Planting trees and amending with waste increases the capacity of mine tailings soils to retain Ni, Pb and Zn

The sorption capacity for Ni, Pb and Zn of mine tailings soil with and without reclamation treatment (tree planting and waste amendment) was evaluated using the batch adsorption technique. It is important to determine the capacity of waste-amended soils to retain Ni, Pb and Zn, as the sludges used usually have high concentrations of these metals. The results obtained in the present study showed that the untreated mine tailings soil had a low capacity for Ni, Pb and Zn retention. The sorption capacity for Pb increased significantly in all of the treated soils, without any significant differences between them. The treatment that most increased the sorption capacity for Ni and Zn was planting with trees and amending with waste simultaneously, as this increased the concentration of both organic and inorganic carbon, exchangeable calcium, soil pH and effective cation exchange capacity

Planting trees and amending with waste increases the capacity of mine tailings soils to retain Ni, Pb and Zn Plantar árboles y enmendar con residuos incrementa la capacidad de los suelos de escombrera de mina para retener Ni, Pb y Zn Plantar árvores e emendar com resíduos incrementa a capacidade dos solos de resíduos orgânicos de mineração pra reter Ni, Pb e Zn

ABSTRACT
The sorption capacity for Ni, Pb and Zn of mine tailings soil with and without reclamation treatment (tree planting and waste amendment) was evaluated using the batch adsorption technique.It is important to determine the capacity of waste-amended soils to retain Ni, Pb and Zn, as the sludges used usually have high concentrations of these metals.The results obtained in the present study showed that the untreated mine tailings soil had a low capacity for Ni, Pb and Zn retention.The sorption capacity for Pb increased significantly in all of the treated soils, without any significant differences between them.The treatment that most increased the sorption capacity for Ni and Zn was planting with trees and amending with waste simultaneously, as this increased the concentration of both organic and inorganic carbon, exchangeable calcium, soil pH and effective cation exchange capacity.

Introduction
Soils in mining areas are usually physically, chemically and biologically degraded.In addition to these problems, soils associated with metal mines are also polluted by metals.A number of strategies have been developed over the years in order to improve the low quality of these soils and to prevent the surrounding areas from becoming polluted.The most popular reclamation treatments for mine soils involve planting vegetation with plants and amending with wastes (Tandy et al. 2009;Dary et al. 2010;Barrutia et al. 2011;Karami et al. 2011).It has been demonstrated that these treatments significantly increase the quality of mine soils by improving their physical structure, concentration of organic matter and nutrients and by promoting microbial growth and functionality (Rodríguez-Jordá et al. 2012;Asensio et al. 2013b;Asensio et al. 2013a;Asensio et al. 2014).However, the waste used to reclaim mine soils (especially sewage sludge) frequently contains considerable concentrations of Ni, Pb and Zn that enter the soils (Nyamangara 1998;Singh and Agrawal 2008;Asensio et al. 2013c).Sorption studies of mine soils that have been, or are going to be, reclaimed using sewage sludge must take into account the extra input of these metals in the soils.
In spite of the numerous articles that have been published on the effect of planting vegetation or amending mine soils with waste (Bendfeldt et al. 2001;Brown et al. 2003;Shrestha and Lal 2008;Baker et al. 2011), the effect of these treatments on the metal sorption capacity of soils has not been properly reported.The term "sorption" involves the loss of a metal ion from an aqueous to a contiguous solid phase, and consists of three important processes: adsorption, surface precipitation and fixation (Apak 2002;Bradl 2004).Planting trees and amending with waste should increase the sorption capacity of mine soils, as they increase the concentration of soil colloids with a known heavy metal retention capacity, such as clay, inorganic carbon and organic matter (Filcheva et al. 2000;Bendfeldt et al. 2001;Alvarenga et al. 2008;Asensio et al. 2013b).On the other hand, the Fe and Mn oxides that are usually found in high concentrations in soils associated with metal mines also have a high sorption capacity (Vega et al. 2006).In order to determine whether this is capable of creating new conditions for retaining metals, it is necessary to test the combination of organic matter, clay, carbonates and Fe and Mn oxides from both the mine soils and the reclaiming treatments.For all the reasons indicated above, the main objective of the present study was to compare the sorption capacity of Ni, Pb and Zn of a mine tailings soil from a depleted Cu mine that was ameliorated by tree planting, waste amendment, or both treatments.Samples of the same soil were taken from a vegetated site, an amended site, a vegetated and amended site, and an untreated site.The trees used were Pinus pinaster Aiton and Eucalyptus globulus Labill.The amendments used were made of sewage sludge and paper mill residues.Previous data have shown that most of the metal found in samples from treated sites was contained in the residual fraction of the soil, i.e. bound to the clay fraction (Asensio et al. 2013c).The present study aimed to go further, by examining whether treated mine tailing soils were capable of retaining relatively high concentrations of Ni, Pb and Zn.This would make it possible to estimate whether the availability of metals could be reduced when mine soils are amended with waste rich in metals.

