Start Submission

Reading: Mineralogical and geochemical variation in stream sediments impacted by acid mine drainage i...


A- A+
Alt. Display

Research Article

Mineralogical and geochemical variation in stream sediments impacted by acid mine drainage is related to hydro-geomorphic setting


David M. Singer ,

Department of Geology, Kent State University, Kent, OH, US
X close

Anne J. Jefferson,

Department of Geology, Kent State University, Kent, OH, US
X close

Eric L. Traub,

Department of Geology, Kent State University, Kent, OH; and Wood PLC., 2000 S. Colorado Blvd., Denver, CO, US
X close

Nicolas Perdrial

Department of Geology, University of Vermont, Burlington, VT, US
X close


Acid mine drainage (AMD) discharge has severe, long lasting impacts on water quality and stream ecology in affected watersheds due in part to the dynamic relationship between toxic metals (e.g. Al, Mn, and Cu) and Fe(III) oxy-hydroxides. Localized areas of biogeochemical activity that can mediate mineralogical transformation changes and cause metal release are potentially linked to stream geomorphology. This relationship has not been previously considered with respect to potential longitudinal variation within an impacted stream. The current work aims to determine how Fe(III) (oxy)-hydroxide speciation and distribution, and pore water chemistry in an AMD-impacted streambed, are affected by the presence of two geomorphic structures (a debris dam and step-pool sequence) in an Ohio watershed impacted by historical coal mining. In terms of solid phase mineralogy and geochemistry, in both the tributary and main stem, goethite was the dominant Fe-bearing phase throughout the AMD deposit depth in cores taken upstream of the geomorphic structures, whereas poorly-crystalline phases dominated downstream of the structures, despite the presence of Fe in the reducible fraction. The concentrations and distribution of extractable Al, Mn, and Cu were also different upstream versus downstream of each structure. Pore water Fe and Mn concentrations were higher downstream of the structures than upstream. Strong downward hydraulic gradients were present above the debris dam and in step-pool 1, whereas weaker upward hydraulic gradients were present below the debris dam and in step-pool 2. This work highlights that AMD deposit speciation and distribution, and pore water chemistry, are not spatially uniform within stream reaches, potentially as a result of groundwater-stream exchange-facilitated interactions in the presence of AMD-derived materials.

Knowledge Domain: Earth & Environmental Science
How to Cite: Singer, D.M., Jefferson, A.J., Traub, E.L. and Perdrial, N., 2018. Mineralogical and geochemical variation in stream sediments impacted by acid mine drainage is related to hydro-geomorphic setting. Elem Sci Anth, 6(1), p.31. DOI:
 Published on 12 Apr 2018
 Accepted on 06 Mar 2018            Submitted on 24 Aug 2017
Domain Editor-in-Chief: Oliver Chadwick; Geography Department, University of California, Santa Barbara, US

1. Introduction

Acid mine drainage (AMD), the legacy of over 200 years of coal mining in mid-Atlantic states, is estimated to impair more than 12,000 km of streams in the eastern USA (Boyer and Sarnoski, 1995). Discharge of AMD from coal and hard rock mining activities is the result of the oxidation of exposed sulfide minerals generating highly acidic, metal-rich fluids that are then released into the local environment. In much of the mid-Atlantic region, this is primarily due to the circulation of rain- and groundwater in abandoned coal mines and unreclaimed coal refuse piles. AMD discharge has severe, long lasting impacts on water quality and stream ecology in affected watersheds due in part to the release of potentially toxic metals such as Al, Mn, As, Cd, Cr, Cu, Ni, Pb, Zn (Cravotta, 2008). These metals are often sequestered by amorphous, poorly-, and crystalline Fe(III) oxy-hydroxides that accumulate on the streambed (Galán et al., 2003). However, biogeochemical processes within the sediments can potentially alter Fe(III) oxy-hydroxides stability resulting in metal release (Gambrell, 1994; Ford et al., 1997). Localized areas of biogeochemical activity that can mediate these changes are potentially linked to stream geomorphology, and this relationship has not been previously considered with respect to potential longitudinal variation within AMD-impacted streams.

Even after remedial actions that stop AMD input, the overall ecology of some impacted areas does not rebound due to the accumulation of clay-sized AMD deposit particles, which is thought to cut off critical geochemical gradients within the streambed (Kimball et al., 1995; Hancock, 2002; Hogsden and Harding, 2011). Iron loadings in streams and subsequent deposition of Fe(III) oxy-hydroxides has been shown to be dependent on seasonally-driven changes in stream discharge; periods of higher discharge (i.e. spring and summer) result in higher Fe(III) oxy-hydroxides formation and deposition due to the flushing of Fe out of historical coal waste and abandoned mine tunnels (August et al., 2002). Few studies have systematically measured the spatial variability of AMD deposits, and possible variations may be related to biogeochemically-driven changes within the streambed or geomorphic patterns of erosion and deposition. For example, at Pinal Creek, Arizona, a stream affected by historical copper mining, many patches of cemented Mn-oxides were present on the streambed and areas where the cemented crust is absent are interpreted as areas of recent deposition (Fuller and Harvey 2000; Bargar et al., 2009).

Mineralogical transformations and the fate of trace metals in AMD deposits located on streambeds can be influenced by biogeochemical processes occurring within the hyporheic zone, a sub-stream region of groundwater-surface water mixing containing at least 10% stream water (Triska et al., 1989). Studies of the hyporheic zone of Silver Bow Creek, Montana, Pinal Creek, Arizona, and a second-order headwater stream in Floyd County, Virginia, support the concept that AMD-derived precipitation results in a finite boundary layer influencing natural chemical gradients and that this zone can act as a natural sink for heavy metals by preventing or slowing the release of contaminants from one regime to the other (Wielinga et al., 1994; Benner et al., 1995; Harvey and Fuller, 1998; Nagorski and Moore, 1999; Fuller and Harvey, 2000; Hancock, 2002; Brown et al., 2007; Bargar et al., 2009; Fuller and Bargar, 2014). An influx of acidity into the hyporheic zone can result in dissolution of Fe(III) (oxy)hydroxides and desorption of toxic metals, causing the hyporheic zone to develop higher concentrations of heavy metals compared to the surface or groundwater (Nagorski and Moore, 1999; Gandy et al., 2007). However, these changes are also dependent upon the assemblage of Fe(III) (oxy)hydroxides present; newly formed amorphous to poorly crystalline phases (e.g. ferrihydrite) are more soluble and susceptible to transformation compared to older, mature phases (e.g. goethite and hematite).

Existing studies of the hydrogeochemistry of AMD-affected streambeds have elucidated reach scale mass balances and vertical geochemical gradients, but have not examined longitudinal variation in mineralogy and geochemistry around geomorphic structures. However, studies in other environments have shown that geomorphic features such as riffle-pool sequences, step pools, and debris dams play a significant role in promoting hyporheic exchange and biogeochemical transformation (e.g., Harvey et al., 1996; Kasahara and Wondzell, 2003; Lautz and Fanelli, 2008; and Sawyer et al., 2011).

The current work is motivated by the following question: How is Fe(III) (oxy)-hydroxide speciation and distribution, and pore water chemistry in an AMD-impacted streambed affected by the presence of geomorphic structures? We examine three hypotheses: (1) AMD deposit and pore waters are spatially uniform in terms of mineralogy and geochemistry (i.e., geomorphic structures have no effect); (2) deposit variability is a function of differences in deposition patterns around geomorphic structures; and (3) deposit and pore water variability is associated with downward and upward hydraulic gradients around geomorphic structures. We focus on an Ohio watershed impacted by historical coal mining, with in-stream geomorphic structures typical of many Appalachian watersheds.

