Introduction

Potentially toxic elements (PTEs) occur naturally in soils. Still, anthropogenic activities associated with metalliferous mining and metallurgical industries, fossil fuels combustion, application of agrochemicals, fertilizers, liming materials, sewage sludge, manure, municipal wastes, and transport emissions are the main sources of soil enrichment with PTEs (Massas et al. 2018). When the concentrations of the PTEs in soils reach or exceed the corresponding toxic thresholds for humans, animals, or plants, cleanup actions are needed (Kalyvas et al. 2018b).

Various physical, chemical, and biological techniques have been proposed to remediate metal(loid) polluted soils. Conventional methods, usually expensive, include in situ vitrification, soil incineration, excavation and landfill, soil spading, soil washing, soil flushing, and chemical fixation (Ali et al. 2013; Yao et al. 2012). Recently, environment friendly and cost-effective methods have emerged like phytoremediation, bioremediation, and ecological remediation (Sarwar et al. 2017; Padmavathiamma and Li 2007). In this context, biochar, a carbon-rich by-product, has gained significant attention as a potential remediation tool of contaminated soils via the immobilization of potentially toxic elements (PTEs) in soil (El-Naggar et al. 2019a; El-Naggar et al. 2020). Moreover, phytoremediation techniques as phytoimmobilization, phytostabilization, phytoextraction, and phytovolatilization can be considered as an alternative, green, cost-efficient, and integrated remediation technology (Sarwar et al. 2017; Padmavathiamma and Li 2007).

Depending on the remediation approach, several chemical reagents and amendments have been used to enhance the mobilization of the contaminants in soils. Ethylenediaminetetraacetic acid (EDTA) chelating agent, sodium dithionite (Na2S2O4) reductive agent, and olive mill wastewater (OMW) that induces both complexation and reductive reactions in soils, are amongst those that have been used to achieve redistribution of contaminants in the soil solid phases in order to increase their mobility/bioavailability (Bolan et al. 2014; Jelusic and Lestan 2014). EDTA is an aminopolycarboxylic acid with extremely high complexing capabilities that efficiently extracts potentially toxic metals from surface sites of different soil solid phases (Manouchehri and Bermond 2009). The new formed metal-EDTA complexes are water soluble and therefore can be leached out of the soil (i.e., soil washing) or in the case of enhanced phytoextraction strategies, the increased metal solubility/availability promotes plant uptake (Udovic and Lestan 2012; Chen et al. 2007). Soil pH, redox potential, metal(loid) speciation, ions competition, and the type of the metal(loid) bond with the soil fractions are amongst the most important factors that control EDTA’s chemical behavior and effectiveness as a chelating agent in the soil media (Bolan et al. 2014; Nowack 2002).

Redox soil conditions are critical parameters that affect the geochemical behavior of the metal(loid)s (Chuan et al. 1996; Gasparatos 2012; El-Naggar et al. 2018). By regulating soil redox conditions, the solubility of the metal(loid)s can be increased (El-Naggar et al. 2019b). Indeed, reduced soil conditions promote the reductive dissolution of Fe-Mn oxyhydroxides resulting to the release of significant amounts of adsorbed metals (Gasparatos 2013). Moreover, reduced conditions weaken the binding of the metals with the soil fractions by lowering their oxidation state and thus increasing their solubility (Kim and Baek 2015; Varadachari et al. 2006; Abumaizar and Smith 1999). The ability of dithionite salts such as sodium dithionite (Na2S2O4) to produce reduced conditions in soils is well known (Kim and Baek 2015; Paul et al. 2003; Mehra and Jackson 1960).