Description of the study area and soil sampling
The sampling site is located in the Touro mine in Galicia, Northwest Spain (Lat/Lon (Datum ETRS89): 8º 20' 12.06'' W 42º 52' 46.18'' N).
Copper was extracted from the Touro mine for 14 years, between 1973 and 1988.The tailings at this mine are materials left over after extracting copper from the ore.The unconsolidated material could be considered a Spolic Technosol according to the latest version of the FAO classification (FAO 2014).
Four sites in the mine tailings zone (M) were selected in order to evaluate the effect of tree vegetation (v) and waste amendments (w): 1. M1 as the control site (untreated).This soil covered an area of 1.20 ha and had an AC horizon 20 cm deep.
2. M2v as the vegetated site, where Pinus pinaster Aiton trees had been planted 21 years before the sampling date.This soil covered an area of 0.60 ha.M2v had an A horizon 4 cm deep (labelled as M2Av) and an AC horizon 20 cm deep (M2Bv).
3. M3w as the amended site, where sewage and paper mill residues were added 6 months before sampling.The different types of waste were added by trucks and then spread on the soil surface.The final depth of this new layer was around 3m, covering an area of 0.8 ha.The amount of waste added was around 158 tons per ha. 4. M4vw as the amended + vegetated site, where eucalyptuses were planted and waste was added at the same time, 10 years before sampling.The amount of added wastes was 297 tons per ha, and the final depth of this new layer was around 70 cm, covering an area of 1.5 ha.
The waste used had pH values of 7-11.5, total organic C of 120-230 g kg -1 , total Cu of 100-500 mg kg -1 , and total Zn of 130-870 mg kg -1 (Camps Arbestain et al. 2008).Sites M3w and M4vw were only amended once, and there were no reapplications.
On 9 March 2010, five soil samples were randomly collected using an Eijelkamp sampler at each selected site from areas spaced sufficiently far apart to be representative.The whole depth of each horizon was sampled and then stored in polyethylene bags, dried at room temperature and sieved to < 2 mm before being analysed.The single horizon of all of the soils was sampled except from M2, in order to observe possible changes in the subsurface horizon.