2. Methods and materials

2.1. Site description

The Huff Run watershed (36.0 km2) is located near Mineral City, OH in the unglaciated portion of the Appalachian Plateau (Figure 1). The watershed is primarily underlain by shale, of the Pennsylvanian Allegheny group, that also contains siltstone, sandstone, and thinly bedded limestone and 0.3–1.5 m thick coal seams (Lamborn, 1956). The regional climate is humid continental with mean annual temperature ~10°C and mean annual precipitation ~100 cm (Poncelet et al., 2014). While the eastern headwaters portion of the watershed was developed for agriculture and remains unmined, the western downstream portion has experienced numerous anthropogenic perturbations that include deep mining (1810–1946), surface mining (1950-present), and oil and gas production (ODNR, 2000). Fourteen reclamation projects completed over the past 14 years have substantially reduced acid loads in the stream and led to increases in biological diversity; however, the Huff Run stream and its tributaries remain impaired by acidic, metal-rich inputs from deep mine discharge, water-filled impoundments, and runoff from unclaimed coal refuse piles (Kinney, 2013). Ongoing reclamation efforts seek to restore water quality and biological diversity to levels necessary to obtain a warmwater habitat designation from the Ohio EPA.

Figure 1 

(a) Huff Run location map; (b) acid loads in portions of the watershed (the black dot shows the location of sub-watershed HR-46 and approximate extent of (c), map of the Huff Run and HR-46 tributary within the study area showing the location of the USGS stream gage (triangle), AMD source (the Farr Tipple Project), culvert and tributary (d), and debris dam (e). The Farr Tipple Project is an unsuccessful passive reclamation project observed to the east of the Huff Run, which drains into the HR-46 tributary. The photograph of the tributary (d) taken from near the culvert shows the location of the two step-pools. (e) A photograph of the debris dam. Survey data from the US EPA and ODNR. DOI:

This study examined two geomorphic settings 2 km upstream from the mouth of Huff Run at Conotton Creek: a mainstem debris dam and a tributary step-pool sequence. At the mainstem debris dam (Figure 1e and 2), the Huff Run has a steep hillslope on its left bank and a wide floodplain on its right bank. The stream was historically channelized and dredged in this reach and is 14 m wide. The debris dam, which was comprised of several key large logs and a mass of smaller woody debris, impounded the Huff Run and decreased channel velocity upstream of the dam. A 3 m wide opening allowed concentrated flow to pass through the dam and created an 8 m long by 4 m wide scour pool below the dam. The scour pool was estimated to be more than 2 m deep, though direct measurements were not possible. A large subaqueous gravel bar occurred 10 m downstream of the debris dam. In the rest of the reach, the surface of the streambed was a mixture of sand and ferric oxyhydroxide particles, underlain by gravels.

Figure 2 

Streambed elevations and piezometer locations in the mainstem debris dam reach. Elevations are relative to an arbitrary datum, and the depth of the scour pool below the dam is likely to be underestimated. Flow is from top right to bottom left. DOI:

In the small HR-46 tributary to Huff Run, 250 m upstream of the debris dam, water emerging from the Farr Tipple AMD site (HR-46) (Figure 1d) flows into a cascade with a slope of 0.2 m/m for ~15 m (Figure 3). In this section, many small steps are formed from both wood and rock, and the channel bed is consolidated iron oxide-bearing sediments. The final ~5 m of the tributary is lower gradient (0.04 m m–1) with two pools separated by a step formed from a sandstone boulder. The stream bed in this section is unconsolidated sediment. The downstream-most pool in the tributary has a bed elevation that is only 0.01 to 0.20 m higher in elevation than the water surface elevation of Huff Run at low flow, and is separated from the main channel by a sediment wedge, with the tributary widening as it approaches the main channel. This pool is likely episodically influenced by backwater flooding from Huff Run.

Figure 3 

Long profile of the tributary reach, with pools 1 and 2 indicated. Elevations are relative to the streambed depth at the confluence with the main channel of Huff Run. DOI:

A USGS stream gage, number 03121850, is located on Huff Run, 500 m upstream of the HR-46 tributary (Figure 1c). During the sampling period (June 27, 2014 through September 12, 2014), the average discharge was 0.39 m3 s–1 with instantaneous maximum and minimum values of 14 and 0.091 m3 s–1, respectively. On the sampling dates, the range of flows was 0.14 to 0.96 m3 s–1. The average flow (0.38 m3 s–1) on the sampling dates is the same as during the sampling period and the gaged period of record (2007–2015). Huff Run is also subject to controlled inundation by the downstream Dover Dam, which causes large sections of Huff Run and its adjacent floodplain to become flooded. During inundation events, flooding extends past the study sites and at least 2.25 km upstream from where the Huff Run empties into the Conotton Creek (USGS, 2015).

2.2. Solid phase mineralogical and geochemical analyses

Characterization of streambed sediments was conducted on extracted core (2.54 cm diameter) samples collected in June 2014. Prior to processing and analysis, sediments were stored at –2°C to prevent biological transformations. Three cores were sectioned into 1 cm samples under ambient conditions. Sediment cores from the step-pool and dam sections were 25–27 cm and 15–17 cm respectively. After sectioning, sediments samples were dried at 40°C to minimize iron oxide phase transformations (Schwertmann and Cornell, 2008) for 32 hours and then pulverized to silt size particles (10–75 micron) with a SPEX-8000M ball mill using tungsten carbide ball bearings. The ball mill was cleaned with ethanol and acetone between samples. The sediment pH of a subset of samples was analyzed on material dried and ground with a mortar and pestle. A suspension of 5 g of sediment and 5 mL of 0.1 M CaCl2 was shaken for 15 minutes and then centrifuged. The pH was measured on triplicate samples using Thermo Scientific Orion Star A111 pH Benchtop Meter and 0.6 pH units was added to the overall average to get the equivalent water pH (Thomas, 1996; Kissel et al., 2009). Milled sediment samples were subsequently analyzed by X-ray diffraction (XRD), loss on ignition (LOI) as a proxy for organic matter (Santisteban et al., 2004), and a sequential extraction procedure for element distribution and abundance.

XRD analysis was performed using a Rigaku DMaxB with a scan range of 3° to 70°, 0.02° per step, 2 seconds per point. Glass slides were washed with acetone between samples. Background subtraction and peak identification was performed using JADE v 6.5 (Materials Data Inc., Livermore, CA) to determine the dominant mineral phases present in the samples. Using the initial peak identification, a quantitative phase analysis was performed using the Rietveld module included in the X’Pert HighScore Plus software. Phase mixtures were then extracted from the literature as well as the AMCSD and COD databases (Downs and Hall-Wallace, 2003; Gražulis et al., 2009; Gražulis et al., 2011) and were modeled for scale factor, preferred orientation, and peak shape (including March-Dollase factor). According to the Highscore Plus Manual, the detection limit is 5 g kg–1.

Loss on ignition was performed with 1 g of milled sample added to a ceramic cup and heated to 550°C for four hours in a muffle furnace and then re-weighed. Given the sediment pH and XRD results (discussed below) it is unlikely that significant amounts of carbonate are present and further loss on ignition at higher temperatures was not performed.

A sequential extraction procedure was used to determine the distribution of elements in four operationally defined fractions: (1) exchangeable (metals bound to the surfaces of minerals and organic matter); (2) carbonates (metals co-precipitated with carbonate minerals); (3) reducible (metals bound within Fe- and Mn-(oxy)hydroxides); and (4) oxidizable (metals bound within complex organic matter and sulfide minerals). All reagents used were analytical grade and solutions were prepared with distilled-deionized water (DDI-H2O) (18.2 MΩ; Milli-Q Direct-Q 3UV-R). The sequential extraction protocol, modified after Tessier (1979) and Gault et al. (2003), used one gram of sediment sequentially exposed to following conditions: (1) 1M sodium chloride at pH 7 for 1 hour with continuous agitation; (2) 1M sodium acetate buffered to a pH of 5 using acetic acid with continuous agitation for five hours; (3) 0.04 M hydroxylamine hydrochloride and 25% acetic acid at a temperature of 95° (±5°C) for six hours; (4) 0.02 M nitric acid with 20% hydrogen peroxide, buffered to a pH of 2 and heated to 85°C (± 2°C) for two hours with occasional agitation, followed by a second allotment of 30% hydrogen peroxide buffered to pH 2 and, heated to 85°C (± 2°C) for three hours with occasional agitation, and a final addition of 3.2 M ammonium acetate in 20% (v/v) nitric acid diluted to 20 mL and agitated for 30 minutes. After each extraction step, the samples were centrifuged for 20 minutes at 10,000 rpm and the decanted solution was passed through a 0.45 µm filter. The remaining sediments were then washed using 8 mL of DDI-H2O and centrifuged again before the next extraction was performed. The fraction of metals tightly bound within aluminosilicate minerals was ignored, as it is unlikely that this pool of metals interacts with materials derived from AMD inputs. Extraction and wash solutions were analyzed for Al, As, Cu, Fe, Mn, Se, and Zn using a Perkin-Elmer 8000 Inductively Coupled Plasma Optical Emission Spectroscopy (ICP-OES), using SPEX CertiPrep trace metal standards. Practical quantification limits for these elements are 0.1 µg L–1, 0.05 µg L–1, 0.05 µg L–1, 1 µg L–1, 1 µg L–1, 0.1 µg L–1, µg L–1, and 0.5 µg L–1, respectively. With the exception of Cu, the other trace elements (As, Se, and Zn) were at or below the detection limits in sequential extraction samples (see Supplementary Data).

