| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departament de Química Analítica-Universitat de Barcelona, Diagonal 647, 08028 Barcelona, Spain
* Corresponding author (fermin.lopez{at}apolo.qui.ub.es).
Received for publication June 23, 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Abbreviations: AES, atomic emission spectrometry • BCR, Community Bureau of Reference • CL, clay loam • CLc, calcareous clay loam • Cs, saline clay • ICP, inductively coupled plasma • L, loam • MS, mass spectrometry • SLhy, hydromorphous sandy loam
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Sequential extractions were designed for the selective extraction of trace elements from operationally defined sediment solid fractions (Tessier et al., 1979; Salomons and Förstner, 1980; Meguellati et al., 1983). Although these procedures are not fully specific in extracting the element bound to a given solid fraction, they may provide comparative information on trace-metal mobility in changing environmental conditions, such as pH or redox potential. The use of sequential extraction procedures in soil analysis as a complement to single extractions has increased. They are applied to soils contaminated by various pollution sources, such as irrigation with wastewater (Flores et al., 1997; Ahumada et al., 1999; Cajuste et al., 2000), mining activity (Ma and Rao, 1997; Maiz et al., 1997; García-Sánchez et al., 1999), automobile emissions (Lee and Touray, 1998), or sewage sludge addition (Canet et al., 1998; Luo and Christie, 1998; Planquart et al., 1999; Walter and Cuevas, 1999). However, few studies deal with sequential extraction of trace elements from soils contaminated by accidental spills (Díaz-Barrientos et al., 1999). The use of a leaching approach based on sequential extractions may help to elucidate the relative contribution of mixed pollution sources (i.e., particulate and/or soluble sources) and may aid in the predictions of trace element mobility. Among the variety of sequential extraction procedures described in the literature, the three-step scheme proposed by the BCR (Ure et al., 1993) and the modified version (Sahuquillo et al., 1999) have become very popular during recent years and their application has increased lately (Pérez Cid et al., 2001; Sutherland and Tack, 2002; Mossop and Davidson, 2003). Moreover, the reproducibility of the modified procedure and its applicability to soils have been tested through interlaboratory exercises (Rauret et al., 1999, 2000). These studies showed a good interlaboratory reproducibility for Cd, Cr, Cu, Ni, Pb, and Zn in the three steps and the suitability of the modified version of the BCR scheme for the analysis of contaminated soil samples.
The extent to which laboratory leaching tests predict mobility in the field is controversial. Soil-to-plant transfer factor (TF) is defined by the ratio of an element concentration in plant tissue versus its concentration in soil. For a given element, TF is usually associated with the root uptake since other pathways are negligible. For similar environmental conditions, the eventual element concentration in the plant depends on its level in the soil solution, corrected by what could be defined as a plant factor (PF). The plant factor includes plant physiological aspects, related to nutrient uptake and selectivity, and depends on the plant, soil solution composition, element, and element species considered (Cushman, 1982; McBride et al., 1997; Roca et al., 1997). The level of an element in the soil solution depends on the total level in the solid phase, the in situ solid–liquid distribution coefficient (Kd), and its available fraction (fav). The term Kd quantifies the ratio of the concentration of a given element in the solid phase versus the concentration in the soil solution, thus being useful to describe how displaced the element equilibrium is between the two phases. It can be estimated by adsorption experiments, or through soil characteristics associated with the sorption pool to be considered for a given element. The fav, which defines the reversibly adsorbed element fraction in the solid phase able to participate in the equilibrium between the soil solution and the solid phase, is usually quantified through leaching tests.
The main factors affecting soil–plant transfer are summarized in the following expression:
 |
From this equation, it is clear that predicting soil-to-plant transfer based only on leaching data is an incomplete approach, since other factors are involved in the final transport. Moreover, the range of variability of these factors, for instance of the solid–liquid distribution coefficient, may exceed that of the extraction yields obtained with leaching tests by several orders of magnitude. However, in several countries leaching data are being included to classify soils within contaminated or noncontaminated groups, or to establish the maximum allowable concentrations. The mobility of the pollutant into the food chain is one of the issues of most concern (Vollmer et al., 1997). This provides justification for comparison of leaching data coming from the application of extraction procedures with data from soil–plant transfer.
