Journal of Environmental Quality 31:822-828 (2002)
© 2002 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
TECHNICAL REPORTS
Heavy Metals in the Environment
Arsenic Remobilization in a Shallow Lake
The Role of Sediment Resuspension
Kathryn L. Linge* and
Carolyn E. Oldham
Centre for Water Research, Univ. of Western Australia, 35 Stirling Hwy., Crawley, Australia, 6009
* Corresponding author (linge{at}cwr.uwa.edu.au)
Received for publication April 26, 2001.
 |
ABSTRACT
|
|---|
Oxic resuspension occurs regularly in shallow lakes, yet its role as a mechanism for contaminant remobilization remains ill defined. This study investigated contaminant remobilization during sediment resuspension and determined whether changes in contaminant sediment partitioning reflected the mechanisms controlling remobilization. Arsenic-contaminated sediment from a shallow wetland was subjected to simulated resuspension under a range of differing initial pH conditions. The effect of resuspension on As partitioning was evaluated using a fractionation scheme targeting the dissolved, ion exchangeable, carbonate, organic, amorphous iron oxide, crystalline iron oxide, and apatite fractions. Rate investigations demonstrated that arsenic remobilization occurred on timescales similar to resuspension events, with concentrations reaching steady state within 24 h. The sediment also buffered slurry pH to 8.3 in experiments where the initial pH was between 4 and 10. This pH regulation was attributed to carbonate dissolution or acid–base equilibria of surface base functional groups, although iron oxide and organic matter dissolution did occur in experiments with an initial pH outside this range. Remobilization caused losses in arsenic associated with the ion exchangeable, organic, and amorphous iron fractions but changes in initial pH have a negligible effect on arsenic remobilization or partitioning within the well-buffered region. Resuspension released approximately 20% of the total sediment arsenic, although calculations indicated that a single resuspension event would not significantly change water column arsenic concentrations. While not conclusively proving the mechanisms of remobilization, fractionation gave valuable insight into the effect of sediment resuspension on contaminant remobilization.
Abbreviations: XRF, X-ray fluorescence
|
INTRODUCTION
|
|---|
SEDIMENT RESUSPENSION in both shallow and deep lakes is common and a well-known mechanism for sediment contaminant remobilization. Contaminant remobilization caused by the resuspension of anoxic sediments into an oxic water column has been widely documented (Förstner, 1995; Simpson et al., 1998; Saulnier and Mucci, 2000). Generally, anoxic sediment resuspension causes the oxidation of sulfide minerals, leading to heavy metal remobilization and a significant decrease in pH. Contaminant remobilization from oxic sediment resuspension is less well understood. Given that oxic resuspension may be ubiquitous in shallow (<3 m) lakes, its role in contaminant cycling must be clarified.
Several studies have shown that the pH of productive waters can significantly fluctuate with changes in algal biomass on both diel and seasonal timescales (Eloranta, 1985; Maberly, 1996). The exposure of sediments to changes in pH during resuspension may additionally promote contaminant remobilization (Meyer et al., 1994). However, the effect of such changes will largely be determined by the mechanisms controlling contaminant remobilization. A balance between binding processes (e.g., adsorption and precipitation) and remobilization processes (e.g., desorption from the sediment surface and mineral dissolution) determines both contaminant availability and its partitioning between different sediment phases, such as organic matter, carbonate minerals, and metal oxides (Förstner, 1990). While adsorption is inherently two-dimensional and precipitation three-dimensional, both cause a loss of material from an aqueous solution phase (Sposito, 1984). Furthermore, the chemical bonds formed in both cases can be very similar and precipitates are often inhomogeneous with one component restricted to a thin outer layer because of poor diffusion. Traditional methods, such as data-fitting to adsorption isotherms and the calculation of ion activity products, do not conclusively distinguish between the two processes. However, insight into the mechanisms controlling remobilization may be gained by investigating changes in contaminant partitioning between different phases before and after sediment resuspension.
The role of oxic resuspension in contaminant remobilization, particularly in conjunction with mild changes in pH, requires further examination. This study describes the remobilization and partitioning of arsenic in Lake Yangebup, a shallow (<3 m), urban lake in Perth, Western Australia (Davis et al., 1993; Arnold and Oldham, 1997). Lake Yangebup exhibits significant metal and nutrient contamination and an elevated pH due to the input of effluent from nearby industry. The lake regularly experiences extensive algal blooms, which are believed to be responsible for the seasonal pH fluctuation of approximately one pH unit. Wind speeds regularly exceed 6 m s-1 for periods of between 6 to 10 h and typically lead to sediment resuspension.