2.3. Hydrogeological analyses

Characterization of streambed hydrogeology was enabled by the installation of nine piezometers in the debris dam reach and four piezometers in the tributary step-pools in June 2014. Four piezometers were installed upstream of the debris dam and five piezometers were installed downstream of the debris dam, while two piezometers were installed in each of the two downstream most pools of the step-pool sequence. In the debris dam reach, piezometers were constructed using 3.8 cm inner diameter PVC with a screened length of 17.8 cm, with screening created by drilling 1.5 mm holes in the PVC. For the step-pools, piezometers were similarly constructed with 2.54 cm inner diameter PVC, screened for a length of 15.2 cm. Debris dam and step-pool piezometers were installed at depths of 57–87 cm and 55–71 cm below the sediment surface, respectively. Piezometers were installed with a sledgehammer and developed using a peristaltic pump prior to data collection. Piezometer locations and other geomorphic features were surveyed using a Topcon Total Station ES.

From June to September 2014, hydraulic head in the piezometers was measured using a Heron Little Dipper water level meter (Heron Instruments, Hamilton, ONT) prior to sampling for water chemistry. On each occasion, water level was measured in the stream adjacent to the piezometer to allow for the calculation of vertical hydraulic gradient between the stream and the screened depth. Slug tests were performed on all piezometers to calculate hydraulic conductivity using the Hvorslev method (Hvorslev, 1951). A 1 L slug of water was added to each piezometer and response was recorded with an Onset Hobo U-20 pressure transducer (Onset Computer Corporation, Bourne, MA) set to log at 1 s intervals. Darcy flux, the specific discharge of groundwater, was calculated using the measured hydraulic conductivity (K) and the average hydraulic gradient (dh̅/dl) for each piezometer:


The Darcy flux measurements should be interpreted as a potential for groundwater discharge to the stream or stream recharge of groundwater, because flowpaths were not directly measured.

In the debris dam reach, mini-piezometers were also installed adjacent to the PVC piezometers at depths of 23–30 cm, following the method of Martinez (Martinez, 2013). Mini-piezometers were made of 6.35 mm inner diameter polyethylene tubing and screened for a length of 4 cm. Mini-piezometers were used only for withdrawing water samples. Their narrow diameter, combined with iron oxide staining of the polyethylene tubing, made accurate measurement of hydraulic head difficult.

2.4. Aqueous phase analyses

Huff Run tributary and main stem surface water and sediment pore water collected from the piezometers were collected approximately bi-monthly from July to November 2014 for chemical analyses. Water samples from the piezometers were collected using a peristaltic pump; three purged sample well volumes were removed before collecting the sample for analysis. All solution samples were passed through a 0.45 µm filter. Dissolved oxygen was measured in the field using an YSI Professional Plus probe (Yellow Springs Instruments, Yellow Springs, OH). Solution pH was measured in the laboratory within five hours of collection. Solution samples were acidified with two drops of concentrated nitric acid and refrigerated prior to ICP-OES analysis for Al, Ca, Fe, Mg, Mn, and Na. Practical quantification limits for Mg and Na are 8 µg L–1 and 10 µg L–1, respectively. Surface and pore water samples were analyzed for trace elements (As, Cu, Se, and Zn) which were below the detection limit of the ICP-OES (see Supplementary Data). Solution extraction materials were cleaned with 10% nitric acid after each sampling trip.

3. Results

3.1. Debris dam reach

3.1.1. AMD deposit thickness

Iron(III) (oxy)-hydroxide-bearing deposits were at least 17 cm thick upstream and downstream of the debris dam. Sediments within the streambed downstream of the dam site became significantly coarser below 17 cm preventing extraction of deeper cores. Particle size in the cores was dominantly coarse sand, based on in-field characterization (see Supplementary Data). Downstream of the debris dam, we were unable to sample in the center of the scour pool, because it was too deep to be accessed, so AMD deposit depth may be smaller in that location. We also observed a gravel bar at the downstream end of the debris dam reach that was not uniformly covered in AMD deposits throughout the study period.

3.1.2. Solid phase chemistry

The solid phase chemistry of the sediment cores extracted from above and below the debris dam were both vertically uniform and do not showed a marked transition with depth (Figure 5), in contrast to the step-pool sediments described below. It is possible that a chemical gradient is present at depth >17 cm in the coarse sediments. The pH of the sediments from above the dam increased gradually from 7 to 7.4 with depth. This also corresponded to an increase in the loss-on-ignition from 10 to 20%. A peak in loss-on-ignition was observed near the base of the core, at 50%. The pH of the sediments from below the dam (pH 6.5) was lower than above the dam (pH 7), and both were relatively invariant with depth. The loss-on-ignition was also relatively consistent with depth at 18%, with a peak value of 50% in the middle of the core.

The mineralogical assemblage of the sediments from above the dam were dominated by quartz (50–71%), feldspar (1–11%), and clays (18–35%), with some goethite (4–26%) throughout the core. In contrast, no crystalline Fe-bearing phases were detected in the sediments below the dam. The amount and distribution of extractable Fe, Al, Mn, and Cu was also dependent on being above or below the dam. Fe was primarily observed in the reducible fraction and average values were higher above the dam (255 mmol kg–1) than below (144 mmol kg–1). Although the solid phase Fe concentration was lower beneath the dam, there was still extractable Fe primarily in the reducible phase. The reducible pool is typically dominated by oxide phases, despite the fact that there were no detectable crystalline Fe-bearing phases from the XRD analyses, which will be discussed below. Al was bound within both the reducible and oxidizable fraction, and average values were also slightly higher above the dam (41 mmol kg–1) than below (23 mmol kg–1). Mn was dominantly observed in the reducible fraction and average values were significantly higher in the above dam sediments (26 mmol kg–1) than below (11 mmol kg–1). Cu was dominantly in the oxidizable fraction and average values were significantly higher in the below dam sediments (0.09 mmol kg–1) than above (0.02 mmol kg–1).

3.1.3. Hydrogeology

Above the debris dam, all four piezometers indicated strong downward hydraulic gradients, while below the debris dam, the five piezometers showed weak upward hydraulic gradients (Figure 4). Above the dam, piezometer A-I had consistently smaller hydraulic gradients than the other three piezometers, which were closer to the center-line of the channel. Below the dam, the largest hydraulic gradients were observed closest to the dam, while the piezometers farther downstream had smaller hydraulic gradients. The below dam piezometers exhibited their smallest hydraulic gradients on August 23, 2014, the measurement date with the highest discharge. Two of the three above dam piezometers measured that day also exhibited suppressed hydraulic gradients relative to surrounding measurement dates.

Figure 4 

Hydraulic gradient for piezometers in the debris dam reach. Piezometer names that begin with A- are above the debris dam, while piezometer names that begin with B- are below the debris dam. Piezometer locations are shown in Figure 2. DOI:

Figure 5 

Solid phase chemical analyses of sediment cores extracted from above and below the debris dam. DOI:

Hydraulic conductivity was fairly uniform above the debris dam, with the highest hydraulic conductivity only 2.5 times higher than the lowest hydraulic conductivity, and a geometric mean of 2.8 × 10–4 m/s (Table 1). Below the dam, hydraulic conductivity had a similar geometric mean (2.1 × 10–4 m/s), but the measured values were more variable, with the highest hydraulic conductivity 48 times greater than the lowest hydraulic conductivity. Because of the smaller hydraulic gradients below the dam, calculated Darcy fluxes were smaller for all piezometers below the dam than any of the piezometers above the dam. The geometric mean of the downward Darcy fluxes above the dam is 15.4 times greater than the geometric mean of the upward Darcy fluxes below the dam.