Soil physical and chemical analyses
Some of the characteristics of the selected soils that generally influence metal sorption are summarized in Table 1.The procedure of Kroetsch and Wang (2008) was used to determine particle size distribution (sand, silt and clay percentages).Mineralogical analysis of the clay fraction was carried out by X-ray diffraction of crystalline powder in a Siemens D-5000 diffractometer (Brindley and Brown 1980).Exchangeable cations (Ca 2+ , K + , Mg 2+ , Na + , Al 3+ , Fe 2+ and Mn 2+ ) were extracted with 0.1 M BaCl 2 (Hendershot and Duquette 1986) and their concentrations determined by ICP-AES (Perkin-Elmer Optima 4300 DV, USA).Effective cation exchange capacity (ECEC) was calculated by adding up the total cation concentrations.Soil pH was determined with a pH electrode in 1:2.5 water to soil extracts.The method developed by Mehra and Jackson (1958) was used to determine the free oxide concentrations.Aluminium, iron and manganese were determined in the extract by ICP-AES.Both soil organic and inorganic carbon (SOC and IC) were determined in a solid module (Shimadzu SSM-5000, Japan) coupled with a TOC analyser (Shimadzu TNM-1, Japan).Dissolved organic carbon (DOC) was extracted with bidistilled H 2 O according to Sánchez-Monedero et al. (1996) and the OC in the supernatant was determined with a TOC analyser.The different chemical organic matter fractions namely humin carbon (Humin C), carbon in the fulvic (C FA ) and the humic acids (C HA ) were extracted first with 0.1M Na 4 P 2 O 7 and then with 0.1M NaOH, by following the method described in De Blas et al. (2010) and Asensio et al. (2014).Pseudototal nickel (Ni), lead (Pb) and zinc (Zn) concentrations were extracted with aqua regia by acid digestion in a microwave oven (Milestone ETHOS 1, Italy).The certified reference material CRM026-050 was also analysed in parallel with samples to check the analysis.

Sorption experiment and construction of isotherms
The sorption capacity was evaluated after carrying out sorption experiments using the batch method described by Alberti et al. (1997) and Gomes et al. (2001) modified by Harter and Naidu (2001).Single-metal solutions of Ni 2+ , Pb 2+ and Zn 2+ were used as nitrates (0.03, 0.05, 0.08, 0.1 and 0.5 mmol L -1 ) containing 0.01M NaNO 3 as background electrolyte (Vega et al. 2009).Tri-metal equimolar solutions of Ni 2+ + Pb 2+ + Zn 2+ as nitrates were also prepared with the same background electrolyte (0.01M NaNO 3 ) to evaluate the competition for binding sites.The concentration of each metal in the tri-metal solution had the same concentration (which was the same as those of the single-metal solutions).We added 25 mL of the "sorption solutions" to 1.5 g of soil samples in polyethylene tubes and then shook them in a rotatory shaker for 24h at 25 ºC.After centrifuging for 10 min at 3000 rpm the supernatant was filtered through Whatman 42 paper (pore size 0.45 μm).The supernatants were analysed for Ni, Pb and Zn concentrations by ICP-AES (Perkin-Elmer Optima 4300 DV).All of the sorption experiments were performed in triplicate, and the data shown is the average of these three replicates.Each concentration sorbed by each soil sample was calculated as the difference between the concentration added (from sorption solution) and the concentration in the solution after equilibration (24 h shaking) with the soil.To minimize the effect of the background electrolyte, the sorbed concentrations were corrected by subtracting the values obtained in additional experiments with a sorption solution only composed of 0.01M NaNO 3 .
We also calculated the distribution of each metal (i) for each sorption solution concentration (c) by following equilibration in each stage.This distribution was expressed in terms of the quotient: (1) where C i,soil is the concentration of metal i on the soil (μmol g -1 ) and Ci,solution is the concentration of metal i in solution (μmol L -1 ) (Anderson and Domsch 1989;Covelo et al. 2008).This K d was calculated for each experimental replicate at each concentration of metal for all of the soils.We then calculated the mean of three replicates in order to obtain one value for each metal concentration.
The obtained isotherms were fitted, wherever possible (R 2 > 0.6), with the Langmuir and Freundlich equations (Vega et al. 2008).The Langmuir equation used is as follows: (2) The linearized Freundlich equation used is the following: (3) where C is the metal concentration in solution at equilibrium (μmol L -1 ); Cs is sorbed metal per gram at equilibrium (μmol g -1 ); β L is the maximum sorption capacity (μmol g -1 ), corresponding to the adsorption of a complete monolayer; K L is the Langmuir constant (L μmol -1 ), which reflects the energy of sorption; K F (L g -1 ) is the Freundlich constant, which reflects sorption capacity; and n is an adimensional parameter reflecting the intensity of sorption.
Finally, the overall capacity of the soils to sorb Ni, Pb or Zn was evaluated as the slope K r according to Vega et al. (2008).This K r is obtained from the regression equation C s,i = K r C p,i , where C s,i is sorbed metal i (μmol) per gram of soil at equilibrium, and C p,i is potentially sorbable metal i (i.e., the amount of metal i in the solution before contact with the soil) also per gram of soil.