Prior investigations have shown the presence of a flocculant sediment layer covering more consolidated sediment (Arnold and Oldham, 1997). Arnold and Oldham (1997) confirmed that rapid arsenic remobilization occurs from the floc sediment of Lake Yangebup under both oxic and anoxic conditions. For contaminant management purposes, it has been proposed that the removal of the floc layer, which has higher arsenic concentrations than the underlying consolidated sediment, could lower dissolved arsenic concentrations. However, further investigation was required to determine the potential for arsenic release from the underlying consolidated sediment and the factors that control this remobilization (e.g., time, pH, Eh).
Consolidated sediment samples were subjected to simulated resuspension to evaluate the effect of pH changes on As remobilization. In addition to As, the behavior of Fe was also investigated to determine its role in arsenic binding. Changes in both As and Fe partitioning were measured before and after resuspension to indicate which phases contributed to remobilization. The rate of As remobilization was also measured to determine whether the sediment could respond to changes in the localized environment caused by resuspension events that potentially last from minutes to hours.
|
METHODS
|
|---|
Sediment Sampling
Sediment samples were collected from Lake Yangebup on two separate excursions (16 Apr. 1999 and 17 Jan. 2000). Subsequent X-ray fluorescence (XRF) analysis found little difference between the elemental composition of sediment collected on each date. On each occasion several sediment cores were collected from the center of the lake (Site 2 in Arnold and Oldham, 1997) using a specially devised corer. The corer consisted of a transparent perspex tube with a removable stainless steel blade attached at one end and a detachable handle at the other. The corer was manually pushed into the lake sediments, the handle removed, the tube filled with lake water and the top end sealed with a plastic screw cap. This seal allowed the corer to be pulled from the sediment without disturbing the sediment core inside. The bottom cap was then attached and the core stored upright to minimize sediment disturbance. When sampling lake sediments, it is widely recommended to store the sediment at low temperatures immediately after sampling. In this case, the length of the perspex corers (approximately 1 m) made this impractical. However, where possible, cores were processed as soon as they reached the laboratory (usually within 5 h of collection) and then stored at 4°C.
The sediment was removed from the corer by first siphoning off the overlying water column. The sediment was then gently pushed out of the perspex corer. The floc layer was discarded and samples of the underlying consolidated sediment from each core were mixed thoroughly to produce a bulk sample that was then subsampled. This was done because preliminary tests had shown that the sediment could show variation in As concentration between cores. Subsampling of the bulk consolidated sample occurred immediately to minimize loss of moisture content.
Fractionation Scheme
A sequential extraction (or fractionation) procedure was used to investigate the partitioning of As and Fe between a number of operationally defined fractions. The scheme was based on that devised by Tessier et al. (1979), although some modifications were made to minimize the interference of organic matter during those steps targeting iron oxide minerals. An aliquot (20 mL) of each extractant was added to a sediment sample of known mass and the resulting slurry extracted under the conditions specified in Table 1. After each extractant, the solid and liquid phases were separated by centrifuge (30 min at 15000 rpm). The supernatant was filtered through a 0.45-µm filter and stored at 4°C before analysis. The residual solid was also washed between extractants with approximately 10 mL of deionized water and centrifuged, and the supernatant was discarded.
View this table:
[in this window]
[in a new window]
|
Table 1. The sequential extractions used to investigate arsenic partitioning between seven operationally defined fractions.
|
|
Despite many inherent problems, fractionation schemes remain one of the more popular tools for investigating the partitioning of trace elements between different phases. Results, however, are affected by a number of factors, including sediment handling procedures, the reagents chosen and extraction time (Louma and Davis, 1983; Slavek et al., 1982), and post-extraction readsorption onto other phases (Rendell et al., 1980; Belzile et al., 1989). In this case the fractionation scheme was used as a tool to investigate the relative effect of sediment resuspension on sediment composition. However, it must be remembered that the operational character of the method means that results cannot be taken independently of the extraction scheme used.
Analyses
To ensure that the extractants used in the fractionation scheme did not affect the results from the analytical methods chosen, individual calibration curves were prepared for each matrix for all analyses.
