土壤中重金属的分布（西班牙）

Journal of Hazardous Materials 162 (2009) 854–859 Contents lists available at ScienceDirect Journal of Hazardous Materials journa l homepage: www.e lsev ier .com Heavy an Madrid flor Eduardo nza Roberto a Departament b Departament a r t i c l Article history: Received 19 D Received in revised form 6 March 2008 Accepted 22 May 2008 Available online 28 May 2008 Keywords: Heavy metals Mining pollution Soil–plant tran Plant screenin Phytoavailabil ution their from a mining valley in NW Madrid (Spain), and total and extractable heavy metals were analysed. Soils affected by mining activities presented total Cd, Cu and Zn concentrations above toxic thresholds. The percentage of extractable element was highest for Cd and lowest for Cu. A highly significant correlation was observed between the total and extractable concentrations of metals in soils, indicating that, among the factors studied, totalmetals concentration is themost relevant for heavymetals extractability in these soils. (NH4)2SO4-extractablemetal concentrations in soils are correlated betterwithmetal concentrations 1. Introdu Thepote heavymeta [1,2]. Sever tion in soils type of freq of metallic condition th uid soil pha toxicity of e als availabil described a the correlat ity, especia induce toxi ∗ Correspon E-mail add 0304-3894/$ – doi:10.1016/j.j sfer g ity in several plant species than total metals in soils, and thus can be used as a suitable and robust method for the estimation of the phytoavailable fraction present in soils. Twenty-five vascular plant species (3 ferns and 22 flowering plants) were analysed, in order to identify exceptional characteristics that would be interesting for soil phytoremediation and/or reclamation. High Cd and Zn concentrations have been found in the aerial parts of Hypericum perforatum (Cd), Salix atrocinerea (Cd, Zn) and Digitalis thapsi (Cd, Zn). The present paper is, to the best of our knowledge, the first report of the metal accumulation ability of the two latter plant species. The phytoremediation ability of S. atrocinerea for Cd and Znwas estimated, obtaining intervals of time that could be considered suitable for the phytoextraction of polluted soils. © 2008 Elsevier B.V. All rights reserved. ction ntial risk for theenvironment andpopulationdue to soil ls arising frommetallic mining has beenwell described al studies have reported a high degree of metal pollu- affected by the oxidation of pyritic materials [3–5], a uent waste in metallic mines. After the accumulation elements in the soil, several physico-chemical factors e transfer of each heavymetal from the solid to the liq- se, causing differences in the availability and, finally, the lements suchasCd, Cu,Mnor Zn. The estimationofmet- ity for plants using unbuffered salt solutions has been s suitable [6], but less information is available about ion between extractable element and phytoavailabil- lly when wild plants are involved. Heavy metals can city in wildlife if the soil level reaches critical values; ding author. Tel.: +34 914974824; fax: +34 914973826. ress: elvira.esteban@uam.es (E. Esteban). also, plant accumulation in above-ground tissues can result in an increase of metal accumulation in top-soil, via leaf deposition, or can create an exposure pathway for metal introduction into