Greece is amongst the biggest producers of olive oil worldwide (FAOSTAT 2013). According to the method used for oil extraction, olive oil processing may result to the generation of significant quantities of olive mill wastewater (OMW) consisted by olive tissue and process water with acidic pH (~ 4–6) and high C/N ratio (Azbar et al. 2004). Since OMW is the effluent of mechanical processing of olives, it is characterized by high organic load while phenolics and nutrients are also present in large amounts (Morillo et al. 2009). According to Pardo et al. (2017), the application of OMWs in soils affects soil redox potential and increases metal(loid)s mobility that may afterwards be available for plant uptake or they may be prone to leaching or to re-adsorption on soil reactive phases. Moreover, the application of diluted OMW that is rich in nutrients can alleviate the impact of the adverse effluent characteristics introducing lower disturbance in the soil environment especially when compared with chemical compounds as EDTA and sodium dithionite. The available literature on the effect of olive oil industry by-products on metal(loid)s mobility in soils is relatively limited and the results are contradictory. Indeed, Romero et al. (2005) concluded that the addition of olive mill solid waste in a soil from mine tailings resulted in increased levels of soluble Pb/Zn after 18 weeks of incubation, while the addition of the water-soluble fraction of fresh solid olive husk in a calcareous soil had no significant result on Pb/Zn solubility and uptake by Beta maritima L. (de la Fuente et al. 2011). The latter is in line with the results of a pot experiment showing that OMW application in a mine soil did not increase Pb and Zn concentration in Pteris vittata fronts (Kalyvas et al. 2018c).

Sequential extraction procedures (SEPs) aim to describe metal(loid)s partitioning in sediments and soils and to evaluate associations between metal(loid)s geochemical forms and different sediment and soil fractions (Kalyvas et al. 2018a, 2018b; Gasparatos et al. 2015; Drahota et al. 2014; Zimmerman and Weindorf 2010). A well accepted method for As sequential extraction from soils and sediments has been developed by Wenzel et al. (2001). According to this scheme, five fractions of As are operationally defined: non-specifically sorbed (easily mobilizable, outer-sphere complexes); specifically sorbed (readily mobilizable, inner-sphere complexes); bound to amorphous and poorly crystalline oxides; bound to well-crystallized hydrous oxides; residual. The well-established BCR method is amongst the methods used to determine metals partitioning in soils (Rauret et al. 1999). Application of the BCR protocol provides three chemical fractions of metals in soils: exchangeable/acid soluble, reducible, and oxidizable.

Recently, Kalyvas et al. (2018c) tested the efficacy of 0.01 M Na2-EDTA and diluted OMW to promote As, Pb, and Zn uptake by Pteris vittata from a heavily polluted mine-affected soil and concluded that the application of EDTA considerably increased As, Pb, and Zn concentrations in P. vittata fronts while OMW increased only As concentration. However, the conducted experiment focused mainly on the effect of EDTA and OMW on metal(loid)s plant uptake and less on metal(loid)s geochemical behavior in the amended soil. Following this and considering that metal(loid)s in natural soil systems tend to transform into more stable chemical forms over time becoming less soluble (Dousis et al. 2013), the present study aims (a) to comparatively evaluate the effects of EDTA, sodium dithionite, OMW, and their combinations on As, Pb, and Zn availability in a mine-affected soil over time; (b) to discuss on the redistribution of the studied elements in the soil fractions as a result of the applied reducing and chelating agents after 90 days of incubation by using two sequential extraction protocols; and (c) to examine the role of Fe/Mn oxyhydroxides on the solubility and partitioning of the As, Pb, and Zn in the studied soil.

Materials and methods

Soil sampling and soil properties

The soil used in this study was collected from a mine-affected area in the outskirts of Lavrion, central Greece (X, 503825.35; Y, 4175571.92/EGSA87). Intensive mining activities and metallurgical processes that lasted for centuries in the wider area of Lavrion, resulted to the enrichment of the surrounding soils with metal(loid)s at levels raising severe environmental concerns that must be addressed for the wealth of the local ecosystem and human health of local residents (Kalyvas et al. 2018b).

A composite top soil sample compiled from subsamples collected from a 2-m2 surface area was transferred to the laboratory and air-dried and the fine earth fraction (< 2 mm) was used for analyses. Samples for soil fractionation were further grounded to pass through a 0.5-mm mesh. Mechanical analysis was performed according to the Bouyoucos hydrometer method (Bouyoucos 1951), the pH was measured in 1:1 (w/v) soil/water slurry, and carbonates (eq. CaCO3) were estimated using the Bernard calcimeter method (NF ISO 10693 1995). The Walkley–Black wet oxidation (Nelson and Sommers 1982) and the sodium acetate methods (Rhoades 1982) were used for the determination of organic matter (OM) content and cation exchange capacity (CEC), respectively. Aqua regia digestion provided the “pseudo-total” metal concentrations (Gasparatos and Haidouti 2001) that in this study are referred as “total.” Soil properties are presented in Table 1.

