ent 378 (2007) 114–118www.elsevier.com/locate/scitotenv

Science of the Total Environm

Effects of two chelating agents (EDTA and DTPA) on theautochthonous vegetation of a soil polluted with Cu, Zn and Cd

Jesús Pastor a,⁎, Alberto Martín Aparicio b,Antonio Gutierrez-Maroto c, Ana Jesús Hernández b

a IRN, Centro de Ciencias Medioambientales, CSIC, Department of Systems Ecology, c/ Serrano 115 dpld., 28006 Madrid, Spainb Dpto. Ecología, Universidad de Alcalá, Madrid, Spain

c Dto. Química Agrícola, Universidad Autonoma, Madrid, Spain

Available online 20 February 2007

Abstract

Chelating agents are thought to be useful for the remediation of metal-polluted soils with adequate organic matter, but theeffects of these substances on the plants or seed bank of plant communities that occur in the soils are relatively unknown. In thiswork, the effects of two chelating substances (EDTA and DTPA) on a wet grassland plant community affected by the presence ofthe abandoned copper mine “Fernandito” (Garganta de los Montes, Madrid) were compared. A microcosm bioassay (6 months)was designed using the soil's top layer containing the grassland's seed bank. This soil showed a high Cu pollution level, significantcontents of Zn and Cd (1120, 190, and 15 ppm, respectively), a pH of 5 and an OM content of 6.2%. The soil was subjected to threedifferent treatments: a) untreated soil (control), b) the addition of 1 g/kg EDTA, or of c) 1 g/kg DTPA. The results presented hereare those related to the plant cover, species richness, aboveground and subterranean biomass and chemical composition of the mostabundant plants. Neither EDTA nor DTPA caused intense negative effects on the plants rather they significantly increased theamount of copper accumulated in aboveground parts and roots. In particular, Agrostis castellana and Corrigiola telephiifoliaextracted high amounts of copper when grown in the soil with added EDTA, although they showed some nutritional imbalances(lower P contents). In contrast, lower metal concentrations were detected in plants grown in the DTPA amended soil.© 2007 Elsevier B.V. All rights reserved.

Keywords: Phytoremediation; Heavy metals; Bioaccumulation; Plant cover; Subterranean and aboveground biomass

1. Introduction

Heavy metal phytoremediation techniques are espe-cially useful in cases of widespread pollution that arelimited to the surface layer of the soil, like those found atabandoned mine sites. However, these techniquesrequire much further research. For instance, chelatingagents are frequently used to increase the bioavailability

⁎ Corresponding author. Tel.: +34 917452500; fax: +34 915640800.E-mail address: jpastor@ccma.csic.es (J. Pastor).

0048-9697/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.scitotenv.2007.01.022

of heavy metals, thus enhancing their uptake by plants,although this may also decrease their biomass becauseof the toxicity of this extra uptake (Madrid et al., 2003).These effects need to be further investigated especiallyin Mediterranean environments where phytoremediationtechniques have been little used, and there is scarceknowledge about the potential usefulness of autochtho-nous flora for this purpose.

Our study was designed to explore the use of severalnative plant species along with two chelating agents forphytoremediation of the soil of an abandoned copper

Table 1aChemical characteristics of the soil used in the bioassay

Macroelements (ppm)

Ca K Na Mg

Total 1965 7800 360 5150Available 1210 245 15 186

Metals (ppm)

Al Fe Mn Zn Cu Cd Ni Pb

Total 28407 16100 515 190.1 1120 4.5 15.1 47.2Available 0.42 7.2 9.5 5.7 71.4 0.56 0.13 1.1

Nitrogen, organic matter and pH

N (%) O. M. (%) pH

0.171 6.22 5.01

115J. Pastor et al. / Science of the Total Environment 378 (2007) 114–118

mine (“Fernandito”) in the Madrid province (centralSpain), which has been previously characterized byEncabo et al. (1997) and Pastor et al. (2003a).

2. Materials and methods

2.1. Study area

Situation, landscape, lithology, mineralogy, soil andvegetation of the site affected by the mine: The“Fernandito” copper mine, abandoned over 40 yearsago, is situated south of the village Garganta de losMontes. It is one of the many mines existing in theCentral System, north of Madrid exploited during theXIXth and XXth centuries. The main metals mined werePb, Cu, Fe, Zn and Ag. The legacy left behind by thesemines is a series of dispersed landfills that act as a fairlyunknown constant source of pollution.

Table 1bSpecies identified in the field from which the soil used in the experiments w

Agrostis castellana Crepis capillarisArmeria arenaria Daucus carota

Bellardia trixago Eryngeum campestre

Brassica sp. Hipochaeris radicataBromus maximus Holcus lanatusCerastium glacile Jasione montana

Cerastium glomeratum Juncus acutiflorusConvolvulus arvensis Juncus

conglomeratusCorrigiola telephiifolia Juncus tenageia

From a perspective of landscape, the Fernandito mineand its surroundings occupy a plain, which is slightlyinclined towards the Lozoya Valley. This promotes thedispersion of mine tailings (“fines”) through the actionsof wind and water that interrupt natural cycles.

