HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Free Alert me when this article is cited Alert me if a correction is posted Services Related articles in JEQ Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (7) Citing Articles via Google Scholar Google Scholar Articles by DeVolder, P. S. Articles by Pandya, K. Search for Related Content PubMed PubMed Citation Articles by DeVolder, P. S. Articles by Pandya, K. Agricola Articles by DeVolder, P. S. Articles by Pandya, K. Related Collections Redox Processes Wetlands and Aquatic Processes Remediation Bioremediation and Biodegradation Heavy Metals

TECHNICAL REPORTS
Heavy Metals in the Environment

Metal Bioavailability and Speciation in a Wetland Tailings Repository Amended with Biosolids Compost, Wood Ash, and Sulfate

^a College of Forest Resources, Box 352100, Univ. of Washington, Seattle, WA 98195-2100
^b National Synchrotron Light Source, Brookhaven National Laboratory, Upton, NY 11973
^c Dep. of Soil Science, Box 7619, North Carolina State Univ., Raleigh, NC 27695-7619

^* Corresponding author (slb{at}u.washington.edu)

Received for publication December 20, 2001.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION DISCUSSION CONCLUSIONS REFERENCES

 
Lead poisoning of waterfowl from direct ingestion of wetland mine tailings has been reported at the Coeur d'Alene River basin in Idaho. A greenhouse study was conducted to evaluate the effects of surface applications of amendments on lead bioavailability in the tailings. Treatments included sediment only, and sediment with three different surface amendments: (i) biosolids compost plus wood ash, (ii) compost + wood ash + a low SO^2-₄ addition as K₂SO₄, and (iii) compost + wood ash + a high SO^2-₄ addition. Measured variables included growth and tissue Pb, Zn, and Cd concentration of arrowhead (Sagittaria latifolia Willd.) and cattail (Typha latifolia L.) and soil pH, redox potential (E_h), pore water Pb, Pb speciation by X-ray absorption spectroscopy, and in vitro Pb bioavailability. The compost + ash amendment alleviated phytotoxicity for both plant species. Bioavailability of Pb as measured by a rapid in vitro extract decreased by 24 to 34% (over control) in the tailings directly below the amendment layer in the compost + SO₄ treatments. The ratio of acid volatile sulfide (AVS) to simultaneously extracted metals (SEM) also indicated a reduction in Pb bioavailability (1:40 control, 1:20 compost, 1:8 compost + low SO₄, and 1:3 compost + high SO₄). Extended X-ray adsorption fine structure (EXAFS) and X-ray absorption near edge structure (XANES) spectroscopy data indicated that lead sulfide was greater after 99 d in the treatments that included additions of sulfate. These results indicated that, under reducing conditions, surface amendments of compost + wood ash (with or without sulfate) decreased the bioavailability of Pb in metal-contaminated mine tailings.

Abbreviations: AVS, acid volatile sulfide • E_h, redox potential • EXAFS, extended X-ray adsorption fine structure • NP, north plot • SEM, simultaneously extracted metals • SP, south plot • XANES, X-ray absorption near edge structure

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION DISCUSSION CONCLUSIONS REFERENCES

 
MINING AND MILLING of metal ores in the Silver Valley in northern Idaho have resulted in widespread distribution of metal contaminants. In many cases tailings were directly discharged into the Coeur d'Alene River system. In some cases, tailings were discarded in natural depressions, including on-site wetlands. This has resulted in elevated metal concentrations in river, wetland, and lake sediment (Bostick et al., 2001; Harrington et al., 1998; Rabe and Bauer, 1977). The widespread contamination has damaged the native ecosystem. Studies have shown toxicity to fish and waterfowl due to elevated metal concentrations in sediment and water (Beyer et al., 1998, Maxfield et al., 1974). Although ecosystem effects due to elevated metal concentrations are not unique to the Coeur d'Alene River basin, the degree and extent of contamination in this region make it an extreme example (Beyer et al., 1985; Dodds-Smith et al., 1992).

The animals that are potentially most severely affected by the elevated metal concentrations in sediment are waterfowl. These include tundra swans (Cygnus columbianus), Canada geese (Branta canadensis), and mallard ducks (Anas platyrhynchos). These birds use the Coeur d'Alene River basin as a feeding and nesting area (Beyer et al., 1998). Several dead tundra swans and other animals found in the area have tested positive for Pb poisoning. A study of waterfowl fecal matter collected within the Basin determined that direct ingestion of sediments is the primary pathway for Pb exposure (Beyer et al., 1998). Beyer et al. (1998) estimated that the diet of tundra swans in the Coeur d'Alene River basin consists of approximately 22% sediment while Canada geese ingest approximately 9%. A direct correlation was found between sediment Pb concentrations and waterfowl exposure to Pb (Beyer et al., 1998).

Recent studies have established a relationship between the mineral forms of metals such as Pb and their bioavailability (Laperche et al., 1997; O'Day et al., 2000; Ruby et al., 1999). In aerobic soil environments, research has focused on the potential to reduce Pb availability through the formation of chloropyromorphite (Ma et al., 1993; Ryan et al., 2001; Traina and Laperche, 1999). In wetland environments, the emphasis has been on mineral changes that are associated with reducing conditions (Simms et al., 2000; Feng and Hsieh, 1998). Although chloropyromorphite is not a redox-sensitive mineral, adding large quantities of P to a wetland would not be an environmentally sound practice (Sharpley et al., 2001). However, a reducing environment may form that supports the presence of sulfate-reducing microorganisms and may cause metals in the tailings to precipitate as highly insoluble sulfides (Simms et al., 2000). Studies have found that a sulfate-reducing environment can be created by adding amendments that provide a food source and sulfur, and raise pH (Feng and Hsieh, 1998).