Statistical analyses
All of the analytical determinations were performed in triplicate, and the data were statistically treated using the program SPSS 15.0 for Windows.Analyses of variance (ANOVAs) and homogeneity of variance tests were carried out.In case of homogeneity, a post-hoc least significant difference (LSD) test was carried out; otherwise Dunnett's T3 test was performed.Student's t-test was carried out to calculate the sorption selectivity sequences.All data was previously subjected to the Kolmogorov-Smirnov test for normality.Bivariate correlation analyses were also carried out according to Pearson's method.The obtained correlation (r) and the used probability (95% confidence = P < 0.05) are given in the results or discussion section.An independent t-test was carried out in order to compare the obtained K r and to be able to establish the selectivity sequences of sorption.

Soil characteristics
The mine soil from the untreated site (M1) is extremely acid according to the USDA (1998) (Table 1).The site vegetated with pines for 21 years (M2v), the site amended with waste (M3w) and the site simultaneously vegetated and amended (M4vw) have a significantly higher pH than M1.
The percentage of clay fraction, crystalline minerals, concentration of soil organic carbon (SOC) and both aluminium and manganese oxides (Al 2 O 3 and MnO) were significantly higher in the treated soils than in M1 (Table 1).On the contrary, the concentrations of iron oxides (Fe 2 O 3 ) were higher in the control soil than in the treated ones.Dissolved organic carbon (DOC) was only detectable in all of the treated soils except in M2Av, probably because it was lixiviated to M2Bv.

Sorption isotherms of Ni, Pb and Zn
The individual and competitive nickel isotherms (Figure 1 Note that the isotherms obtained for all of the treated soils were similar.The H-type indicates a high initial affinity of the treated soils for this metal.The high affinity of soils for Pb has been previously reported by other authors (Sauvé et al. 2000;Fontes and Gomes 2003;Vega et al. 2006).
The individual and competitive zinc isotherms were H-type for all of the treated soils and L-type for the control.These isotherms indicate that M4vw soils had the highest affinity for Zn and M1 the lowest.
It is interesting to observe that the isotherms of the subsurface horizon at the vegetated site (M2Bv) were similar to the ones from its topsoil (M2Av) instead of the ones from the control soil (Figure 1).According to the calculated K r , M4vw showed the highest sorption capacity of Ni and Zn for both individual and competitive experiments (Table 4).For the individual sorption of Pb, all of the treated soils statistically had the same K r , which was significantly higher than in the untreated soil.The highest K r values for competitive Pb were from the amended soils (M3w and M4vw) with values ranging from 0.98 to 1.00 (K r = 1 means the maximum sorption capacity).The untreated soil (M1) showed the lowest K r in all cases (Table 4).The K r values in the subsurface horizon at the vegetated site were similar to the values for its topsoil, instead of those from the control soil.
According to the calculated K r , the general soil sequence according to the both the individual and competitive sorption capacity for Ni, Pb and Zn was M4vw ≥ M3w ≥ M2Av ≥ M2Bv > M1.The small differences among them can be observed in Table 4.

Influence of soil properties on metal sorption
The data obtained for evaluating the sorption capacity (β L , K F and K r ) were correlated with the data for a number of soil physico-chemical soil characteristics (Table 1).There was a significantly positive correlation between the β L , K F and K r (for Ni, Pb and Zn, in both individual and competitive sorption) and the following soil characteristics: pH, exchangeable Ca 2+ , CECe, Al and Mn oxides, organic C (SOC), and humin C (P < 0.05 in all cases).The concentration of inorganic C (IC) had a significantly positive correlation with the values of K r and K F for Ni.

