al. (2015) demonstrated that road salts caused an urban kettle lake in Michigan to transition to meromixis (no annual mixing). Both Novotny et al. (2008) and Sibert et al. (2015) reported hypolimnetic anoxia in these lakes, in part due to their failure to turnover. Eutrophic lakes have high levels of nutrients and labile organic matter. In these lakes, the water column may become redox stratified as labile organic matter remaining after oxygen depletion is oxidized via a sequence of increasingly less energetically favorable terminal electron acceptors including nitrate, Mn4+, Fe3+ and SO4−2. This results in the production and accumulation of reduced solutes including NH4+,
Corresponding author. 1903 West Michigan Ave, Kalamazoo, MI, 49008, USA. E-mail address: [email protected] (C.M. Koretsky). Present address: Geology Department, University of Illinois, Champaign, IL, 61,821.
https://doi.org/10.1016/j.apgeochem.2017.12.023 Received 10 August 2017; Received in revised form 21 December 2017; Accepted 22 December 2017 Available online 27 December 2017 0883-2927/ © 2017 Published by Elsevier Ltd.
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Fig. 1. Bathymetric map of Asylum Lake with sampling location noted by star. Map was created in 1991 by William Sauck (Institute for Water Sciences), William Laton (Department of Geology), and Bryan Allen (Department of Geology) using a Lowrance Model X-16 with 192 kHz transducer along numbered transects. Contour interval is 0.61 m.
Mn2+, Fe2+, H2S and CH4 in the hypolimnion (Boehrer and Schultze, 2008; Koretsky et al., 2012; Sibert et al., 2015; Dupuis, 2017). If a eutrophic lake fails to turnover due to road salt contamination, hypolimnetic anoxia and redox-stratification may become permanent. Whether a lake transitions to meromixis due to road salt influx likely depends on the concentration of salt, the residence time of water in the lake, and the lake depth and surface area (Rodrigo et al., 2001). In this study, the water column chemistry of a primarily groundwater-fed, eutrophic kettle lake located in urban Kalamazoo, MI was studied over a period of nineteen months to assess the influence of road salt contamination on seasonal turnover and the development of redox stratification in the lake.
0.5% bare soil. 3. Materials and methods Water column sampling was conducted at approximately the deepest point in Asylum Lake (Fig. 1) at 0.5 m intervals for in situ measurements of dissolved oxygen, pH, conductivity, and temperature and at 1 m intervals for ex situ samples. Water samples were collected approximately once a month from September 2012 to March 2014, with the exception of the period between mid-December 2012 and March 2013. During this time, ice thickness was such that it was unsafe to traverse the lake and no samples were collected. During periods of potential turnover, in situ samples were collected more frequently. Water samples were transported back to Western Michigan University and analyzed for Fe2+, Mn2+, orthophosphate, total NH4+, alkalinity, total sulfide (ΣH2S), Cl−, SO42−, Ca2+, Mg2+, K+, and Na+ according to the methods described in Sibert et al. (2015).
2. Study site Asylum Lake is an urban kettle lake located in Kalamazoo, MI, with a surface area of approximately 19.8 ha, a mean depth of 7.2 m, and a maximum depth of 15.8 m (Sauck and Barcelona, 1992; Kieser and Associates, 2008; Koretsky et al., 2012, Fig. 1). Water enters the lake primarily through groundwater discharge on the west side of the lake and exits the lake through groundwater recharge to the east, as well as via a small culvert leading to Little Asylum Lake. The residence time of water in the lake is estimated to be 0.68 years (Kieser and Associates, 2008). Asylum Lake is bordered by wetlands to the west, prairie and woods to the south, and by residential neighborhoods to the north. The soils around the lake are primarily Houghton and Sebewa soils characterized by excess humus and wetness (USDA, 1979). South of Asylum Lake, the soils are Kalamazoo loam with a 6–12% slope; north and west of Asylum Lake, the soils are Urban Oshtemo complex with slopes of 12–25%; southeast and southwest of Asylum Lake, soils are Oshtemo sandy loam with 18–35% slope and 12–18% slope, respectively (USDA, 1979). Kieser and Associates (2008) completed a topographic assessment of the surrounding watershed and concluded that it is comprised of ∼34% forest or open herbaceous land; 30% farmland; 18% high/low density urban; 10% roads/parking lots; 4% wetlands; 3% water and
4. Results 4.1. Temperature, dissolved oxygen, conductivity, and pH Asylum Lake temperature profiles show strong thermal stratification in summer and early fall, weak stratification in late fall and early spring, and reverse stratification during all periods of ice cover. In December 2012, April 2013, November 2013, and April 2014 the lake is isothermal, indicative of potential turnover events (Fig. 2A). The epilimnion is supersaturated with respect to dissolved oxygen in all seasons except during periods of ice cover (Fig. 2B). In the summer, the hypolimnion is hypoxic to anoxic below 6–9 m depth. In fall and spring, oxic conditions persist to depths of between 10 m and 13 m, except during isothermal conditions in late November 2013, when dissolved oxygen increases slightly with depth to 14 m and then decreases at 14.5 m (Fig. 2B). In all seasons, specific conductivity in Asylum Lake increases with 266
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
the lake is nearing isothermal conditions, the pH varies less with depth. 4.2. Major ions Alkalinity in Asylum Lake increases with depth in all seasons, ranging from approximately 250 to 300 ppm in the epilimnion and from approximately 350 to 420 ppm in the hypolimnion (Fig. S2). During all ice-free periods, alkalinity variations from month to month are small. However, in February 2013, during maximum ice cover, alkalinity concentrations are elevated throughout the entire water column, ranging from 420 ppm in the epilimnion to 510 ppm in the hypolimnion. Concentrations of dissolved calcium increase with depth in all seasons with concentrations ranging from ∼30 to 40 ppm at the surface to ∼60–80 ppm at depth (Fig. S3). The only period during which dissolved calcium does not increase with depth is December 2012, when concentrations remain steady between 60 and 65 ppm. Dissolved magnesium and potassium concentrations increase only slightly with depth during all seasons with dissolved magnesium concentrations typically ∼25–30 ppm and dissolved potassium concentrations ∼1–5 ppm (Figs. S4 and S5). 4.3. Redox sensitive solutes and nutrients In summer and fall, total ammonium and orthophosphate accumulate in the water column between 10 and 12 m depth, coincident with the depletion of dissolved oxygen (Fig. 3A and B). During potential turnover in December 2012, small but detectable levels of total ammonium are also present in the epilimnion. During periods of ice cover, orthophosphate and total ammonium reach concentrations as high as 50 ppb orthophosphate and 200 ppb total ammonium in the epilimnion. Dissolved Mn2+ is always present at levels of ≥400 ppb below 11 m depth, accumulating in spring and winter at depths where dissolved oxygen is depleted (Fig. 4A). In November 2012, December 2012, and August 2013, Mn2+ is detectable throughout the lake water column. Fe2+ concentrations remain below detection limits of 0.2 ppm during all sampling periods. Dissolved ΣH2S accumulates in the hypolimnion during summer and fall, reaching concentrations approaching 4 ppm (Fig. 4B). During early spring and periods of ice cover, ΣH2S concentrations remain below detection limits throughout the water column. During summer and fall months, sulfate concentrations increase slightly between the surface and 12 m depth and then decrease with depth, often coinciding with increasing dissolved ΣH2S concentrations (Fig. 4C). 4.4. Chloride and sodium December 2012 chloride concentrations are fairly uniform (∼165 ppm) with depth (Fig. 5A). After snowmelt in 2013, chloride concentrations steadily increase with depth in both the epilimnion and hypolimnion, peaking at > 330 ppm at 14 m depth in September 2013. Chloride concentrations decrease slightly throughout the lake from September until March 2014. For all sampling dates in 2012, sodium levels in Asylum Lake remain fairly uniform with depth at ∼70–85 ppm (Fig. 5B). After snowmelt in 2013, sodium concentrations increase in the water column to a maximum of ∼130 ppm in the hypolimnion. During periods of ice cover (January through March 2014), sodium concentrations in the epilimnion remain stable whereas concentrations in the hypolimnion rise to ∼150 ppm.