Table 1

Hydraulic conductivity and Darcy fluxa. DOI:

Piezometer Hydraulic Conductivity (m/s) Average Darcy flux (m/s)

Step-Pool Sequence 1–A 8.5 × 10–4 –5.2 × 10–5
1–B 1.9 × 10–4 –1.5 × 10–5
2–C Not measured
2–D 5.7 × 10–5 4.6 × 10–6

Above Debris Dam A–F 2.5 × 10–4 –7.5 × 10–5
A–G 2.3 × 10–4 –6.8 × 10–5
A–H 5.1 × 10–4 –1.4 × 10–4
A–I 2.1 × 10–4 –4.5 × 10–5

Below Debris Dam B–A 5.2 × 10–5 5.0 × 10–7
B–B 7.9 × 10–4 1.4 × 10–5
B–C 1.3 × 10–4 3.8 × 10–6
B–D 2.0 × 10–3 5.9 × 10–5
B–E 4.2 × 10–5 1.9 × 10–6

aDarcy flux is calculated following Eqn. 1. Positive Darcy fluxes indicate upward movement of water.

3.1.4. Surface and pore water chemistry

The average pH value of the Huff Run main stem surface water above and below the debris dam was 6.78 (Table 1). Well water from the sediments above and below the dam had average pH values of ~ 6–7. The dissolved oxygen concentrations decreased to 2.4 mg L–1 in the deepest sampling depth. Average concentrations of Ca, Mg, and Na were fairly constant between the surface water and well water within the sediments above and below the debris dam. Surface water concentrations of Fe and Mn at the dam site were 0.04 mM and 0.11 mM, respectively. Average Fe and Mn concentrations in sediment pore water from above the debris dam sediment, where downwelling was dominant, were 0.09 mM and 0.08 mM, respectively. In contrast, below the debris dam where upwelling dominates, Fe increased to 0.33 mM and Mn stayed relatively constant at 0.12 mM. These higher Fe concentrations with depth suggest a chemical gradient is present that was not observed in the shallower solid samples, similar to the step-pool sediments discussed below. For each sampling location, variation in water chemistry was small across sampling date (Traub, 2016). As a point of contrast, average surface water concentrations of Fe and Mn in the non-AMD impacted upper reach of the Huff Run (site HRR01, over the period of 2005–2017 were 0.009 mM and 0.003 mM, respectively.

3.2. Step-pool sequence

3.2.1. AMD deposit thickness

Iron(III) (oxy)-hydroxide-bearing deposits were >25 cm thick upstream and downstream of the step, exceeding the length of the corer. Particle sizes in the cores were dominantly silt to fine sand, based on in-field characterization (see Supplementary Data). Changes in the exposed piezometer pipe length in the lower pool indicate ~6 cm erosion of pool sediments over the course of the summer. Episodic flooding from the Huff Run main stem may contribute to a dynamic deposition and scour pattern at this location.

3.2.2. Solid phase chemistry

The solid phase chemistry of the sediment core extracted from step-pool 1 was dominated by a transition point at 7–8 cm in depth (Figure 7). Above this depth, the sediment pH was ~5 and the loss-on-ignition values were 10–12%; below this depth the pH dropped to 4 and the loss-on-ignition increased to 18–20%. The mineralogical assemblage also changed at this depth; above the transition point, the sediments were dominated by quartz (35–55%), clays (24–34%), and goethite (20–40%). Below the transition point, the sediments were nearly 100% goethite, with minor amounts of quartz (< 5%), magnetite (1–4%), and jarosite (0–10%). The distribution of extractable Fe, Al, and Mn was also dependent on the transition with depth. In general, the reducible fraction (dominated by Fe- and Mn-(oxy)hydroxides) accounted for most of the extractable Fe, Al, and Mn. The amount of extractable Fe increased at the transition point (from ~150 mmol kg–1 to 300 mmol kg–1 and then slowly decreased with depth), which was consistent with the increase in the amount of goethite and other Fe-bearing crystalline phases based on the XRD results. In contrast, small amounts of extractable Al (< 30 mmol kg–1) and Mn (< 20 mmol kg–1) were only detected in the shallow sediments. Some of the Al (~ 30%) was also found to be associated with the oxidizable fraction. Minor amounts of copper (near the detection limit) were observed throughout the sediment core, primarily in the exchangeable fraction.

The solid phase chemistry of the sediment core extracted from step-pool 2 was also characterized by a transition point, at 12–15 cm in depth (Figure 7). However, in contrast to step-pool 1, the pH was relatively invariant (~ pH 4.5) throughout the core. Below the transition point, the loss-on-ignition decreased from 15% to 2%. Above the transition depth, the mineral assemblage was dominated by goethite (11–71%), quartz (20–44%), clays (8–36%), and some feldspars (0–12%). Below the transition depth, no crystalline Fe-bearing phases were detected, and the sediments were dominated by quartz (58–86%), clays (10–29%) and feldspars (04–16%). The distribution of extractable Fe, Al, Mn, and Cu was also strongly dependent on this transition. In the shallow sediments, extractable Fe was primarily in the reducible fraction; in the deeper sediments, Fe was nearly absent. The maximum extractable Fe was also higher in step-pool 2 (660 mmol kg–1) compared to step-pool 1 (440 mmol kg–1). The amount of extractable Al was also much higher in step-pool 2 (338 mmol kg–1) than in step-pool 1 (28 mmol kg–1), with a maximum just above the transition depth. In the shallow sediments, Al was primarily found in the reducible fraction, whereas it was found primarily in the oxidizable fraction in the deeper sediments. Extractable Mn was dominated by the exchangeable and carbonate fractions with a maximum concentration of 35 mmol kg–1. Like Fe, Mn was only observed in the shallow sediments of step-pool 2. In contrast, Cu was observed only in the deeper sediments with a maximum concentration 1.2 mmol kg–1, primarily in the oxidizable fraction (likely bound as sulfides, discussed below). The maximum extractable Cu was also higher in step-pool 2 compared to step-pool 1.

3.2.3. Hydrogeology

In the upper pool (step-pool 1), the two piezometers (1-A, 1-B) both indicated downward hydraulic gradients for the first six of eight measurement sets between June and September 2014 (Figure 6). In the lower pool (step-pool 2) piezometers (2-C, 2-D), hydraulic gradients were upward on all measurement dates. In both pools, hydraulic gradients were smallest on August 23, 2014, which was the highest streamflow recorded by the USGS gage on Huff Run on any measurement date.

Figure 6 

Hydraulic gradient for piezometers in the tributary step-pool sequence. Piezometers 1-A and 1-B are in the upper pool, while piezometers 2-C and 2-D are in the lower pool. On July 30, 2014, the streambed adjacent to piezometer 2-C was dry, so no hydraulic gradient is calculated. DOI:

Figure 7 

Solid phase chemical analyses of sediment cores extracted from step-pools 1 and 2. DOI:

Hydraulic conductivity varied by a factor of 15 across the three piezometers in which it was measured, with a geometric mean of 2.1 × 10–4 m/s (Table 1). Darcy fluxes were calculated for each piezometer, based on the hydraulic conductivity and the average hydraulic gradient. Downward Darcy flux in step-pool 1 is approximately seven times greater than the upward Darcy flux in step-pool 2.