The sites lithology is dominated by migmaticgneisses. Its main mineralogy is calcopirite in paragen-esis with small amounts of pyrite, blend and marcasite.There is some galena and quartz gangue, and scarceamounts of arsenopyrite. Metal minerals infill quartzfractures without replacing the quartz (IGME, 1991).

The soils of the mine area are cambisols and lepto-sols (Monturiol and Alcala del Olmo, 1990). Those ofthe study area lie at heights below 1210 m and have beenclassified as humic and distric cambisols by the FAO(Encabo et al., 1997). On these siliceous soils, poor inbases with a subhumid ombroclimate, the vegetation iscomposed of small oakwood formations includingQuercus pyrenaica and in the wetter zones, ash woods(Fraxinus angustifolia) appear on somewhat deepersoils, avoiding the more polluted sites. Here, thevegetation corresponds to different ecosystems of wetand temperate grasslands and tilled pastures mainly usedby grazing cattle and wild animals. Grazing promotesthe presence of a high number of plant species.

This mine has a landfill of fines from the Cuextraction process, whose volume is about 3500 m3. Inthe material deposited in the landfills and adjacent soils,Al, Ba, Ca, Cd, Cr, Cu, Fe, K, Mg, Mn, Ni, Pb and Znlevels were determined by X-ray fluorescence andatomic absorption. Among the heavy metals, we onlyobserved high Cu levels and at some points high Cdconcentrations, with contents of Zn worth mentioningdetected on occasion. In the fractions in which Cu couldbe considered theoretically more available, we notedthat samples obtained from the greatest heights widelysurpassed the threshold total soil copper content

as obtained

Molineriella laevis Trifolium campestreOrnithopuscompressus

Trifolium dubium

Petrorragia sp. Trifoliumglomeratum

Plantago lanceolata Trifolium pratenseRumex acetosella Trifolium repensRumexbucephalophorus

Trifolium striatum

Silene sp. Vulpia, sp.SpergulariapurpureaTrifolium arvense

Table 2Cover values (%) for all the species growing in the microcosms

Control 1 g EDTA 1 g DTPA

Agrostis castellana 18±5 22±19 5±9Anthemis sp. 0 0 1±2Corrigiola telephiifolia 31±17 43±31 54±20Molineriella laevis 2±2 3±3 3±3Spergularia purpurea 1±1 2±2 1±0Trifolium dubium 2±3 2±3 1±1

Table 4Dry weight biomass (g/dm2) for the microcosms at the end of thebioassay

Treatment Aboveground Subterranean Total A/S ratio

Control 1.91±0.26 a 1.34±0.27 a 3.26±0.27 a 1.47±0.38 a1 g EDTA 1.84±0.12 a 0.91±0.33 a 2.74±0.33 a 2.29±1.08 a1 g DTPA 1.87±0.40 a 0.78±0.27 a 2.66±0.40 a 2.59±0.95 a

116 J. Pastor et al. / Science of the Total Environment 378 (2007) 114–118

stipulated for soils of pH under 7 by Spanish legislation(BOE, 1990). These soils can thus be considered toxic.

2.2. Sampling

Prior to the experiment, we screened the soil's toplayer at 10 different points along an altitudinal gradient.Plant species were identified at each point, and thechemical characteristics of the soil samples (fertility andheavy metal content) were determined. The results ofour analysis of the soil finally used in the experiment areshown in Table 1a. The species present in this zone ofthe mine are listed in Table 1b.

The soil was placed in nine microcosm pots toprepare three replicates for each of the three treatments:unamended soil, soil plus 1 g/kg EDTA and soil plus1 g/kg DTPA (a dose equivalent to 4 Tm/ha). The potswere watered with deionized water to promote thegrowth of their autochthonous seed bank. Several plantand soil variables were measured in each pot. After30 weeks, the experiment was ended by harvesting theplants.

Species richness was determined by counting thenumber of different plant species in each pot. TheShannon–Wiener diversity index (H′) was calculatedfrom species richness. The advantage of this index usedto measure biodiversity is that it takes into account thenumber of species and the evenness, or numericalequality, of the species populations. The index isincreased either by having more unique species, or byhaving a greater species evenness. A higher value of H′means a higher level of diversity.

Soil and plant samples were prepared according tothe procedures described in Hernández and Pastor

Table 3Plant diversity and plant cover values recorded at the end of thebioassay

Treatment Plant cover (%) Sp. richness Shannon index

Control 16±10 a 4.33±0.58 a 1.18±0.36 a1 g EDTA 22±14 b1 4.33±0.58 a 0.93±0.37 a1 g DTPA 19±14 a 4,33±0.58 a 0.74±0.41 a

(1989). Plant samples (aboveground parts and roots)were washed in deionized water, dried in an oven for48 h at 70 °C and weighed. Plants were chemicallyanalyzed when there was enough biomass of a species ina microcosm. The subterranean biomass of each pot wasanalyzed without separating the roots of differentspecies. These plants were ground, digested in HNO3

and HClO4 (3:1) and analyzed by ICP spectroscopy.Prior to analysis, they were subjected to intensivewashing, rinsing the roots with water under pressurefollowed by successive submerging and washing byhand in a series of recipients containing deionized water.This procedure was repeated until no soil particlesappeared in the washing recipients.