Studies on metal-contaminated sediments along the Coeur d'Alene River and into Lake Coeur d'Alene have demonstrated that metal sulfides are the dominant form of Pb and Zn in the lake, where a water cap is present year round. In the lake, the correlation between Fe, Pb, and Zn concentrations was weak, indicating that these metals are not necessarily associated with each other. In contrast, a large fraction of both Pb (49.3%) and Zn (63.3%) was associated with sulfides (Harrington et al., 1998). In area wetland sediments, however, the fraction of total Zn present as Zn sulfide varied seasonally with changes in water depth and temperature (Bostick et al., 2001). Bostick et al. (2001) found that the mineral species of Zn was readily altered in response to changes in environmental conditions. A study on the rate of metal sulfide dissolution found that a significant fraction of CdS, PbS, and ZnS oxidized after 300 min of exposure to oxygenated conditions (Simpson et al., 1998). These results, along with field observations, indicate that formation of sulfides may be an effective means to limit metal availability under static conditions.

Additionally, reducing conditions alone may not be sufficient to ensure that the metals of concern would be present as sulfides. An additional factor is the effect of plants on soil parameters. A study examined the influence of plant roots on metal speciation in the rhizosphere. Metal precipitates on the root surfaces of reed canary grass (Phalaris arundinacea L., an indigenous aquatic plant in northern Idaho) included organo–Pb complexes and Zn carbonates (Hansel et al., 2001). This observation indicates that plants can cause localized changes in redox conditions that affect metal speciation. Metal speciation may not be accurately predicted by redox measurements on bulk soil.

West Page Swamp is an 11-ha, naturally occurring wetland located in the Coeur d'Alene River basin near Pinehurst, ID. The area was used as a tailings repository from 1918 to 1929. The depth of tailings within West Page Swamp varies from approximately 45 cm to more than 3 m, with metal concentrations in the tailings ranging from 16 to 249 mg kg^-1 Cd, 1740 to 18400 mg kg^-1 Zn (Huston, 1999), and 4670 to 20700 mg kg^-1 Pb (McCulley, Frick & Gilman, Inc., personal communication, 1993). Organic carbon in the tailings averaged 1.06%. Effluent from the local municipal wastewater treatment plant is scheduled to be diverted into the wetland in the near future. This will maintain a constant water depth over the exposed tailings. This was done to limit wildlife exposure to the tailings. Within one year before sample collection for this study, the wetland had been excavated to accommodate the treatment plant effluent. One to two meters of surface material was removed. The excavation resulted in the removal of all organic matter that had accumulated since the tailings impoundment had closed. In addition, the excavation exposed pure mine tailings that had been buried for more than 50 years.

The present study was conducted to determine the effectiveness of surface amendments of biosolids compost and wood ash, with and without SO₄, for restoring a plant cover and lowering the bioavailable metal concentrations in the tailings. The experiments were conducted in a controlled greenhouse environment with a constant water cover. The amendment cap was anticipated to reduce metal availability through two different mechanisms. On a basic level, application of amendments to the surface of the tailings would provide a physical barrier to the contaminants that would be conducive to plant growth. Second, there is the potential that a surface application of the amendment mixture could alter the mineral form and bioavailability of metals in the underlying tailings. This would reduce the availability of Pb for sediments that are ingested.

Amendments of biosolids compost and wood ash are expected to facilitate SO₄ reduction by organotrophic anaerobes (e.g., Desulfomonas and Desulfovibrio), potentially providing an environment where Pb and Zn are converted to less available forms such as galena (PbS) (solubility product constant [K_sp] = 10^-27.5) and sphlalerite (ZnS) (Paul and Clark, 1996). The initial phase of anoxic conditions would involve dissolution of Fe and Mn oxides. This effect has been observed in a laboratory study with C and N added to anoxic sediment collected from the Coeur d'Alene River (LaForce et al., 1998). However, as E_h continues to decrease, sulfates would also expect to be reduced. If a portion of the metals in the system are associated with Fe or Mn oxides, these should come into solution as the reducing conditions are maintained. If the reducing conditions are maintained, the initial flux of metals into solution, concomitant with dissolution of Fe oxides, should subsequently be reduced with the precipitation of metal sulfides.

The objectives of this research were to determine how various biosolids, ash, and sulfate amendments affected (i) revegetation of tailings-affected sediments, (ii) bioavailability of sediment Pb, and (iii) molecular-scale speciation of Pb. Different, complementary techniques were used to assess changes in Pb bioavailability and speciation induced by the treatments, and included a rapid in vitro extract, AVS to SEM ratio, sediment E_h and pH measures, pore water Pb concentrations, and EXAFS spectroscopy. While in vivo tests are an absolute measure of metal bioavailability to a given organism, several surrogate laboratory procedures have been developed. These include the in vitro test, which was developed to estimate the bioavailable portion of soil Pb to humans (Ruby et al., 1999) and the ratio of AVS to SEM, which has been related to the toxicity of metals to aquatic organisms (Allen et al., 1993; DiToro et al., 1992). The in vitro extraction is conducted at an acidic pH, with a high ratio of solution to soil. The exposure pathway for waterfowl involves ingestion of much higher rates of sediment in combination with foodstuffs. This would alter gastric pH as well as the potential for Pb adsorption. However, this extraction is the closest to a standard method for predicting Pb availability through direct ingestion of soils or sediments.

The underlying concept in analysis of the AVS to SEM ratio is that reactive sulfides are capable of forming stable precipitates with metals in sediments that have a very low solubility, thus controlling their bioavailability (Lee et al., 2000). While neither of these in vitro methods were developed to mimic the behavior of waterfowl, observed reductions using these techniques may reflect lower metal bioavailability for other organisms. It may also be possible to evaluate changes in bioavailability using techniques that document an actual change in the mineralogy of the contaminants to less soluble species. Mineral forms that are stable under equilibrium conditions may be predicted through the use of E_h–pH diagrams (Lindsay, 1979). Metal sulfide minerals can also be detected using spectroscopic techniques (Sayers and Bunker, 1988; Hesterberg et al., 1997).