Fig. 2. Depth-time isoplots of (A) temperature (°C), (B) dissolved oxygen (mg·L−1), and (C) specific conductance (mS·cm−1) at Asylum Lake. Thick rectangles at 0 m depth indicate approximate periods of ice cover.
depth (Fig. 2C). A distinct increase in conductivity in both the epilimnion and hypolimnion is apparent after snowmelt in 2013 and 2014. The maximum vertical differential in conductivity increases from 0.126 in 2012 to 0.443 mS cm−1 in 2014. During isothermal conditions in late November 2013, conductivity was uniform in the upper 14 m with a sharp increase at 14.5 m depth. The pH in Asylum Lake decreases with depth in all seasons (Fig. S1). Typically, pH in the epilimnion ranges from 8.0 to 8.7 and hypolimnion pH ranges from 7.1 to 7.3. Epilimnion and hypolimnion pH values increase during periods of ice cover. During fall and spring periods, when
5. Discussion 5.1. Salinization In an unpublished report, Engemann (1977) noted chloride concentrations of up to 126 ppm in Asylum Lake. Since this time, elevated 267
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Fig. 3. Depth-time isoplots of (A) total ammonium (ppb) and (B) orthophosphate (ppb) at Asylum Lake. Thick rectangles at 0 m depth indicate approximate periods of ice cover.
(Koretsky et al., 2012; Dupuis, 2017). Chloride concentrations were compared to conductivity in Asylum Lake and found to consistently correlate, with a coefficient of determination (R2) of 0.8 or greater in May and June 2013 and at least 0.9 from July 2013 through the end of the study (Table 1). This is comparable to correlations between chloride and conductivity observed by Sibert et al. (2015) and Novotny et al. (2008) in other urban lakes impacted by road salt deicers. Sodium concentrations were also compared to conductivity. However, the R2 for these correlations is less consistent, generally ranging from 0.76 to 0.97, with a few lower values. This is likely because sodium is a less conservative element than chloride due to ion exchange reactions between sodium and clay minerals or organic matter in the soils of the groundwater aquifer that discharges into Asylum Lake (Kehew, 2001). The strong correlations between sodium and conductivity and especially between chloride and conductivity suggest that the dissolved ion concentrations in Asylum Lake are dominated by road salt input.
chloride concentrations at Asylum Lake have been documented in three additional studies spanning three decades. Buening (1994) measured maximum chloride concentrations in Asylum Lake of 140 ppm, which had increased to 175 ppm in 2006–07 (Kieser and Associates, 2008) and to 225 ppm another 15 years later (Koretsky et al., 2012). In the present study, we found maximum chloride concentrations of > 330 ppm, over 2.5 times the amount of chloride reported by Engemann (1977). Elevated chloride levels are also found in wells near Asylum Lake. Wyman (2014) sampled a well screened at 20.7–22.2 m depth, located to the southwest of Asylum Lake and just to the east of a heavily salted interstate and high use local road. Chloride concentrations in the well ranged from 105 to 153 ppm, suggesting that significant quantities of road salt enter the lake via groundwater, in addition to the salt that enters via overland flow. In contrast to the high chloride concentrations that have accumulated in Asylum Lake, recently reported chloride concentrations in two nearby rural kettle lakes, Brewster Lake and North Lake, are consistently < 15 ppm throughout the water column 268
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Asylum Lake (Kieser and Associates, 2008), and thus, chloride levels may fluctuate significantly from year to year, depending on the severity of winter conditions and the resulting amount of road salt use. 5.2. Lake mixing Dimictic lakes turnover in the fall when surface waters cool and become denser and again in spring after ice out (Wetzel, 2001). However, it has been demonstrated that lakes influenced by increased road salt may transition to monomixis or meromixis due to changes in the density structure of the water column (Judd, 1970; Novotny et al., 2008; Novotny and Stefan, 2012; Sibert et al., 2015). Isothermal conditions were observed in Asylum Lake during four sampling events in this study: December 2012, April 2013, November 2013, and April 2014 (Fig. 2A). If water column mixing occurs, the water column should be isochemical as well as isothermal. However, nearly isochemical conditions were only apparent during one of the four isothermal sampling events. A full sampling of the lake was completed on November 16, 2013; on this date, the lake was still chemically and thermally stratified. One week later, on November 24, 2013, only in situ data was collected. On this date, the lake was nearly isothermal and isochemical to a depth of 14 m. At 14.5 m, there was a large decrease in dissolved oxygen and pH, and a large increase in conductivity. Turnover may have occurred just before or after this date, or it is possible that the lake did not entirely turnover at the deepest point in the lake in fall 2013. It is also important to note that turnover probably did occur at shallower locations. Our data implies that the lake has transitioned to monomixis or perhaps even periodic meromixis, although this cannot be determined definitively without continuous year-round monitoring of the lake. The delayed or absent mixing is likely due to increased density at depth due to the elevated levels of chloride from application of road salt to nearby roads (Fig. 5A). 5.3. Redox stratification Redox stratification can develop in lake water columns as a result of oxidation-reduction reactions that occur as terminal electron acceptors are used in sequence to oxidize organic matter (Stumm and Morgan, 1996; Wetzel, 2001). Terminal electron accepters are used approximately in order of thermodynamic yield; if a terminal electron acceptor is exhausted by reaction with organic matter or other reducing agents and not replenished through mixing or other chemical reactions, then organic matter oxidation will proceed using the next available terminal electron acceptor that will produce the greatest energetic yield. Because of the high quantities of labile organic matter in eutrophic lakes, redox stratification can be especially pronounced, particularly in the absence of significant wind mixing or influx of oxic waters (Stumm and Morgan, 1996; Wetzel, 2001). In Asylum Lake, redox stratification can be delineated from dissolved oxygen, Mn2+ and ΣH2S profiles during all seasons (Figs. 2B, 4A and 4B). In April 2013, the oxic zone stretches from the surface to 12 m depth, beneath which hypoxic conditions are present and Mn2+ begins to accumulate. As organic matter productivity increases in spring and summer 2013, subsurface peaks in dissolved oxygen are observed as photosynthetic plants produce oxygen in the photic zone. pH peaks at depths 0.5–1 m deeper are further evidence of photosynthetic production of dissolved oxygen because photosynthesis consumes carbon dioxide, raising the pH (Fig. S1). With increasing organic matter productivity and temperature, dissolved oxygen is depleted at much shallower depths. Hypoxic conditions are present at depths as shallow as 6–7.5 m from August to October 2013 and Mn2+ concentrations are detectable at depths as shallow as 1 m in August 2013 (Fig. 4A). Dissolved ΣH2S is observed at depths below those where Mn2+ first appears, with detectable concentrations noted at and below 13 m depth in July 2013 and persistent concentrations of ∼1–3.5 ppm between 12 and 13 m depth throughout
Fig. 4. Depth-time isoplots of (A) manganese (ppb), (B) total sulfide (ppb), and (C) sulfate (ppm) at Asylum Lake. Thick rectangles at 0 m depth indicate approximate periods of ice cover.
Kalamazoo climate data from October 2011 through April 2014 shows an increase in the number of days below freezing and an increase in the amount of snowfall between 2011–2012 and 2013–2014 (Table 2). In 2011–2012, there were only 20 days below freezing, compared to 78 in 2013–2014. Total snowfall accumulation ranged from 1446 mm in 2011–2012 to 2877 in 2013–2014. Due to the mild 2011–2012 winter season, less deicer was applied to roads compared to the subsequent winter seasons (Table 2). This may explain the poorer correlations of chloride with conductivity in 2012 as compared to 2013 and 2014 (Table 1). Prior work suggests a residence time of < 1 year in
269
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Fig. 5. Depth-time isoplots of (A) chloride (ppm) and (B) sodium (ppm) at Asylum Lake. Thick rectangles at 0 m depth indicate approximate periods of ice cover.