Total As concentrations were measured using hydride generation with detection by inductively coupled plasma mass spectrometry (ICP–MS) (PerkinElmer, 1996). Arsine gas (AsH3) was produced by reacting samples with a borohydride reducing solution [NaBH4 (2 g L-1), NaOH (0.5 g L-1)] in a flow injection system–hydride generator (PerkinElmer [Wellesley, MA] FIAS 400) that was directly connected to the ICP–MS (PerkinElmer SCIEX ELAN 6000). An antifoam reagent (Dow Corning [Midland, MI] Antifoam Emulsion 110 A, 10 mL L-1) was added to the borohydride reducing solution when samples contained large amounts of organic matter (e.g., extractions from Step III), as they were found to react excessively during hydride generation. To ensure the efficient production of AsH3, all samples were pretreated to reduce As(V) to As(III). An ascorbic acid–potassium iodide reducing solution [9 mL, ascorbic acid (15 g L-1), KCl (15 g L-1), concentrated HCl (100 mL L-1)] was added to each sample (1 mL) before reacting in a water bath (50°C) for 1 h. Samples were analyzed within 12 h to minimize reoxidation to As(V). After the pretreatment, calibration curves produced using As(V) were indistinguishable from those produced using As(III) and the recovery of As(V) spikes from any of the extractant matrices was never less than 93%.
Iron(II) concentrations were measured using the colorimetric 1,10 phenanthroline method (Clesceri et al., 1998). All samples were allowed to react for 15 min for color development except those containing oxalate (e.g., extractants from Step VI), which were left overnight. Preliminary investigations showed that this extended time ensured that the presence of oxalate did not affect the extent of color development.
Moisture content was estimated by drying sediment samples overnight at 105°C. Total sediment concentrations of a suite of elements (As, Si, Al, Fe, Ca, S, P) were measured using X-ray fluorescence (XRF) on oven-dried sediment.
Resuspension Tests
To determine how oxic sediment resuspension affected dissolved As concentrations over time, sediment (50 g) was added to deionized water (1.5 L) and stirred using a magnetic stirrer at a rate sufficient to keep most sediment in suspension. At specified time intervals, water samples (10 mL) were collected and the pH and Eh of the slurry measured. Water samples were immediately filtered (0.45 µm) and later analyzed for As.
To determine how remobilization was affected by the response of sediment to solutions of varying initial pH (pHinit = 0, 1, 2, 3...14), triplicate sediment samples (3 g) were mixed with solutions varying in concentration of either HCl or NaOH (45 mL). The slurries were placed on an orbital mixer (Ratek Instruments, Boronia, Australia) to ensure continual agitation and their pH measured at fixed times (1, 8, and 21 d). After the final pH reading, the slurries were centrifuged, filtered, and analyzed for As and Fe.
This experiment was then repeated to investigate changes in As partitioning in the pH range most typical of natural waters. Sediment was again equilibrated in triplicate with solutions of varying pH (pHinit = 4, 4.5, 5, 5.5...10) but the sediment to water ratio was increased to 4 g in 35 mL. The increased mass decreased the variability between replicate slurries. Slurries were placed on the orbital mixer for 24 h and then centrifuged, filtered, and analyzed for As. Partitioning was investigated in residual sediment from replicate experiments of pHinit 4, 5.5, 7, 8.5, and 10 by applying Steps II to VII of the fractionation scheme (see Table 1). Dissolved As (Step I) was assumed to be removed in the remobilization experiment. In addition to measuring the As concentration in each extractant, Fe concentrations were measured in those extractions targeting the amorphous and crystalline fractions (Steps V and VI). Color was measured by light absorbance at 460 nm (Louma and Bryan, 1981) in Step III, which targets the organic fraction. The Fe and color measurements were used as a measure of fraction abundance.
|
RESULTS
|
|---|
Initial Sediment Arsenic Partitioning and Elemental Composition
The initial sediment As partitioning prior to resuspension was measured on three replicate sediment samples and confirmed that more than half the As extracted in the fractionation scheme was extracted in Steps V and VI, which target amorphous and crystalline Fe oxide phases, respectively (Fig. 1)
. These findings are in agreement with other studies that have found correlations between sediment As and Fe concentrations (e.g., Crecelius, 1975; Masscheyleyn et al., 1991; La Force et al., 1998). There was little As remaining in the residual sediment and the sum of As extracted from all fractions (21.4 ± 0.61 SD mg kg-1 dry sediment) corresponded to the total sediment As measured by XRF (23.3 ± 2.89 SD mg kg-1 dry sediment).