The goals of the study were to determine (i) if a surface amendment application of compost + wood ash, with or without supplemental SO₄, could provide a fertile barrier to limit access to mine tailings in an anaerobic environment; (ii) if surface amendment application could alter the bioavailability of Pb in the underlying tailings as measured through in vitro procedures; and (iii) if there would be an associated change in the mineral form of Pb.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION DISCUSSION CONCLUSIONS REFERENCES

 
Experimental Design and Treatments
This study was conducted at the University of Washington Center for Urban Horticulture greenhouse in Seattle, WA from May through December 1999. Sediment samples for the study were collected in March 1999 in the north plot (NP) and south plot (SP), within the West Page Swamp in northern Idaho. For each sample collection area, multiple cores were collected and combined. Samples were stored in a cold room at 2°C before the study initiation. There was no standing water at the time the samples were collected. No effort was made to initiate anaerobic conditions after sample collection and before experimental setup. The two sediment sources corresponded to two experimental blocks (NP and SP), with a minimum of two replicates per block.

Four treatments were tested: (i) control, (ii) compost + ash, (iii) compost + ash + low SO₄, and (iv) compost + ash + high SO₄. For the control treatment, pots were filled to a depth of 20 cm with tailings (control). For all other treatments, pots were filled to 8 cm with tailings and covered with 12 cm of biosolids compost and wood ash amendment (compost). The amendment mixture that served as the base for all treatments was made by combining biosolids compost and wood ash at a ratio of 3:1 compost to ash by volume. Biosolids compost was obtained from the City of Coeur d'Alene Wastewater Division Compost Facility. This material was characterized by a pH of 6.0 and a C to N ratio of 20:1 (220 g C kg^-1 and 12 g N kg^-1). Wood ash was obtained from Avista Utilities, a wood-fired electricity generating facility in northeastern Washington. The ash was characterized by a pH of 10.3 and a C to N ratio of 200:1 (186 g C kg^-1 and 1 g N kg^-1). Elemental concentrations of the amendment mixtures and the sediment are reported in Table 1.

Amendment materials were combined in a large plastic container and saturated with H₂O. The H₂O was decanted to reduce salinity. This was done to better replicate the conditions at West Page Swamp, where moving fresh water is able to flush soluble salts out of the system. Electrical conductivity of the amendment mixture was measured with an automatic conductivity bridge and cell (Model 31 conductivity bridge; YSI, Yellow Springs, OH) to ensure that its soluble salt concentration was reduced to within the range typically found in soils (0–4 dS m^-1; Brady and Weil, 1996). The mixture was then divided among three containers. Potassium sulfate (K₂SO₄) was mixed thoroughly into two of the containers at the low and high rates. It was not possible to remove excess salts from the sulfate treatments, as this would have removed a portion of the sulfate amendment.

Plugs of two obligate wetland plant species native to the Coeur d'Alene River basin, arrowhead and cattail, were grown in 3.6-L closed-bottom polyethylene pots (approximately 24 cm high and 14 cm in diameter) containing tailings or tailings covered with each amendment. An illustration of the experimental pots and water and sediment sampling areas within the pot is presented in Fig. 1 . Arrowhead and cattail were selected due to their importance as food and nesting material for wetland birds and small mammals (Stevens and Vanbianchi, 1993). Plant material was obtained from Wildlife Habitat Institute in Princeton, ID and stored in a cold room at 2°C until planting. Arrowhead plants were planted as dormant rhizomes and cattail plants were planted as 16-cm³ plugs with approximately 15 cm of shoot growth. Half of the control and compost-treated pots contained two arrowhead plugs per pot. The remaining half contained two cattail rhizomes per pot. All of the compost + low SO₄– and compost + high SO₄–treated pots contained one arrowhead rhizome and two cattail plugs. In all pots, both species were planted within the top 16 cm of each pot.

View larger version (61K): [in this window] [in a new window]   Fig. 1. A diagram of the experimental units used for the study. The diagram shows the depth of the tailings, compost amendment, and water cover. It also shows the amendment, interface, and tailings sampling points.

The pots were submerged in 19-L containers of water to provide a 5-cm water cover over the top of each substrate. These containers were arranged in a completely randomized design. Water was added as necessary throughout the duration of the experiment to maintain a constant water level. Porous ceramic water samplers (2.2-cm o.d., 6.4-cm length) attached to plastic tubes (2.2-cm-o.d., 30-cm length; Soilmoisture Equipment Corp., Santa Barbara, CA) were installed in four pots, one of each treatment (control, compost, compost + low SO₄, and compost + high SO₄), at three depths: 12 cm (high), 7.5 cm (middle), and 4 cm (low) from the bottom of the pot. To facilitate comparisons, all pots selected for water samplers contained SP sediment. In those pots containing a layer of amendment, the high sampler was located within the amendment layer, the low sampler was located within the tailings layer, and the middle sampler was located just beneath the interface between the amendment and tailings layers (Fig. 1).

Analysis of Soil Pore Water
Water samples were collected from the ceramic water samplers once a week with a 2-mm tube attached to a syringe. Although interactions between metal ions and the ceramic cups may occur, the relative concentration of Pb in the measured solution facilitates valuable comparisons. Samples were acidified with concentrated HCl and stored at 2°C before analysis by flame atomic absorption spectrometry (Model 5100 atomic absorption spectrometer; PerkinElmer, Wellesley, MA) to estimate pore water Pb concentrations.

Redox Potential and pH Measurements
Reduction–oxidation (i.e., redox) potential (E_h) was measured three times per week using an Orion 520A pH/mV/ORP/temperature meter with an Orion 96-78 redox platinum electrode (Thermo Orion, Beverly, MA). This same meter with an Orion Model 91-07 platinum electrode was used to measure pH once per week. All measurements were taken in the top 2 cm of the substrate. Reported E_h values are relative to a standard hydrogen electrode.