Fe2+ in Asylum Lake is due to a lower delivery of fresh particulate iron than in Woods Lake, which likely has greater influx of particulates from stormwater sewers and steeper slopes surrounding the lake (Sibert et al., 2015). We hypothesize that any Fe2+ produced by reductive dissolution of iron oxides in the lake water or sediment columns of Asylum Lake reacts with sulfide and is removed from the water column before it can accumulate to levels above detection limits (Davison, 1993; Postma and Jakobsen, 1996; Stumm and Morgan, 1996; Wetzel, 2001; Rickard and Luther, 2007). Phosphate only occurs in a single oxidation state but it is influenced by changes in redox conditions because it binds readily to iron oxide minerals and is released into solution as they are reductively dissolved (Wetzel, 2001). Therefore, increases in phosphate concentrations are commonly coincident with increases in dissolved Fe2+ concentrations in redox-stratified systems. However, phosphate can also be released from decaying organic material in the hypolimnion, causing an increase in phosphate without a corresponding increase in dissolved Fe2+.
the remainder of summer 2013 (Fig. 4B). In fall 2013, cooling of water in the epilimnion results in at least some water column mixing, and oxic conditions to depths of at least 12 m (Fig. 2A and B). During this period Mn2+ and ΣH2S concentrations decrease slightly but continue to persist at depths below 10 and 12 m, respectively. After January 2014, ΣH2S concentrations are below detection limits at all depths and remain so throughout the winter. The water column is hypoxic from 12 to 14 m depth throughout the winter, with persistent accumulation of Mn2+ at concentrations between ∼400 and 1800 ppb. No detectable Fe2+ was observed in the water column during any sampling period in this study. This is interesting, particularly given the high levels of dissolved Fe2+ in the hypolimnion of nearby Woods Lake, which is also impacted by road salt deicer (Sibert et al., 2015). In addition, dissolved Fe2+ concentrations in excess of 20 ppm have been reported in the sediment pore waters of shallow sediments fringing Asylum Lake (Koretsky et al., 2006). Asylum Lake is primarily groundwater fed, suggesting that the lack of accumulation of dissolved 270
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Mn2+ or sulfide. While more work must be done to fully ascertain the origin of phosphate in Asylum Lake, this suggests that phosphate is also released by organic matter decay. Ammonium is an energetically favorable source of nitrogen for aquatic plants and is rapidly assimilated when present in the photic zone. Additionally, ammonium oxidizes quickly and rarely accumulates in oxic waters. Therefore, ammonium only accumulates in lakes when organic material begins to decay in anoxic hypolimnia (Wetzel, 2001). In Asylum Lake, the presence of ammonium was observed in all seasons in the anoxic hypolimnion (Fig. 3A). It was only observed at levels above detection limits at shallower depths during December 2012, January 2014, and February 2014. During these winter months, assimilation of ammonium by aquatic plants was likely slowed due to colder temperatures and reduced sunlight. Also, dissolved oxygen saturation (Fig. 1) was reduced to below 60% in the epilimnion during these months. These combined effects likely allowed ammonium to persist at shallower depths in the water column.
Table 1 Coefficient of determination (R2) between chloride and conductivity and sodium and conductivity for each month of sampling in Asylum Lake. Date
Cl and Conductivity
Na and Conductivity
9/14/12 9/29/12 11/18/12 12/8/12 4/6/13 5/31/2013 6/3/2013 6/27/13 7/25/13 8/27/13 9/26/13 10/22/13 11/16/13 1/10/14 2/8/14 3/21/14 4/8/14
No data No data No data 0.66 0.45 0.80 0.80 0.88 0.94 0.95 0.97 0.95 0.89 0.94 0.98 0.96 0.97
0.13 0.17 0.11 0.00 0.93 No data No data 0.46 0.76 0.006 0.87 0.85 0.83 0.51 0.95 0.97 0.97
6. Conclusions Throughout the nineteen-month study period, Asylum Lake chloride concentrations in the hypolimnion exceed background levels by approximately two orders of magnitude. Chloride and sodium strongly correlate with conductivity after the 2013 spring snowmelt, indicating that dissolved ion concentrations in the water column are dominated by sodium and chloride influxes from road salt deicers. Lakes in the Midwest are typically dimictic (Wetzel, 2001); however, the high concentrations of chloride in Asylum Lake and corresponding increases in the density of hypolimnetic waters appears to have resulted in transition of the lake from dimixis to monomixis and perhaps even periodic meromixis. The lack of consistent fall and spring turnover leads to persistent anoxia in the hypolimnion, with accumulation of Mn2+, ΣH2S, and NH4+, which undoubtedly impacts the aquatic ecosystem. The lack of Fe2+ accumulation in the hypolimnion is particularly interesting, especially given the high concentrations observed in a similar nearby kettle lake that has also been impacted by road salt, and warrants further study. Methane concentrations were not measured in this study. However, the compressed redox stratification suggests that significant levels of methane may also accumulate in the hypolimnion of eutrophic lakes impacted by road salt deciers. This could result in an elevated efflux of methane from the lake column to the atmosphere and should also be studied further. This study clearly demonstrates that the anthropogenic input of deicers to freshwater kettle lakes changes mixing dynamics and biogeochemical cycles, and may exacerbate existing eutrophication issues.
Table 2 Climate data for the city of Kalamazoo from the NOAA climate data center (NESDIS/ NOAA/NCEI). Number of days with maximum temperature below 0°C and average monthly temperature were collected at the Kalamazoo Battle Creek International Airport. Total snowfall accumulation was collected at Western Michigan University. The amount of road salt applied per winter season was obtained through personal communication with a city of Kalamazoo employee. Date
Number of Days with Maximum Temperature Below 0 °C
2011–2012 Oct-2011 Nov-2011 Dec-2011 Jan-2012 Feb-2012 Mar-2012 Apr-2012 Total 2012–2013 Oct-2012 Nov-2012 Dec-2012 Jan-2013 Feb-2013 Mar-2013 Apr-2013 Total 2013–2014 Oct-2013 Nov-2013 Dec-2013 Jan-2014 Feb-2014 Mar-2014 Apr-2014 Total
Winter Season 0 0 3 11 4 2 0 20 Winter Season 0 1 5 14 14 7 0 41 Winter Season 0 5 18 21 23 11 0 78
Average Monthly Temperature (°C)
Total Snowfall Accumulation (mm)
Road Salt Applied (tons)
11.3 6.8 1.4 −1.7 −0.1 10.5 9.1 NA
0 0 106 828 398 114 0 1446
NA NA NA NA NA NA NA 4438
9.9 4.1 1.7 −2.4 −3.5 0 7.8 NA
0 0 146 519 776 258 18 1717
NA NA NA NA NA NA NA 6212
11.5 3.2 −3 −8.3 −8.3 −2.5 8.9 NA
0 67 781 1091 638 300 0 2877
NA NA NA NA NA NA NA 10,132
Acknowledgements We deeply appreciate field and laboratory assistance from Ryan Sibert, Tom Howe, Ann Gilchrist, Denisha Griffey, Rebecca Kiekhaefer, and Derrick Lingle. William Sauck is acknowledged for assisting in the creation of Fig. 1. The WMU Graduate College is also acknowledged for providing a research grant to Allie Wyman, which helped to support this work. Steve Keto and the Asylum Lake Preservation Council are acknowledged for providing access to Asylum Lake.