the food chain [7,8]. On the other hand, plants living in metalliferous soils can have exceptional properties which make them interest- ing for phytoremediation [9]. Thus, many authors have reported plant screening in sites enriched with metals and have identified interesting plants for further studies [8,10,11], aiming to find self- sustainable plants that could clean polluted environments [12]. Moreover, other studies focus on the possibilities of revegetation and phytostabilisation of mine tailings with Iberian tolerant plants [13–15]. In our study site, theMo´nica pyritemine, mining activities were carried out from 1427 until 1980 [16], and a group of galleries and pyritic dumps remain, close to the village of Bustarviejo, across the La Mina stream gorge (Sierra de Guadarrama, NW Madrid, Spain). Many plants species inhabit in this valley and some of them could show properties useful in phytotechnolgies. These plants present the advantage of being adapted to the edaphoclimatic conditions of this area. see front matter © 2008 Elsevier B.V. All rights reserved. hazmat.2008.05.109 metals distribution in soils surrounding (Spain) and their transference to wild Moreno-Jime´neza, Jesu´s M. Pen˜alosaa, Rebeca Ma Gamarrab, Elvira Estebana,∗ o de Qu´ımica Agr´ıcola, Universidad Auto´noma de Madrid, E-28049 Madrid, Spain o de Biolog´ıa, Universidad Auto´noma de Madrid, E-28049 Madrid, Spain e i n f o ecember 2007 a b s t r a c t The present work concerns the distrib surrounding soils of a mine site and / locate / jhazmat abandoned mine in NW a noa, Ramo´n O. Carpena-Ruiza, and mobility of heavy metals (Fe, Mn, Cu, Zn and Cd) in the transfer to wild flora. Thus, soils and plants were sampled E. Moreno-Jime´nez et al. / Journal of Hazardous Materials 162 (2009) 854–859 855 Table 1 pH, dichromate-oxidable organic matter and metal concentrations in soils surrounding Mo´nica mine (Bustarviejo, Spain) Soils close to mining dumps Soils affected by mine drainage Unaffected soils an pH 0 % OM 1 Total Fe (%) 3 Total metal co Cd 5 Cu Mn Zn Extractable me Fe 5 Cd 8 Cu 5 Mn Zn Relative extrac Fe 15 Cd Cu Mn Zn Mean, standar Our stud the vicinity als for the p transfer to phytoremed 2. Materia 2.1. Site des Soils and mine, close studied site valley, betw Y=4524302 between M were collec (top 0–30 c La Barranca referencedb dividing soi soils affecte The min zones, a d and Cytisus oromediter pyrenaica W naicae); alo atrocinerea Roth and Fr 2.2. Analyti Soils we homogenise pH of a 1:2. protocols o concentrati an au tract the face w ded t soo tma ed at f mi dry a and uted met d pla ry (P Mean (S.E.) Median Range Mean (S.E.) Medi 4.87 (0.25) 4.17 3.89–5.99 5.27 (0.13) 5.2 4.33 (1.00) 3.19 0.69–8.61 5.91 (0.94) 5.4 2.02 (0.01) 1.86 1.32–3.38 2.16 (0.02) 2.0 ncentration (mgkg−1) 16.35 (3.15) 13.72 1.78–34.98 8.68 (1.37) 9.7 308.7 (56.4) 301.7 17.3–605.3 182.2 (38.9) 186.7 353.6 (47.8) 298.9 184.7–658.4 432.6 (62.2) 405.3 845.7 (186.2) 604.2 92.4–2243.6 571.1 (111.4) 566.4 tal concentration (mgkg−1) 4.49 (0.99) 3.57 1.07–12.48 3.64 (0.29) 3.3 2.01 (0.72) 1.32 0.06–7.23 1.36 (0.26) 1.3 2.68 (0.59) 2.21 0.46–6.32 2.08 (0.58) 1.3 18.2 (3.9) 17.2 1.4–46.6 34.3 (7.1) 34.2 51.0 (15.9) 42.8 1.24–149.4 38.6 (8.5) 36.5 table metal concentration (%) 0.023 (0.004) 0.017 0.005–0.049 0.020 (0.003) 