View larger version (68K):
[in this window]
[in a new window]
|
Fig. 1. The partitioning of arsenic between different operationally defined phases in Lake Yangebup sediment. Average of three replicates, errors represent the standard deviation.
|
|
Substantial amounts of As were also associated with Steps III and IV, targeting organic matter such as humic acids and carbonate minerals, respectively. Liming with calcium minerals, such as calcium carbonates, has been found to bind arsenic in heavily contaminated sites (Bothe and Brown, 1999) and investigations have shown that As adsorption on such minerals is possible and may, after Fe, be the major control of dissolved As concentration (Sadiq, 1997). Less is known about the adsorption of As on organic matter. Thanabalasingam and Pickering (1986) found that phosphate competed strongly with both As(V) and As(III) in adsorption studies using humic acids. However, significant proportions of As were also displaced by other anions, such as carbonate and fluoride ions, which suggests that the humic acid was acting as an anion exchanger to some extent. Ultimately, the extent of specific adsorption in the organic fraction will depend on the surface functional groups present.
The initial Fe partitioning (not shown) indicated that 95.4% of extracted Fe was present in either Step V (42.2%) or Step VI (53.2%), which targeted amorphous and crystalline iron oxides. Only nominal amounts (<2%) were extracted in any of the other steps.
X-ray fluorescence (XRF) showed that the sediment contained significant concentrations of Si, Al, Ca, and Fe, indicating the likely presence of iron and calcium minerals, aluminosilicate clays, and quartz (Table 2). The high loss on ignition value also indicated that there was likely to be a significant amount of organic matter in the sediment.
The Rate of pH Buffering and Arsenic Remobilization
The kinetic response of sediment to resuspension was measured to investigate how quickly the sediment could respond to changes in the localized environment (Fig. 2)
. The results suggest that the sediment is able to buffer the pH of added equilibrating solutions to between 8 and 8.5 and does so faster than the rate of As release. From an initial value of 6.3, average slurry pH increased to a maximum of 8.5 within 30 min but then steadily decreased to 8.1 after 24 h and 7.7 after 1 wk. This decrease is attributed to carbonic acid buffering from the dissolution of atmospheric carbon dioxide into solution (Stumm and Morgan, 1996). Similar behavior was noted in experiments using water initially at pH 8.7 but with no sediment added. The Eh (not shown) remained oxic for the duration of the experiment, decreasing from 550 to 350 mV upon addition of the sediment and then increasing slowly to 460 mV at the end of the mixing period. However, further interpretation of Eh is extremely difficult. The measurement relies on the assumptions that the redox system measured is reversible and that the system is in equilibrium (Stumm and Morgan, 1996). A further complication is the fact that any number of redox reactions could be contributing to the measured value.
View larger version (18K):
[in this window]
[in a new window]
|
Fig. 2. Changes in pH and dissolved arsenic over time in a sediment–water slurry. Slurry pH reached a steady state within 30 min; dissolved arsenic reached a steady state after approximately 20 h. The solid line indicates the remobilization expected from Step I of the fractionation scheme. Each point represents the average of three samples. The error bars represent the 95% confidence interval. The average starting pH was 6.3.
|
|
After an initial rapid release, the average dissolved As concentration increased at a slower rate and reached a steady concentration of about 32.4 µg L-1 (4.0 mg kg-1 dry sediment) after approximately 24 h. The amount of As involved in the initial rapid release equals the amount of As that is defined as dissolved in Step I of the fractionation scheme and is attributed to porewater mixing with the overlying water. Other experiments using an initial pH of 4 (not shown) also indicated that As remobilization reached a steady state within 24 h.
Sediment Buffering and Arsenic Remobilization
The experiments investigating the response of sediment to water at varying pH showed the same buffering behavior as the rate tests and, after 24 h, sediment slurries of pHinit 4 to 10 were buffered to 8.28 ± 0.06 (Fig. 3A)
. However, final slurry pH showed increasing deviation from this value outside this range and those of pHinit 0, 1, 13, or 14 showed almost no shift from pHinit. Over 21 d, a gradual decrease in pH was observed in sediment slurries with pH > 6 and, as in the rate experiments, this was attributed to the dissolution of atmospheric carbon dioxide.