Analysis of Plant Biomass
Aboveground, living plant tissue was harvested when plants exhibited toxicity symptoms (e.g., chlorosis or senescing leaves). Plants growing in control sediment were harvested from 38 to 59 d. All plants growing in compost-treated sediment were harvested after 94 d. Following harvest, plant tissues were washed in a mild sodium lauryl sulfate solution, rinsed in deionized water, and dried at 70°C. Dried samples were weighed and ashed at 480°C in a muffle furnace for 16 h, then digested in concentrated HNO₃ and dissolved in 3 M HCl. Extracts were diluted to 25 mL using 0.1 M HCl, and subsequently analyzed by inductively coupled argon plasma emission spectroscopy (ICP) (Thermo Jarrell Ash Model 61 ICAP–AES; Thermo Elemental, Franklin, MA) to determine metal concentrations. For quality control, National Institute of Standards and Technology (NIST) standards were routinely integrated into plant and soil analyses.

Analysis of Sediment
One half of the pots for each treatment were dismantled after 99 d. The remaining pots were maintained for a total of 207 d. This was done to compare changes in contaminant availability over time. At the time that the pots were dismantled, soil samples were collected from each of the respective levels: sediment, interface, and amendment. Interface samples were collected from the area just beneath the junction of the sediment and amendment layers. For each pot, samples from each layer were subdivided. Half of each sample was air-dried and sieved. The remaining portion was kept wet and stored in a cold room at 2°C. Total metal concentrations were determined for the dry <2-mm and <250-µm soil fractions by aqua regia digestion (McGrath and Cunliffe, 1985) followed by analysis by ICP.

Rapid In Vitro Analysis
The in vitro extraction was conducted with a pH 2.5 solution for 1 h at 37°C using <250 µm dry soil and a 1:100 soil to solution ratio with a 0.4 M glycine solution (Brown et al., 2003; Ruby et al., 1999, 2001). John Drexler (Dep. of Geological Sciences, Univ. of Colorado at Boulder) performed the in vitro analysis for samples collected after 99 d. Sample analysis for samples collected at 207 d, using the same method at pH 2.2, was conducted at the University of Washington. There was no difference in results as a function of extractant pH or time, so results have been averaged over both extraction times. In addition, the in vitro procedure was conducted using moist samples collected after 207 d. Percent moisture was calculated for the samples and extractable Pb was adjusted accordingly.

Analysis of Acid Volatile Sulfide to Simultaneously Extracted Metals Ratio
The analysis of AVS to SEM ratio was conducted by Savannah Laboratories in Savannah, GA. In preparation for shipping, wet sediment was placed in 60-mL glass vials sealed with septa lids. Samples were packed in an ice chest with dry ice and shipped overnight. Samples were collected at plant harvest. A minimum of four samples for the bottom and interface depth of the control and compost samples were analyzed. However, because we wanted to maintain the experiment, only one sample at each depth for both the low-S and high-S treatments were analyzed. For analysis, 10-g samples were placed in a vessel that had been purged with N₂. Twenty milliliters of 6 M HCl were added. Sulfur, volatilized as H₂S, was trapped in a scrubber containing NaOH. The scrubber solution was then analyzed for total S colorimetrically. Simultaneously extracted metals in the sediment and HCl solution were analyzed by ICP after filtering.

X-Ray Absorption Spectroscopy
Sample Preparation and X-Ray Absorption Spectroscopy Data Collection
Extended X-ray absorption fine structure (EXAFS) and X-ray absorption near edge structure (XANES) spectroscopy analysis of Pb speciation was conducted on eight sediment samples at Beamline X-11A of the National Synchrotron Light Source at Brookhaven National Laboratory in Upton, NY. To facilitate comparisons, all samples were collected from pots containing SP sediment. These included one replicate of each treatment from pots dismantled after 99 d (all from the interface level), and one of each treatment from pots dismantled after 207 d (all from the interface level). Wet sediments were passed through a 500-µm stainless steel sieve and placed in high density polyethylene scintillation vials. To prevent desiccation, individual vials were wrapped in moist paper towels and placed in glass jars. The jars were subsequently purged with N₂ gas and sealed with Teflon ribbon–lined lids to exclude O₂. Samples were packed in an ice chest with dry ice and shipped overnight for X-ray absorption preparation. On receipt, samples were mounted under an Ar atmosphere in special acrylic sample holders, covered with Kapton tape, and placed in cold (3°C) storage until X-ray absorption data were collected.

The X-ray absorption data were collected on moist sediment samples at ambient temperature across the lead L_III absorption. Sample data were collected in transmission mode using ionization chamber detectors. A metallic Pb foil placed between two detectors behind each sediment sample was used as a reference to correct for any energy shifts due to beamline optics. At least two EXAFS spectra collected on samples and standards were ensemble-averaged (one spectrum was collected for the 99-d control sample).

Extended X-Ray Adsorption Fine Structure Data Analysis
Data processing was done using the computer program MacXAFS (Bouldin et al., 1995). The EXAFS spectra were baseline-corrected using a linear function between -200 and -50 eV relative energy, where the edge energy (E₀) was taken as the maximum in the first derivative spectrum and occurred between 13 055 and 13 058 eV. The EXAFS background normalization was done across a relative energy range between about 13 and 550 eV (wave vector k = 1.9–12.4 Å^-1) using a cubic spline function with five equally spaced knots (Sayers and Bunker, 1988). An exception was the compost-treated sediment sample, for which five knots were placed near the nodes of the EXAFS oscillations to better remove low-frequency background oscillations as indicated by a false peak in the Fourier-transformed spectrum (radial structure function [RSF]) at a radial distance of 1.2 Å. The radial structure functions were produced by Fourier transformation of k³ weighted (w = 3) chi [(k)] data over a wave vector range between 2.4 and 10.5 Å^-1, with the exact endpoints for a given spectrum chosen at nodes ( = 0) in the EXAFS spectrum (Mansour and Melendres, 1998). No windowing functions were used in EXAFS transformations.

