1 Introduction

Under natural conditions, the concentration of trace elements in the soil is relatively low, but certain human actions, such as agriculture, (Kabata-Pendias 1995), the application of tailings with residues when used as an organic amendment (Plaquart et al. 1999; McBride 2003), the application of solid urban waste as fertiliser, atmospheric particle transport (Alloway 1995; Föstner 1995), as well as mining (Egger 1994; Ripley et al. 1996; Vartanyan 1989), mining accidents (Cabrera et al. 1999; Simón et al. 1998) or continuous atmospheric depositions (Nicholson et al. 2003) can raise contamination above toxic levels. In these cases, the soil is capable of attenuating the mobility of diverse contaminants by, for example, filtration, neutralisation, adsorption and precipitation; these processes inactivate potential toxicity, blocking undesired elements from passing to more sensitive systems such as the air or water. The capacity of each soil to retain metals depends on the soil properties, mainly texture, organic-matter content, ion-exchange capacity, oxide contents, pH, specific surface area and carbonate content (Alloway 1995; Ross 1994). When this capacity is exceeded, the soil ceases to be effective as the protector of the ecosystem (sink) and can even function as a source of toxic substances. The critical load represents “the maximum quantity of a given contaminant that can be supplied to a soil without causing chemical changes leading to long-term harmful effects on ecosystem structure and function” (Hettelingh et al. 1991).

This study was aimed at estimating the critical load of soils with different properties withstanding a contaminating solution resulting from the oxidation of pyrite tailings with a pH value lower than 2.0 and containing a high concentration in trace elements such as Cu, Zn, Cd and Pb. This estimation can be made as a function of: (a) maximum concentration of Cu, Zn, Cd and Pb that precipitate in each soil within the range of pollution considered in this study (MPC, maximum precipitated concentration), (b) concentration of precipitated elements at which action values (AV, action values) or values above which plants and soil solution should be watched (Prüeβ 1997), are reached (CAV).

When the soil concentration of a certain contaminant approaches the MPC, not only does it exceed its critical load and affect the functioning of the ecosystem, but new additions could reach surface waters and groundwater whereupon the contamination would spread to other ecosystems. At the same time, given that the precipitated elements may remain in bioavailable form, before the MPC is reached, the AV may be reached, at which point the soil would be expected to begin to display harmful effects. Therefore, the structure and functioning of the ecosystem would begin to be visibly affected when the soil reaches the AV, whereas when it reaches the MPC the surrounding ecosystems would also be visibly affected.

2 Materials and Methods

2.1 Soil Properties

The soil samples belong to the provinces of Granada and Jaen located in the southeast of Spain. The selected soil samples represent a heterogeneous range of soils with different chemical properties and developed over different parent materials (Table 1).

Table 1 Physico-chemical properties of the studied soils
Full size table

Soil samples were air dried and then screened to 2 mm to analyse them. Particle size distribution was determined by the pipette method after elimination of organic matter with H2O2 and dispersion with sodium hexametaphosphate (Loveland and Whalley 1991). The pH was measured potentiometrically in a 1:2.5 soil–water and soil 0.1 mol l-1 KCl suspension. The calcium carbonate equivalent content (CaCO3, %) was determined by the method of Bascomb (1961). Total carbon was analysed by dry combustion with a LECO SC-144DR instrument. Organic carbon (OC) was determined by the difference between total carbon and inorganic carbon from CaCO3. The cation-exchange capacity (CEC) was determined with 1 M Na acetate at pH 8.2, measuring sodium in a Meteor NAK-II flame-photometer. Exchange bases were determined with 1 M H4N acetate at pH 7.0, measuring sodium and potassium in a Meteor NAK-II flame-photometer and calcium and magnesium by atomic absorption spectrometry in a Varian SpectrAA 220FS machine, the specific surface area (SA) determined by weighing the water adsorbed by the sample from a solution saturated with CaCl2 (Keeling 1961); total iron (Fed) and aluminium (Ald) oxides and oxi-hydroxides of the soil samples were extracted with citrate-dithionite (Holmgren 1967).