In 1998, a pyritic spill from a mine in Aznalcóllar (southern Spain) covered approximately 2000 ha of agricultural soils with a layer of mud. It was composed mainly of fine pyrite particles (with small amounts of clay minerals, quartz, calcite, and gypsum), together with acid wastewater (López-Pamo et al., 1999). The amounts of mud and wastewater on the soils varied with the distance from the pollution source. One month after the accident, the sludge layer was removed by heavy machinery. The removal was not uniform: in some areas, the sludge and the topsoil layer were taken, and in others, sludge was left.
The present article describes the application of the modified BCR three-step sequential extraction procedure to soil samples and sludge from the Aznalcóllar area. The trace elements considered included Cd, Cu, Pb, and Zn, and other significant elements in the Aznalcóllar accident, such as As, Bi, and Tl. The major elements (Al, Ca, Fe, Mg, and Mn) released in each step by applying the same scheme were also determined to obtain information on the main solid fractions that interact with the elements of concern. The data from sequential extractions are compared with trace element concentration in plants from the contaminated area. The validity and limitations of trace element mobility predictions based solely on leaching data are discussed.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
As the objective of the sampling was to study soil–trace element interaction, and not to derive a concentration map, the sampling strategy followed was judgmental, thus samples were chosen to be representative of the entire range of soils found in the affected area. In this area, there is a high degree of heterogeneity in the sludge content in the soils. Where possible, noncontaminated reference soils were taken from areas apparently unaffected by the spill, slightly contaminated soils were taken from spots where sludge had been removed almost completely, and highly contaminated soils were taken from "hot spots" where sludge particles remained. For soils from near the marshlands (Cs), which were affected only by acidic wastewater (Cabrera et al., 1999), only a reference and an affected soil were available. For the SLhy soils, no reference soil was available. As previous studies had shown that the composition of the sludge was homogeneous throughout the affected area, a single sample of sludge was taken near the area of L soils (Alastuey et al., 1999).
Soil samples were taken with an auger down to 20 cm, sampling four individual samples in an area of 25 x 25 m. After air-drying and breaking down the aggregates, soils were passed through a 2-mm nylon sieve. To integrate spatial variability due to the sludge removal, composite samples were prepared from all four individual samples and stored in polyethylene containers for analysis, during three weeks at controlled temperature (22 ± 3°C) and humidity (<50%) conditions. The parameters determined were pH (in water), cation exchange capacity and particle size distribution (Ministerio de Agricultura, Pesca y Alimentación, 1994), organic carbon and carbonate content (Cadahía, 1973), and total sulfur (Fiedler et al., 1999). The major elements in soils were determined by X-ray fluorescence spectrometry (XRF) and the measurements were performed on a PW2400 X-ray spectrophotometer equipped with Rh excitation tubes (Philips, Eindhoven, The Netherlands). Samples were diluted (1:20) with lithium tetraborate and melted in a Philips PERL'X2 microprocessing system to obtain 30-mm-diameter pearls. Fifty-six geological international reference samples (mainly from the Institute of Geophysical and Geochemical Prospection, National Institute of Standards and Technology, and Association Nationale de la Recherche Technique) were used for calibration.
Plant Samples
Plant samples were collected, shortly after the soil collection, from three of the five sites where the soils had been sampled (areas of SLhy, CL, and Cs soils) because there were plants growing at these sites only. The exactly location where soils had been sampled was determined by using a global positioning system (GPS) receptor. To calculate the soil-to-plant transfer factors, additional soil samples were also taken along with plants at each site.
The plant samples included dominant species pertaining to the Umbelliferae family that had grown in the sampling area. Four sampling areas (50 x 50 cm) were defined for each site. The four individual plant samples were pooled to create a composite sample in each site. Then, after removing roots, shoots were carefully rinsed in deionized water, dried at 60°C, and ground in a mill. The total digestion of plant samples (0.5 g) was performed with a mixture of HNO3 (70%), H2O2 (30%), HF (40%), and HClO4 (70%) using a four-step program in a focused microwave oven (Sastre et al., 2002). The final solution obtained was filtered through an ashless Whatman (Maidstone, UK) 42 filter, then diluted to 25 mL with HNO3 (2%), and stored in polyethylene bottles at 4°C until trace element analysis.