In Asylum Lake, phosphate is present in the water column while dissolved Fe2+ remains below detection limits in all seasons (Fig. 3B). Except during September 2013 and January through April 2014, we observed increasing phosphate concentrations at the same depth or one meter below the first detection of dissolved Mn2+ and above the first detection of ΣH2S (Figs. 3B, 4A and 4B) These trends are consistent with Fe(III) reduction occurring just below Mn(IV) reduction, releasing both dissolved PO43− and Fe2+, but with the Fe2+ immediately removed via reactions with sulfide. In contrast, during September 2013 and January through April 2014 phosphate accumulates at shallower depths than
Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx. doi.org/10.1016/j.apgeochem.2017.12.023. REFERENCES Boehrer, B., Schultze, M., 2008. Stratification of lakes. Rev. Geophys. 46, RG2005. Buening, Vincent E., 1994. Investigation of Glacial-Drift Aquifer Interactions Using Hydrogeology, Geochemistry, and Stable Isotopes at Asylum Lake Michigan (Master's
271
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Koretsky, C.M., MacLeod, A., Sibert, R.J., Snyder, C., 2012. Redox stratification and salinization of three kettle lakes in southwest Michigan, USA. Water Air Soil Pollut. 223, 1415–1427. Likens, G.E., Buso, D.C., 2010. Salinization of mirror lake by road salt. Water Air Soil Pollut. 205, 205–214. Mullaney, J.R., Lorenz, D.L., Arntson, A.D., 2009. Chloride in Groundwater and Surface Water in Areas Underlain by the Glacial Aquifer System, Northern United States: U.S. Geological Survey Scientific Investigations Report 2009-5086. 41p. Novotny, E.V., Murphy, D., Stefan, H.G., 2008. Increase of urban lake salinity by road deicing salt. Sci. Total Environ. 406, 131–144. Novotny, E.V., Stefan, H.G., 2012. Road salt impact on lake stratification and water quality. J. Hydraul. Eng. 138, 1069–1080. Postma, D., Jakobsen, R., 1996. Redox zonation: equilibrium constraints on the Fe(III)/ SO4- reduction interface. Geochem. Cosmochim. Acta 60, 3169–3175. Rickard, D., Luther, G.W., 2007. Chemistry of iron sulfides. Chem. Rev. 107, 514–562. Rodrigo, M.A., Miracle, M.R., Vicente, E., 2001. The meromictic Lake La Cruz (Central Spain). Patterns of stratification. Aquat. Sci. 63, 406–416. Rogora, M., Mosello, R., Kamburska, L., Salmaso, N., Cerasino, L., Leoni, B., Garibaldi, L., Soler, V., Lepori, L., Columbo, L., Buzzi, F., 2015. Recent trends in chloride and sodium concentrations in the deep subalpine lakes (Northern Italy). Environ. Sci. Pollut. Res. 22, 19013–19026. Sauck, W.A., Barcelona, M.J., 1992. Final Project Report: Long-Term Hydrogeological Research and Education Test Site. Western Michigan University Unpublished. Retrieved from: https://wmich.edu/asylumlake/pastresearch. Sibert, R.J., Koretsky, C.M., Wyman, D.A., 2015. Cultural meromixis: effects of road salt on the chemical stratification of an urban kettle lake. Chem. Geol. 395, 126–137. Stumm, M., Morgan, J.J., 1996. Aquatic Chemistry. John Wiley and Sons, New York, NY. Thunqvist, E.-L., 2004. Regional increase of mean chloride concentration in water due to the application of deicing salt. Sci. Total Environ. 325, 29–37. United States Department of Agriculture, 1979. Soil Conservation Service. Soil Survey of Kalamazoo County, Michigan. National Cooperative Soil Survey. U.S. Geological Survey, 2014. Salt statistics. In: Kelly, T.D., Matos, G.R. (Eds.), Comps. Historical Statistics for Mineral and Material Commodities in the United States, U.S. Geological Survey Data Series 140, available online at: http://pubs.usgs.gov/ds/ 2005/140/, Accessed date: 17 July 2015. Wetzel, R.G., 2001. Limnology: Lake and River Ecosystems, third ed. Academic, San Diego. Wyman, D.A., 2014. Effects of Road Salt on Asylum Lake Geochemistry (Master's Thesis). Western Michigan University.
Thesis). Western Michigan University. Chapra, S.C., Dove, A., Rockwell, D.C., 2009. Great Lakes chloride trends: long-term mass balance and loading analysis. J. Great Lake. Res. 35, 272–284. Chapra, S.C., Dove, A., Warren, G.J., 2012. Long-term trends in Great Lakes major ion chemistry. J. Great Lake. Res. 38, 550–560. Corsi, S.R., De Cicco, L.A., Lutz, M.A., Hirsch, R.M., 2015. Chloride trends in snow-affected urban watersheds: increasing concentrations outpace urban growth rate and are common among all seasons. Sci. Total Environ. 508, 488–497. Davison, W., 1993. Iron and manganese in lakes. Earth Sci. Rev. 34, 119–163. Dailey, K.R., Welch, K.A., Lyons, W.B., 2014. Evaluating the influence of road salt on water quality of Ohio rivers over time. Appl. Geochem. 47, 25–35. Dugan, H.A., Bartlett, S.L., Burke, S.M., Doubek, J.P., Krivak-Tetley, F.E., Skaff, N.K., Summers, J.C., Farrell, K., McCullough, I., Morales-Williams, A.M., Roberts, D.C., Ouyang, Z., Scordo, F., Hanson, P.C., Weathers, K.C., 2017. Salting our freshwater lakes. Proc. Natl. Acad. Sci. Unit. States Am. 114, 4453–4458. Dupuis, D., 2017. The Influence of Road Slat on Seasonal Mixing and Redox Stratification in Three Southwest Michigan Kettle Lakes. Western Michigan University (Master's Thesis). Engemann, J.G., 1977. Observations on Asylum Lake. Western Michigan University. Department of Biological Sciences, unpublished report, Kalamazoo, Michigan Retrieved from: http://www.wmich.edu/asylumlake/pastresearch. Environment Canada, 2001. Priority Substance List Assessment Report: Road Salts. 170 pp. Judd, J.H., 1970. Lake stratification caused by runoff from street deicing. Water Res. 4, 521–532. Kaushal, S.S., Groffman, P.M., Likens, G.E., Belt, K.T., Stack, W.P., Kelly, V.R., Band, L.E., Fisher, G.T., 2005. Increased salinization of fresh water in the northeastern United States. Proc. Natl. Acad. Sci. U. S. A. 102, 13517–13520. Kaushal, S.S., McDowell, W.H., Wollheim, W.M., 2014. Tracking evolution of urban biogeochemical cycles: past, present and future. Biogeochemistry 121, 1–21. Kehew, A., 2001. Applied Chemical Hydrogeology, first ed. Prentice Hall. Kelly, W.R., Panno, S.V., Hackley, K.C., 2012. Impacts of road salt runoff on water quality of the Chicago, Illinois, region. Environ. Eng. Geosci. XVIII (No. 1), 65–81. Kieser, Associates, 2008. Water Quality Evaluation of Asylum Lake and Little Asylum Lake with Management Recommendations. Unpublished Report for Western Michigan University Asylum Lake Policy and Management Council. Retrieved from: https://wmich.edu/asylumlake/pastresearch. Koretsky, C.M., Haas, J.R., Miller, D., Ndenga, N.T., 2006. Seasonal variations in pore water and sediment geochemistry of littoral lake sediments (Asylum Lake, MI, USA). Geochem. Trans. 7.
272
Contents lists available at ScienceDirect
Applied Geochemistry journal homepage: www.elsevier.com/locate/apgeochem
Effects of road salt deicers on an urban groundwater-fed kettle lake 1
D. Allie Wyman , Carla M. Koretsky
T
∗
Department of Geosciences, Western Michigan University, Kalamazoo, MI, 49008, USA
A R T I C L E I N F O
A B S T R A C T
Handling Editor: Prof. M. Kersten
Road salt deicers significantly influence the chemistry and physical mixing of urban lakes, even causing transition from dimixis to monomixis or meromixis. In this study, the water column geochemistry of Asylum Lake, a primarily groundwater-fed, eutrophic kettle lake in urban Kalamazoo, MI, was monitored for over a year to determine the extent of road salt deicer influence on the lake chemistry and the physical mixing of the lake. Water column samples from the surface to the deepest part of Asylum Lake were analyzed monthly for nineteen months for a suite of parameters including dissolved oxygen, pH, conductivity, temperature, alkalinity, Fe2+, Mn2+, orthophosphate, total NH4+, alkalinity, total sulfide, Cl−, SO42−, Ca2+, Mg2+, K+, and Na+. During the study period, spring mixing was never observed and a nearly complete fall turnover was observed only in November 2013. The hypolimnion of Asylum Lake was always hypoxic or anoxic and redox-stratified, with seasonal development of suboxic and sulfidic zones, which were disrupted in fall and winter following partial fall turnover and subsequent ice cover. This study suggests that road salt deicers have caused Asylum Lake to transition from dimixis to meromixis or periodic monomixis with significant consequences for biogeochemical cycles in the lake waters.