Multishell EXAFS fitting analysis was done using a combination of theoretical and physical standards following the general approach described by Hesterberg et al. (1997). Up to three shells (Pb–O, Pb–S, and Pb–Pb) were fit simultaneously. The mineralogical purity of the physical standards was verified using X-ray diffraction analysis. In order to determine bonding parameters (radial distance, coordination numbers, and Debye–Waller factors [²]), the amplitude, phase shift, and theoretical EXAFS calculations were done with the University of Washington FEFF computer programs (Rehr et al., 1992). Theoretical EXAFS spectra for first-shell Pb–S or Pb–O coordination and higher-shell Pb–Pb coordination were generated using structural parameters for PbS (galena) and –PbO (litharge), which have uniform first-shell coordination distances (Moller et al., 1989; Manceau et al., 1996). These spectra were used in fitting analysis to determine average bonding parameters for Pb species in the sediment samples. Amplitude reduction factors for Pb–S coordination and Pb–Pb coordination were obtained by fitting the measured spectrum for PbS (galena) with their respective theoretical EXAFS spectra (Hesterberg et al., 1997), and taking the ratio of the fitted coordination numbers to the known crystallographic values of 6 (Pb–S) and 12 (Pb–Pb), respectively (Moller et al., 1989). The amplitude reduction factor of 0.83 determined for Pb–S bonding was also used to correct coordination numbers for Pb–O bonding because fitting of the ß–PbO standard yielded a reduction factor (0.47) that was considered to be unrealistically low. The amplitude reduction factor is deemed to be more a function of electronic effects on the central Pb atom than on the coordinating atoms (Sayers and Bunker, 1988). Furthermore, Pb L_III–EXAFS can underestimate the total number of oxygen neighbors, particularly for Pb species having a high degree of anharmonic static disorder (multiple Pb–O bond lengths), as is typical of oxygen-bonded Pb(II) (Manceau et al., 1996).

The EXAFS fitting analysis was done only on sediment samples collected at 99 d because data from the 209-d sampling showed a monochromator glitch across the wave vector range k = 7.3 to 8.0 Å^-1 that could not be reliably removed. Spectra for the four 99-d samples were fit in r space over a k range between 2.5 and 10.1 Å^-1 and a radial distance range between 0.9 and 4.7 Å. The exact k range is shown by the fits to the chi spectra in Fig. 1 (discussed below), and the exact r range for fitting depended on the positions of peaks in the radial structure functions (Hesterberg et al., 1997). Simultaneous three-shell fitting (Pb–O, Pb–S, Pb–Pb) was tried on all samples, and the Pb–O shell was eliminated when fitting yielded a negative Pb–O coordination number. In two cases where an initial fitting analysis yielded a slightly negative ² (Debye–Waller factor), suggesting better structural order than the theoretical standard with ² = 0 Ų, ² was fixed at a low value (0.0001) to complete the fitting analysis. The energy shift (E) was constrained in the fitting to be equivalent for all shells, so that the number of fitted parameters was always less than the number of independent points (Fendorf, 1999).

X-Ray Absorption Near Edge Structure Data Analysis
To determine whether we could detect changes in the speciation of sediment Pb between samples collected at 99 and 207 d, lead L_III-XANES data were analyzed. X-ray absorption data that had been baseline-corrected were normalized to an edge step of 1 at 13 175 eV. To more quantitatively estimate differences in Pb speciation between samples, we performed nonlinear least squares fitting (linear combination fitting) analyses on the XANES data as described in Hutchison et al. (2001). The fitting analysis was done using XANES data for four mineral standards: PbS (galena), PbO (massicot), PbCO₃ (cerrusite), and PbSO₄ (anglesite). Only fits that included the PbS were considered valid, because EXAFS data indicated the presence of PbS in all sediment samples. Only binary combinations of the standards yielding the best goodness of fit (lowest ² value) are reported to avoid an excess number of fitted parameters. Three energy ranges were used in the fitting (13 030–13 130, 13 030–13 090, and 13 030–13 075 eV), and the results were averaged.

Data Analysis
Data were statistically analyzed using SPSS Version 10.0 (SPSS, 1999). Sediment elemental concentrations were compared using a t test. Plant tissue Fe and Pb data was log-transformed before statistical analysis to obtain normally distributed data. For ease of interpretation, nontransformed values are presented. Plant tissue and rapid in vitro data were tested for treatment and block effects with analysis of variance (ANOVA) followed by multiple comparison testing using the Waller–Duncan t test. A significance level of p < 0.05 was selected for determination of statistical significances. Since no statistical difference was found between rapid in vitro samples collected on two different sampling dates (Days 99 and 207), these values were combined for each depth of each treatment.

Field Application of Amendment
After samples had been collected from the West Page Swamp, the swamp itself was amended with compost and wood ash. Amendments were added to the surface of the swamp in October 1998 and September 2000. Compost and wood ash were mixed with a front-end loader in the same proportions as used in the greenhouse study. The amendment was applied using an aerospread in 1998 and with a blower truck in 2000. A road was constructed through the swamp from woody waste from local log yards to facilitate amendment application. Volunteer plant species were allowed to revegetate the swamp. Plant data from the actual swamp restoration is also presented. Samples were collected from the excavated portion of the swamp before the amendment application in 1998 and from the amended portion of the swamp in 2000. For the control, two plants were sampled and 14 plants from five different portions of the treated areas of the swamp were sampled in 2000.