Aqua Regia Extraction Procedure
The trace element content in sludge and soil samples was estimated from aqua regia extractions, following the procedure recommended by the International Organization for Standardization (1995). Samples (3 g) were digested with a HCl (37%) and HNO3 (70%) (3:1) mixture (28 mL) at room temperature for 16 h. Thereafter, the suspension was digested at 130°C for 2 h under reflux conditions. The obtained suspension was then filtered through an ashless Whatman 41 filter, diluted to 100 mL with 0.5 mol L-1 HNO3, and stored in polyethylene bottles at 4°C for element analysis.
Sequential Extraction Procedure
The modified BCR sequential extraction procedure was applied to sludge and soils sampled in May and July 1998. The procedure is described in the following paragraphs (Rauret et al., 1999).
First Step: Extraction with 0.11 mol L-1 Acetic Acid (Exchangeable and Weak Acid Soluble Fraction)
A 1-g soil sample was treated with 40 mL of 0.11 mol L-1 acetic acid solution (prepared by diluting Suprapur acetic acid; Merck, Darmstadt, Germany). This mixture was shaken in a mechanical, end-over-end shaker at 30 ± 10 rpm at 22 ± 5°C for 16 h. The extract was separated from the solid residue by centrifugation at 3000 x g for 20 min. The supernatant was decanted, collected in polyethylene bottles and stored at 4°C until analysis. The residue was washed by adding 20 mL of double-deionized water, shaking for 15 min on the end-over-end shaker, and centrifuging for 20 min at 3000 x g. The supernatant was decanted and discarded, taking care not to discard any of the solid residue.
Second Step: Extraction with 0.5 mol L-1 Hydroxylammonium Chloride (Reducible Fraction)
Forty milliliters of 0.5 mol L-1 hydroxylammonium chloride (Merck Pro-Analysis) solution, adjusted to pH 1.5 by the addition of a fixed volume of 2 mol L-1 HNO3, was added to the residue from the first step. This mixture was shaken in a mechanical, end-over-end shaker at 30 ± 10 rpm at 22 ± 5°C for 16 h. The extract was separated and stored as in the first step. The residue was washed as in the previous step and the supernatant was decanted and discarded.
Third Step: Digestion with 8.8 mol L-1 Hydrogen Peroxide and Extraction with 1 mol L-1 Ammonium Acetate (Oxidizable Fraction)
Ten milliliters of 8.8 mol L-1 hydrogen peroxide (Merck Suprapur) solution was carefully added to the residue from the second step. The mixture was digested for 1 h at 22 ± 5°C and for 1 h at 85 ± 2°C, and the volume of liquid was reduced to less than 3 mL. A second aliquot of 10 mL of H2O2 was added, the mixture was digested for 1 h at 85 ± 2°C, and the volume was reduced to about 1 mL. Finally, 50 mL of 1 mol L-1 ammonium acetate (Merck Pro-Analysis and pH adjusted to 2.0 ± 0.1 with concentrated HNO3) solution was added to the cool, moist residue. The mixture was shaken in a mechanical, end-over-end shaker at a speed of 30 ± 10 rpm and a room temperature of 22 ± 5°C for 16 h. The extract was separated from the solid residue by centrifugation and decantation as in previous steps, collected in polyethylene bottles, and stored at 4°C until analysis. The residue was washed as in previous steps and the supernatant was decanted and discarded.
Residue from the Third Step: Extraction with Aqua Regia (Residual Fraction)
The residue from Step 3 was digested with aqua regia, following International Organization for Standardization (1995) guidelines. In this case, the amount of acid used to attack 1 g of sample was reduced to keep the same volume to mass ratio of 7.0 mL of HCl (37%) and 2.3 mL of HNO3 (70%).
Three independent replicates were performed for each sample and blanks were measured in parallel for each set of analyses using the extraction reagents described above. The moisture content of each sample was determined by drying a separate 1-g sample in an oven (105 ± 2°C) to constant weight. From this, a correction to dry mass was obtained that was applied to all analytical concentrations reported.