0.0 9.7 (2.27) 7.9 3.7–25.1 14.9 (1.41) 14.2 1.4 (0.25) 1.5 0.4–2.7 1.3 (0.21) 0.9 6.2 (1.42) 6.0 0.8–15.6 7.2 (0.89) 6.8 6.5 (1.65) 3.9 1.3–14.6 6.4 (0.80) 6.9 d error (S.E.), median and range (n=12–16). y focuses on the dispersion of heavy metals in soils in of the Mo´nica mine and the availability of these met- lant community at this site. Metal accumulation and natural flora was also studied, in order to evaluate the iation ability of these plant species. ls and methods cription plantswere sampled in the surroundings of theMo´nica to the village of Bustarviejo (Madrid) (Fig. S1). The extends across 200000m2 within the La Mina stream een the following UTM coordinates: 30T—X=0438606, ; X=0437797, Y=4523518. Sampling was carried out tion in was ex for 4h; Sur was ad ysed as Plan anddri 10mL o to 0.5 g 1500P and dil The soil an tromet ay and June 2006. Shoots of 25 vascular plant species ted (Table S1), as well as the soil below the plants m soil layer). Moreover, water from the La Mina and streams was sampled. All sampling points were geo- yGPS. Fig. S1 shows the samplingpointswithin the site, ls into three groups: (1) soils close tomining dumps, (2) d bymine drainage and (3) potentially unaffected soils. ing area shows three types of vegetation: in the upper ense scrub dominated by Genista cinerascens Lange oromediterraneus Rivas Mart. et al. (Genisto-Cytisetum ranei); in the lower zones, an open forest of Quercus illd. with a grass pasture (Luzulo-Quercetum pyre- ng the stream, a riparian community (Rubo-Salicetum e) with Salix atrocinerea Brot., Athyrium filix-femina (L.) angula alnus Miller. cal procedures re dried at 50 ◦C for 7 days, sieved to 2mm and d. Dichromate-oxidable organic matter (OM) and the 5 (soil:water) suspension were measured following the f the Spanish Ministry of Agriculture [17]. Pseudo-total ons of elements were assayed after HNO3:H2O2 diges- were measu 2.3. Statisti The data dows. Stati using the no regression, nents analy data. 3. Results 3.1. Metal d Both pH in the same northweste fected soils and no sign Levels of affected by fected soils Range Mean (S.E.) Median Range 4.24–6.09 5.08 (0.11) 5.16 4.01–5.90 1.83–13.04 6.28 (0.70) 6.62 2.11–11.66 1.01–3.45 1.36 (0.01) 1.42 0.94–1.65 2.44–15.29 2.91 (0.50) 2.89 n.d.–6.15 15.5–444.3 16.6 (2.3) 13.3 6.2–32.5 158.9–1002.0 427.8 (49.7) 517.3 76.9–647.8 75.2–1437.5 96.7 (14.0) 92.9 30.8–200.3 2.22–5.52 2.67 (0.21) 2.67 2.15–4.73 0.10–2.66 0.19 (0.06) 0.11 n.d.–1.00 0.30–4.30 0.34 (0.04) 0.33 0.05–0.71 3.7–73.7 48.7 (15.8) 28.1 3.21–255.9 4.95–84.5 1.9 (0.7) 0.8 0.1–13.1 0.009–0.036 0.022 (0.002) 0.022 0.013–0.040 3.6–22.4 6.0 (1.33) 4.0 0–17.9 0.7–3.1 2.3 (0.29) 2.0 0.3–4.3 2.3–10.5 10.5 (2.68) 7.1 1.5–21.8 1.9–12.3 2.9 (0.54) 3.2 0.4–8.1 toclave [18]. The extractablemetals content of the soils ed by shaking 2g of soil with 20mL of 0.1M (NH4)2SO4 suspension was filtered and the filtrate analysed [19]. aters were sampled in plastic flasks and 1mL of HNO3 o 40mL of water. Samples were stored at 4 ◦C and anal- n as possible. terial waswashed thoroughly in tap and distilledwater 50 ◦C for 7days. For acidmineralisation of plant tissues, li-Q water, 3mL of HNO3 and 2mL of H2O2 were added weight (DW) of tissue and digestion was performed at 125 ◦C [20], in an autoclave. The extract was