	RESULTS AND DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION DISCUSSION CONCLUSIONS REFERENCES

 
Soil Redox Potential
Soil redox potential (E_h) values are shown in Fig. 2 . Redox potential was consistently highest in the control treatments, with measurements ranging between 100 and 400 mV for the duration of the study. Sulfide-reducing (<-160 mV) conditions (Bohn et al., 1985) were generally reached in the compost + low SO₄ treatment by 20 d, and in the compost + high SO₄ treatment after 8 d. The E_h values in these treatments remained less than -160 mV for the duration of the experiment. Compost (no SO₄) treatments produced the greatest redox fluctuations. The E_h readings in these pots were initially similar to those in the compost + SO₄ treatments. Starting on Day 27, however, E_h in the compost treatment increased to >-160 mV while E_h in the treatments with SO₄ additions remained highly reducing.

View larger version (31K): [in this window] [in a new window]   Fig. 2. Sediment redox potential (Eh) values averaged over all replicates of the north plot (NP) and south plot (SP) sediments for each sampling period for the greenhouse study. Measurements were taken in the top 2 cm of substrate.

Soil pH
Soil pH values are shown in Fig. 3 . Average sediment pH in control pots ranged from 6 to 7 for the duration of the study, while all other treatments exhibited pH > 7. The compost + low SO₄ and compost + high SO₄ pots had the highest average pH readings, which ranged between pH 8 and 9. Average pH in the compost pots was initially in this same range, but began to decrease after approximately 28 d to pH < 8, where it remained for the rest of the study.

View larger version (27K): [in this window] [in a new window]   Fig. 3. Sediment pH values for the greenhouse study. Measurements were taken in the top 2 cm of substrate. Data points represent the means of all replicates averaged over north plot (NP) and south plot (SP) sediment for each sampling date.

View larger version (25K): [in this window] [in a new window]   Fig. 4. Redox potential (Eh) and pH data from the control, compost, compost + low SO4, and compost + high SO4 treatments plotted on an Eh–pH diagram for a range of Pb minerals. This diagram does not take into account changes in stability based on variable sulfur concentrations.

View this table: [in this window] [in a new window]   Table 2. Average pore water Pb concentrations by depth within pot, averaged over all sampling times.

All of the plants in the compost + low SO₄ and compost + high SO₄ treatments died almost immediately following planting. Salt toxicity as a result of the high rates of SO₄ addition was the likely cause of their demise. Average electrical conductivity readings of the compost + low SO₄ and compost + high SO₄ treatments were 17 and 41 dS m^-1, respectively. The average electrical conductivity readings of the control and compost treatments, in contrast, were 1.0 and 0.6 dS m^-1, respectively.

In the amended portion of the wetland, volunteer plant species had completely colonized the treated areas within one year after application. The most likely seed source was an adjacent undisturbed wetland. Wind-blown seed and seed from bird droppings were the probable sources of plant material. Vigorous growth has been maintained, as was evidenced in a site visit in October 2002.

Elemental Concentrations in Aboveground Plant Tissue
Mean elemental concentrations in aboveground plant tissue are shown in Table 3. The compost amendment was sufficient to alleviate the phytotoxic conditions in the unamended sediment. Concentrations of Cd, Pb, and Zn in plants grown in this treatment were significantly (p < 0.05) reduced over those grown in the control sediment. Arrowhead and cattail plants in the control pots averaged 50 and 13 mg Cd kg^-1 while treated plants averaged 3 and 4 mg Cd kg^-1, respectively. Zinc concentrations in control plant tissue (900–1600 mg kg^-1) exceeded the phytotoxicity threshold (>400 mg kg^-1; Knezek and Ellis, 1980); concentrations in compost-treated plants (100–300 mg kg^-1), in contrast, were below phytotoxic range. Lead concentrations were reduced from 80 to 300 mg kg^-1 in control plants to 1 to 2 mg Pb kg^-1 in treated plants. The observations in the wetland itself mirrored those in the greenhouse study. Plant tissue Zn was reduced from 1000 mg kg^-1 in the unamended wetland to an average value of 115 mg kg^-1 in the years following amendment addition. These data suggest that consumption of aboveground plant tissue grown in compost-treated, metal-contaminated sediments would result in lower food chain transfer of these elements.

At both the interface and bottom sampling depths, total SEM values for all treatments were similar, averaging 78.7 ± 0.9 mmol kg^-1. For both SO₄ treatments, there was no detectable AVS at the bottom depth. The AVS in both the control and compost alone treatments at this depth averaged 1 ± 0.33 mmol kg^-1. For all treatments, the ratio of AVS to SEM clearly indicated potentially toxic conditions (Fig. 5) . At the interface depth, however, there were significant increases in AVS as a result of amendment addition. The AVS increased in the control treatment to 2 ± 0.5 mmol kg^-1. This increase may have been the result of increased microbial activity closer to the surface. In the compost alone treatment, AVS also increased to 4 ± 1.5 mmol kg^-1. The most significant increases in AVS were observed in the compost + SO₄ treatments. This is expected, due to the higher concentrations of sulfur in these treatments. The AVS increased to 8 mmol kg^-1 in the low SO₄ treatment and to 25 mmol kg^-1 in the high SO₄ treatment. It should be noted that E_h, measured about 12 cm above the samples collected for AVS/SEM, was highly reducing in both sulfur treatments at the time the samples were collected (99 d). The E_h in the control measured 122 mV, with E_h in the compost equal to -66 mV. For the low SO₄ treatment, E_h measured -235 mV and was -245 in the high SO₄ treatment. The observed increase in AVS makes sense both because of increased SO₄ concentration as well as increased microbial activity as evidenced by the increased electron activity. Although this data suggests a decrease in toxicity as measured by AVS/SEM primarily in the treatments that contained SO₄, the implications for waterfowl that ingest sediment directly are not clear.

View larger version (44K): [in this window] [in a new window]   Fig. 5. Acid volatile sulfide (AVS) to simultaneously extracted metals (SEM) ratio from the tailings at the interface and bottom level (99 d) from amended and control treatments in the greenhouse study. Different letters above bars indicate significantly different values using the Duncan–Waller means separation procedure (p < 0.05).

