1. INTRODUCTION

The ore field Libiola, situated on steep slopes of the Gromolo valley was in the past the most important Cu-mine in the Liguria district. It was exploited since the Copper/Bronze Age, mining activity is documented at the end of the 4th millennium BC (Maggi & Pearce, 2005) and later in the 17th century, as well as from 1864 to 1961. The large open-scale quarry operations in the fifties and sixties of 20th century overlapped part of the ancient mining localities and substantially changed the country relief (Kienlin & Pearce, 2008).

The mineralization is formed on the western boundary of the Valgraveglia ophiolite complex, in noth-east direction from Sestri Levante. More than 1.5 Mt of Cu-ore was between 1864 and 1961 exploited (Galli, 1996). The sulfide mineralization forms massive lenses at the top parts of a huge pillow-basalt dome (Ferrario & Garuti, 1980). These basalts underlain and locally overthrust by serpentinite and gabbro bodies (Garuti et al., 2008). The second type of sulphide mineralization is the serpentinite hosted low-grade one, connected with shear zones. According to Bertolani (1952) it was formed by syntectonic remobilization of original magmatic sulfides.

The ore mineralization of quartz-calcite veins (with chlorite) consists most of all of pyrite and chalcopyrite, accompanied by less represented sphalerite and pyrrhotite (Marescotti et al., 2007; Garuti et al., 2008; Di Lisi et al., 2020; Andráš et al., 2021).

The aim of the presented article is describe the geochemical characteristics of the dump matter and visualize the spatial distribution of the selected PTEs.

Two types of mine waters were distinguished within the studied area. The first one is represented by red acid waters (pH 2.4 - 2.8), and the second type by neutral waters (pH 7.0 - 7.5) of  blue colour, which  is  caused  by precipitated phases. Also, a mixture of these two specific types of mining waters is present (Marini et al., 2003).

The studied dump-field (Figures 1-3) is localized on a steep slope without vegetation above Gromolo creek and topped by a quarry at the top of the hill. At the central part, there are several natural outcrops of the underlying rock complex. The aim of the article is to characterize the dump field's technosol contamination and present its spatial distribution using GIS tools.

2. MATERIAL AND METHODS

The technosol is very shallow, only several centimeters deep on the slope. The depth overreaching 1 m is only at the foot of the dump-hill. The sampling was realized in regular grid network (50x50 m) with respect to the terrain conditions. Each sample by homogenization of 10 smaller samples weighing 500 g was created. All of them were dried at laboratory temperature to minimize loss of volatile elements and both rinse and paste pH/Eh was determined according to Sobek et al. (1978) from distilled water and 1M KCl lixivium, using pH-metre EUTECH Instruments device. The obtained data was calculated for standard hydrogen electrode. 

To assess the PTEs concentrations in soil, a dry sample was ground to fine powder. For the performance of microwave digestion (Ethos plus 2, MLS), 100 mg of powder was dissolved in a solution of nitric and hydrochloric acid (aqua regia) in a ratio of 1:3 according to Krachler et al. (2002). The elemental concentration was measured by mass spectrometry with inductively coupled plasma (ICP-MS), single quadropole (Agilent 7900, USA).

X-ray powder diffraction analysis were performed by a Bruker D8 Advance diffractometer (Earth Sciences Institute in Banská Bystrica) using CuKα radiation generated at 40 kV and 40 mA and Sol-X SDD detector. Experimental data were compared with the PDF2 database. The X-ray powder diffraction analysis was performed to identify the mineralogical composition of the technosol. Analyses were performed by a Bruker D8 Advance diffractometer (Earth Sciences Institute in Banská Bystrica) using CuKα radiation generated at 40 kV and 40 mA and position sensitive detector SSD 160. Experimental data were compared with the PDF2 database. 

The Energy-dispersive micro X-ray fluores-cence spectrometry (ED-XRF) analysis was perfor-med using an M4 TORNADO (Bruker) spectrometer. Element distribution maps were obtained from the selected area within a polished grainy specimen. The excitation current 600 μA at 50 kV of Rh anode was used. The analyses were carried out in a vacuum of 20 mbar, the distance between the two measurement points was 15 μm, at a speed of 5 ms/pixel (spot size was 20 μm). The SDD detector that collects the fluorescent signal has an active area of 30 mm2 and a spectral resolution of 145 eV. Element concentrations were computed by the fundamental parameter method, concentrations were integrated across homogenous areas selected from elemental maps.