Influence of reclamation treatments on Ni, Pb and Zn retention
The parameters used to estimate the sorption capacity of soils (K r , β L and K F ) indicated that the untreated soil (M1) had a very low capacity for Ni, Pb and Zn retention (Tables 3 and 4).
Nevertheless, it can be seen that this capacity increases in the vegetated (M2v), amended (M3w) and vegetated + amended soil (M4vw).
The metal selectivity sequences according to K r for individual and competitive sorption can be also observed in detail in Table 4.The general selectivity sequence is Pb ≥ Zn ≥ Ni.
The treatment that increased the sorption capacity the most for Ni and Zn was simultaneous tree planting and amendment with waste, as the soil with both treatments (M4vw) had the highest K r and K F values for the sorption of these metals (Tables 3 and 4).In the case of Pb, amendment with waste was the treatment that most increased the sorption capacity of soils, since there were no significant differences between the K r of M3w and M4vw for this metal.Moreover, the β L for Pb was higher in M3w than in M4vw.This higher sorption capacity of the amended soils in comparison with the soils that were only vegetated or untreated was due to the highest concentration of soil components with high retention capacities in M3w and M4vw.It has previously been demonstrated that the soil components with the highest sorption capacity are usually organic matter, clay minerals and carbonates (Stahl and James 1991;Temminghoff et al. 1997;Vega et al. 2006;Covelo et al. 2007b).Soil M4vw had one of the highest SOC concentrations and the highest IC and Mn oxide concentrations of all of the soils.M4vw had a higher retention capacity for Ni, Pb and Zn than M2Av, despite having a lower percentage of clay than M2Av.These results indicate that organic matter and inorganic C played a key role in the retention of metals in the reclaimed mine soils.
The statistical data support this hypothesis with the significantly positive correlation obtained between β L , K F and K r for Ni, Pb and Zn (in both individual and competitive sorption) and the Planting trees also significantly increased the sorption capacity for Ni, Pb and Zn in the mine soil, since M2Av had significantly higher values of β L , K F and K r than the untreated site M1 (Tables 3 and 4).The maximum sorption values were significantly much lower in M2Av than in M4vw, in spite of the highest clay percentage in M2Av.This is probably because of the lower SOC and pH in this soil than in M4vw, which seem to be more important parameters than clay in the retention capacity of metals by the studied soils.

Sorption capacity and selectivity sequences
The metal that was sorbed the most in the individual and competitive sorption was Pb for all of the soils except for M4vw, whose retention capacity was the same for Pb and for Zn.This is the opposite of what we expected, as lead is usually the most strongly sorbed metal by soils and sediments (Fontes and Gomes 2003;Fan et al. 2007;Vega et al. 2008;Seo et al. 2008).
Therefore, there may be some factor promoting Zn sorption in M4vw compared to the other studied soils.This vegetated + amended soil can be distinguished from the other soils by its much higher concentration of exchangeable calcium (Table 4), and it being the only one where calcite (CaCO 3 ) was detected.It is possible that Pb 2+ and Zn 2+ precipitated as PbCO 3 and ZnCO 3 , sorbed onto calcite, exchanged by Ca 2+ or that all of the processes occurred in M4vw.It is known that both Pb and Zn can be sorbed onto calcite and can even precipitate if their ion activity product is exceeded (Papadopoulos and Rowell 1989;Zachara et al. 1989;Elkhatib et al. 1991;Izquierdo et al. 2013).It is also known that Zn unavailability is due to the sorption of Zn by carbonates, precipitation of Zn in form of hydroxide or carbonates, or even the formation of insoluble calcium zincate (Adriano 2001).The high SOC concentration in M4vw also influenced the higher sorption capacity compared to the other soils.Several statistical correlations help to support this hypothesis: the positive correlations obtained between the exchangeable Ca 2+ and the K r for Pb and Zn in both the individual and competitive sorption (P < 0.01 in all cases), and the positive correlations obtained between the metal sorbed in each experiment (Ni, Pb and Zn, individual and competitive sorption) and the corresponding released Ca 2+ (P < 0.01 in all cases).However, this exchange between Pb or Zn and Ca may occur when the concentration of these metals reaches around 0.5 M, because at lower concentrations, the amount of Ca released was undetectable (Figure 2).It is likely that low concentrations of Pb 2+ or Zn 2+ in the soil solution were sorbed by the soil elements (organic matter, minerals or oxides) and when their concentration is high enough, the exchange with Ca 2+ occurred.
All of these arguments suggest that when amending degraded soils, it is important to use wastes that not only have a high concentration of organic carbon or clays, but also of carbonates.
The amount of Zn sorbed was always higher than the amount of Ni in all of the treated soils in the individual experiments, while the opposite occurred in the untreated soils, where Ni was sorbed more than Zn.
As with the individual sorption experiments, the same happened in the competitive experiments, except in soils M1 and M2Av, where the sorption capacity of Ni and Zn was statistically the same.This implies that the sorption of each studied metal was not significantly affected by the presence of the others.