Major and Trace Element Determination in Extracts
The concentration of major and trace elements in the extracts was determined by inductively coupled plasma–atomic emission spectrometry (ICP–AES) or inductively coupled plasma–mass spectrometry (ICP–MS) depending on the concentration level.
A TJA Model 25 system (equipped with a cross-flow nebulizer) with a TJA-300 autosampler (Thermo Elemental, Franklin, NJ) was used for the ICP–AES determinations. The analytical wavelengths (nm) used were: Al, 308.215; Ca, 317.933; Cd, 228.802; Cu, 324.754; Fe, 259.940; Mg, 279.079; Mn, 257.610; Pb, 220.353; and Zn, 213.856.
An ELAN 6000 ICP–MS (equipped with a cross-flow nebulizer and a quadrupol mass spectrometer) with an AS-91 autosampler (PerkinElmer, Wellesley, MA) was used for ICP–MS determinations. Various isotopes were measured for each element to detect and control possible isobaric and polyatomic interferences: 75As; 209Bi; 111Cd, 112Cd, 113Cd, and 114Cd; 63Cu and 65Cu; 206Pb and 208Pb; 203Tl and 205Tl; and 64Zn, 66Zn, and 68Zn. In hydroxylammonium chloride and aqua regia extracts, arsenic was measured with hydride generation (HG) with inductively coupled plasma–mass spectrometry (ICP–MS), due to the 40Ar35Cl interference. The trace element concentration in digested plants was determined by ICP–MS and results were corrected by the Ti content in the corresponding soil sample, because plant Ti was assumed to come from adhered soil particles in the plant tissue.
The extracts were previously diluted to minimize matrix interferences. The calibration was performed using an external calibration graph prepared in the extracting reagent for ICP–AES measurements and in 1% HNO3 with Rh as internal standard for ICP–MS measurements. The limit of detection, calculated as 3s/m (where s is the standard deviation of the blank [1% HNO3] and m is the slope of the calibration curve), for each element determined was (µg L-1): Al, 6.2; Ca, 3.8; Cd, 1.9; Cu, 2.2; Fe, 2.0; Mg, 10; Mn, 0.5; Pb, 22; and Zn, 4.6 for ICP–AES determinations, and As, 0.26; Bi, 0.03; Cd, 0.05; Cu, 0.18; Pb, 0.21; Tl, 0.05; and Zn, 0.71 for ICP–MS determinations.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
The quality of the analytical data for the sequential extraction procedure was assessed by carrying out analyses of the certified reference material BCR-701, a lake sediment certified for extractable metal contents in the three steps of the modified BCR sequential extraction procedure and indicative values for aqua regia extraction (Pueyo et al., 2001). The results and the certified and indicative values of BCR-701 are shown in Table 1. Comparable results were obtained for Cd, Cu, Pb, and Zn in relation to the certified and indicative values. Only significant positive deviations were observed for Cd in the second and third step. Moreover, the sum of the extracted metals from the three steps plus residual fraction compared well with the aqua regia–extractable contents, indicating that laboratory working conditions were under control.
|
|
Sulfur and iron concentrations (see Table 2) can be used as indicators of the contribution of sludge particles to soils, as levels increased from reference to highly contaminated samples, with the exception of the Cs soil, which was only affected by acid wastewater.
The concentration ranges obtained for the trace elements in the samples studied are shown in Table 3. Detailed information can be found elsewhere (Vidal et al., 1999). Although soil quality criteria are not in agreement internationally, some conclusions can be derived from the comparison of the obtained concentration ranges with data derived from national regulations and directives. Reference soils show similar and even higher trace element concentrations (especially for Cu and Zn) than those considered as target levels by regulations in Belgium and the Netherlands, with the exception of Cd (Adriano et al., 1997). This suggests that previous exploitation of the mine nearby contributed to the high metal concentration levels found in our samples (González et al., 1990; Ramos et al., 1994). For several elements (e.g., As, Cu, Pb, and Zn), contaminated soils show higher trace element concentrations than those established as target and maximum allowable concentrations (MAC) in agricultural soils, and even higher than intervention limits given in the directives in various countries (Adriano et al., 1997; Kabata-Pendias and Pendias, 2001). This fact indicates the severity of the accident and the need for studies to derive trace element mobility predictions in the contaminated area.
|
Aluminum results indicated, as expected, that this element was dominantly associated with the residual fraction, regardless of the soil sample and level of contamination, with percentages higher than 95% of the total Al content. The same pattern has been described for soils contaminated by other pollution sources (Stalikas et al., 1999; Sutherland and Tack, 2000).