filtered to 25mL. als (Cd, Cu, Fe, Mn and Zn) in the surface water and the nt extracts were analysed by atomic absorption spec- erkinElmer AAnalyst 800). Three analytical replicates red for each sample. cal analysis were analysed statistically using SPSS 14.0® for Win- stical differences among soil groups were analysed n-parametric Kruskal–Wallis or Wilcoxon tests. Linear simple and bivariate correlations and principal compo- sis (PCA) were performed for the soil and plant analysis and discussion istribution and mobility in soils and dichromate-oxidable organicmatter (Table 1)were range as values reported as normal in soils from the rn mountains of Madrid [21]. Both affected and unaf- were acid and contained high levels of organic matter, ificant differences were observed among soil groups. Cd, Cu and Zn in soils close to mining dumps and soils mine drainageweremuch higher than those fromunaf- (P<0.001), but Mn levels were in the same range in the 856 E. Moreno-Jime´nez et al. / Journal of Hazardous Materials 162 (2009) 854–859 Table 2 Results of the principal component analysis (factor loadings) of the total metal concentration and other properties in soils, and of the metals in the shoots of all plant species Soils Plants Comp. 1 (47%) Comp. 2 (27%) Comp. 1 (34%) Comp. 2 (31%) Comp. 3 (17%) Cd 0.88 0.89 Zn 0.95 0.84 Cu 0.91 0.76 Mn 0.77 0.82 Fe 0.70 0.07 0.77 OM 0.84 – – – pH 0.62 – – – Factor loadings smaller than 0.5 were omitted. entire valley (Table 1). The Cd, Cu and Zn levels in unaffected soils and the Mn in all soils were quite similar to the geochemical back- ground of the soils from this orophilous region ofMadrid (Table S2) [21]. More than 80% of the soils close to themining dumps and 50% of the soils affected by mine drainage exceeded the toxic concen- tration in soils (Table S2) [3,22,23] for at least onemetal (Zn, Cu and Cd), posing an important environmental risk. (NH4)2SO4-extractableCd, CuandZnconcentrationsweremuch higher in affected soils than in unaffected ones (P<0.001), while the extractable Mn and Fe concentrations were similar in all soils (Table 1). Although the total metal concentrations are usually used as the primary pollution reference in legislation, the occurrence of toxic elements in soils, especially in disused mining areas, needs further analysis [24]. The extractable heavy metals show a better correlation with the metal concentration in plant tissues [6], so it appears to be a better indicator of the environmental risk caused by the pres The perc metals, sho extractable low extract The relative 6.5 and 4% as normal than Cu [2 to those for other soils close to pyritic mines [3,27]. The rela- tive abundance of total heavy metals at the polluted sites was Fe>Zn≈Mn>Cu�Cd, while for the extractable metal concentra- tions it was Zn≈Mn� Fe >Cu≈Cd. The degree of mobility of the metals, given as the percentage of extractable metal with respect to the total, was: Cd≈Mn�Zn>Cu� Fe, showing the higher envi- ronmental risk posed by Cd, despite its lower total concentration in the soil. After a bivariate correlation analysis, extractable and total Cd, Cu and Zn were strongly associated (Pearson’s coefficients from 0.65 to 0.99; P<0.001), indicating their simultaneous presence in the mine wastes. Total Fe was moderately correlated