The protocol for this extraction calls for materials to be air-dried. It was initially developed to measure Pb availability in aerobic soil systems with a human endpoint. As the primary concern in this case is for waterfowl that ingest the reduced sediment, it may be appropriate to carry out this extraction using wet materials. For wet samples, the observed decrease in bioavailability was more pronounced. The high SO₄ treatment showed a 41 and 24% reduction in bioavailability at the interface and bottom levels, respectively. In addition, the compost-alone amendment reduced bioavailability in the NP sediment by 18% over the control. There were not sufficient wet tailings to run the rapid in vitro extract procedure for the low SO₄ for all experimental units. However, for the single samples that were extracted, this treatment showed a 70 and 35% reduction in bioavailability at the interface and bottom depths. These data suggest that the addition of a compost layer to tailings may result in reduced sediment Pb bioavailability, and supplementary SO₄ decreases bioavailability even further.

The rapid in vitro extraction was also performed on the amended portion of the columns. Here, total concentrations of Pb were significantly lower than the Pb in the sediment. In addition, a smaller fraction of total Pb was bioavailable. All of the treatments were effective at this level, with reductions in bioavailability over the control ranging from 42 to 50% in the compost-alone treatment to 76 to 80% in the compost + SO₄ treatments. We found that the percent of available Pb in the tailings located in the bottom and interface levels was between 44 and 63%, whereas tailings in amendment level was about 96 and 99%, or almost all available. In sediment or tailings that contain very high concentrations of Pb, the location of the sediment within the soil profile may be very important both in terms of accessibility and availability. In addition, the compost amendment may serve as an effective barrier to the contaminated sediment.

X-Ray Absorption Spectroscopy
Extended X-Ray Adsorption Fine Structure Results
Figure 6 shows experimental lead L_III–EXAFS spectra (chi data) overlaid with model fits for interface-level pot samples collected after 99 d and two mineral standards. The model fits represent the multishell, EXAFS fitting results shown in Table 5. Radial structure functions derived from Fourier transformation of the k space data are shown in Fig. 7 . A comparison of EXAFS results for samples with those for the PbS (galena) standard indicates that lead sulfide was present in all four sediment samples. This result is particularly evident from the correspondence of the EXAFS oscillations between the sediment samples and galena in Fig. 6, and the high-magnitude peak at about 2.4 Å (uncorrected for phase shift) in Fig. 7. Fitting analysis (Table 5) yielded an average Pb–S bond length between 2.85 and 2.91 Å for sediment samples, which was comparable with the Pb–S bond length determined for galena (2.92 Å). The discrepancy between these results and the published Pb–S bond length of 2.97 Å for galena (Moller et al., 1989) is consistent with underestimation of Pb(II) coordination distances by ambient temperature EXAFS analysis (Manceau et al., 1996). Higher-shell Pb–Pb distances for all sediment samples (4.12–4.17 Å) also corresponded (within 0.08 Å) with Pb–Pb interatomic distances of galena, which was about 0.6 Å greater than the Pb–Pb distance determined for ß–PbO (Table 5).

View larger version (38K): [in this window] [in a new window]   Fig. 6. Stacked, normalized k3–weighted LIII–extended X-ray adsorption fine structure (EXAFS) spectra (chi data) for lead in sediment samples taken after 99 d of incubation with various treatments, and mineral standards. Experimental spectra (solid lines) are overlaid with spectral fits (dashed lines) based on the fitting parameters reported in Table 5.

View this table: [in this window] [in a new window]   Table 5. Lead LIII–extended X-ray adsorption fine structure (EXAFS) spectroscopy fitting results for sediment samples collected after 99 d of incubation with various treatments, and selected standards.

View larger version (30K): [in this window] [in a new window]   Fig. 7. Stacked radial structure function (RSF) magnitudes derived by Fourier transformation of k3–weighted, lead LIII–extended X-ray adsorption fine structure (EXAFS) spectra shown in Fig. 6. Radial distances are not corrected for phase shift, and therefore they do not reflect true interatomic distances.

X-Ray Absorption Near Edge Structure Results
To evaluate whether any detectable changes in Pb speciation of sediments from the potting study occurred over time, we compared lead L_III–XANES results between samples taken at 99 and 207 d. Results showed detectable changes in Pb speciation between samples for the same treatment taken at different times (Fig. 8) . In comparing possible oxygen-bonded and sulfur-bonded mineral species of Pb, XANES data for the standards used in this study showed that PbSO₄ and PbCO₃ had a more distinct white-line (WL) peak at 13 065 eV than PbS. The WL peak of ß–PbO was uniquely shifted to a greater energy. The shapes and intensities of the WL peaks for the spectra for the sediment samples were intermediate between those of the oxygen-bonded Pb species (particularly PbCO₃ and PbSO₄) and PbS. Compared with the SO₄–amended samples (low SO₄ and high SO₄ treatments), the XANES spectra for the control and compost samples exhibited a more distinct WL peak that was more characteristic of spectra for PbCO₃ and PbSO₄ than PbS.

View larger version (31K): [in this window] [in a new window]   Fig. 8. Stacked, normalized LIII–X-ray absorption near edge structure (XANES) spectra for lead in sediment samples and standards. Sediment samples were collected at 99 d (solid line) or 207 d (dashed line) for each treatment. Data for standards are represented as solid lines for PbCO3 and PbSO4, and dashed lines for PbS and PbO.

View this table: [in this window] [in a new window]   Table 6. Lead LIII–X-ray absorption near edge structure (XANES) spectroscopy fitting results showing the relative proportions (mole percentages) of PbS, PbCO3, and PbSO4 standards yielding the best fits to the soil XANES data in binary (or ternary) linear combination fitting.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION DISCUSSION CONCLUSIONS REFERENCES

The EXAFS and XANES analyses indicated that a portion of the Pb in all treatments was present as PbS. However, the surface application of compost + ash + sulfate (at each of two levels) was especially effective in promoting an increase in the proportion of total Pb bound as lead sulfide in the underlying sediment, apparently through sulfate reduction. These observations were in accord with the pH and E_h measurements that indicated that PbS would be stable in these treatments. For the compost and ash treatments, it is not clear if increased redox potential and the reduced rate of PbS formation is an artifact of the pot study or represents what is occurring in the field. It is likely that the plant roots grew into the contaminated sediment because of the small size of the pots. In the field, root growth may be limited to the amended horizon. If this is the case, it is likely that reducing conditions would persist in the underlying tailings as they did in the 40- to 90-d portion of the greenhouse study.