The values of the rH2 factor, defined by (Richter & Hlušek, 2003) given by the formula: rH2 = Eh/30 + 2pH, vary for well aerated soils in the range 28 – 34, and the not aerated soils are defined by values < 20.

The DpH factor represents the value that can be obtained by substraction of pHH2O from pHKCl. A positive DpH factor determines the presence of soil colloids with negative surface charge, while negative DpH values indicate the presence of colloids with positive charge (McNeill, 1992).

2 g of well homogenized fine-grained samples (80 mesh) were wetted with a distilled H2O and dissolved in H2O-HF-HClO4-HNO3 mixture. The limpid solution was refilled by HCl to exact volume and analysed for Fe, Cu, Pb, Zn, As, Sb, Ni, Co, Cd, Sr, Bi, U, Th, Mn, Mo, Ag and Au, by Inductively Coupled Plasma Mass Spectro-metry (ICP-MS) in Acme Laboratories (Vancouver, Canada) by Perkin-Elmer Elan 6000 device. 

The sulphur content (total sulphur – Stot., sulphidic sulphur – SS and sulphate sulphur – SO4) was measured in Ecologic Laboratories in Spišská Nová Ves by Inductively coupled plasma atomic emission spectro-scopy (ICP-AES) in 0.5 g of sample. The carbon content (Ctot. and Corg.) was measured in the Geological Institute of the Slovak Academy of Sciences in Banská Bystrica by IR spectroscopy using a Ströhlein C-MAT 5500 device in 0.5 g of sample. The Cinorg. was calculated as a difference of Ctot. – Corg.

The Spearman rank correlation coefficients between couples Ss/pH(H2O) as well as Cu and Ni /pH were calculated from all metal pairs in software application SPSS Statistics (version 19) provided by IBM. The correlations characterized by p ≤ 0.05 values were considered as statistically significant.

In order to estimate the spread of contamination by selected elements, GIS tools were applied. The base-map and the boundaries of dump-field were generated with respect to the original ortophoto documentations. Distri-bution was modelled in ArcGIS application (version 10.3.1), which is one of the most popular and universal software used by numerous experts. The models of contamination spread were generated separately for each element using the IDW interpolation method which uses the measured values surrounding the prediction locations.

3. RESULTS

At Libiola dump-field is the dump-matter very acidic, the pH_H2O varies from 2.76 to 4.94 (x̄ 4.03) and the pHKCl from 3.37 to 6.47 (x̄ 4.82). The redish-brown colour of the sediments indicates oxic conditions (EhH2O 12 – 146, x̄ 67.71; EhKCl 32.5 – 223.3, x̄ 128.8). The DpH factor (McNeill, 1992) which fluctuates between  -1.71 to 0.05 (x̄ -0.79; Table 1) indicates soil colloids with negative surface charge.

The values of rH₂ factor (Richter & Hlušek, 2003) 13.32 to 14.58 (x̄ 14.00; Table 1), are characteristic for not well aerated soils. The mineralogical composition of the technosol is as follows: the dominant minerals are phyllo-silicates (chlorite, chlorite-smectite and tri-smectite/ saponite; categorically are present also talc and minerals from the serpentinite group), quartz, plagioclase, magne-tite, hematite, and goethite, as well as some not exactly defined micas (Figures 4-5).

The area of the dump-field is shown on Figure 1. The distribution of the main contaminants is presented on Figures 6 – 13. The highest Fe and Pb contents were detected on the northern border of the dump-field on the foot of the hill and in the open-scale quarry on the top of the landfill itself (Figures 6-7). The highest Cu contamination (Figure 8, as well as of the As and Zn is in the quarry (Figures 9-10). Chromium and nickel are concentrated mainly under the slope in the depression above Gromolo creek (Figures 11-12).

The cumulative contamination given by the sum of all PTEs is presented on Figure 13. The highest correlation factor r was confirmed for metal pairs Co/Mn (0.79), As/Ag (0.67), Fe/Mo (0.67), Fe/Ag (0.61), Pb/As (0.76) and Pb/Sr (0.66), Mo/Ag (0.78), Zn/Cd (0.67), Zn/As (0.60) and Zn/Sr (0.65); correlation between two most risk metals Cu/Ni is very low (0.18).