Figure 1 shows the distributions obtained for Ca, Mg, Mn, and Fe in sludge and soil samples. Magnesium distributions in SLhy soils and sludge were not determined. Extraction yields are expressed as a percentage of the respective total element concentration (measured by X-ray fluorescence spectrometry).
|
As seen in Fig. 1b, Mg partitioning was similar to that of Ca (see the L soils), but more shifted to the residual fraction. In Cs soils, the partitioning of Mg differed, with higher extractability in the first two steps of the procedure and lower residual fractions, which can be explained by the high contribution of Mg from saline origin.
For Mn (Fig. 1c), partitioning depended on the level and type of contamination rather than on the physicochemical properties of the soils. In all reference and slightly contaminated samples, the most dominant fraction was the Mn extracted with hydroxylammonium chloride, with extraction yields ranging between 40 and 70%, thus indicating that Mn was mainly associated with reducible fractions, such as Mn oxides (Zhang et al., 1998; Maiz et al., 2000). Distributions changed for the highly contaminated soils, since the fraction soluble in acetic acid increased. The Cs soils, not affected by sludge, were an exception to this pattern. This can be explained by the Mn distribution in the sludge, where 70% of total Mn was extracted with acetic acid. This fact indicates that Mn in highly contaminated soils was also associated with fractions more easily solubilized than oxides, such as sulfides of Mn(II) (da Silva et al., 2002) because of the presence of pyritic sludge particles.
The fingerprint of the sludge particles was evident for Fe (Fig. 1d), which despite being a major element acted in this case as a marker of the pollution source. In reference and low-level contaminated soils, only a small fraction of Fe (10%) was solubilized by reduction and no extraction was observed at the oxidation step. However, the distribution of the sludge sample showed that Fe was partially solubilized after digestion with H2O2 at the third step, with the rest remaining in the residual fraction. A similar pattern was observed for Fe distribution in the highly contaminated samples, where a higher fraction (up to 60%) was extracted in oxidant conditions due to the oxidation of the iron sulfide from the sludge. In contrast, clay Cs soils, where the origin of contamination was acid wastewater and not sludge, showed similar distributions to those of reference soils. Therefore, the partitioning of Fe, especially the mobilization at the third step, allowed us to confirm the presence of sludge particles as a source of contamination.
Changes in pH after Each Step of the Community Bureau of Reference Sequential Extraction Procedure
The changes in pH that occurred after each step of the BCR sequential extraction procedure were examined to determine the role of pH in the solubilization of the different soil fractions. Table 4 shows the changes in pH (expressed as 
pH) obtained for each step of the procedure. The final pH of extractant solution was determined for each sample after each step of the procedure. When pH values were lower than double the standard deviation of the final pH, changes were considered not significant.
|
For the second step, the pH in Cs soils may be due to the solubilization of the remaining carbonates not previously dissolved in the first step. For the rest of soils, a significant pH was only seen in a few contaminated samples, possibly due to the H+ consumption in the solubilization of several soil fractions as ferric hydroxides (Bermond, 2001).
For the third step, the reaction conditions allow the oxidation of several solid fractions, specifically organic matter and sulfides. The oxidation of sulfides to sulfates is related to a decrease in pH due to the protonic balance of the reaction (Singer and Stumm, 1970). This process, which is dominant here considering the nature of the source of pollution and the low organic matter content in all samples, was observed when comparing the changes in pH after this extraction, where a decrease in pH was observed only for the highly contaminated soils with an initial neutral or basic pH, thus indicating partial oxidation of the pyrite particles.