Keywords: Eutrophication Road salt Anoxia Redox stratification Mixing
1. Introduction The widespread application of road salt deicers has the potential to significantly influence the salinity of freshwater systems. In the United States, an estimated 8 Mt of salt was used for deicing in 1975, increasing to 55 Mt in 2003 (U.S. Geological Survey, 2014). Rising chloride levels related to the use of road salts have been documented in groundwater and surface waters in the United States, Canada and elsewhere (e.g. Environment Canada, 2001; Chapra et al., 2009, 2012; Mullaney et al., 2009; Kaushal et al., 2014; Rogora et al., 2015; Dugan et al., 2017). For example, Novotny et al. (2008) showed that urban lakes in Minnesota have 10 to 25 times greater sodium and chloride levels as compared to rural Minnesota lakes, with chloride concentrations in the urban lakes increasing at an annual rate of 1.8%. Other studies have demonstrated increasing chloride levels in streams and rivers (e.g. Kaushal et al., 2005; Kelly et al., 2012; Dailey et al., 2014; Corsi et al., 2015), the Great Lakes (Chapra et al., 2009, 2012), as well as in lakes of New Hampshire (Likens and Buso, 2010), Michigan (Koretsky et al., 2012; Sibert et al., 2015), Italy (Rogora et al., 2015), Canada (Environment Canada, 2001), and Sweden (Thunqvist, 2004) due to influx of road salt deicers. With continued use of road salt deicers, chloride concentrations in urban freshwater systems are likely to continue increasing.
∗
1
Levels of chloride in many surface water systems now periodically or even permanently exceed the chronic toxicity threshold for freshwater species of 250 mg/L (e.g., Environment Canada, 2001; Kaushal et al., 2005; Mullaney et al., 2009; Corsi et al., 2015; Dugan et al., 2017). Increasing chloride levels can also affect the density structure, and thus potentially change the mixing dynamics, of freshwater lakes. For example, Judd (1970) found that wind action was insufficient to cause seasonal turnover of First Sister Lake in Michigan because of the increased salinity, and thus increased density of the lake bottom waters, created by the influx of road salts. Similarly, Novotny et al. (2008) observed that road salt contamination caused two urban lakes in Minnesota to transition from dimixis (twice per year mixing) to monomixis (once per year mixing), and Sibert et al. (2015) demonstrated that road salts caused an urban kettle lake in Michigan to transition to meromixis (no annual mixing). Both Novotny et al. (2008) and Sibert et al. (2015) reported hypolimnetic anoxia in these lakes, in part due to their failure to turnover. Eutrophic lakes have high levels of nutrients and labile organic matter. In these lakes, the water column may become redox stratified as labile organic matter remaining after oxygen depletion is oxidized via a sequence of increasingly less energetically favorable terminal electron acceptors including nitrate, Mn4+, Fe3+ and SO4−2. This results in the production and accumulation of reduced solutes including NH4+,
Corresponding author. 1903 West Michigan Ave, Kalamazoo, MI, 49008, USA. E-mail address: [email protected] (C.M. Koretsky). Present address: Geology Department, University of Illinois, Champaign, IL, 61,821.
https://doi.org/10.1016/j.apgeochem.2017.12.023 Received 10 August 2017; Received in revised form 21 December 2017; Accepted 22 December 2017 Available online 27 December 2017 0883-2927/ © 2017 Published by Elsevier Ltd.
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Fig. 1. Bathymetric map of Asylum Lake with sampling location noted by star. Map was created in 1991 by William Sauck (Institute for Water Sciences), William Laton (Department of Geology), and Bryan Allen (Department of Geology) using a Lowrance Model X-16 with 192 kHz transducer along numbered transects. Contour interval is 0.61 m.
Mn2+, Fe2+, H2S and CH4 in the hypolimnion (Boehrer and Schultze, 2008; Koretsky et al., 2012; Sibert et al., 2015; Dupuis, 2017). If a eutrophic lake fails to turnover due to road salt contamination, hypolimnetic anoxia and redox-stratification may become permanent. Whether a lake transitions to meromixis due to road salt influx likely depends on the concentration of salt, the residence time of water in the lake, and the lake depth and surface area (Rodrigo et al., 2001). In this study, the water column chemistry of a primarily groundwater-fed, eutrophic kettle lake located in urban Kalamazoo, MI was studied over a period of nineteen months to assess the influence of road salt contamination on seasonal turnover and the development of redox stratification in the lake.
0.5% bare soil. 3. Materials and methods Water column sampling was conducted at approximately the deepest point in Asylum Lake (Fig. 1) at 0.5 m intervals for in situ measurements of dissolved oxygen, pH, conductivity, and temperature and at 1 m intervals for ex situ samples. Water samples were collected approximately once a month from September 2012 to March 2014, with the exception of the period between mid-December 2012 and March 2013. During this time, ice thickness was such that it was unsafe to traverse the lake and no samples were collected. During periods of potential turnover, in situ samples were collected more frequently. Water samples were transported back to Western Michigan University and analyzed for Fe2+, Mn2+, orthophosphate, total NH4+, alkalinity, total sulfide (ΣH2S), Cl−, SO42−, Ca2+, Mg2+, K+, and Na+ according to the methods described in Sibert et al. (2015).
2. Study site Asylum Lake is an urban kettle lake located in Kalamazoo, MI, with a surface area of approximately 19.8 ha, a mean depth of 7.2 m, and a maximum depth of 15.8 m (Sauck and Barcelona, 1992; Kieser and Associates, 2008; Koretsky et al., 2012, Fig. 1). Water enters the lake primarily through groundwater discharge on the west side of the lake and exits the lake through groundwater recharge to the east, as well as via a small culvert leading to Little Asylum Lake. The residence time of water in the lake is estimated to be 0.68 years (Kieser and Associates, 2008). Asylum Lake is bordered by wetlands to the west, prairie and woods to the south, and by residential neighborhoods to the north. The soils around the lake are primarily Houghton and Sebewa soils characterized by excess humus and wetness (USDA, 1979). South of Asylum Lake, the soils are Kalamazoo loam with a 6–12% slope; north and west of Asylum Lake, the soils are Urban Oshtemo complex with slopes of 12–25%; southeast and southwest of Asylum Lake, soils are Oshtemo sandy loam with 18–35% slope and 12–18% slope, respectively (USDA, 1979). Kieser and Associates (2008) completed a topographic assessment of the surrounding watershed and concluded that it is comprised of ∼34% forest or open herbaceous land; 30% farmland; 18% high/low density urban; 10% roads/parking lots; 4% wetlands; 3% water and
4. Results 4.1. Temperature, dissolved oxygen, conductivity, and pH Asylum Lake temperature profiles show strong thermal stratification in summer and early fall, weak stratification in late fall and early spring, and reverse stratification during all periods of ice cover. In December 2012, April 2013, November 2013, and April 2014 the lake is isothermal, indicative of potential turnover events (Fig. 2A). The epilimnion is supersaturated with respect to dissolved oxygen in all seasons except during periods of ice cover (Fig. 2B). In the summer, the hypolimnion is hypoxic to anoxic below 6–9 m depth. In fall and spring, oxic conditions persist to depths of between 10 m and 13 m, except during isothermal conditions in late November 2013, when dissolved oxygen increases slightly with depth to 14 m and then decreases at 14.5 m (Fig. 2B). In all seasons, specific conductivity in Asylum Lake increases with 266
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
the lake is nearing isothermal conditions, the pH varies less with depth. 4.2. Major ions Alkalinity in Asylum Lake increases with depth in all seasons, ranging from approximately 250 to 300 ppm in the epilimnion and from approximately 350 to 420 ppm in the hypolimnion (Fig. S2). During all ice-free periods, alkalinity variations from month to month are small. However, in February 2013, during maximum ice cover, alkalinity concentrations are elevated throughout the entire water column, ranging from 420 ppm in the epilimnion to 510 ppm in the hypolimnion. Concentrations of dissolved calcium increase with depth in all seasons with concentrations ranging from ∼30 to 40 ppm at the surface to ∼60–80 ppm at depth (Fig. S3). The only period during which dissolved calcium does not increase with depth is December 2012, when concentrations remain steady between 60 and 65 ppm. Dissolved magnesium and potassium concentrations increase only slightly with depth during all seasons with dissolved magnesium concentrations typically ∼25–30 ppm and dissolved potassium concentrations ∼1–5 ppm (Figs. S4 and S5). 4.3. Redox sensitive solutes and nutrients In summer and fall, total ammonium and orthophosphate accumulate in the water column between 10 and 12 m depth, coincident with the depletion of dissolved oxygen (Fig. 3A and B). During potential turnover in December 2012, small but detectable levels of total ammonium are also present in the epilimnion. During periods of ice cover, orthophosphate and total ammonium reach concentrations as high as 50 ppb orthophosphate and 200 ppb total ammonium in the epilimnion. Dissolved Mn2+ is always present at levels of ≥400 ppb below 11 m depth, accumulating in spring and winter at depths where dissolved oxygen is depleted (Fig. 4A). In November 2012, December 2012, and August 2013, Mn2+ is detectable throughout the lake water column. Fe2+ concentrations remain below detection limits of 0.2 ppm during all sampling periods. Dissolved ΣH2S accumulates in the hypolimnion during summer and fall, reaching concentrations approaching 4 ppm (Fig. 4B). During early spring and periods of ice cover, ΣH2S concentrations remain below detection limits throughout the water column. During summer and fall months, sulfate concentrations increase slightly between the surface and 12 m depth and then decrease with depth, often coinciding with increasing dissolved ΣH2S concentrations (Fig. 4C). 4.4. Chloride and sodium December 2012 chloride concentrations are fairly uniform (∼165 ppm) with depth (Fig. 5A). After snowmelt in 2013, chloride concentrations steadily increase with depth in both the epilimnion and hypolimnion, peaking at > 330 ppm at 14 m depth in September 2013. Chloride concentrations decrease slightly throughout the lake from September until March 2014. For all sampling dates in 2012, sodium levels in Asylum Lake remain fairly uniform with depth at ∼70–85 ppm (Fig. 5B). After snowmelt in 2013, sodium concentrations increase in the water column to a maximum of ∼130 ppm in the hypolimnion. During periods of ice cover (January through March 2014), sodium concentrations in the epilimnion remain stable whereas concentrations in the hypolimnion rise to ∼150 ppm.