3.2.4. Surface and pore water chemistry

Surface water and pore water chemistry, averaged over the sampling period July–November 2014 (Table 2), were fairly constant across sampling dates (Traub, 2016). The average pH value of the tributary surface water within the step-pools was 6. Circumneutral pH values of tributaries within the watershed are often observed, despite their proximity to abandoned coal mine works, due to the presence of limestone bedrock and clasts which buffer the pH. Pore water from the AMD-dominated sediments within the step-pools had average pH values between 3–3.3, and the dissolved oxygen concentrations decreased to 2.4 mg L–1 in the deepest sampling depth. Average concentrations of Ca, Mg, and Na were fairly constant (within 10%, see Supplementary Data) between the surface water and well water within the step-pool sediments throughout the sampling period. In contrast, Mn increased from 0.15 mM in the surface water to 0.23 mM and 0.27 mM in step-pool 1 and 2, respectively, and Fe increased from 0.27 mM in surface water to 1.74 mM and 3.33 mM in step-pool 1 and 2, respectively. In general, Al was below the detection limit in nearly all aqueous samples.

Table 2

Surface water and well water measurements from the Huff Run main stem and tributary, averaged over the sampling period July–November 2014a. DOI:

SAMPLING metal concentrations (mM)

SITE DEPTH T (°C) pH DO (mg L–1) Al Ca Fe Mg Mn Na

STEP-POOL surface water 16.1 ± 0.2 6.07 ± 0.02 7.1 ± 0.5 < d.l. 3.57 ± 0.35 0.27 ± 0.06 1.35 ± 0.09 0.15 ± 0.02 0.76 ± 0.11
STEP-POOL 1 66 cm 15.4 ± 0.1 3.03 ± 0.03 3.9 ± 0.3 0.39 ± 0.06 2.92 ± 0.17 1.74 ± 0.58 1.36 ± 0.10 0.23 ± 0.03 0.63 ± 0.03
STEP-POOL 2 63 cm 16.8 ± 0.1 3.31 ± 0.01 2.4 ± 0.1 < d.l. 3.28 ± 0.07 3.33 ± 0.24 1.29 ± 0.09 0.27 ± 0.04 0.48 ± 0.09

DAM surface water 18.5 ± 0.8 6.78 ± 0.03 6.6 ± 0.1 < d.l. 3.25 ± 0.25 0.04 ± 0.01 2.02 ± 0.29 0.11 ± 0.01 0.85 ± 0.12
ABOVE DAM 23 cm 19.1 ± 0.7 6.83 ± 0.03 3.4 ± 0.1 < d.l. 3.09 ± 0.28 0.09 ± 0.02 1.98 ± 0.30 0.08 ± 0.01 0.81 ± 0.12
ABOVE DAM 60 cm 18.9 ± 0.6 6.73 ± 0.03 2.4 ± 0.1 < d.l. 2.81 ± 0.29 0.09 ± 0.02 1.84 ± 0.29 0.09 ± 0.01 0.72 ± 0.11
BELOW DAM 28 cm 18.9 ± 0.7 7.02 ± 0.03 2.8 ± 0.1 < d.l. 2.56 ± 0.26 0.33 ± 0.04 1.69 ± 0.27 0.12 ± 0.02 0.70 ± 0.09
BELOW DAM 67 cm 18.0 ± 0.4 6.95 ± 0.03 2.7 ± 0.1 < d.l. 2.43 ± 0.25 0.21 ± 0.03 1.66 ± 0.24 0.12 ± 0.01 0.68 ± 0.08

aReported values include one standard deviation of the averaged values over the sampling period.

4. Discussion

Using mineralogical, geochemical, and hydrogeological data, we show consistent differences in AMD deposits and pore waters with respect to position upstream or downstream of two geomorphic structures. In terms of solid phase mineralogy and geochemistry, in both the tributary and main stem, goethite was the dominant Fe-bearing phase throughout the AMD deposit depth in cores taken upstream of the geomorphic structures. Downstream of the structures, no crystalline Fe-bearing minerals were detected at any depth at the debris dam site or below 13 cm at the step-pool site despite the presence of Fe in the reducible fraction, typically associated with Fe (oxy)hydroxides. The concentrations and distribution of extractable Al, Mn, and Cu were also different upstream versus downstream of each structure. Pore water Fe and Mn concentrations were higher downstream of the structures than upstream, and pore water Fe concentrations were much higher than in the surface water. Strong downward hydraulic gradients were measured upstream of the structures, while weaker upward hydraulic gradients were measured downstream of the structures. It is therefore apparent that AMD deposit speciation and distribution and pore water chemistry are not spatially uniform within the tributary or the mainstem of the stream, and we suggest that geomorphic structures affect streambed mineralogy and geochemistry in AMD-impacted streams. The variation in solid phase speciation and distribution and pore water chemistry may be explained by deposition patterns or hydraulic gradients, and we examine each of these hypotheses below.

4.1. AMD deposit speciation and distribution in relation to depositional patterns

In locations where water surface slopes flatten upstream of obstructions, like steps or debris dams, deposition of suspended material and bed load tends to occur (e.g., Keller and Swanson, 1979; Lisle and Hilton, 1999). Downstream of the obstructions, pools are maintained by scouring (e.g., Bilby and Ward, 1989; Zimmermann and Church, 2001). Thus, AMD deposits might accumulate slowly, or not all, downstream of geomorphic structures, while accumulating relatively quickly upstream of the structures. However, the limited observations made in this study suggest that similar thicknesses of deposits are present upstream and downstream of the step and debris dam.

At the debris dam, solid-phase Fe and Al concentrations are similar in the cores above and below the dam, but goethite is missing from the core below the dam. It is difficult to explain the similar geochemistry but divergent mineralogy through the physical processes of deposition and erosion alone. In the step-pool cores, the mineralogy and geochemistry are broadly similar for the upper 10–15 cm of each core, but there is marked divergence in the lower portion of the cores. The abrupt change in mineralogy and geochemistry of the Pool 2 core below 15 cm could indicate scour to that depth associated with flooding of the mainstem Huff Run, then gradual refilling by deposition of fresh AMD-bearing material. The dramatic decrease in slope between Pool 2 and the mouth would favor deposition by tributary AMD-bearing material in that area, despite the erosional energy associated with the head drop at the step. This interpretation of the Pool 2 core as representing scour and refilling is consistent with observations of sediment dynamics in the pool, but does not explain the geochemistry or mineralogy in the lower part of the core.

The abundance of organic matter, as determined by the LOI values, provides further evidence that there is no relationship between sediment deposition and changes in Fe speciation and distribution. The LOI values above and below debris dam are not correlated with solid phase Fe, and at depths where sudden increases in LOI values (up to 50%) occur, there is no corresponding change in solid phase Fe. The LOI values in step-pool 1 increase with depth while the amount of solid phase Fe decreases, whereas the LOI values in step-pool 2 decrease while the amount of solid phase Fe increases to a depth of 15 cm. Below this depth, only (alumino)silicate minerals were present, suggesting a lack of AMD deposition. The source of the organic matter could be from allochthonous carbon inputs or accumulation via in-situ biological activity (Bilby and Likens, 1980). Although the LOI values cannot be used to distinguish the dominant source of organic matter, it would be expected that sudden changes in allochthonous carbon and/or biological activity within the sediments would result in changes to Fe speciation and distribution (Schwertmann et al., 1986; Lovley, 1987), which is not observed.

The data collected here suggest that deposition patterns, especially in dynamic scour/fill environments, may play a role in determining deposit thickness and variation in the ratio of AMD-derived phases to other detrital phases. However, other differences in AMD deposit mineralogy and geochemistry, particularly near the bottom of the deposit layer, are unlikely to be the result of purely physical processes.

4.2. AMD deposit speciation and distribution in relation to hydraulic gradients

In both the debris dam and step-pool, measured downward hydraulic gradients in the piezometers are interpreted as reflecting downwelling of stream water, while measured upward hydraulic gradients are interpreted as reflecting upwelling of groundwater toward the stream. This interpretation is consistent with flow directions that have been observed in similar geomorphic settings (e.g., Kasahara and Wondzell, 2003; Hester and Doyle, 2008). Steep downward hydraulic gradients are typical for short distances above log, beaver, and debris dams, whereas upwelling downstream of the dams tends to be more spatially diffuse or is not detected (e.g., Lautz et al., 2006; Fanelli and Lautz, 2008; Briggs et al., 2013). In the absence of tracer tests showing surface water return in the upwelling zones, we cannot determine to what extent upwelling represents generalized groundwater gaining conditions versus hyporheic exchange.