with total Cu and Zn (P<0.01). Total and available Mn were not correlated with the other heavy metals, but showed moderate correlation with organic matter (P<0.01) and low correlation with pH (P<0.05). Metal availability in soils is generally controlled by total metal trati etal ion e m first arian nd F ct, to supp Table 3 Metal concent rid, Sp Plant species g g−1 Seedless vascu Equisetum ra (7.8– Pteridium aq (5.5–2 Athyrium fili (10.3– Annual and pe Centaurea ni (9.9–1 Hypericum p (5.21 Digitalis thap (3.3–7 Aira caryoph (2.68 Glyceria fluit (9.2–3 Diplotaxis er (10.8 Daucus carot (4.45 Silene latifoli (4.36 Woody plants Cytisus scopa Cytisus orom Genista ciner Adenocarpus Thymus mas Santolina ros Frangula aln Betula pendu Erica arborea Salix atrocin Means (range) ence of metals in soils [2]. entages of extractable metals, in relation to total soil w that Cd and Mn were significantly more easily than the other metals (P<0.001), while Cu showed ability and Fe was strongly retained in soils (Table 1). proportions of extractable Cd, Cu, Mn and Zn (10, 1.5, , respectively in average) were in the range reported [25], Cd and Zn usually being more mobile in soils 6]. Total Fe concentrations in the soils are similar concen total m regress from th The total v (0.91) a This fa ments, rations in shoots of plants growing in the areas surrounding the Mo´nica mine (Mad Cd (�gg−1 shoot DW) Zn (�gg−1 shoot DW) Cu (� lar plants mosissimum 1.53 (1.24–1.83) 172.9 (172.7–173.1) 8.22 uilinum 0.81 (n.d.–1.74) 91.3 (9.5–191.1) 11.2 x-femina 1.83 (1.13–2.34) 204.9 (126–247) 12.0 rennial herbs gra 1.28 (0.24–2.92) 150.6 (19.0–277.9) 13.4 erforatum 10.22 (6.29–20.32) 59.18 (24.09–102.73) 16.11 si 13.28 (0.85–22.04) 245.3 (107–351) 23.5 yllea 2.16 (1.64–2.37) 65.8 (62.7–68.8) 2.77 ans 5.64 (1.16–11.02) 364.9 (77.4–663.0) 18.6 ucoides 14.51 (1.90–19.04) 581.1 (101.3–1048.0) 15.24 a 4.06 (3.01–5.11) 91.2 (79.6–102.9) 7.11 a 6.91 (0.48–13.59) 228.1 (21.3–440.0) 6.42 rius 0.87 (n.d.–2.48) 200.1 (15.2–902.9) 19.1 (4.9–5 editerraneus 0.98 (n.d.–3.90) 158.9 (13.1–509.1) 4.59 (3.6– ascens 1.07 (n.d.–2.74) 108.5 (26.1–425.8) 6.16 (4.52 complicatus 2.86 (1.56–3.98) 299.3 (246–358) 4.55 (2.52 tichina 1.82 (0.35–3.70) 30.3 (8.3–58.3) 9.9 (7.3–1 marinifolia 19.43 (18.5–20.4) 349.8 (345.9–353.6) 12.89 (12.7 us 0.56 (0.31–0.72) 129.8 (30.5–235.6) 7.81 (6.25 la 4.75 (3.23–6.27) 708.5 (629–787) 6.17 (6.03 0.20 (n.d.–0.37) 35.2 (20.4–70.7) 7.92 (4.65 erea 33.26 (13.1–53.1) 667.7 (310–936) 7.13 (4.71 , n=2–12. on, pH and organic matter [2], but in our study only significantly explained metal extractability after linear analysis, so that organic matter and pH were omitted odel. PCA component for soils (Table 2) explained 47% of the ce, and the factor loadings of Cd (0.88), Zn (0.95), Cu e (0.70) showed the highest values for this component. gether with the high correlation among all these ele- orts the hypothesis that metal pollution is mainly due ain) shoot DW) Mn (�gg−1 shoot DW) Fe (�gg−1 shoot DW) 8.6) 49.2 (47.5–50.9) 55.0 (53.4–56.6) 2.9) 85.3 (23.8–196.1) 100.0 (119.2–232.4) 13.4) 19.8 (14.9–22.2) 117.2 (76.5–141.8) 7.0) 86.9 (42.6–108.5) 89.9 (72.7–109.6) –26.50) 230.2 (134.4–344.5) 108.11 (92.8–119.2) 0.2) 182.7 (83.4–362.4) 422.1 (107.1–831.6) –2.86) 163.9 (138.2–189.5) 140.9 (120.7–161.2) 