2.2 Contaminating Solution (CS)

Ten grams of pyrite tailings from the mining spill at Aznalcóllar (Seville, Spain) were placed in contact for 3 days with 1,000 cm3 of H2O2 at 15%. The contaminating solution (CS) presented a pH of 1.7 and a concentration in Cu, Pb, Zn and Cd of 14.5, 3.2, 55.4 and 0.21 mg dm−3, respectively (more details of the contaminating solution composition is given in Simón et. al. 1999).

2.3 Spiked Soils

According to their properties, different quantities of CS were added to soil samples. Thus, 1, 1.5, 2, 3 and 5 cm3 CS g−1 were added to the sandy and acid soils (pH < 7.0); 1, 2, 4, 6 and 10 cm3 CS g−1 were added to soils with a sandy-loam texture or finer and pH > 7.0 and CaCO3 <2%; 2, 6, 10, 15 and 20 cm3 g−1 were added to soils with a loamy texture or finer, pH > 7.5 and CaCO3 between 2% and 30%; and 10, 20, 30, 40 and 50 cm3 g-1 were added to soils with a loamy texture or finer, pH > 8.0 and CaCO3 > 30%. After 72 h of shaking (Alonso et al. 1997), each soil extract–CS was measured for pH and centrifuged at 3,000 rpm for 15 min, separating the solid and liquid fractions.

2.4 Total Water and NH4NO3 Extractable Trace Element Concentrations

To determine water soluble forms, soil–water extracts were prepared in a ratio 1:10 (Norma DIN 38 414-4) and then Cu, Pb, Zn and Cd water extractable concentrations were analysed. Trace element precipitated concentrations (MPC) were determined as the difference between trace element concentration in the CS and water extractable concentration measured in soil water extracts. Finally, the pH value measured in the soil–water extracts. Soils were also analysed for the trace element concentrations extractable by NH4NO3 1 M (DIN 19730; Prüeβ 1997). In all cases, the concentration of the different metals (Cu, Zn, Cd and Pb) was measured by ICP-MS in a PE SCIEX ELAN-5000 A spectrometer.

Distribution coefficients (K d) represent the sorption affinity of the soil solid phase for trace elements in solution and can be used as a valuable tool to study metal-cation mobility and retention in soil systems. The distribution coefficients (K d), ratios between the quantity of precipitated element (mmol kg−1) and the quantity remaining in solution (mmol dm−3), were calculated according to Alloway (1995). According to Anderson and Christensen (1988), high K d values indicate that the metal has been retained by the solid phase through sorption reactions, while low values of K d indicate that a large fraction of the metal remains in solution.

2.5 Determination of Buffering Capacity of Soils

The pH-buffering capacity of soils at pH = 3.5 was evaluated (BC3.5) as the cmol of H+ needed to reduce soil pH value up to 3.5. The titration curves were drawn following the method of Hartikainen (1986), soil samples (air dried, <2 mm) were equilibrated with solutions of increasing concentrations of 10−3 M HCl (soil-solution ratio of 1:10) while maintaining the ionic strength constant. The suspensions were shaken for 60 min and then the pH measured after 4 days of settling; then the amount of H+ added were plotted against the equilibrium pH values (titration curves). From these curves, the values for the acid neutralisation (BC) were determined graphically, this being referred to as the quantity of acid needed to bring the pH of the soil to a given value (Van Breemen et al. 1983). In the present study, the pH value of 3.5 (BC3.5) was taken as the reference value, like one of the intervals defined by Ulrich (1981) for soils and reflects the aluminium and iron buffer region. Finally, the data were statistically processed using SPSS 12.0 software package.

3 Results and Discussion

3.1 pH-Buffering Capacity

The main property that marked a significantly different behaviour among the soils was the BC3.5, which was related (P < 0.005) to the calcium carbonate content (% CaCO3) and the sum of the exchangeable calcium and magnesium content in soils ([Ca + Mg] expressed in cmolc kg−1):

$${\text{BC}}_{3.5} \left( {{\text{cmol}}\,{\text{H}}^ + \,{\text{kg}}^{ - 1} } \right) = 0.792\,{\text{CaCO}}_{\text{3}} + 0.870\left[ {{\text{Ca}} + {\text{Mg}}} \right]\quad r^2 = 0.903$$

indicating that the weathering of the carbonates and the exchange between basic cations (among which Ca and Mg are the most abundant) and protons are the main mechanisms controlling soil acidification. In any case, soils with different BC3.5 values show similar pH-buffering capacity pattern in such a way that, based on this pattern, the soils were assigned to three groups (Fig. 1):