View larger version (30K):
[in this window]
[in a new window]
|
Fig. 3. (A) The pH response of sediment samples equilibrated with solutions of varying pHinit after 1, 8, and 21 d. (B) Released As and Fe in sediment samples equilibrated with solutions of varying pHinit for 21 d. Each As measurement is the average of three samples; Fe measurements are from single samples. The error bars represent the 95% confidence interval.
|
|
The results in Fig. 3A were used to calculate the sediment's ability to buffer solution pH. We calculated that 0.6 mol H+ kg-1 dry sediment was required to lower lake pH to 4. Correspondingly, 0.2 mol OH- kg-1 dry sediment would be required to increase lake pH to 10. These buffering capacities are comparable with other values found in the literature (Federer and Hornbeck, 1985; Starr et al., 1996) and will indefinitely fix water column pH between 4 and 10, assuming that environmental conditions do not change catastrophically.
Deviation from buffer pH was also associated with increases in dissolved As, which was measured in all slurry experiments, apart from those using pHinit greater than 11 (Fig. 3B). Iron concentrations were measured in all solutions and also showed increased remobilization when pHinit depleted the sediment's buffering ability (i.e., pHinit < 2 and pHinit > 10) compared with slurry experiments of more neutral pHinit. Solutions of pHinit greater than 11 showed increasing color that caused a precipitation upon acidification, interfering with the As analysis. The color was ascribed to humic acid or iron oxide dissolution and therefore it would be reasonable to expect increased As concentrations in these solutions (Thanabalasingam and Pickering, 1986; Raven et al., 1998). Adsorption envelopes for both As(III) and As(V) indicate that the release of As from Fe oxides or clays usually increases after about pH 8 or 9 (Manning and Goldberg, 1997; Raven et al., 1998) and additional As remobilization could also occur in this region if hydroxyl ions compete for anion adsorption sites. In order to confirm the increase in As remobilization at high pHinit values, an attempt to remove the color interference was made by diluting the samples before acidification. However, this did not stop the precipitation and the potential development time required for other methods (see, for example, Samanta et al., 1999) was deemed excessive for the analysis of three samples.
Despite the fact that no trends in As release with pH were established within the well-buffered region, 19% of the total sediment As (18 mg kg-1 dry sediment) was remobilized in the well-buffered region (Fig. 3B) in the initial buffer experiment. An average of 19% of the total sediment As (23.3 mg kg-1 dry sediment) was also remobilized when the equilibration experiment was repeated over the pHinit range of 4 to 10, supporting the 24-h equilibration time indicated by the rate experiments. However, any trend of As release with pH in the well-buffered region again remained imperceptible.
Arsenic Partitioning
The initial sediment partitioning was compared with partitioning measured in the residual sediment of pHinit 7 slurries equilibrated for 24 h (Fig. 4A)
to ascertain which phases contributed to remobilization. Equilibration caused no measurable change in the As concentration of extractions targeting the carbonate, crystalline Fe oxide, or apatite phases. However, decreases in As concentration were observed in extractions targeting the ion exchangeable, organic, and amorphous iron oxide fractions.
View larger version (25K):
[in this window]
[in a new window]
|
Fig. 4. (A) The effect of equilibration time on sediment arsenic partitioning. Comparison is made between initial sediment partitioning and partitioning in sediment shaken for 24 h. (B) Iron concentrations of selected fractions before and after resuspension. Initial partitioning was determined using the average of results from three soil samples, partitioning in resuspended sediment is from the average of two. The error bars represent the standard deviation.
|
|
Iron concentrations (Fig. 4B) were used to measure the change in fraction abundance for the iron oxide fractions. There was no measurable change in iron concentration of either fraction. Organic fraction color remained similarly unchanged.
The effect of pH on As partitioning was investigated by comparing the partitioning measured in the residual sediment of slurries equilibrated at different pHinit values (Fig. 5)
. While there was possibly a small decrease in ion exchangeable and organic As extracted as pH increased from 4 to 10, there was little change in As partitioning. Similarly, abundance measurements indicated no trends with pHinit.