Hydraulic conductivities, as measured in the piezometer slug tests, are in the range of gravel to silty sand (10–3 to 10–5 m/s), consistent with our limited field observations of sediments in this depth range. In situ measurements of hydraulic conductivity within the deposits overlying the gravel were not feasible, but the conductivity of this layer would likely be substantially lower (10–6 to 10–12 m/s) because of its clay-sized particles (Freeze and Cherry, 1979). The decreased fluxes and long residence times associated with such low hydraulic conductivity sediments might limit the exchange between the stream and the gravels (Hester et al., 2008). On-going AMD production and deposition of Fe-oxide-rich particles on the surface of the stream bed would also be expected to increase flowpath tortuosity and create stagnant zones in the fine sediment layer (Chen et al., 2009). Nonetheless, the presence of both downward and upward hydraulic gradients around the step-pool and debris dam in the manner often associated with such geomorphic features is indicative of the potential for surface-groundwater exchange, despite the capping fine sediments.

Downwelling and upwelling flowpaths have distinct impacts on geochemical gradients within streambeds, which does not appear to be impeded by the presence of AMD deposits. Given the potential for ground- and surface-water to mix, biogeochemical reactions within this zone may explain the heterogeneous mineralogical abundance and distribution and metal speciation and distribution observed in the cores. For example, downwelling above the debris dam causes surface water to become deoxygenated as it passes through and interacts with sediments containing freshly deposited material derived from AMD generation. If downwelling is not strong, it is also possible that low dissolved oxygen groundwater could migrate into these sediments. Either process would result in the transformation of more crystalline Fe-bearing materials (dominated by goethite) to poorly crystalline or amorphous material. This is consistent with the nearly constant amount of goethite in the above dam sediment core, whereas the amount of reducible Fe, Mn, and Al increased at depth. During upwelling below the debris dam, deoxygenated water alters local geochemical conditions through reductive dissolution of Fe-bearing phases and subsequent re-precipitation of poorly crystalline to amorphous material. These results are also consistent with the increase in soluble Fe observed within this deoxygenated region. Similar results were observed within the step-pool sequence; downwelling in step-pool 1 resulted in the accumulation and transformation of Fe-bearing material into magnetite and jarosite. Upwelling in step-pool 2 resulted in the accumulation and transformation of Fe-bearing material into poorly crystalline to amorphous phases and an increase in soluble Fe. However, in contrast to the debris dam, the more proximal location to the AMD source resulted in a greater abundance of freshly deposited AMD material (i.e. the presence of significant goethite at the top of step pool 2, in contrast to no goethite present at the top of the below dam section). The mineralogical sequence in step-pool 2 is also potentially complicated by periodic flooding of the Huff Run main stem, which can result in both scour and additional deposition of AMD-derived material from upstream.

These results highlight the dynamic nature of these regions, and are in contrast to the common geochemical behavior of Fe-bearing sediments following AMD deposition. Typically, hydrolysis of Fe(III) resulting from AMD production results in the formation of initially poorly crystalline and/or amorphous Fe(III) (oxy)-hydroxides (Cheng et al., 2009). Transformation of these phases through dissolution-reprecipitation, solid-state crystallization, and aggregation-based crystal growth into more thermodynamically stable mineral forms ultimately results in the formation of lepidocrocite (γ-FeO(OH)), goethite (α-FeO(OH)) and finally hematite (α-Fe2O3) (Cudennec and Lecerf, 2006; Wang et al., 2006; Marescotti et al., 2012). These past results would suggest that more crystalline Fe-bearing material would accumulate with depth. However our results indicate that geomorphic feature-dependent flowpaths result in mixing of surface and groundwater that upend expected mineralogical trends. Ultimately, Fe-bearing materials accumulating in areas of strong upwelling show evidence of re-working and mineralogical transformations that cannot be accounted for by variations in deposition.

The speciation and distribution of Mn is also related to the distribution of Fe-bearing phases within the sediment cores. The average solid phase Mn concentration above the debris dam (32 mmol kg–1) was significantly higher than below (11 mmol kg–1). In both sediment cores, solid phase Mn is primarily associated with the reducible fraction; however, there were no crystalline Fe-bearing phases in the below dam sediment core. In the step-pool sequence, solid phase Mn is dominated by the reducible fraction in step-pool 1 which is dominated by goethite. In contrast, in step-pool 2 where upwelling has resulted in the transformation of the Fe-(oxy)hydroxide assemblage, solid phase Mn is dominated by more weakly bound phases (exchangeable and carbonate). These results suggests that the presence of Mn is controlled by Fe-(oxy)hydroxide mineralogical transformations. Previous work has shown that the formation of ferrihydrite in the presence of Mn results in low uptake of Mn, however, Mn uptake increases during transformation to goethite (Ford et al., 1997; Alvarez et al., 2005). The presence of these more crystalline Mn-bearing Fe-(oxy)hydroxides can further influence the fate and transport of other trace metals compared to Mn-free phases (Sun et al., 1999; Singer et al., 2013). Ultimately, our results indicate that the speciation and distribution may also be a function of upwelling and downwelling related to geomorphic features within the stream.

In both geomorphic settings, detectable solid-phase Cu was only observed when upwelling of deoxygenated water occurred. Cu speciation at these locations was dominated by the oxidizable fraction, which could be the result of Cu bound to organic matter and/or sulfide phases. These observations are consistent with the reductive dissolution of primary Fe-(oxy)hydroxides resulting in the release of metals into solution. Adsorption to organic matter and/or precipitation with sulfide phases would likely sequester other chalcogenic elements, such as As and Se which were below the detection limit in the sequential extraction experiments, but are present within the primary sulfide minerals in the shales from which AMD is generated (Cahill and Singer, 2014). The presence of an additional pool of metals other than Fe-(oxy)hydroxides is a further departure from the typically observed speciation and distribution of metals in AMD-impacted sediments. Poorly ordered Fe-(oxy)hydroxides typically have high metal affinities and act as efficient sink for trace metals due to their high specific surface area (Johnson, 1986), and transformations to more ordered, mature phases can result in desorption and structural incorporation of sorbed metals, resulting in metal-specific, nonreversible partitioning (Cornell, 1988; Ford et al., 1997). However, the results shown here indicate that mixing of surface and groundwater can result in an additional sequestration pathway that has potentially been overlooked, resulting in a pool of metals that may be more easily released during subsequent disturbances to flowpaths.

5. Implications and conclusions

We have shown that in-stream geomorphic features are associated with spatially variable AMD deposit speciation in an Ohio stream impacted by historical coal mining. While there is some evidence that dynamic depositional environments may play a role in creating this spatial variability, it is difficult to explain all of the observed changes between upstream and downstream positions through deposition and scour alone. Upward and downward hydraulic gradients in the streambed, and associated geochemical gradients, provide a mechanism to explain the observed geochemical and mineralogical variation. Thus, interaction between groundwater-stream exchange patterns and streambed mineralogy and metal sequestration potentially impacts the ability of stream biota to rebound from ecosystem degradation caused by AMD.

Watershed managers need to account for heterogeneous AMD deposition related to geomorphic features when assessing stream health, especially if such features are created or destroyed. Geomorphic features within the streambed may be able to be designed to favor biogeochemical conditions within the surface-groundwater mixing zone that favor recalcitrant metal-bearing phases, in contrast to more transient and reactive phases which might continue to release toxic metals during AMD mineral transformations. This, in turn, may influence choices of restoration design and implementation. For example, large-scale projects such as dredging a long stream reach may be unnecessary if smaller-scale projects that modify or create localized geomorphic features can promote sufficient groundwater-stream exchange through low permeability AMD deposit-bearing sediments. To compensate for the low permeability of the streambed, a high hydraulic gradient may need to be induced.

Ultimately, we have described a more complicated picture of AMD-derived sediment mineralogy in the streambed than previously observed and found support for groundwater-stream exchange as a potential mechanism driving the variability. Further work is needed to closely associate fine scale variations in hydraulic conductivity with water and sediment chemistry and mineralogy within the vertical profile. Future work should focus on determining hydraulic conductivity and head gradient within low-flow/high-residence time microenvironments and tracer tests to determine if surface water is penetrating to high permeability zones below the AMD deposits and residence time distributions of water in each zone. Further, geoelectrical techniques or sub-centimeter-scale in-situ geochemical analyses can potentially passively characterize pore water above and below these low to high permeability transitions to further refine the fate and transport of dissolved and colloid-associated metals across this region.