Changes in Pb speciation induced by sediment amendments should directly affect the environmental effects of sediment Pb. Previous in vivo studies indicated that the bioavailability of Pb to swine (Sus scrofa; a surrogate for humans) is directly affected by the mineral form of Pb present in the soil (Ruby et al., 1994, 1996). Furthermore, the portion of total Pb that was bioavailable in PbS (galena) was less bioavailable Pb than in other Pb minerals, including PbSO₄ (anglesite) and PbCO₃ (cerrusite). Thus, an increase in lead sulfide as a proportion of total Pb in the sediment is expected to decrease the bioavailability of Pb from ingested sediment. Indeed, our results from indirect chemical (in vitro) assessments (rapid in vitro extract lead and AVS to SEM ratio) indicated that the compost + ash + SO₄ treatments (and compost + ash alone) should decrease bioavailable Pb, particularly at the sediment–cap interface level (Table 4, Fig. 5). It should be noted, however, that each of the bioavailability procedures used was developed to predict changes in toxicity for very different exposure pathways. The rapid in vitro extract was designed to mimic the human gastric system. The AVS/SEM procedure was developed to predict toxicity to benthic organisms, such as clams and aquatic worms, whose feeding habits and digestive systems differ substantially from vertebrates. While the results of this assay may not relate directly to vertebrates other than humans, indications of decreasing Pb bioavailability to other organisms may nonetheless be pertinent. In addition, plant tissue data (Table 3) indicated that a compost + ash treatment decreased the amount of Pb taken up by wetland plants. It should also be noted that these mineral shifts and reductions in in vitro assays were achieved only when the treatments had a consistent water cap. Results would be expected to differ with a fluctuating water table and similar reductions would not be expected.

	CONCLUSIONS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION DISCUSSION CONCLUSIONS REFERENCES

 
This potting study indicated that the risk of Pb poisoning to waterfowl from sediments at the West Page swamp could be reduced by amendments of compost + ash (with or without supplementary sulfate addition). This risk reduction may be achieved by one or a combination of several mechanisms: (i) an overlying cap of a compost + ash amendment would serve as a physical barrier to diminish ingestion of contaminated sediment by waterfowl; (ii) the amendments chemically alter Pb in the tailings at the amendment–sediment interface such that a greater proportion of the Pb is bound as lead sulfide, which results suggest would be less bioavailable if the sediment were ingested; and (iii) a compost + ash treatment without supplementary sulfate allowed vigorous plant growth, probably by providing a more favorable rooting medium for the plants. Furthermore, the Pb content of the plant tissue was diminished in the compost + ash–amended sediment compared with a control treatment, which would reduce Pb exposure via ingestion of plant tissue.

Field observations since this potting study confirm the effectiveness of the compost + wood ash treatment. In the portion of the West Page Swamp that was treated with a surface application of compost and ash in 1998, volunteer plants restored a lush cover to the area by 1999, and this cover persists. Plant tissue data show concentrations of Pb, Cd, and Zn comparable with what was observed in the greenhouse (Table 3). Continued monitoring of the swamp indicates that wildlife have returned to the area (M. Sprenger, USEPA Environmental Response Team, personal communication, 2001).

	ACKNOWLEDGMENTS

 
This research was carried out (in part) at the National Synchrotron Light Source (NSLS), Brookhaven National Laboratory, which is supported by the U.S. Department of Energy, Department of Materials Sciences and Division of Chemical Sciences. Financial support for the EXAFS part of this work came from a DuPont Educational Aid Grant. The authors also thank John Drexler at the University of Colorado for conducting rapid in vitro analysis and Chuck Henry (University of Washington) for assistance in collecting sediment. The authors are grateful to Dr. Dale E. Sayers, North Carolina State Department of Physics, for helpful discussions. Funding for this project was also provided by the USEPA Environmental Response Team.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS AND DISCUSSION DISCUSSION CONCLUSIONS REFERENCES

 

• Allen, H.E., G. Fu, and B. Deng. 1993. Analysis of acid-volatile sulfide (AVS) and simultaneously extracted metals (SEM) for the estimation of potential toxicity in aquatic sediments. Environ. Toxicol. Chem. 12:1441–1453.
• Beyer, W.N., O.H. Pattee, L. Sileo, D.J. Hoffman, and B.M. Mulhern. 1985. Metal contamination in wildlife living near two zinc smelters. Environ. Pollut. Ser. A 38:63–86.
• Beyer, W.N., D.J. Audet, A. Morton, J.K. Campbell, and L. LeCaptain. 1998. Lead exposure of waterfowl ingesting Coeur d'Alene River basin sediments. J. Environ. Qual. 27:1533–1538.[Abstract/Free Full Text]
• Bohn, H., B.L. McNeal, and G.A. O'Conner. 1985. Soil chemistry. John Wiley & Sons, New York.
• Bostick, B.C., C.M. Hansel, M.J. LaForce, and S. Fendorf. 2001. Seasonal fluctuations in zinc speciation within a contaminated wetland. Environ. Sci. Technol. 35:3823–3829.[Medline]
• Bouldin, C.E., L. Furenlid, and W.T. Elam. 1995. MACXAFS—An EXAFS analysis package for the Macintosh. Physica B (Amsterdam) 208/209:190–192.
• Brady, N.C., and R.R. Weil. 1996. The nature and properties of soil. 11th ed. Prentice Hall, Upper Saddle