This high degree of correlation corresponds to the close geochemical properties of the bound-ed element pairs that migrate during weathering (Polański & Smulikowski, 1978; Aiello et al., 2021).

As a consequence of the not well aerated technosol matter, the Eh–pH diagram for the Fe-S-H2O system (Osseso-Asare et al., 1984; Figure 14) shows the presence of not oxidized (or only slightly oxidized) iron in Fe2 form. 

Application of pH-Eh diagram according to Ryu et al. (2002) show that the substantial proportion of As belongs to the stability field As2S3, it means that it is present in the high toxic As3+ form. It is caused by the low degree of the technosol aeration (rH2 20; Table 1). Also the higher content of sulphidic sulphur (Ss) in comparison to sulphate sulphur (SSO4) presented in Table 3 explain the low degree of the As oxidation.

The studied technosols samples contain 0.07 – 2.22 % of Stot. (Table 3). In the dump-field matter sulphidic sulphur (SS  in average 0.35 %) prevalence on sulphate sulphur (SSO4 0.10 %) was recognized, thus the dump matter still contain a great portion of not oxidized primary sulphidic minerals.

The correlation between sulphidic sulphur Ss content and pH (r = -0.46) show moderate (week) relationship. These data indicates a certain degree of dependence between the content of sulfides and their ability to produce acidity.

The correlation between pH and two most risk metal content show that Cu/pH (r = -0.19) Cu migrate in acid conditions much more better as Ni (r = 0.43). This finding is supported also by week correlation betweer Cu and Ni (r = 0.18).

Copper in technosol in pH/Eh plot according to Fairthorne et al. (1997) belongs predominantly to the stability field Cu2+ and only a very small part of the samples tends to the Cu (Fe2O3) field (Figure 15).

The total carbon content is with respect to the character of the bedrocks (ultrabasic rocks) very low (in average 0.67 %) and the substantial part of this content depends on the concentration of the organic carbon (in average 0.64 %).

4. DISCUSSION

Several authors published basic descriptions of the PTEs contamination at Libiola dump field (Dinelli, et al., 1998; Marescotti et al., 2012; Buccheri et al., 2014, 2019; Andráš et al., 2021) but first contour maps of selected compounds and elements  (Fe2O3, MgO, Cr2O3, Cu, Ni), net acid-producing potential, and several selected minerals (sulphides, goethite, silicates) were published only by Marescotti et al. (2012, 2015). Our findings mostly correspond with data published in this work. 

The PTEs distribution depends on the content of the individual metals in the bedrocks and by their migration, which is controlled by weathering, oxidation, geochemical properties of PTEs, by their solubility, influenced mainly by activity of acid drainage water, as well as by sorption capacity of the secondary minerals (e.g. clay minerals, Fe-hyro-oxides, described by Marescotti et al. (2012) and by great erosion process on the steeply deeping slope (mostly without any vegetation cover) of the hill above the Gromolo creek. 

Also the character of the soil solutions containg negative surface charge contribute to the PTEs immobilization. The equilibrium between the PTEs release from the technosol and of their adsorption processes during transport control the river pollution.

The pH data are identical with those of (Zotti et al., 2014). More authors studied the PTEs content in technosol at the Libiola deposit in a different context. Our results are comparable with those published by Dinelli et al. (1998) and Buscaroli et al., (2017), except Ni (244 – 455 mg·kg-1 vs. 165 – 3455 mg·kg-1 tab. 2. Zotti et al. (2014) presents comparable data for Cu content but lower data for the other metals.The highest mobility was observed in the case of PTEs, which are accumulated in the lower parts of the dump: Fe, Ni, Cr and Co. The generally very mobile Zn (Benvenuti et al., 1997, 2000; Lu et al., 2005) show the highest contents in the quarry on the top of the dump-hill. The rest of Zn was probably washed away along a steep slope into the Gromolo creek. This assumption is also confirmed by Zn concentration data, which reach in the creek up to 2105 mg·L-1. Also, high concentrations of Cu and Ni (1057 and 44.78 mg·L-1) were in the creek drainage at the valley below the heap described (Singovszká et al., 2015; Buccheri et al., 2019; Buccheri & Andráš, 2022). These results document that the dump represents a significant source of metal pollution for the aquatic environment and confirm the link between geochemical conditions in technosol and surface water quality.