Conclusions
Regardless of the possible supply of Ni, Pb and Zn to degraded soils from the sewage or paper mill residues used as amendments, the soils treated with these types of waste had a high sorption capacity to retain these metals.This would prevent Ni, Pb and Zn from occurring in mobile form in the amended sites, even if they are present in a high concentration in the soils.
The sorption capacity of the soils amended with waste can be attributed to their high concentration of organic and inorganic carbon, as well as their high pH and effective cation exchange capacity.The high concentration of calcium in the amended and vegetated soil, which came from the sewage and paper mill residues that were added, considerably increased the capacity of the soils to retain metals as Ca 2+ can be exchanged with Ni, Pb or Zn.Therefore, the treatment that most increased the capacity of the mine soils to retain Ni, Pb and Zn was the amendment with organic wastes rich in calcium and planting of vegetation.
) were L-type for the control (M1) and the vegetated soils (M2Av and M2Bv), and C-type for the amended (M3w and M4vw) according toGiles et al. (1974).These isotherms indicate that the amended soils had the highest affinity for Ni.The individual and competitive Pb isotherms (Figure1) were L-type for M1 and H-type for all of the treated soils according toGiles et al. (1974).

Figure 2 .
Figure 2. Concentrations of Pb and Zn sorbed in the individual batch experiment as well as Ni + Pb + Zn sorbed in the competitive experiment, together with the concentrations of Ca 2+ displaced in each situation.

Table 1 .
Selected physical and chemical properties of the studied soils Note: Mean and 95% confidence interval (CI) for three independent replications.Values followed by different letters in each row of each mine area differ significantly with P < 0.05.u.l.: undetectable level.EC: electrical conductivity.ECEC: effective cation exchange capacity.IC: inorganic carbon.SOC: soil organic carbon.DOC: dissolved organic carbon.Humin C: carbon in the humin.C FA and C HA : carbon in fulvic and humic acids.[PLANTINGTREES AND AMENDING WITH WASTE INCREASES THE CAPACITY OF MINE TAILINGS SOILS TORETAIN Ni, Pb AND Zn ]

Table 2 .
Fitting sorption isotherms for Ni 2+ , Pb 2+ and Zn 2+ according to the best fitting model PLANTING TREES AND AMENDING WITH WASTE INCREASES THE CAPACITY OF MINE TAILINGS SOILS TO RETAIN Ni, Pb AND Zn ] Note:Values in bold type indicate that R 2 < 0.7.[

Table 2 )
can be used for evaluating the sorption capacities.However, the sorption data did not fit well with either of the models in all of the cases, i.e., Ni, Pb and Zn in both individual and competitive sorption.For this reason, it was only possible to use the Freundlich model for comparing Ni and Zn sorption while the Langmuir model was only useful for the Pb sorption data.Therefore,

Table 3 .
Maximum capacity of sorption for Ni 2+ , Pb 2+ and Zn 2+ from the studied soils according to the best fitting isotherm model [ PLANTING TREES AND AMENDING WITH WASTE INCREASES THE CAPACITY OF MINE TAILINGS SOILS TORETAIN Ni, Pb AND Zn ]