Table 1 Soil physicochemical properties
Full size table

Experimental design and procedure

A two-factor (treatment × time) experiment in four replicates was carried out on a heavily polluted mine soil, to test the effect of four amendments and incubation time on As, Pb, and Zn availability. The amendments used for the experiment were deionized water, Na2-EDTA 0.1 M (referred as EDTA in the text), sodium dithionite 0.1 M, and fresh olive mill wastewater with chemical properties as shown in Table 2, diluted in deionized water at a 40% v/v. From these amendments, eight treatments were produced: (i) deionized water (DW), (ii) EDTA (E), (iii) sodium dithionite (SD), (iv) olive mill wastewater (OMW), (v) E+SD, (vi) E+OMW, (vii) SD+OMW, (viii) E+SD+OMW. All treatments were applied on 1 g of soil placed in 50-mL falcon bottles and in a volume equal to the 80% of the soil water-holding capacity. The treated soil samples were incubated at 20 °C in a dark growth chamber for 1 day (1d), 3 days (3d), 7 days (7d), 15 days (15d), 30 days (30d h), 60 days (60d), and 90 days (90d), maintaining the initial moisture constant. The experiment was run in duplicate and consisted by batch I for As and batch II for Pb and Zn. At the end of each incubation period, easily mobilizable As and exchangeable Pb/Zn were extracted by (NH4)2SO4 0.05 mol L−1 and by acetic acid 0.11 mol L−1 from the treated soil samples of batch I and batch II, respectively. These two reagents are used in the first step of the Wenzel and BCR sequential extraction methods respectively, and are considered to extract the most available forms of the studied elements. Aiming to determine the treatment effect on the As, Pb, and Zn partitioning into the tested soil at the end of the 90 days incubation period, the Wenzel (Wenzel et al. 2001) and the BCR (Rauret et al. 1999) sequential extraction protocols (SEPs) were applied to batch I for As and batch II for Pb and Zn. Both SEPs were also applied to the control soil (untreated). The experimental conditions of SEPs are summarized in Table 3. Finally, to test the hypothesis that Fe-Mn oxyhydroxides may affect As, Pb, and Zn mobility, Fe and Mn concentrations were also determined in the acetic acid extracts at the end of each incubation period and in all BCR extracts. The experimental design is provided as supplementary material (Figure S1).

Table 2 Chemical properties of the olive mill wastewater used in this study (Karpouzas et al. 2009)
Full size table
Table 3 The BCR and Wenzel sequential extraction schemes
Full size table

Analytical determinations

Elements concentrations were determined by ASS, using a Varian-spectra A300 system whereas for low As concentrations, the hydride generator Varian model VGA 77 was used. The relative standard deviation of three measurements was less than 3% and 5% for AAS and HGAAS determinations respectively and the reproducibility was checked by reanalyzing 30% of the samples. Mean As, Pb, Zn, Fe, and Mn recovery from ERM-CC141 European Reference Material (loam soil) certified for aqua regia procedure was 98, 95, 104, 102, and 98%, respectively.

Statistical analysis

The effects of time and treatment on As, Pb, Zn, Fe, and Mn availability as well as the effect of treatment on elements partitioning were evaluated by analysis of variance (ANOVA) followed by Tukey’s honestly significant difference test. STATISTICA (StatSoft, Inc., USA, 1984–2011, Version 10) and IBM SPSS Statistics 20 were used for the performed statistical analysis.