View larger version (14K):
[in this window]
[in a new window]
|
Fig. 5. Changes in arsenic partitioning due to equilibration with solutions of varying pHinit. Arsenic concentrations are expressed as the percentage of total arsenic extracted. The seven sediment phases are grouped into three separate graphs according to magnitude. Each point represents the average of two samples. The error bars represent the standard deviation.
|
|
|
DISCUSSION
|
|---|
Mechanisms of pH Buffering and Arsenic Remobilization
The relationship between Fe and As concentrations observed in slurries below pHinit 2 suggests that iron oxide dissolution caused a release of Fe-associated As from the consolidated sediment. Ulrich (1983) has suggested that pH-initiated iron oxide dissolution occurred when pH < 3.8, which agrees with the increased release seen in low pHinit slurries. The high Fe concentrations and increased color measured in slurries of high pHinit suggested that iron oxides and organic matter may also dissolve at high pH values, presumably causing increased As remobilization as discussed in the results section.
The mechanisms controlling pH in the buffered region are less obvious. After 24 h, pH in well-buffered slurries was 8.3. Carbonate mineral dissolution is known to occur between pH 6.2 to 8 and buffers to approximately pH 8.3 (Ulrich, 1983; Stumm and Morgan, 1996). However, while carbonate abundance was not measured, the fact that "carbonate" As did not remobilize is at variance with the theory that carbonate minerals dissolved. It is also possible that pH could be regulated by an acid–base equilibria of surface functional groups present on oxide minerals and organic molecules that dissolve into solution (Schindler and Stumm, 1987; Sposito, 1984).
Superficially, the hypothesis that As remobilization will be facilitated via desorption is supported by the partitioning results. Equilibration caused a release of As into solution from the operationally defined ion exchangeable, organic, and amorphous iron oxide sediment fractions, without measurable loss in fraction abundance. However, calculations show that the decrease in Fe that would accompany the measured loss in As via dissolution of an Fe–As precipitate could be smaller than the error measured in the iron abundance measurements. While the conclusion that As binding is controlled by adsorption–desorption mechanisms rather than dissolution–precipitation agrees with results of other studies (Livesey and Huang, 1981; Masscheyleyn et al., 1991), these studies rely on interpretation of results investigating adsorption isotherms and ion activity products and are themselves open to question.
Sediment As partitioning was not affected by variations of pHinit within the well-buffered region, although partitioning presumably deviates substantially outside this region. While most release mechanisms are pH dependent, if pH equilibrates before dissolved As concentrations reach steady state then the value of pHinit becomes unimportant.
The Effect of Sediment Resuspension in Lake Yangebup
The rate experiments indicated that As remobilization occurs on timescales comparable with resuspension events. These results can be used estimate the potential for As release from consolidated sediment. The amount of sediment As remobilized during a single resuspension event will depend on the length of time that the sediment remains suspended in the water column. In the most extreme case remobilization from sediment particles will attain steady state, releasing 4 mg kg-1 dry sediment of As. The maximum suspended solids concentration expected in lakes of this size and morphology is approximately 200 mg L-1 (Kristensen et al., 1992; Bailey and Hamilton, 1997). If we assume that the amount of As released varies linearly with mass of sediment resuspended, the dissolved As concentration of the lake is calculated to increase by 0.8 µg L-1. This figure represents just over two percent of the measured dissolved As concentration in the lake (approximately 35 µg L-1) and is much smaller than the concentration of 13.4 µg L-1 estimated to remobilize from the floc layer by Arnold and Oldham (1997). It should be noted that Arnold and Oldham measured remobilization using an elutriate test developed to assess the potential contaminant release during the disposal of dredged aquatic sediments in a water body. The sediment to water ratio used in their experiments and calculations was much higher than that expected in the lake, leading to a probable overestimation of As release. Recalculating their results using a suspended solids concentration of 200 mg L-1 leads to an estimation of only 0.24 µg L-1 released from the floc sediment. This suggests that the floc layer actually releases less As than the underlying consolidated sediment.
|
CONCLUSION
|
|---|
The results from this study suggest that between pHinit 4 to 10 both As remobilization and buffering capacity are able to respond quickly to rapid changes in localized conditions. However, the results also show that, in this case, small pH fluctuations are unlikely to alter the amount of As remobilized during sediment resuspension. Alterations in pHinit also caused little change in As partitioning in this range. The effect of changing pHinit was seen more clearly in regions where phase dissolution definitely occurred.