Italy has two different law limits for the content of PTEs given by the Italian Law Decree 152/2006 (Italiana, 2006) the first approving the Code on the Environment for industrial/ commercial use and the second one, more strict, for public and green areas. As the dump-field is several km away far from the village of Libiola, it is suitable to compare the PTEs contents with the law limits for industrial/commercial use.

The Cu content, except 3 samples, exceeds the limits allowed by law (600 mg·kg-1). Pb contents exceed the stricter limits for residential and green areas only in three samples (100 mg·kg-1). On the contrary, the Zn contents do not meet either of the limits (150 and 1,500 mg·kg-1), with exception of 4 samples that meet both. The Cd content in all samples respects the less strict standard (15 mg·kg-1) and except for 11 samples which exceed the stricter standard (2 mg·kg-1). Contents of Co don’t exceed both the law limits (20 and 250 mg·kg-1). On the other hand, the Ni contents exceed the less strict limits (500 mg·kg-1) in half of the samples. Except for one single sample, As contents don’t exceed even the stricter standard (20 and 50 mg·kg-1). There are no limits established for the other studied elements in the Law Decree 152/2006.

The great part of the total sulphur content (0.45 %) falls on the sulphidic sulphur (0.35 %) and the sulphate sulphur forms only the rest (0.10 %). It means that there is relatively great presence of undecomposed sulphides in the tailings which could possibly form increased acidic conditions, as a consequence of weathering processes.

The ultrabasic bedrocks at Libiola contain no carbonates, so therefore, the total carbon contents (Ctot.) are very low (in average 0.67 %) and the substantial part of this carbon corresponds to the organic carbon (x̄ = Corg. 0.64 %), whereas the inorganic carbon forms only the negligible part of the total carbon (0.02 %). The lack of the carbonates significantly reduce the neutralization of the dump material and cause its high acidity, as described by Andráš et al. (2021) and others. 

The GIS visualization of the contamination (Figs. 6 –13) indicates, that the drainage water results first of all, the release of Ni and Zn to the environ-ment, while the rest of the PTEs is less mobile and partly adsorbed by dump-material.

On the other hand, until the drainage water is rich in the inhabited area, its pH increase to value 6.91 and the contents of the PTEs are in comparison with the contents near the dump field generally ten times lower (Fe 5.1, Cu 1057, Pb 1,91, Zn 2105, As 7, Sb 3.51, Ni 44.78, Co 16.06, Cd 0.17 mg·L-1; Buccheri et al., 2014) and the environmental risk of such a water is relatively limited.

5. CONCLUSION

The main contaminants at the investigated dump-field are the following PTEs: Fe, Cu, Mn, Ni, Zn, Co, Pb and Sr. Only the Cu and Ni exceed substantially limits for soil, given by the Italian Legislativo Decree No. 152/06 (Italiana, 2006). The soil reaction is acid and the technogenous sediment/technosol is intensely oxidized and poorly aerated.

As sulphidic sulphur prevails in the technosol and carbonates absent, continuation of AMD/ARD formation is expected. The contrast between PTEs release and retention mechanisms underlines the high sensitivity of the system to changes in geochemical conditions. The relationship between increased sulfide content, the oxidation of which causes acidity, and low pH values ​​has been confirmed. Higher acidity causes more significant Cu migration than Ni.

 On the other hand, the negative charge of the soil solutions enables the immobilization of part of the metals released during the weat-hering process. Based on the obtained results, it is evident which parts of the dump field are most critical from the viewpoint of the individual PTEs' contamination.

The described environmental hazards corres-pond to the high PTEs contents (mainly of Cu and Ni) to the high toxicity of As3+ as well as to the weathering effectiveness and by the potential of the dump matter acidity production that could potential-ly intensify the release of the PTEs to the country elements. The risk is limited by the relatively great distance of the residential area (village Libiola).

These findings support the necessity of long-term monitoring and environmental risk assessment to minimize negative impacts on surface waters and soils in the Gromolo creek basin.

Acknowledgments

This work was supported by the VEGA under Grant number 1/0220/23. The autors thanks also to Dr. Stanislava Millovská and Dr. Adrian Biroň for the Energy-dispersive micro X-ray fluorescence spectro-metry (ED-XRF) analysis and to Dr. Jozef Krnáč for technical works.