Extractable Trace Elements Using the Community Bureau of Reference Three-Step Sequential Extraction Procedure
Trace element concentration (mg kg-1) and relative standard deviation (%) obtained for the three steps of the sequential extraction procedure are reported in Table 5. Trace element partitioning in sludge and soil samples is shown in Fig. 2
, which shows the distribution of Cd, Zn, Cu, Pb, As, Bi, and Tl. Arsenic, Bi, and Tl distributions were not determined for SLhy soils. Extraction yields are expressed as a percentage of the respective aqua regia–extractable content.
|
|
Extraction yields in the second step were high for all the elements studied in most soils, especially for Pb. Similar findings have been reported in the literature for soils of similar composition (Ramos et al., 1994; Chlopecka et al., 1996; García-Sánchez et al., 1999), although Pb extractability may be overestimated due to a possible readsorption process after the first step (Gilmore et al., 2001). The relevant extraction yields of Zn and Cu in most soils indicate that the influence of the soil–metal interaction was significant, since extraction yields in the second step were low for these elements in the sludge. For the sludge sample, extractability was important for Pb, Bi, As, and Cd, indicating a leaching process mainly due to the low pH of the extracting agent, at which certain secondary minerals derived from pyrite are solubilized (Yanful and Orlandea, 2000). This explains the increase in As and Bi extractability from reference to highly contaminated soils, since the Cs soils were again the exception to this behavior.
The extraction yields in the third step for the soil samples were low for Bi, Tl, and Pb. For Cu, Zn, and As, extraction yields were low in reference samples, although they increased with the level of contamination in samples affected by the sludge, because of the oxidation of sulfides from the sludge particles (Singer and Stumm, 1970; Díaz-Barrientos et al., 1999). As observed for Fe, this pattern was not found in the Cs samples. In contrast, the opposite pattern was observed for Pb, for which the highly contaminated samples showed the lowest extraction yields, which may be explained by the low soluble secondary minerals formed by the oxidation of the sulfides (Yanful and Orlandea, 2000).
From the leaching data obtained from the application of the sequential extraction procedure, Cd, followed by Zn and to a lesser extent by Cu, were predicted to be the most mobile trace elements, with a significant extraction at low pH, which agrees with other results found in the literature (Ramos et al., 1994; Ahumada et al., 1999). In general, the residual percentages of Cd, Zn, and Cu decreased with the level of contamination, indicating that highly contaminated soils may require immediate intervention. The low mobility observed for Tl in this work is in good agreement with the speciation of Tl obtained by Villar et al. (2001) after applying the BCR scheme to soils and sediments. Arsenic and Bi were more mobile in highly contaminated soils than in the other samples. For most of the trace elements, this pattern was repeated in the Cs soils despite the fact that there was no influence of the sludge in the contaminated Cs soil. All these results agree with the view that metals from anthropogenic sources are more mobile than those from soil parent materials (Chlopecka et al., 1996).
Comparison between Leaching Test and Plant Uptake Data
In our study, to compare the mobility predictions from the ex situ extraction approach with in situ data is not an easy task because of the difficulty of performing controlled soil-to-plant transfer experiments in field conditions. However, the relative scale of mobility predicted from the use of the extraction procedure could be compared with results obtained from plant samples that had grown in the contaminated soils, in the areas of SLhy, CL, and Cs soils.
The trace element concentrations in shoots are given in Table 6. The values were mostly within the normal range in plants. The values were within the range of critical plant concentrations proposed by Kabata-Pendias and Pendias (2001) for Zn only. Arsenic results for plant analyses are not presented in Table 6, because they were under the detection limit in all plant samples. Arsenic in soluble forms may be passively taken up by plants (Kabata-Pendias and Pendias, 2001), but here it was mostly associated with particulate contamination (i.e., sludge particles), as was demonstrated with the sequential extraction procedure. Therefore, the uptake of arsenic by plants in this scenario was drastically reduced.
|
Table 6 also includes data from leaching tests. Data from the sequential extractions are structured following the sequence of the reagents. The values are the mean of the low- and high-contamination soils to pool two potential soil scenarios for plant growth, considering the high heterogeneity in the sludge content for soil samples. To facilitate the comparison between soil–plant transfer data and extraction data, extraction yield ratios (calculated from the extraction yield of a given element versus that of Cu) have also been calculated in all cases to establish a relative sequence of extraction yields of the trace elements in a given soil.