While we were not able to determine the mechanism of remobilization conclusively, the fractionation method was shown to be able to resolve small changes in As partitioning in Lake Yangebup sediment and indicated which the fractions were affected by remobilization. The measured Fe partitioning supported the importance of Fe in As binding. However, it could not be resolved to the same extent as the As partitioning and the method would benefit from improved measurements of fraction abundance, particularly the ion exchangeable, carbonate, and apatite fractions.
Our calculations indicate that remobilization from consolidated sediment is unlikely to significantly contribute to Lake Yangebup's dissolved As concentration after a single resuspension event. However, the consequence of multiple resuspension events requires investigation. It is unknown whether the consolidated sediment has the capacity to resorb As or if repeated resuspension may cause greater increases in dissolved As. Revisiting the As remobilization from floc sediment resuspension also suggests that further investigation is required into the effect of dissolved organic matter and biological activity on dissolved As concentrations. The consolidated sediment has been shown to buffer pH values between 4 and 10 and should indefinitely accommodate small changes in pH in the water column.
With the appropriate adjustments to ensure oxygen-free conditions, the simple techniques used in this paper are equally suitable for investigating the effects of anoxia on sediments, a process that is also likely to play an important role in remobilization in shallow eutrophic lakes.
|
ACKNOWLEDGMENTS
|
|---|
This work was funded by an Australian Research Council Small Grant. Kathryn Linge was supported financially by an Australian Postgraduate Award. We thank F. Lincoln and A. Rate for their comments on the manuscript. This paper is Centre for Water Research Report ED1506KL.
|
REFERENCES
|
|---|
- Arnold, T.N., and C.E. Oldham. 1997. Trace-element contamination of a shallow wetland in Western Australia. Mar. Freshwater Res.48:531–539.
- Bailey, M.C., and D.P. Hamilton. 1997. Wind induced resuspension, a lake wide model. Ecol. Modell. 99:217–228.
- Belzile, N., P. Lecomte, and A. Tessier. 1989. Testing readsorption of trace elements during partial chemical extractions of bottom sediments. Environ. Sci. Technol. 23:1015–1020.
- Bothe, J.V., and P.W. Brown. 1999. Arsenic immobilization by calcium arsenate formation. Environ. Sci. Technol. 33:3806–3811.
- Clesceri, L.S., A.E. Greenberg, and A.D. Eaton (ed.) 1998. Standard methods for the examination of water and wastewater. 20th ed. Am. Public Health Assoc., Washington, DC.
- Crecelius, E.A. 1975. The geochemical cycle of arsenic in Lake Washington and its relation to other elements. Limnol. Oceanogr. 20:441–451.
- Davis, J.A., R. Rosich, J.S. Bradley, J.E. Growns, L.G. Schmidt, and F. Cheal. 1993. Wetlands of the swan coastal plain. Volume 6. Wetland classification on the basis of water quality and invertebrate community data. Water Authority of Western Australia, Environmental Protection Agency, Perth, Western Australia.
- Eloranta, P. 1985. Relations between physical and chemical variables of water quality and phytoplankton in one eutrophic pond. Aqua Fenn. 15:105–113.
- Federer, C.A., and J.W. Hornbeck. 1985. The buffer capacity of forest soils in New England. Water Air Soil Pollut. 26:163–173.
- Förstner, U. 1990. Inorganic sediment chemistry and elemental speciation. p. 61–105. In R. Baudo et al. (ed.) Sediments, chemistry and toxicity of in-place pollutants. Lewis Publ., Ann Arbor, MI.
- Förstner, U. 1995. Non-linear release of metals from aquatic sediments. p. 247–307. In W. Salomens and W.M. Stigliani (ed.) Biogeodynamics of pollutants in soil and sediments. Springer–Verlag, Berlin.
- Kristensen, P., M. Søndergaard, and E. Jeppesen. 1992. Resuspension in a shallow eutrophic lake. Hydrobiologia 228:101–109.
- La Force, M.J., S.E. Fendorf, G.C. Li, G.M. Schneider, and R.F. Rosenzweig. 1998. A laboratory evaluation of trace element mobility from flooding and nutrient loading of Coeur d'Alene River sediments. J. Environ. Qual. 27:318–328.[Abstract/Free Full Text]
- Livesey, N.T., and P.M. Huang. 1981. Adsorption of arsenate by soils and its relation to selected chemical properties and anions. Soil Sci. 131:88–94.