Results

Effects of time and treatment on elements availability

As presented in Table 4, incubation time and treatment significantly affected As, Pb, and Zn available concentrations whereas significant interaction between the two factors was observed for all the studied elements. Independently from incubation time, the treatment effect on the availability of the studied elements was almost identical for Pb and Zn, while a slightly different pattern was observed for As (Fig. 1a, c, e). The highest availability of As, up to 41 mg kg−1 on average, was observed for the EDTA treatment followed by E+OMW (mean value 32 mg kg−1) and E+SD+OMW (mean value 27 mg kg−1). Treatments OMW, E+SD, and SD+OMW resulted to the release of similar amounts of As up to 14–16 mg kg−1, while SD and DW exhibited the lowest available As concentrations (mean value 6 and 8 mg kg−1, respectively). The significantly highest Pb and Zn available concentrations were obtained by the EDTA and E+OMW treatments (mean values up to 2532 mg kg−1 and 1427 mg kg−1 for Pb and Zn respectively) followed by the E+SD (mean value 2259 and 1333 mg kg−1 for Pb and Zn, respectively) and E+SD+OMW (mean value 2388 and 1340 mg kg−1 for Pb and Zn, respectively) treatments that showed a similar impact on both metals. The other treatments resulted in much lower Pb and Zn availability and showed similar strength to release Pb and Zn from the tested soil.

Table 4 ANOVA results. Treatment and time effects on As, Pb, and Zn mobility
Full size table
Fig. 1
figure 1

Time and treatment effects on the available concentrations of As (a, b), Pb (c, d), and Zn (e, f). Comparisons of means were performed by Tukey’s HSD test (a ≤ 0.05). The presence of common letters implies no significant difference

Full size image

Incubation time on As, Pb, and Zn availability had its highest effect at 7d, 3d, and 90d, respectively (Fig. 1b, d, f). Three different patterns can be observed for the studied elements. Arsenic available concentration increased from 1d to 7d and then decreased to the initial levels at 30d, remaining approximately constant thereafter. The effect of incubation time on Pb availability followed a different pattern. The observed highest Pb availability at 3d was accompanied by a sharp decline at 7d, at a level significantly lower than that of 1d. This lowest Pb availability remained up to the 15d, increased again at 30d reaching the initial concentration of 1d and then remained almost constant until 90d without any significant variation. No significant alteration observed on Zn mobility until 7d. Afterwards, this hysteresis was gradually withdrawn and available Zn increased until 90d.

To better visualize the effects of treatment and time on As, Pb, and Zn availability, three diagrams produced showing the variation of W1/B1 fractions of the studied elements for each treatment and incubation interval (Fig. 2). It is apparent that for all the incubation times, E, E+OMW, E+SD+OMW, and E+SD treatments resulted to distinctly increased Pb and Zn solubility than the treatments without EDTA (i.e., DW, SD, OMW, and SD+OMW) (Fig. 2b, c). This is more pronounced for Pb, since for the treatments with EDTA available Pb concentrations ranged between 2100 and 2772 mg kg−1, markedly higher than for the non EDTA treatments that ranged between 207 and 419 mg kg−1. For As, the treatments with EDTA were generally demonstrated higher extraction ability, significant only for E, E+OMW, and E+SD+OMW after 7 and 15 days of incubation (Fig. 2a).

Fig. 2
figure 2

Treatment effect on the available concentrations of As (a), Pb (b), and Zn (c) for all incubation intervals (mean plots with standard error bars). The error bars represent the least significant difference at the 95% confidence level

Full size image

Effect of treatment on elements partitioning

To investigate the effect of treatment on the distribution of As, Pb, and Zn in the different soil phases, two sequential extraction procedures (SEPs) were applied to the soil samples after the 90 days incubation period. The Wenzel SEP specifically designed for As (Wenzel et al. 2001) and the well-established BCR SEP for cations (Rauret et al. 1999). For each treatment, the results are presented as the percentage of each fraction to total (sum of the fractions) (Fig. 3).

Fig. 3
figure 3

Treatment effect on the percentage of As (a) in non-specifically sorbed (W1), specifically sorbed (W2), amorphous and poorly crystalline hydrous oxides of Fe and Al (W3), well-crystallized hydrous oxides of Fe and Al (W4), and residual fractions (WRF) and on the percentages of Pb (b) and Zn (c) in exchangeable/acid soluble (B1), reducible (B2), oxidizable (B3), and residual fractions (BRF). Comparisons of means between treatments for every fraction were performed by Tukey’s HSD test (a ≤ 0.05). The presence of common letters implies no significant difference

Full size image

In the control and the treated soils, incubated for 90 days, As mean percentage into the various soil fractions was in the order W3>W4>WRF>W2>W1 (Fig. 3a). Compared with the control, As W1 fraction was significantly increased after 90 days of incubation in all treatments containing EDTA, from 0.6% for control to 2.1, 2.5, 1.5, and 1.6% for E, E+OMW, E+SD, and E+SD+OMW, respectively. EDTA and EDTA combined with OMW (E and E+OMW treatments) showed the significantly higher increase of As W1 fraction while no significant effect of DW, SD, OMW treatments was noticed.