Data Accessibility Statement

All data are uploaded as online supporting information.

Supplementary Files

The supplementary file for this article can be found as follows:

  • Supplementary Data S1. Mineralogical and geochemical variation in stream sediments impacted by acid mine drainage is related to hydro-geomorphic setting. DOI:


The authors thank the following Kent State University undergraduate and graduate students for field and laboratory assistance: Yuchen Shen, Krista Brown, Stuart Baker, Sarah Morrison, Owen Jensen, and Laura Zemanek. We also thank David Costello and Nicholas Johnson (KSU Department of Biological Sciences) for ICP-OES support, and Marissa Lautzenheiser (Huff Run Watershed Restoration Partnership) for logistical support at the study site. We also thank Christopher Rowan (KSU Department of Geology) for creating the location map shown in Figure 1. Comments from the editor-in-chief, Oliver Chadwick, and an anonymous reviewer improved this manuscript.

Funding informations

We gratefully acknowledge support from Kent State University (KSU) and the KSU Department of Geology (DMS start-up) and the KSU Graduate Student Senate and KSU chapter of Sigma Gamma Epsilon (ET field work and travel support).

Competing interests

The authors have no competing interests to declare.

Author contributions

  • Contributed to conception and design: DMS, AJJ, and ELT
  • Contributed to acquisition of data: ELT
  • Contributed to analysis and interpretation of data: DMS, AJJ, ELT, and NP
  • Drafted and/or revised the article: DMS and AJJ Approved the submitted version for publication: DMS, AJJ, ELT, and NP.


  1. Alvarez, M, Sileo, EE and Rueda, EH 2005 Effect of Mn(II) incorporation on the transformation of ferrihydrite to goethite. Chem. Geol. 216: 89–97. DOI: 10.1016/j.chemgeo.2004.11.004

  2. August, EE, McKnight, DM, Hrncir, DC and Garhart, KS 2002 Seasonal Variability of Metals Transport through a Wetland Impacted by Mine Drainage in the Rocky Mountains. Environ. Sci. Technol. 36: 3779–3786. DOI: 10.1021/es015629w

  3. Bargar, JR, Fuller, CC, Marcus, MA, Brearley, AJ, Perez De la Rosa, M et al. 2009 Structural characterization of terrestrial microbial Mn oxides from Pinal Creek, AZ. Geochim Cosmochim Acta 73(4): 889–910. DOI: 10.1016/j.gca.2008.10.036

  4. Benner, SG, Smart, EW and Moore, JN 1995 Metal behavior during surface-groundwater interaction, Silver Bow Creek, Montana. Environ Sci Technol 29(7): 1789–1795. DOI: 10.1021/es00007a015

  5. Bilby, RE and Likens, GE 1980 Importance of organic debris dams in the structure and function of stream ecosystems. Ecology 61: 1107–1113. DOI: 10.2307/1936830

  6. Bilby, RE and Ward, JW 1989 Changes in characteristics and function of woody debris with increasing size of streams in western Washington. Transactions of the American Fisheries Society 118: 368–378. DOI: 10.1577/1548-8659(1989)118<0368:CICAFO>2.3.CO;2

  7. Boyer, J and Sarnoski, B 1995 Progress report—statement of mutual intent strategic plan for the restoration and protection of streams and watersheds polluted by acid mine drainage from abandoned coal mines. US Environmental Protection Agency .

  8. Briggs, MA, Lautz, LK, Hare, DK and González-Pinzón, R 2013 Relating hyporheic fluxes, residence times, and redox-sensitive biogeochemical processes upstream of beaver dams. Freshwater Science 32(2): 622–641. DOI: 10.1899/12-110.1

  9. Cahill, M and Singer, DM 2014 Speciation and Distribution of Trace Metals in Iron Sulfide-Bearing Shales. Synchrotron Environmental Science VI Argonne National Laboratory .

  10. Chen, C, Lau, BLT, Gaillard, J-F and Packman, AI 2009 Temporal evolution of pore geometry, fluid flow, and solute transport resulting from colloid deposition. Water Resour Res 45(6): W06416. DOI: 10.1029/2008WR007252

  11. Cheng, H, Hu, Y, Luo, J, Xu, B and Zhao, J 2009 Geochemical processes controlling fate and transport of arsenic in acid mine drainage (AMD) and natural systems. J Hazard Mater 165(1–3): 13–26. DOI: 10.1016/j.jhazmat.2008.10.070

  12. Cornell, RM 1988 The influence of some divalent cations on the transformation of ferrihydrite to more crystalline products. Clay Minerals 23(3): 329–332. DOI: 10.1180/claymin.1988.023.3.10

  13. Cravotta, CA 2008 Dissolved metals and associated constituents in abandoned coal-mine discharges, Pennsylvania, USA. Part 1: constituent quantities and correlations. Appl Geochem 23(2): 166–202. DOI: 10.1016/j.apgeochem.2007.10.011

  14. Cudennec, Y and Lecerf, A 2006 The transformation of ferrihydrite into goethite or hematite, revisited. J Solid State Chem 179(3): 716–722. DOI: 10.1016/j.jssc.2005.11.030

  15. Downs, RT and Hall-Wallace, M 2003 The American Mineralogist crystal structure database. Am. Mineral. 88: 247–250.

  16. Fanelli, RM and Lautz, LK 2008 Patterns of Water, Heat, and Solute Flux through Streambeds around Small Dams. Ground Water 46(5): 671–687. DOI: 10.1111/j.1745-6584.2008.00461.x

  17. Ford, RG, Bertsch, PM and Farley, KJ 1997 Changes in Transition and Heavy Metal Partitioning during Hydrous Iron Oxide Aging. Environ Sci Technol 31(7): 2028–2033. DOI: 10.1021/es960824

  18. Freeze, RA and Cherry, JA 1979 Groundwater. Prentice-Hall, Englewood Cliffs, NJ.

  19. Fuller, CC and Bargar, JR 2014 Processes of Zinc Attenuation by Biogenic Manganese Oxides Forming in the Hyporheic Zone of Pinal Creek, Arizona. Environ Sci Technol 48(4): 2165–2172. DOI: 10.1021/es402576f

  20. Fuller, CC and Harvey, JW 2000 Reactive Uptake of Trace Metals in the Hyporheic Zone of a Mining-Contaminated Stream, Pinal Creek, Arizona. Environ Sci Technol 34(7): 1150–1155. DOI: 10.1021/es990714d

  21. Galán, E, Gómez-Ariza, JL, González, I, Fernández-Caliani, JC, Morales, E and Giráldez, I 2003 Heavy metal partitioning in river sediments severely polluted by acid mine drainage in the Iberian Pyrite Belt. Appl. Geochem. 18: 409–421. DOI: 10.1016/S0883-2927(02)00092-6

  22. Gambrell, R 1994 Trace and toxic metals in wetlands—a review. J. Environ. Qual. 23: 883–891. DOI: 10.2134/jeq1994.00472425002300050005x

  23. Gandy, CJ, Smith, JWN and Jarvis, AP 2007 Attenuation of mining-derived pollutants in the hyporheic zone: A review. Sci Total Environ 373(2–3): 435–446. DOI: 10.1016/j.scitotenv.2006.11.004

  24. Gault, AG, Polya, DA, Lythgoe, PR, Farquhar, ML, Charnock, JM et al. 2003 Arsenic speciation in surface waters and sediments in a contaminated waterway: an IC–ICP-MS and XAS based study. Appl Geochem 18(9): 1387–1397. DOI: 10.1016/S0883-2927(03)00058-1

  25. Gražulis, S, Chateigner, D, Downs, RT, Yokochi, A, Quirós, M, Lutterotti, L, Manakova, E, Butkus, J, Moeck, P and Le Bail, A 2009 Crystallography Open Database–an open-access collection of crystal structures. J. Appl. Crystallogr. 42: 726–729. DOI: 10.1107/S0021889809016690