  1. 1.

    soils with BC3.5 < 3 cmol H+ kg−1, in which the pH value was reduced rapidly with the addition of CS, reaching pH values of 3.0 for additions of ≤2 cm3 CS g−1 of soil;

  2. 2.

    soils with BC3.5 between 3 and 20 cmol H+ kg−1, in which the pH decreased gradually, reaching values of 3.0 for relatively lower concentration of contaminating solution (between 2 and 4 cm3 CS g−1 of soil);

  3. 3.

    soils with BC3.5 > 20 cmol H+ kg−1, in which the decline of the pH was very gradual and in no case reached values of less than 6.5.

Fig. 1
figure 1

Changes in pH values measured in H2O of soils with different pH-buffering capacity at pH = 3.5 (BC3.5) with increasing amounts of contaminating solution added (cm3 SC g−1 soil). (A) BC3.5 > 20 cmol H+ kg−1; (B) BC3.5 between 3 and 20 cmolH+ kg−1; (C) BC3.5 < 3 cmol H+ kg−1

Full size image

3.2 Water-extractable Concentrations and Coefficient of Distribution

The coefficient of distribution (K d), the ratio between the quantity of the precipitate of each element (mmol kg−1) and the quantity that remained in solution (mmol dm−3), revealed that the precipitation rate of the different trace elements was related to the BC3.5. Thus, the greatest proportion of water extractable trace elements concentration was found in soils with BC3.5 < 3 cmol H+ kg−1 (Fig. 2 A), followed by those with BC3.5 between 3 and 20 cmol H+ Kg−1 (Fig. 2 B); meanwhile, the greatest precipitation rate was registered for soils with BC3.5 > 20 cmol H+ Kg−1 (Fig. 2 C). Similarly, the K d values showed high solubility of Cd and Zn as opposed to Cu and Pb.

Fig. 2
figure 2

Distribution coefficient for Cu, Pb, Cd and Zn in the studied soils, calculated at a dosage of 500 mg kg-1 of SC (K d500). In the graph, we can distinguish between (A) K d500 of soils with BC3.5 < 3 cmol H+ kg-1; (B) K d500 of soils with BC3.5 between 3 and 20 cmolH+ kg-1; (C) K d500 of soils with BC3.5 > 20 cmol H+ kg-1

Full size image

In general, within each of the defined groups, the quantity of precipitated metal tended to increase progressively with the ratio CS–soil extract until the maximum retention capacity of each soil (MPC), above which the precipitated metal concentration either tended to remain constant or tended to diminish. The first behaviour was noted in the soils with BC3.5 > 20 cmol H+ kg−1 (Fig. 3, Gr13), whereas the soils with BC3.5 < 20 cmol H+ kg−1 displayed the second behaviour (Fig. 3, Gr3). Therefore those soils with BC3.5 < 20 cmol H+ kg−1 could thus become a source of contamination when the MPC is exceeded (chemical time bomb; Stigliani 1988). In general, this process takes place in the studied soils when the pH of the soil extract–CS is less than or equal to 3.0.

Fig. 3
figure 3

Graph showing the precipitated concentration of copper with the increasing amount of metal (copper) added with the contaminating solution in soils Gr-3 (with BC3.5 < 20 cmol H+ kg-1) and Gr-13 (with BC3.5 > 20 cmol H+ kg-1)

Full size image

In any case, these two types of behaviour should be understood within the range of the studied pollution, since at higher contamination rates even the pH of those soils with BC3.5 > 20 cmol H+ kg−1 could fall to values below 3.0, releasing precipitated elements and thus becoming in a potential threat to the ecosystem exposed to them.