Fig. 2. Depth-time isoplots of (A) temperature (°C), (B) dissolved oxygen (mg·L−1), and (C) specific conductance (mS·cm−1) at Asylum Lake. Thick rectangles at 0 m depth indicate approximate periods of ice cover.
depth (Fig. 2C). A distinct increase in conductivity in both the epilimnion and hypolimnion is apparent after snowmelt in 2013 and 2014. The maximum vertical differential in conductivity increases from 0.126 in 2012 to 0.443 mS cm−1 in 2014. During isothermal conditions in late November 2013, conductivity was uniform in the upper 14 m with a sharp increase at 14.5 m depth. The pH in Asylum Lake decreases with depth in all seasons (Fig. S1). Typically, pH in the epilimnion ranges from 8.0 to 8.7 and hypolimnion pH ranges from 7.1 to 7.3. Epilimnion and hypolimnion pH values increase during periods of ice cover. During fall and spring periods, when
5. Discussion 5.1. Salinization In an unpublished report, Engemann (1977) noted chloride concentrations of up to 126 ppm in Asylum Lake. Since this time, elevated 267
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Fig. 3. Depth-time isoplots of (A) total ammonium (ppb) and (B) orthophosphate (ppb) at Asylum Lake. Thick rectangles at 0 m depth indicate approximate periods of ice cover.
(Koretsky et al., 2012; Dupuis, 2017). Chloride concentrations were compared to conductivity in Asylum Lake and found to consistently correlate, with a coefficient of determination (R2) of 0.8 or greater in May and June 2013 and at least 0.9 from July 2013 through the end of the study (Table 1). This is comparable to correlations between chloride and conductivity observed by Sibert et al. (2015) and Novotny et al. (2008) in other urban lakes impacted by road salt deicers. Sodium concentrations were also compared to conductivity. However, the R2 for these correlations is less consistent, generally ranging from 0.76 to 0.97, with a few lower values. This is likely because sodium is a less conservative element than chloride due to ion exchange reactions between sodium and clay minerals or organic matter in the soils of the groundwater aquifer that discharges into Asylum Lake (Kehew, 2001). The strong correlations between sodium and conductivity and especially between chloride and conductivity suggest that the dissolved ion concentrations in Asylum Lake are dominated by road salt input.
chloride concentrations at Asylum Lake have been documented in three additional studies spanning three decades. Buening (1994) measured maximum chloride concentrations in Asylum Lake of 140 ppm, which had increased to 175 ppm in 2006–07 (Kieser and Associates, 2008) and to 225 ppm another 15 years later (Koretsky et al., 2012). In the present study, we found maximum chloride concentrations of > 330 ppm, over 2.5 times the amount of chloride reported by Engemann (1977). Elevated chloride levels are also found in wells near Asylum Lake. Wyman (2014) sampled a well screened at 20.7–22.2 m depth, located to the southwest of Asylum Lake and just to the east of a heavily salted interstate and high use local road. Chloride concentrations in the well ranged from 105 to 153 ppm, suggesting that significant quantities of road salt enter the lake via groundwater, in addition to the salt that enters via overland flow. In contrast to the high chloride concentrations that have accumulated in Asylum Lake, recently reported chloride concentrations in two nearby rural kettle lakes, Brewster Lake and North Lake, are consistently < 15 ppm throughout the water column 268
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Asylum Lake (Kieser and Associates, 2008), and thus, chloride levels may fluctuate significantly from year to year, depending on the severity of winter conditions and the resulting amount of road salt use. 5.2. Lake mixing Dimictic lakes turnover in the fall when surface waters cool and become denser and again in spring after ice out (Wetzel, 2001). However, it has been demonstrated that lakes influenced by increased road salt may transition to monomixis or meromixis due to changes in the density structure of the water column (Judd, 1970; Novotny et al., 2008; Novotny and Stefan, 2012; Sibert et al., 2015). Isothermal conditions were observed in Asylum Lake during four sampling events in this study: December 2012, April 2013, November 2013, and April 2014 (Fig. 2A). If water column mixing occurs, the water column should be isochemical as well as isothermal. However, nearly isochemical conditions were only apparent during one of the four isothermal sampling events. A full sampling of the lake was completed on November 16, 2013; on this date, the lake was still chemically and thermally stratified. One week later, on November 24, 2013, only in situ data was collected. On this date, the lake was nearly isothermal and isochemical to a depth of 14 m. At 14.5 m, there was a large decrease in dissolved oxygen and pH, and a large increase in conductivity. Turnover may have occurred just before or after this date, or it is possible that the lake did not entirely turnover at the deepest point in the lake in fall 2013. It is also important to note that turnover probably did occur at shallower locations. Our data implies that the lake has transitioned to monomixis or perhaps even periodic meromixis, although this cannot be determined definitively without continuous year-round monitoring of the lake. The delayed or absent mixing is likely due to increased density at depth due to the elevated levels of chloride from application of road salt to nearby roads (Fig. 5A). 5.3. Redox stratification Redox stratification can develop in lake water columns as a result of oxidation-reduction reactions that occur as terminal electron acceptors are used in sequence to oxidize organic matter (Stumm and Morgan, 1996; Wetzel, 2001). Terminal electron accepters are used approximately in order of thermodynamic yield; if a terminal electron acceptor is exhausted by reaction with organic matter or other reducing agents and not replenished through mixing or other chemical reactions, then organic matter oxidation will proceed using the next available terminal electron acceptor that will produce the greatest energetic yield. Because of the high quantities of labile organic matter in eutrophic lakes, redox stratification can be especially pronounced, particularly in the absence of significant wind mixing or influx of oxic waters (Stumm and Morgan, 1996; Wetzel, 2001). In Asylum Lake, redox stratification can be delineated from dissolved oxygen, Mn2+ and ΣH2S profiles during all seasons (Figs. 2B, 4A and 4B). In April 2013, the oxic zone stretches from the surface to 12 m depth, beneath which hypoxic conditions are present and Mn2+ begins to accumulate. As organic matter productivity increases in spring and summer 2013, subsurface peaks in dissolved oxygen are observed as photosynthetic plants produce oxygen in the photic zone. pH peaks at depths 0.5–1 m deeper are further evidence of photosynthetic production of dissolved oxygen because photosynthesis consumes carbon dioxide, raising the pH (Fig. S1). With increasing organic matter productivity and temperature, dissolved oxygen is depleted at much shallower depths. Hypoxic conditions are present at depths as shallow as 6–7.5 m from August to October 2013 and Mn2+ concentrations are detectable at depths as shallow as 1 m in August 2013 (Fig. 4A). Dissolved ΣH2S is observed at depths below those where Mn2+ first appears, with detectable concentrations noted at and below 13 m depth in July 2013 and persistent concentrations of ∼1–3.5 ppm between 12 and 13 m depth throughout
Fig. 4. Depth-time isoplots of (A) manganese (ppb), (B) total sulfide (ppb), and (C) sulfate (ppm) at Asylum Lake. Thick rectangles at 0 m depth indicate approximate periods of ice cover.