Except the oxidizable fraction B3, the treatment effect on Pb and Zn partitioning in the different soil fractions was similar (Fig. 3b, c). Treatments containing EDTA significantly increased Pb and Zn percentages in exchangeable/acid soluble fraction (B1) and that was accompanied by a significant decrease of the metals percentages in the reducible fraction (B2). A different pattern was observed for the oxidizable fraction (B3) of the two metals. For Pb, all treatments contained EDTA resulted to small but significant decrease of the metal percentages associated with the oxidizable fraction. Unlike Pb, treatments E and OMW resulted to a significant increase of Zn oxidizable fraction, i.e., from 19.8% in the control sample to 21.8 and 22.9%, respectively.

Discussion

Effects of time and treatment on elements availability

Soil treatments with EDTA resulted to the highest available concentrations of all the studied elements (Fig. 1a, c, e). Depending on soil properties and the prevailing conditions in soil, EDTA can extract exchangeable metal(loid)s forms and a fraction of total metal(loid)s concentration bound to carbonates, amorphous Fe-Mn oxides, and organic matter (Manouchehri and Bermond 2009). The ability of EDTA to increase As, Pb, and Zn mobility in polluted soils has been reported by many authors. In particular, EDTA addition in a highly contaminated sediment (Meers et al. 2005) and in contaminated arable and grassland soils (Mühlbachová 2011) significantly increased Zn and Pb mobility, respectively, while the application of various EDTA concentrations in a soil from an industrial area strongly enhanced AB-DTPA and water-soluble As concentrations and As uptake by maize plants (Abbas and Abdelhafez 2013).

Sodium dithionite was used to produce reducing conditions that could lead to the dissolution of Fe oxides and to the subsequent release of the associated As, Pb, and Zn. Compared with deionized water however, dithionite did not have any significant effect on the availability of the studied elements (Fig. 1a, c, e). As it is supported by Kim and Baek (2015), this poor As, Pb, and Zn solubility during the experimental period may be due to the partial dissolution of Fe3+-oxides and the re-formation of new reactive phases (Fe2+-oxide forms) that sequestered the initially released studied elements. Following this, the observed lower Pb availability in 7d and 15d than in 1d may have occurred due to the re-adsorption of Pb onto newly formed soil reactive phases (Fig. 1d).

In relation to deionized water and sodium dithionite, OMW resulted to significantly higher available concentrations only for As (Fig. 1a). According to Madrid and Diaz-Barrientos (1994) and de la Fuente et al. (2011), the addition of fresh OMWs in soils can reinforce the solubility of metals. This may happen by the reductive dissolution of metal oxides and the consequent release of bound metal(loid)s as a result of free oxygen consumption by the soil bacteria that mineralize excess organic matter introduced in the soil system and the oxidation of phenolic compounds. Afterwards, dissolved organic material and phenols can form organometallic complexes with the released metal(loid)s that do not re-adsorb onto the soil solid phases increasing metal(loid)s availability. However, the results of our study did not show signs of increased Pb and Zn solubility probably because diluted OMW was not capable to dissolve metal oxides. Similar results were observed by de la Fuente et al. (2011), reporting that Pb and Zn availability (CaCl2 extraction) did not substantially increase in a calcareous soil after the application of water-soluble fraction of fresh solid olive husk. The increased availability of As can be explained by the competitive behavior between arsenates and phosphates as they both exist as oxyanions in the soil environment in the common range of soil pH and redox potential values (Violante and Pigna 2002; Anawar et al. 2018). Thus, it is highly possible that a portion of the significant amounts of phosphates in OMW (Table 1) replaced arsenates in soil colloids leading to the higher As available concentrations.