5.2) 55.5 (22.3–69.2) 200.1 (46.6–448.9) 7–22.98) 46.5 (30.1–60.1) 367.8 (256.4–540.1) –9.77) 66.7 (49.1–84.3) 39.5 (39.4–39.7) –9.66) 95.0 (27.0–155.9) 160.8 (52.6–399.8) 7.6) 104.7 (26.5–215.6) 57.5 (28.5–140.1) 5.7) 85.0 (48.2–201.4) 94.9 (24.5–226.6) –8.69) 90.8 (15.8–212.0) 85.8 (39.6–128.9) –6.03) 86.9 (43.9–170.1) 98.9 (88.2–136.8) 2.3) 77.3 (52.3–109.8) 344.4 (229.2–430.0) –13.0) 44.6 (39.5–49.7) 50.7 (49.6–51.7) –9.68) 199.6 (104–171) 62.4 (57.2–69.2) –6.31) 102.4 (83.6–121.3) 35.5 (34.5–36.5) –10.76) 125.7 (60.9–306.9) 108.7 (40.9–223.2) –11.86) 163.2 (61.8–400.6) 53.9 (21.0–95.9) E. Moreno-Jime´nez et al. / Journal of Hazardous Materials 162 (2009) 854–859 857 Table 4 Correlations between metal concentrations in plants and soils (total and available) for the dominant plant species (n=6–12; *P<0.05; **P<0.01; ***P<0.001) Total Available Cd Zn Cu Mn S. atrocinerea 0.75* 0.88** 0.31 0.22 Cytisus scopari 0.75** 0.59* 0.56* 0.15 G. cinerascens 0.83** 0.99*** 0.82* 0.49 E. arborea 0.45 0.71* 0.40 0.94*** P. aquilinum 0.68* 0.45 0.28 0.38 to dispersio nent (27%), and pH (0.6 The proxim enrich the s particles vi mobilised b Stream both stream across the f 0.196mgZn 0.125–0.487 higher than 3.2. Metal c phytoremed Three se them mono Manyof the centrations [3,22,23], b hyperaccum woody plan S. atrocinere concentrati (30%higher ment with p high levels tissue levels lar effectsw have stood certain met perforatum [3]. Thus, m lators, altho ecotypes an In order metal conc in the shoot tive correla their extrac tration, ind extractable data, fromv mentwith t grown unde soils, corrob and robustm of metals pr To evalu transfer fac tration of e surroundin d (4.3) andZn (1.5), butnot forCu.H. perforatum andD. thapsi gher transfer factors than the other herbaceous species for etals, although Silene latifolia also showed a high TF for Zn . All the seedless vascular plants showed low transfer ofmet- he aerial parts. Moreover, Erica arborea and F. alnus showed ransfer factors (except for Mn) than the other woody plants, entaurea nigra and Aira caryophyllea had lower Cd and Zn r factors than the other herbaceous species. ough the shootmetal concentrations clearly differed among , the data were studied by a factorial analysis. The first PCA nent (Table 2) explained 34% of the variance, and the factor s of Cd (0.89) and Zn (0.84) showed the highest values. Both ts showed a high availability in soils after mine pollution 1) and the highest differences in metal transfer to plants the species (Fig. 1), i.e. between S. atrocinerea (4.3 for Cd for Zn) and E. arborea (0.03 for Cd and 0.09 for Zn). Both ele- Cd a , and the 0.76) re pr were n. T rpre , wh oadi dep e of erial Cd Zn Cu Mn 0.68* 0.81* 0.10 0.06 us 0.70* 0.36 0.24 0.02 0.70 0.76* 0.82* 0.32 0.17 0.44 0.06 0.56 0.51 0.11 0.15 0.50 n of mining wastes in these soils. The second compo- associatedmainlywithMn (0.77), organicmatter (0.84) 2), was not correlated directly with mining pollution. ity to mining dumps or to mine drainage seemed to oils in metals, probably due to an enrichment in pyritic a transfer from the dumps and/or retention of metals y waters. water was also sampled along the site (Fig. S1). In s (La Mina and La Barranca), heavy metals ranged ollowing concentrations: n.d.–0.024mgCdL−1; 0.078– L−1; 0.073