Kalamazoo climate data from October 2011 through April 2014 shows an increase in the number of days below freezing and an increase in the amount of snowfall between 2011–2012 and 2013–2014 (Table 2). In 2011–2012, there were only 20 days below freezing, compared to 78 in 2013–2014. Total snowfall accumulation ranged from 1446 mm in 2011–2012 to 2877 in 2013–2014. Due to the mild 2011–2012 winter season, less deicer was applied to roads compared to the subsequent winter seasons (Table 2). This may explain the poorer correlations of chloride with conductivity in 2012 as compared to 2013 and 2014 (Table 1). Prior work suggests a residence time of < 1 year in
269
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Fig. 5. Depth-time isoplots of (A) chloride (ppm) and (B) sodium (ppm) at Asylum Lake. Thick rectangles at 0 m depth indicate approximate periods of ice cover.
Fe2+ in Asylum Lake is due to a lower delivery of fresh particulate iron than in Woods Lake, which likely has greater influx of particulates from stormwater sewers and steeper slopes surrounding the lake (Sibert et al., 2015). We hypothesize that any Fe2+ produced by reductive dissolution of iron oxides in the lake water or sediment columns of Asylum Lake reacts with sulfide and is removed from the water column before it can accumulate to levels above detection limits (Davison, 1993; Postma and Jakobsen, 1996; Stumm and Morgan, 1996; Wetzel, 2001; Rickard and Luther, 2007). Phosphate only occurs in a single oxidation state but it is influenced by changes in redox conditions because it binds readily to iron oxide minerals and is released into solution as they are reductively dissolved (Wetzel, 2001). Therefore, increases in phosphate concentrations are commonly coincident with increases in dissolved Fe2+ concentrations in redox-stratified systems. However, phosphate can also be released from decaying organic material in the hypolimnion, causing an increase in phosphate without a corresponding increase in dissolved Fe2+.
the remainder of summer 2013 (Fig. 4B). In fall 2013, cooling of water in the epilimnion results in at least some water column mixing, and oxic conditions to depths of at least 12 m (Fig. 2A and B). During this period Mn2+ and ΣH2S concentrations decrease slightly but continue to persist at depths below 10 and 12 m, respectively. After January 2014, ΣH2S concentrations are below detection limits at all depths and remain so throughout the winter. The water column is hypoxic from 12 to 14 m depth throughout the winter, with persistent accumulation of Mn2+ at concentrations between ∼400 and 1800 ppb. No detectable Fe2+ was observed in the water column during any sampling period in this study. This is interesting, particularly given the high levels of dissolved Fe2+ in the hypolimnion of nearby Woods Lake, which is also impacted by road salt deicer (Sibert et al., 2015). In addition, dissolved Fe2+ concentrations in excess of 20 ppm have been reported in the sediment pore waters of shallow sediments fringing Asylum Lake (Koretsky et al., 2006). Asylum Lake is primarily groundwater fed, suggesting that the lack of accumulation of dissolved 270
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Mn2+ or sulfide. While more work must be done to fully ascertain the origin of phosphate in Asylum Lake, this suggests that phosphate is also released by organic matter decay. Ammonium is an energetically favorable source of nitrogen for aquatic plants and is rapidly assimilated when present in the photic zone. Additionally, ammonium oxidizes quickly and rarely accumulates in oxic waters. Therefore, ammonium only accumulates in lakes when organic material begins to decay in anoxic hypolimnia (Wetzel, 2001). In Asylum Lake, the presence of ammonium was observed in all seasons in the anoxic hypolimnion (Fig. 3A). It was only observed at levels above detection limits at shallower depths during December 2012, January 2014, and February 2014. During these winter months, assimilation of ammonium by aquatic plants was likely slowed due to colder temperatures and reduced sunlight. Also, dissolved oxygen saturation (Fig. 1) was reduced to below 60% in the epilimnion during these months. These combined effects likely allowed ammonium to persist at shallower depths in the water column.
Table 1 Coefficient of determination (R2) between chloride and conductivity and sodium and conductivity for each month of sampling in Asylum Lake. Date
Cl and Conductivity
Na and Conductivity
9/14/12 9/29/12 11/18/12 12/8/12 4/6/13 5/31/2013 6/3/2013 6/27/13 7/25/13 8/27/13 9/26/13 10/22/13 11/16/13 1/10/14 2/8/14 3/21/14 4/8/14
No data No data No data 0.66 0.45 0.80 0.80 0.88 0.94 0.95 0.97 0.95 0.89 0.94 0.98 0.96 0.97
0.13 0.17 0.11 0.00 0.93 No data No data 0.46 0.76 0.006 0.87 0.85 0.83 0.51 0.95 0.97 0.97
6. Conclusions Throughout the nineteen-month study period, Asylum Lake chloride concentrations in the hypolimnion exceed background levels by approximately two orders of magnitude. Chloride and sodium strongly correlate with conductivity after the 2013 spring snowmelt, indicating that dissolved ion concentrations in the water column are dominated by sodium and chloride influxes from road salt deicers. Lakes in the Midwest are typically dimictic (Wetzel, 2001); however, the high concentrations of chloride in Asylum Lake and corresponding increases in the density of hypolimnetic waters appears to have resulted in transition of the lake from dimixis to monomixis and perhaps even periodic meromixis. The lack of consistent fall and spring turnover leads to persistent anoxia in the hypolimnion, with accumulation of Mn2+, ΣH2S, and NH4+, which undoubtedly impacts the aquatic ecosystem. The lack of Fe2+ accumulation in the hypolimnion is particularly interesting, especially given the high concentrations observed in a similar nearby kettle lake that has also been impacted by road salt, and warrants further study. Methane concentrations were not measured in this study. However, the compressed redox stratification suggests that significant levels of methane may also accumulate in the hypolimnion of eutrophic lakes impacted by road salt deciers. This could result in an elevated efflux of methane from the lake column to the atmosphere and should also be studied further. This study clearly demonstrates that the anthropogenic input of deicers to freshwater kettle lakes changes mixing dynamics and biogeochemical cycles, and may exacerbate existing eutrophication issues.
Table 2 Climate data for the city of Kalamazoo from the NOAA climate data center (NESDIS/ NOAA/NCEI). Number of days with maximum temperature below 0°C and average monthly temperature were collected at the Kalamazoo Battle Creek International Airport. Total snowfall accumulation was collected at Western Michigan University. The amount of road salt applied per winter season was obtained through personal communication with a city of Kalamazoo employee. Date
Number of Days with Maximum Temperature Below 0 °C
2011–2012 Oct-2011 Nov-2011 Dec-2011 Jan-2012 Feb-2012 Mar-2012 Apr-2012 Total 2012–2013 Oct-2012 Nov-2012 Dec-2012 Jan-2013 Feb-2013 Mar-2013 Apr-2013 Total 2013–2014 Oct-2013 Nov-2013 Dec-2013 Jan-2014 Feb-2014 Mar-2014 Apr-2014 Total
Winter Season 0 0 3 11 4 2 0 20 Winter Season 0 1 5 14 14 7 0 41 Winter Season 0 5 18 21 23 11 0 78
Average Monthly Temperature (°C)
Total Snowfall Accumulation (mm)
Road Salt Applied (tons)
11.3 6.8 1.4 −1.7 −0.1 10.5 9.1 NA
0 0 106 828 398 114 0 1446
NA NA NA NA NA NA NA 4438
9.9 4.1 1.7 −2.4 −3.5 0 7.8 NA
0 0 146 519 776 258 18 1717
NA NA NA NA NA NA NA 6212
11.5 3.2 −3 −8.3 −8.3 −2.5 8.9 NA
0 67 781 1091 638 300 0 2877
NA NA NA NA NA NA NA 10,132
Acknowledgements We deeply appreciate field and laboratory assistance from Ryan Sibert, Tom Howe, Ann Gilchrist, Denisha Griffey, Rebecca Kiekhaefer, and Derrick Lingle. William Sauck is acknowledged for assisting in the creation of Fig. 1. The WMU Graduate College is also acknowledged for providing a research grant to Allie Wyman, which helped to support this work. Steve Keto and the Asylum Lake Preservation Council are acknowledged for providing access to Asylum Lake.