REFERENCES

• Aiello, G., René, M., Ambrosio, G.A. & Bahariya, G.E.l. 2021. Geochemistry. London ISBN 978-1-83962-850-4, 314 p.
• Andráš, P., Midula, P., Matos, J.X., Buccheri, G., Drímal, M., Dirner, V., Melichová, Z. & Turisová, I., 2021.Comparison of soil contamination at the selected European copper mines. Carpathian Journal of Earth and Environmental Sciences, 16, 163–174. 
• Benvenuti, M., Mascaro, I., Corsini, F. Lattanzi, P. Parrini, P. & Tanelli, G., 1997. Mine waste dumps and heavy metal pollution in abandoned mining district of Boccheggiano (Southern Tuscany, Italy). Environmental Geology, 30, 238–243. 
• Benvenuti, M., Mascaro, I., Corsini, F., Ferrari, M., Lattanzi, P., Parrini, P., Costagliola, P. & Taneli, G., 2000. Environmental mineralogy and geochemistry of waste dumps at the Pb(Zn)-Ag Bottino mine, Apuane Alps, Italy. European Journal of Mineralogy, 12, 2, 441-453.
• Bertolani, M., 1952. I giacimenti cupriferi nelle ofioliti di Sestri Levante. Periodico di Mineralogia, 21, 149–170.
• Buccheri, G., Andráš, P., Astolfi, M.L., Canepari, S., Ciucci, M. & Marino, A., 2014. Heavy metal contamination in water at Libiola abandoned copper mine, Italy. Romanian Journal of Mineral Deposits, 87, 1, ISSN 1220-5648, 65-70.
• Buccheri, G., Andráš, P., Vajda, E., Midula, P. & Dirner, V., 2019. Soil contamination by heavy metals at Libiola abandoned copper mine, Italy. Acta Montanistica Slovaca, 23, 1, ISSN 1335-1788, 337-345.
• Buccheri, G. & Andráš, P., 2022. Inquinamento dell’area mineraria di Ľubietová (Slovacchia) e valutazione delle possibilità di bonifica Pollution at Ĺubietová mining area (Slovakia) and evaluation of remediation possibilities. Geologia dell Ambiente, Suplemento al. N. 4, ISSN 1591-5352, 128-130.
• Buscaroli, A., Zannoni, D., Menichetti, M. & Dinelli, E., 2017. Assessment of metal accumulation capacity of Dittrichia viscosa (L.) Greuter in two different Italian mine areas for contaminated soils remediation, Journal of Geochemical Exploration, 182, 123–131.
• Di Lisi, S. Di, Morino, A. & Muzio, C., 2020. Oro sui chiodi, Un classico della mineralogia ligure della miniera di Libiola (GE). Rivista Mineralogica Italiana, 44, 3 (3-2020) 186-189.
• Dinelli, E., Morandi, N. & Tateo, F. 1998. Fine-grained weathering products in waste disposal from two sulphide mines in the northern Apennines, Italy, Clay Minerals, 33, 423–433. 
• Dinelli, E., Lucchini, F., Fabbri, M. & Cortec, G., 2001. Metal distribution and environmental problems related to sulfide oxidation in the Libiola copper mine area (Ligurian Apennines, Italy). Journal of Geochemical Exploration, 74, 1-3, 141-152.
• Fairthorne, G., Fornasiero, D. & Ralston, J., 1997. Effect of oxidation on the collectorless flotation of chalcopyrite. International Journal of Mineral Processing, 49, 1, 31–48. 
• Ferrario, A. & Garuti, G., 1980. Copper deposits in the basal breccias and volcano-sedimentary sequences of the eastern ligurian ophiolites (Italy). Mineralium Deposita, 15, 3, 291–303. 
• Galli, M., 1996. Le miniere di rame e di manganese della Liguria Orientale. Atti dell’Accademia Ligure di Scienze e Lettere, 53, 215–247.
• Garuti, G., Bartoli, O., Scacchetti, M. & Zaccarini, F., 2008. Geological setting and structural styles of Volcanic Massive Sulfide deposits in the northern Apennines (Italy): Evidence for seafloor and sub-seafloor hydrothermal activity in unconventional ophiolites of the Mesozoic Tethys. Boletin de la Sociedad Geologica Mexicana, 60, 1, ISSN 1405-3322, 121–145. 