As indicated in Fig. 1a, c, e, E+SD treatment resulted to significantly lower As, Pb, and Zn available concentrations than EDTA, indicating that SD restricted the chelating ability of EDTA. This is in line with the results obtained by Kim et al. (2016) that showed lower extraction ability of EDTA for As in the presence of sodium dithionite at concentrations > 0.05 M and in alkaline pH, conditions as those of the present study. Contrary to E+SD treatment, the combination SD+OMW resulted to slightly higher available concentrations of Pb and Zn than SD and OMW and of As than SD, suggesting a positive effect of OMW amendment on the solubility of metal(loid)s. According to Table 1, OMW contributed positively to the SD reductive effect by offering soluble phenols and low molecule organic substances with high chelating ability that complex Fe released by the Fe oxides dissolution, thus preventing the formation of new reactive phases and the re-immobilization of As, Pb, and Zn. Finally, the combination of EDTA and OMW (E+OMW) considerably reduced As available concentration when compared with the single EDTA treatment. Considering that olive mill wastewaters contain significant amounts of soluble Ca+2 (Table 1) and that (NH4)2SO4 was used to extract available As, it is highly possible that some of the SO42− co-precipitated with the Ca+2 as insoluble gypsum and led to suppressed extraction of arsenates from the reactive soil solid surfaces.

Effect of treatment on elements partitioning

Compared with the control, As mean percentage in W2 fraction (readily mobilizable, inner-sphere complexes) was significantly increased only with the E+OMW treatment (from 9.9% in the control sample to 12.5%) indicating that OMW enhanced EDTA’s capability to extract As from the less labile W2 fraction. As discussed earlier, E+OMW treatment also significantly increased As in W1 fraction (Fig. 3a). The observed higher presence of As in W1 and W2 fractions after 90 days of incubation was accompanied by a significant reduction of As concentration in W3 fraction. Indeed, for fraction W3 that corresponds to the extraction of As chemical forms from amorphous oxides of Fe, Mn, and Al, the As percentages decreased significantly from 50.5% in the control sample to 44.1%, respectively. An almost identical distribution pattern of As between the W1, W2, and W3 was observed for the EDTA treatment. Distribution of As in fractions W4 and WRF showed minor differences between the treatments, indicating that all amendments used neither effectively dissolved the crystallized oxides of Fe, Mn, and Al nor solubilized the residual soil phases. Sequential extraction results clearly show that the increased concentrations of As in W1 and W2 fractions were obtained because EDTA extracted As from the amorphous metal oxides and this was further enhanced by the presence of OMW. This is in line with Borggaard (1982), who report that EDTA demonstrated selectivity in extracting As from artificial mixtures of synthetic amorphous iron oxides and natural or synthetic crystalline iron oxides.

Though the two sequential extraction protocols use different fractionation scheme, a similar to As but much clearer distribution pattern amongst the chemically defined soil fractions was observed for Pb and Zn. High amounts of Pb were found in the B2 fraction (bound to Fe/Mn oxides) of control soil. The addition of EDTA and of amendments containing EDTA (E+SD, E+OMW, E+SD+OMW) exhibited strong mobilizing effect on Pb and Zn in the soil. In the case of Pb, the B2 fraction decreased and B1 fraction increased markedly, almost 40%. For Zn, the B1 fraction also increased with EDTA treatments and this change of Zn exchangeable fraction was opposite to that of Zn bound to Fe/Mn oxides, a pattern similar to the distribution of Pb fractions. These results clearly indicate that EDTA is a very efficient agent in releasing heavy metals, particularly Pb from soils (Neugschwandtner et al. 2008). The decreased concentration in the Fe/Mn oxide bound fraction of Pb and Zn due to EDTA promoted dissolution of Fe oxides, contributing to a significant increase of these metals concentration in the mobile fraction (B1). However, in BCR sequential extraction procedure, no step is included to differentiate between amorphous and crystalline metal oxides. According to Rodriguez et al. (2003) and more recently Kalyvas et al. (2018a), hydroxylamine hydrochloride—the reagent of BCR second step—cannot effectively breakdown well-crystallized phases and mainly dissolves amorphous Fe and Mn oxides. Hence, EDTA single or combined with OMW increased the available concentrations of all studied elements that were mainly originated from the amorphous Fe oxides. In a metal-contaminated calcareous soil from a former Pb-Zn area in Spain, Clemente and Bernal (2006) found that addition of humic acids isolated from a compost (mixture of olive leaves and the solid fraction of OMW) increased Zn concentrations in the EDTA extractable fraction confirming metal mobilization after a 28-week period of incubation.