If we compare the transfer factor ratios with the ratios of the extraction yields from the first step of the sequential extractions, the relative sequence of mobility was well predicted for the SLhy and Cs soils, whereas Pb mobility was overpredicted in the CL soil. However, the extraction yield ratios were quite different from those from transfer factor ratios, especially in the Cs soil. Therefore, the qualitative information derived from the extraction with acetic acid was correct, although the quantitative information overpredicted the mobility of Cd, Zn, and Pb with respect to that of Cu and of the elements with a low mobility, such as Tl and Bi.
The inclusion of the second step of the sequential extractions in the mobility predictions (use of NH2OH·HCl) completely changed the sequence of the predicted mobility, since the trace element mobility was predicted to level out. This fact advised against the use of the extraction yields coming from the second step for predictions of soil–plant transfer, although further studies are needed to verify the usefulness for long-term mobility predictions if changes in redox potential occur.
The prediction capacity of this sequential-extraction procedure, especially of the first step, can be compared with several single-extraction procedures described in the literature (Kennedy et al., 1997; Rauret, 1998). Some of them, such as those using CH3COOH, give similar information to the first step of this procedure (Vidal et al., 1999), whereas others such as those using CaCl2 may give a better prediction of the amount of trace element available in the soil–plant system, especially because the pH of the extraction is kept near to the initial soil pH (Novozamsky et al., 1993; Vidal et al., 1999). Therefore, data from single extractions with 0.01 and 1 mol L-1 CaCl2 have also been included in Table 6 (unpublished data, 2002), and the extraction yields ratios are also calculated.
Although the use of single extractions at a similar pH to that of the soil was hypothesized to be a better approach for soil–plant transfer predictions than the use of an acidic extraction, the mobility predictions were in general quite similar to those derived from the first step data. Predictions were slightly better with 0.01 than 1 mol L-1 CaCl2 , because the relative sequence of mobility of the set of trace elements in a given soil was well predicted in the three studied soils, with the exception of Zn in the Cs soil, where their bioavailability was clearly underestimated. As for the first step, the quantitative information was not as useful as the qualitative, since transfer factor ratios were only similar to extraction yield ratios in some cases.
Finally, those leaching tests showing a better prediction of the relative sequence of transfer of the trace elements in a given soil were only poor estimators of the differences between soils for a given trace element. In some cases (e.g., prediction of the different transfer of Zn between SLhy and CL soils with CH3COOH; Cd, Zn, Cu, and Pb between SLhy and Cs soils with 0.01 mol L-1 CaCl2) the differences between the relative transfer between soils and prediction from leaching test data were of several orders of magnitude. This fact can be explained by the greater influence of the changes in the soil solution composition and related plant factor when comparing different soils, thus the prediction solely on the basis of leaching data is even more limited.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Based on the results from the extractability in the first step, the trace elements studied could be classified as mobile elements (Cd, Zn, and Cu) and less-mobile elements (Pb, As, Tl, and Bi). The third step of the sequential extraction procedure was especially useful for the detection of sludge pyrite particles due to their distinctive oxidation pattern. Therefore, sequential extraction is useful for assessing the long-term mobility of trace elements released to the environment when changes in redox potential or pH occur.
The relative mobility of the trace elements considered here was in general well predicted when comparing the yields of the first step with soil–plant transfer factors, although the quantitative information derived from the procedure overestimates the element availability to plants. However, this estimation could be improved if the plant uptake studies were performed over several years, and if other soil parameters (sorption pattern, soil solution composition) are included in predictions.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONCONCLUSIONSREFERENCES |
|---|
Related articles in JEQ:
This article has been cited by other articles:
|
|
 |
|
  J. C. Seaman, V. M. Vulava, A. G. Sowder, B. P. Jackson, S. A. Aburime, and P. M. Bertsch Metal extractability from contaminated SRS sediments: Comparison of column and batch results Environmental Geosciences, December 1, 2005; 12(4): 235 - 242. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. D. Burton, I. R. Phillips, and D. W. Hawker Geochemical Partitioning of Copper, Lead, and Zinc in Benthic, Estuarine Sediment Profiles J. Environ. Qual., January 1, 2005; 34(1): 263 - 273. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| The SCI Journals | Agronomy Journal | Crop Science | |||
| Journal of Natural Resources and Life Sciences Education |
Vadose Zone Journal | ||||
| Soil Science Society of America Journal | Journal of Plant Registrations | The Plant Genome | |||