• Italiana, 2006. Decreto Legislativo 3 aprile 2006, n. 152, Gazzetta Ufficiale, 1(c), p. 172.
• Kienlin, T.L. & Pearce, M., 2008. Bright Blades and Red Metal. Essays on North Italian Prehistoric Metalwork. European Journal of Archaeology, 2017/01/25, 11(2–3), 288–290.
• Krachler M., Mohl, C., Emons, H. & Shotyk, W., 2002. Influence of digestion procedures on the determination of rare earth elements in peat and plant samples by USN-ICP-MS. Journal of Analytical Atomic Spectrometry, 17, 8, DOI: 10.1039/b200780k, 844-851.
• Lu, L., Rucheng, W., Fanrong, Ch., Jiue, X. & Peihua, Z., 2005. Element mobility during pyrite weathering: Implications for acid and heavy metal pollution at mining-impacted sites, Environmental Geology, 49, 1, 82–89.
• Maggi, R. & Pearce, M., 2005. Mid fourth-millennium copper mining in Liguria, north-west Italy: The earliest known copper mines in Western Europe’, Antiquity, 79.
• Marescotti, P., Carbone ÆC., De Capitani ÆL., Grieco, G. Lucchetti, ÆG. & Servida, ÆD., 2007. Mineralogical and geochemical characterisation of open-air tailing and waste-rock dumps from the Libiola Fe-Cu sulphide mine (Eastern Liguria, Italy). Environmental Geology, DOI 10.1007/s00254-007-0769-8, 1-14.
• Marescotti, P, Bullet, E., Bullet, A., Servida, ÆD., Carbone, ÆC., Grieco, G., De, L., Bullet, C. & Lucchetti, G., 2012. Mineralogical and geochemical spatial analyses of a waste-rock dump at the Libiola Fe–Cu sulphide mine (Eastern Liguria, Italy). Environmental earth sciences, 187–199.
• Marescotti, P, Solimano, M., Marin, V., Salmona, P., Brancucci, M. & Lucchetti, G., 2015. Mineralogical and geochemical spatial analysis of waste-rock dumps°a case study from the abandoned Rio Bansigo sulphide mine (eastern Liguria, Italy). Rendiconti Online della Società Geologica Italiana, 35, Supplemento, 2, ISSN 2035-8008.
• Marini, L., Saldi, G., Cipolli, F., Ottonello, G., Zuccolini, M. V., 2003. Geochemistry of water discharges from the Libiola mine, Italy. Geochemical Journal, 37, 199–216.
• McNeill, J.D., 1992. Rapid, Accurate Mapping of Soil Salinity by Electromagnetic Ground Conductivity Meters. Advances in Measurement of Soil Physical Properties: Bringing Theory into Practice. (SSSA Special Publications), 209–229.
• Osseo-Asare, K., Xue, T. & Ciminelli, V.S.T., 1984. Solution chemistry of cyanide leaching systems. Metallurgical Society of AIME, 173–197.
• Polański, A. & Smulikowski, K., 1978. Geochemistry. Slovenské pedagogické nakladateľstvo. Bratislava, 607 p. (in Slovak).
• Richter, R. & Hlušek, J., 2003. Soil floor. 2nd edn. Praha: Ústav zemědelských a potravinářských informací, Ministerstvo zemědelství, (in Czech).
• Ryu, J., Gao, S., Dahlgren, R.A. & Ziernberg, R.A., 2002. Arsenic distribution, speciation and solubility in shallow groundwater of Owens Dry Lake, California. Geochimica et Cosmochimica Acta, 66, 17, 2981-2994.
• Singovszka, E., Balintova, M. & Holub, M., 2015. Heavy metal contamination and its indexing approach for sediment in Smolnik creek (Slovakia). Clean Technologies and Environmental Policy, 18. 
• Sobek, A.A., Schuller, W.A., Freeman, J.R. & Smith, R. M. 1978. Field and laboratory methods applicable to overburden and minesoils. EPA 600/2-78-054, 203 p.
• Zotti, M., Simone, Di P., Roccotiello, E., Lucchetti, G., Mariotti, M.G., Marescotti, P., 2014. Microfungi in highly copper-contaminated soils from an abandoned Fe-Cu sulphide mine: Growth responses, tolerance and bioaccumulation, Chemosphere, 117C, 471–476.