The role of Fe/Mn oxides on As, Pb, and Zn mobility

Partitioning results clearly showed that the increased concentrations of As, Pb, and Zn in the more mobile fractions (i.e., in W1 for As and B1 for Pb and Zn) were mainly derived from the reducible soil chemical phases. Therefore, it was assumed that is the product of EDTA (mainly) and of the other amendments to complex with metal(loid)s bound to Fe/Mn oxides. To further elucidate the role of metal oxides on As, Pb, and Zn mobility under the experimental conditions of this study, the results of correlation and principal component analysis (PCA) analyses were evaluated. Correlation analysis was performed to test for any relations between As, Pb, Zn, Fe, and Mn mobile fraction concentrations (W1 or B1 depending on the element) for all treatments and incubation intervals. Input data for the PCA were W1/B1 concentrations of As, Pb, and Zn and B1, B2, B3, and BRF concentrations of Fe and Mn after 90 days incubation (Table 5).

Table 5 Fe and Mn fractions (mg kg−1) determined by BCR method in various treatments at the end of incubation period (90 days)
Full size table

According to the correlation coefficient matrix presented in Table 6, fractions W1/B1 of As, Pb, and Zn were strongly correlated with the B1 fraction of Fe indicating a common source of all elements in the labile fraction that most probably was originated from the partial dissolution of Fe oxides. Manganese concentration in B1 fraction was significantly correlated with the corresponding Pb and Zn concentrations and not with that of As. This however was rather an expected result, considering that As mainly adsorbs on Fe and Al oxides and not on Mn oxides (Wenzel et al. 2001). PCA analysis showed that two main factors explained 70% of the total variability. As it is presented in Fig. 4, AsW1, PbB1, ZnB1, and FeB1 fractions form a group that is negatively associated with the reducible B2 fraction of Fe, providing strong evidence that Fe oxides were the main pool for the increased available amounts of As, Pb, and Zn at the end of the incubation period. The other significant negative association found was between B1 and B2 fractions of Mn pointing to that Mn oxides dissolution increased Mn availability but did not seriously affect the availability of As, Pb, and Zn. Following this and considering that hydroxylamine hydrochloride, the extractant used in the reducible phase of BCR, dissolves primarily, if not only, the amorphous Fe and Mn oxides, PCA results also strongly suggest that amorphous Fe oxides is the main soil reactive phase that regulates As, Pb, and Zn mobility.

Table 6 Correlation coefficient matrix of As, Pb, Zn, Fe, and Mn available fractions (W1/B1) for all treatments and incubation times (n = 224)
Full size table
Fig. 4
figure 4

Principal component analysis results for As, Pb, and Zn available fractions (W1/B1) and exchangeable/bound to carbonates (B1), reducible (B2), oxidizable (B3), and residual (BRF) fractions of Fe and Mn

Full size image

Conclusions

An incubation experiment was conducted using a mine-affected soil in order to investigate the effect of Na2-EDTA, sodium dithionite, olive mill wastewater, and their combinations on the mobility of As, Pb, and Zn, over time. Olive mill wastewater increased As solubility, probably due to competition with phosphates added with OMW for sorption sites on amorphous Fe oxides. Combined with the other amendments however, olive mill wastewater produced contradictory results in terms of As, Pb, and Zn mobility and functioned either by reinforcing or by suppressing the extracting ability of dithionite and EDTA. EDTA and its combinations increased all studied elements mobility while dithionite demonstrated poor results. Moreover, sodium dithionite when combined with EDTA inhibited the chelating strength of EDTA. The evaluation of the sequential extractions results indicated that EDTA efficiently released As, Pb, and Zn from amorphous Fe-Mn oxides mainly. Furthermore, PCA results clearly demonstrated that As, Pb, and Zn increased available concentrations in fractions W1 and B1 of the Wenzel and the BCR SEPs, respectively, were mainly originated from the partial dissolution of amorphous Fe oxides. Though OMW at the tested dilution rate mobilized effectively As sorbed on amorphous Fe oxides, more work under different experimental conditions is needed to support the findings of this study.