  26. Gražulis, S, Daškevič, A, Merkys, A, Chateigner, D, Lutterotti, L, Quiros, M, Serebryanaya, NR, Moeck, P, Downs, RT and Le Bail, A 2011 Crystallography Open Database (COD): an open-access collection of crystal structures and platform for world-wide collaboration. Nucleic Acids Res. 40: D420–D427. DOI: 10.1093/nar/gkr900

  27. Hancock, PJ 2002 Human Impacts on the Stream–Groundwater Exchange Zone. Environmental Management 29(6): 763–781. DOI: 10.1007/s00267-001-0064-5

  28. Harvey, JW and Fuller, CC 1998 Effect of enhanced manganese oxidation in the hyporheic zone on basin-scale geochemical mass balance. Water Resources Research 34: 623–636. DOI: 10.1029/97WR03606

  29. Harvey, JW, Wagner, BJ and Bencala, KE 1996 Evaluating the reliability of the stream tracer approach to characterize stream-subsurface water exchange. Water Resources Research 32(8): 2441–2451. DOI: 10.1029/96WR01268

  30. Hester, ET and Doyle, MW 2008 In-stream geomorphic structures as drivers of hyporheic exchange. Water Resour Res 44(3): n/a–n/a. DOI: 10.1029/2006WR005810

  31. Hogsden, KL and Harding, JS 2011 Consequences of acid mine drainage for the structure and function of benthic stream communities: a review. Freshwater Science 31(1): 108–120. DOI: 10.1899/11-091.1

  32. Hvorslev, MJ 1951 Time lag and soil permeability in ground-water observations: U.S. Waterways Experimental Station, U.S. Army Corps of Engineers, Vicksburg, Ms., 49 p.

  33. Johnson, CA 1986 The regulation of trace element concentrations in river and estuarine waters contaminated with acid mine drainage: The adsorption of Cu and Zn on amorphous Fe oxyhydroxides. Geochim Cosmochim Acta 50(11): 2433–2438. DOI: 10.1016/0016-7037(86)90026-8

  34. Kasahara, T and Wondzell, SM 2003 Geomorphic controls on hyporheic exchange flow in mountain streams. Water Resour Res 39(1): 1005. DOI: 10.1029/2002WR001386

  35. Keller, EA and Swanson, FJ 1979 Effects of large organic material on channel form and fluvial processes. Earth Surf. Processes Landforms 4: 361–380. DOI: 10.1002/esp.3290040406

  36. Kimball, BA, Callender, E and Axtmann, EV 1995 Effects of colloids on metal transport in a river receiving acid mine drainage, upper Arkansas River, Colorado, USA. Appl Geochem 10(3): 285–306. DOI: 10.1016/0883-2927(95)00011-8

  37. Kinney, C 2013 Huff Run Acid Mine Drainage Abatement and Treatment Plan Addendum (Available from

  38. Kissel, DE, Sonon, L, Vendrell, PF and Isaac, RA 2009 Salt concentration and measurement of soil pH. Commun Soil Sci Plant Anal 40(1–6): 179–187. DOI: 10.1080/00103620802625377

  39. Lamborn, RE 1956 Geology of Tuscarawas County . Division of Geological Survey.

  40. Lautz, L and Fanelli, R 2008 Seasonal biogeochemical hotspots in the streambed around restoration structures. Biogeochemistry 91: 85–104. DOI: 10.1007/s10533-008-9235-2

  41. Lautz, LK, Siegel, DI and Bauer, RL 2006 Impact of debris dams on hyporheic interaction along a semi-arid stream. Hydrol Processes 20(1): 183–196. DOI: 10.1002/hyp.5910

  42. Lisle, TE and Hilton, S 1999 Fine bed material in pools of natural gravel bed channels. Water Resour. Res. 35: 1291–1304. DOI: 10.1029/1998WR900088

  43. Lovley, DR 1987 Organic matter mineralization with the reduction of ferric iron: a review. Geomicrobiol. J. 5: 375–399. DOI: 10.1080/01490458709385975

  44. Marescotti, P, Carbone, C, Comodi, P, Frondini, F and Lucchetti, G 2012 Mineralogical and chemical evolution of ochreous precipitates from the Libiola Fe-Cu-sulfide mine (Eastern Liguria, Italy). Appl Geochem 27(3): 577–589. DOI: 10.1016/j.apgeochem.2011.12.024

  45. Martinez, CJ 2013 Mini-piezometers for Measuring Groundwater to Surface Water Exchange. USDA, University of Florida IFAS Extension Online at:

  46. Nagorski, SA and Moore, JN 1999 Arsenic mobilization in the hyporheic zone of a contaminated stream. Water Resour Res 35(11): 3441–3450. DOI: 10.1029/1999WR900204

  47. ODNR 2000 Huff Run Watershed Acid Mine Drainage Abatement and Treatment Plan; Prepared for Ohio DNR by Gannett Fleming. Available from

  48. Poncelet, DM, Cavender, N, Cutright, TJ and Senko, JM 2014 An assessment of microbial communities associated with surface mining-disturbed overburden. Environ Monit Assess 186(3): 1917–1929. DOI: 10.1007/s10661-013-3505-8

  49. Santisteban, JI, Mediavilla, R, López-Pamo, E, Dabrio, CJ, Zapata, MBR et al. 2004 Loss on ignition: a qualitative or quantitative method for organic matter and carbonate mineral content in sediments? J Paleolimnol 32(3): 287–299. DOI: 10.1023/B:JOPL.0000042999.30131.5b

  50. Sawyer, AH, Bayani Cardenas, M and Buttles, J 2011 Hyporheic exchange due to channel-spanning logs. Water Resour. Res. 47.

  51. Schwertmann, U and Cornell, RM 2008 Iron oxides in the laboratory: preparation and characterization. John Wiley & Sons.

  52. Schwertmann, U, Kodama, H and Fischer, W 1986 Mutual Interactions Between Organics and Iron Oxides1. Interactions of soil minerals with natural organics and microbes , 223–250.

  53. Singer, DM, Fox, PM, Guo, H, Marcus, MA and Davis, JA 2013 Sorption and Redox Reactions of As(III) and As(V) within Secondary Mineral Coatings on Aquifer Sediment Grains. Environmental Science & Technology 47: 11569–11576. DOI: 10.1021/es402754f

  54. Sun, X, Doner, HE and Zavarin, M 1999 Spectroscopy study of arsenite [As(III)] oxidation on Mn-substituted goethite. Clays Clay Miner. 47: 474–480. DOI: 10.1346/CCMN.1999.0470409

  55. Tessier, A, Campbell, PGC and Bisson, M 1979 Sequential extraction procedure for the speciation of particulate trace metals. Anal Chem 51(7): 844–850. DOI: 10.1021/ac50043a017

  56. Thomas, G 1996 Soil pH and soil acidity. In ‘Methods of soil analysis. Part 3. Chemical methods’ . Sparks, DL (Ed.) 475–490. Madison, WI: Soil Science Society of America.

  57. Traub, EL 2016 The Effects of Biogeochemical Sinks on the Mobility of Contaminants in an Area Affected by Acid Mine Drainage, Huff Run, Ohio . Kent, Ohio: Kent State University, 2016.

  58. USGS 2015 National Water Information System data available on the World Wide Web (USGS Water Data for the Nation), at URL Accessed July 14, 2015.

  59. Wang, H, Bigham, JM and Tuovinen, OH 2006 Formation of schwertmannite and its transformation to jarosite in the presence of acidophilic iron-oxidizing microorganisms. Materials Science and Engineering: C 26(4): 588–592. DOI: 10.1016/j.msec.2005.04.009

  60. Wielinga, B, Benner, S, Brick, C, Moore, J and Gannon, J 1994 Geomicrobial profile through the hyporheic zone of a historic mining flood plain. Proceedings of the Second International Conference on Ground Water Ecology. Bethesda, Maryland.

  61. Zimmermann, A and Church, M 2001 Channel morphology, gradient profiles and bed stresses during flood in a step–pool channel. Geomorphology 40: 311–327. DOI: 10.1016/S0169-555X(01)00057-5

comments powered by Disqus