- Louma, S.N., and G.W. Bryan. 1981. A statistical assessment of the form of trace metals in oxidized estuarine sediments employing chemical extractants. Sci. Total Environ. 17:165–196.
- Louma, S.N., and J.A. Davis. 1983. Requirements for modelling trace metal partitioning in oxidised estuarine sediments. Mar. Chem. 12:159–181.
- Maberly, S.C. 1996. Diel, episodic and seasonal changes in pH and concentrations of inorganic carbon in a productive lake. Freshwater Biol. 35:579–598.
- Manning, B.A., and S. Goldberg. 1997. Arsenic(III) and arsenic(V) adsorption on three California soils. Soil Sci. 162:886–895.
- Masscheyleyn, P.H., R.D. Delaune, and W.A. Patrick, Jr. 1991. Arsenic and selenium chemistry as affected by sediment redox potential and pH. J. Environ. Qual. 20:522–527.[Abstract/Free Full Text]
- Meyer, J.S., W. Davison, B. Sundby, J.T. Oris, D.J. Lauren, U. Förstner, J. Hong, and D.G. Crosby. 1994. Synopsis of discussion session: The effects of variable redox potentials, pH and light on bioavailability in dynamic water–sediment environments. p. 155–170. In J. Hamelink (ed.) Bioavailability—Physical, chemical and biological interactions. Lewis Publ., Boca Raton, FL.
- PerkinElmer. 1996. Flow injection mercury/hydride analyses: Recommended analytical conditions and general information. Release 4.0. Publ. B350510. PerkinElmer, Wellesley, MA.
- Raven, C.P., A. Jain, and R.H. Loeppert. 1998. Arsenite and arsenate adsorption on ferrihydrite, kinetics, equilibrium and adsorption envelopes. Environ. Sci. Technol. 32:344–349.
- Rendell, P.S., G.E. Bately, and A.J. Cameron. 1980. Adsorption as a control of metal concentrations in sediment extracts. Environ. Sci. Technol. 14:314–318.
- Sadiq, M. 1997. Arsenic chemistry in soils: An overview of thermodynamic predictions and field observations. Water Air Soil Pollut. 93:117–136.
- Samanta, G., T.R. Chowdhury, B.K. Mandal, B.K. Biswas, U.K. Chowdhury, G.K. Basu, C.K. Chanda, D. Lodh, and D. Chakraborti. 1999. Flow injection hydride generation atomic adsorption spectrometry for determination of arsenic in water and biological samples from arsenic-affected districts of West Bengal, India and Bangladesh. Microchem. J. 62:174–191.
- Saulnier, I., and A. Mucci. 2000. Trace metal remobilization following the resuspension of estuarine sediments, Saguenay Fjord, Canada. Appl. Geochem. 15:191–210.
- Schindler, P.W., and W. Stumm. 1987. The surface chemistry of oxides, hydroxides and oxide minerals. p. 83–110. In W. Stumm (ed.) Aquatic surface chemistry, chemical processes at the particle–water interface. John Wiley & Sons, New York.
- Simpson, S.L., S.C. Apte, and G.E. Batley. 1998. Effect of short-term resuspension events on trace metal speciation in polluted anoxic sediments. Environ. Sci. Technol. 32:620–625.
- Sposito, G. 1984. The surface chemistry of soils. Oxford Univ. Press, New York.
- Starr, M., C.J. Westman, and J. Ala-Reini. 1996. The acid buffer capacity of some Finnish forest soils, results of acid addition laboratory experiments. Water Air Soil Pollut. 89:147–157.
- Stumm, W., and J.J. Morgan. 1996. Aquatic chemistry. 3rd ed. John Wiley & Sons, New York.
- Tessier, A., P.G.C. Campbell, and M. Bisson. 1979. Sequential extraction procedure for the speciation of particulate trace metals. Anal. Chem. 51:844–851.
- Thanabalasingam, P., and W.F. Pickering. 1986. Arsenic sorption by humic acids. Environ. Pollut. Ser. B. 12:233–246.
- Ulrich, B. 1983. Soil acidity and its relation to acid deposition. p. 127–146. In B. Ulrich and J. Pankrath (ed.) Effects of accumulation of air pollutants in forest ecosystems. Proc. of a Workshop, Göttingen, Germany. 16–19 May 1982. D. Reidel Publ., Dordrecht, the Netherlands.
TOP
ABSTRACT
INTRODUCTION