Although MPCCu, MPCPb, MPCZn and MPCCd were significantly related (p < 0.01) to the BC3.5, this relationship presented certain particularities, depending on the element. Thus, in the case of Cu, Zn and Cd, the relationship between the two parameters was linear:

$$\matrix {{\text{MPC}}_{{\text{Cu}}} \left( {{\text{mg}}\,{\text{kg}}^{ - 1} } \right) = 11.84 + 3.11\,{\text{BC}}_{3.5} \left( {{\text{cmol}}\,{\text{H}}^ + \,{\text{kg}}^{ - 1} } \right)}{r^2 = 0.965} \\ {{\text{MPC}}_{{\text{Zn}}} \left( {{\text{mg}}\,{\text{kg}}^{ - 1} } \right) = 13.03 + 9.35\,{\text{BC}}_{3.5} \left( {{\text{cmol}}\,{\text{H}}^ + \,{\text{kg}}^{ - 1} } \right)}{r^2 = 0.955} \\ {{\text{MPC}}_{{\text{Cd}}} \left( {{\text{mg}}\,{\text{kg}}^{ - 1} } \right) = 0.124 + 0.02\,{\text{BC}}_{3.5} \left( {{\text{cmol}}\,{\text{H}}^ + \,{\text{kg}}^{ - 1} } \right)}{r^2 = 0.976} \ $$

while Pb presented a different behaviour, depending on the value of the BC3.5 (Fig. 4). When the BC3.5 < 70 cmol H+ kg−1, MPCPb was related logarithmically to BC3.5 and tended to reach a constant value. When BC3.5 > 70 cmol H+ kg−1, as in the soils with a very high CaCO3 content, the MPCPb tended to increase linearly.

Fig. 4
figure 4

Relationship between the maximum quantity of Pb precipitated (MPCPb) and the pH-buffering capacity (BC3.5)

Full size image

3.3 Copper, Lead, Zinc and Cadmium NH4NO3 Extractable Concentrations

The relationship between the precipitated and the NH4NO3 1 M extractable concentration for each metal was established. From this relationship, we can deduce the quantity of precipitated metal at which the action level is reached (CAV) accepted by the DIN 19730 guidelines where AV are reported as 1 mg Cu kg−1 soil (AVCu), 3 mg Pb kg−1 soil (AVPb), 5 mg Zn kg−1 (AVZn) and 0.08 mg Cd kg−1 soil (AVCd) (Prüeβ 1997). In all cases, the CAV value was also significantly related (p < 0.01) to the BC3.5, although with certain particularities (Fig. 5). Thus, the CAV of Cu and Zn were related to the BC3.5 through second-degree equations, differentiating among soils with BC3.5 < 70 cmol H+ kg−1 and BC3.5 > 70 cmol H+ kg−1. It is worth noting that the soils with BC3.5 > 70 cmol H+ kg−1 presented relatively low CAV values, indicating that, although the highly carbonate soils can precipitate high quantities of Cu and Zn, their bioavailability is high and can easily get into the food chain.

Fig. 5
figure 5

Relationship between the quantity of metal precipitated (CAVCu and CAVZn) at which the action value (AV) is reached and the pH-buffering capacity at pH 3.5 (BC3.5). Relationship between the quantity of metal precipitated (CAVCd and CAVPb) at which the action value (AV) is reached and the pH-buffering capacity at pH 3.5 (BC3.5)

Full size image

On the contrary, Cd and Pb did not differ. Thus, in all the soils, the CAVCd was related to the BC3.5 according to the logarithmic equation in which there was a tendency for a steady value to be reached (Fig. 5c); this indicates, in general terms, that when the soils presented a Cd concentration of around 2.5 mg kg−1, they exceeded the critical bioavailability level. In the case of Pb (Fig. 5d), the relation between CAVPb and BC3.5 was exponential, so that an increase in the neutralisation capacity of soil acids allows a considerable increase in Pb before the critical bioavailability level is reached.

4 Conclusions

The calcium carbonate content together with exchangeable calcium and magnesium content are the main parameters regulating the buffering capacity as well as the critical load of the soils polluted by pyrite tailings. However, these parameters differ between the studied elements, Cu, Zn, Cd and Pb in soils. Thus, while the NH4NO3 extractable Cu and Zn concentration tends to increase in highly carbonated soils, the Pb reduces drastically and the Cd tends to remain constant regardless of the soil characteristics.