In Asylum Lake, phosphate is present in the water column while dissolved Fe2+ remains below detection limits in all seasons (Fig. 3B). Except during September 2013 and January through April 2014, we observed increasing phosphate concentrations at the same depth or one meter below the first detection of dissolved Mn2+ and above the first detection of ΣH2S (Figs. 3B, 4A and 4B) These trends are consistent with Fe(III) reduction occurring just below Mn(IV) reduction, releasing both dissolved PO43− and Fe2+, but with the Fe2+ immediately removed via reactions with sulfide. In contrast, during September 2013 and January through April 2014 phosphate accumulates at shallower depths than
Appendix A. Supplementary data Supplementary data related to this article can be found at http://dx. doi.org/10.1016/j.apgeochem.2017.12.023. REFERENCES Boehrer, B., Schultze, M., 2008. Stratification of lakes. Rev. Geophys. 46, RG2005. Buening, Vincent E., 1994. Investigation of Glacial-Drift Aquifer Interactions Using Hydrogeology, Geochemistry, and Stable Isotopes at Asylum Lake Michigan (Master's
271
Applied Geochemistry 89 (2018) 265–272
D.A. Wyman, C.M. Koretsky
Koretsky, C.M., MacLeod, A., Sibert, R.J., Snyder, C., 2012. Redox stratification and salinization of three kettle lakes in southwest Michigan, USA. Water Air Soil Pollut. 223, 1415–1427. Likens, G.E., Buso, D.C., 2010. Salinization of mirror lake by road salt. Water Air Soil Pollut. 205, 205–214. Mullaney, J.R., Lorenz, D.L., Arntson, A.D., 2009. Chloride in Groundwater and Surface Water in Areas Underlain by the Glacial Aquifer System, Northern United States: U.S. Geological Survey Scientific Investigations Report 2009-5086. 41p. Novotny, E.V., Murphy, D., Stefan, H.G., 2008. Increase of urban lake salinity by road deicing salt. Sci. Total Environ. 406, 131–144. Novotny, E.V., Stefan, H.G., 2012. Road salt impact on lake stratification and water quality. J. Hydraul. Eng. 138, 1069–1080. Postma, D., Jakobsen, R., 1996. Redox zonation: equilibrium constraints on the Fe(III)/ SO4- reduction interface. Geochem. Cosmochim. Acta 60, 3169–3175. Rickard, D., Luther, G.W., 2007. Chemistry of iron sulfides. Chem. Rev. 107, 514–562. Rodrigo, M.A., Miracle, M.R., Vicente, E., 2001. The meromictic Lake La Cruz (Central Spain). Patterns of stratification. Aquat. Sci. 63, 406–416. Rogora, M., Mosello, R., Kamburska, L., Salmaso, N., Cerasino, L., Leoni, B., Garibaldi, L., Soler, V., Lepori, L., Columbo, L., Buzzi, F., 2015. Recent trends in chloride and sodium concentrations in the deep subalpine lakes (Northern Italy). Environ. Sci. Pollut. Res. 22, 19013–19026. Sauck, W.A., Barcelona, M.J., 1992. Final Project Report: Long-Term Hydrogeological Research and Education Test Site. Western Michigan University Unpublished. Retrieved from: https://wmich.edu/asylumlake/pastresearch. Sibert, R.J., Koretsky, C.M., Wyman, D.A., 2015. Cultural meromixis: effects of road salt on the chemical stratification of an urban kettle lake. Chem. Geol. 395, 126–137. Stumm, M., Morgan, J.J., 1996. Aquatic Chemistry. John Wiley and Sons, New York, NY. Thunqvist, E.-L., 2004. Regional increase of mean chloride concentration in water due to the application of deicing salt. Sci. Total Environ. 325, 29–37. United States Department of Agriculture, 1979. Soil Conservation Service. Soil Survey of Kalamazoo County, Michigan. National Cooperative Soil Survey. U.S. Geological Survey, 2014. Salt statistics. In: Kelly, T.D., Matos, G.R. (Eds.), Comps. Historical Statistics for Mineral and Material Commodities in the United States, U.S. Geological Survey Data Series 140, available online at: http://pubs.usgs.gov/ds/ 2005/140/, Accessed date: 17 July 2015. Wetzel, R.G., 2001. Limnology: Lake and River Ecosystems, third ed. Academic, San Diego. Wyman, D.A., 2014. Effects of Road Salt on Asylum Lake Geochemistry (Master's Thesis). Western Michigan University.
Thesis). Western Michigan University. Chapra, S.C., Dove, A., Rockwell, D.C., 2009. Great Lakes chloride trends: long-term mass balance and loading analysis. J. Great Lake. Res. 35, 272–284. Chapra, S.C., Dove, A., Warren, G.J., 2012. Long-term trends in Great Lakes major ion chemistry. J. Great Lake. Res. 38, 550–560. Corsi, S.R., De Cicco, L.A., Lutz, M.A., Hirsch, R.M., 2015. Chloride trends in snow-affected urban watersheds: increasing concentrations outpace urban growth rate and are common among all seasons. Sci. Total Environ. 508, 488–497. Davison, W., 1993. Iron and manganese in lakes. Earth Sci. Rev. 34, 119–163. Dailey, K.R., Welch, K.A., Lyons, W.B., 2014. Evaluating the influence of road salt on water quality of Ohio rivers over time. Appl. Geochem. 47, 25–35. Dugan, H.A., Bartlett, S.L., Burke, S.M., Doubek, J.P., Krivak-Tetley, F.E., Skaff, N.K., Summers, J.C., Farrell, K., McCullough, I., Morales-Williams, A.M., Roberts, D.C., Ouyang, Z., Scordo, F., Hanson, P.C., Weathers, K.C., 2017. Salting our freshwater lakes. Proc. Natl. Acad. Sci. Unit. States Am. 114, 4453–4458. Dupuis, D., 2017. The Influence of Road Slat on Seasonal Mixing and Redox Stratification in Three Southwest Michigan Kettle Lakes. Western Michigan University (Master's Thesis). Engemann, J.G., 1977. Observations on Asylum Lake. Western Michigan University. Department of Biological Sciences, unpublished report, Kalamazoo, Michigan Retrieved from: http://www.wmich.edu/asylumlake/pastresearch. Environment Canada, 2001. Priority Substance List Assessment Report: Road Salts. 170 pp. Judd, J.H., 1970. Lake stratification caused by runoff from street deicing. Water Res. 4, 521–532. Kaushal, S.S., Groffman, P.M., Likens, G.E., Belt, K.T., Stack, W.P., Kelly, V.R., Band, L.E., Fisher, G.T., 2005. Increased salinization of fresh water in the northeastern United States. Proc. Natl. Acad. Sci. U. S. A. 102, 13517–13520. Kaushal, S.S., McDowell, W.H., Wollheim, W.M., 2014. Tracking evolution of urban biogeochemical cycles: past, present and future. Biogeochemistry 121, 1–21. Kehew, A., 2001. Applied Chemical Hydrogeology, first ed. Prentice Hall. Kelly, W.R., Panno, S.V., Hackley, K.C., 2012. Impacts of road salt runoff on water quality of the Chicago, Illinois, region. Environ. Eng. Geosci. XVIII (No. 1), 65–81. Kieser, Associates, 2008. Water Quality Evaluation of Asylum Lake and Little Asylum Lake with Management Recommendations. Unpublished Report for Western Michigan University Asylum Lake Policy and Management Council. Retrieved from: https://wmich.edu/asylumlake/pastresearch. Koretsky, C.M., Haas, J.R., Miller, D., Ndenga, N.T., 2006. Seasonal variations in pore water and sediment geochemistry of littoral lake sediments (Asylum Lake, MI, USA). Geochem. Trans. 7.
272