Soil cover data were analyzed by Kruskal–WallisANOVA, comparing means using the Mann–WhitneyU-test. The rest of the data were Box Cox transformedfor one-way ANOVA, with means compared by Fisher'sLSD method. However, the data provided in the tablesare untransformed. In all the tables, different lettersascribed to values in a column indicate significantdifferences. When followed by 1, the significance levelwas pb0.1; otherwise it was pb0.05. All statistical testswere performed using the program STATISTICA 6.0.

3. Results and discussion

Table 2 shows the mean plant cover values recorded.The microcosms prepared with amended soil showedhigher cover values than controls, with higher valuesobserved for the EDTA-treated soil than the DTPA-treated soil. This may be explained by the fertilizingaction of these chelators, which can mobilize soilnutrients and thus increase their bioavailability to theplants (Li and Suman, 1996; Adiloglu, 2003).

Corrigiola telephiifolia was the predominant speciesin the ecosystem, providing greater cover values inresponse to soil treatment with both chelators, whileAgrostis castellana, another abundant species, exhibitedreduced cover values when DTPAwas added to the soiland was little affected by EDTA (see Table 2).

Kruskal–Wallis ANOVA revealed significantlygreater plant cover values (though only at a pb0.1)after EDTA treatment, perhaps again as the result of the

Table 5Macronutrients (ppm) determined in the subterranean biomass and shoots of the most abundant species for each treatment (⁎ only one sampleanalyzed due to insufficient data for ANOVA)

Ca K Na Mg P

Subterranean biomassControl 2405±361 a 6198±1229 a 683±47 a 786±290 a 266±29 a1 g EDTA 3784±2538 a 2685±1866 a 1154±407 a 757±447 a 234±24 a1 g DTPA 3483±1027 a 2969±1966 a 2082±1142 a 1086±574 a 243±11 a

Agrostis castellanaControl 4322±569 a 21024±5557 a 445±112 a 2078±242 a 2396±525 a1 g EDTA 4509±1061 a 16375±6791 a 1171±545 b 2647±305 a 238±30 b1 g DTPA 4280⁎ 16378⁎ 1291⁎ 2601⁎ 210⁎

Corrigiola telephiifoliaControl 3973±396 a 19933±513 a 486±94 a 2018±405 a 2580±268 a1 g EDTA 4378±1431 a 16721±2065 a 1163±404 b 2264±524 a 621±684 b1 g DTPA 3747±238 a 19388±1970 a 1285±183 b 2336±184 a 1211±305 b

117J. Pastor et al. / Science of the Total Environment 378 (2007) 114–118

fertilizing action of chelating agents (Table 3). However,the number of species in the plant community was low,and the amendments seemed not to affect this variable.The Shannon index was lower, though not significantly,for the treated soils, possibly due to an increasedavailability of contaminants in the presence of chelatingagents leading to the dominance of the most resistantspecies (Table 3). No significant differences in biomasswere detected (Table 4).

Copper accumulation in the roots was enhanced inthe amended soils, although the significance level wasonly pb0.1 for the EDTA treatment. This could meanthat high amounts of copper may stabilize in the roots(Table 6). The Cu concentrations of the shoots andleaves of A. castellana increased significantly for theEDTA-treated soil, but there was a significant decrease

Table 6Heavy metals (ppm) determined in the subterranean biomass and shoots of thedue to insufficient Data for ANOVA)

Al Fe Mn

Subterranean biomassControl 653±466 a 2280±576 a 315±31 g EDTA 428±36 a 1618±1008 a 282±11 g DTPA 371±37 a 1934±473 a 238±1

Agrostis castellanaControl 28±9 a 66±14 a 695±91 g EDTA 42±24 a 723±367 b 405±11 g DTPA 19⁎ 502⁎ 544⁎

Corrigiola telephiifoliaControl 48±27 a 108±22 a 136±31 g EDTA 223±117 a 561±413 b 149±31 g DTPA 132±88 a 268±91 a 133±1

in P contents (Table 5). Similarly, increased above-ground plant Fe concentrations for the EDTA treatment,were accompanied by a significant decrease in Mnlevels. The copper concentrations observed in thetissues of A. castellana were sufficiently high for itspotential use as an EDTA-enhanced phytoextractionspecies, despite the changes in the chemical composi-tion of this plant possibly affecting its vigor due tonutritional imbalances (Table 5). The biomass of thisspecies growing in the DTPA treated soil wasinsufficient for statistical analysis although we providethe data for one sample in the table.

The chelating agents led to similar changes in C.telephiifolia, such as a significant decrease in planttissue P concentrations (only for EDTA) and increase Fecontents (see Table 6). Na and Cu concentrations also

most abundant species for each treatment (⁎ only one sample analyzed

Zn Cu Cd

9 a 92±13 a 568±221 a 11±6 a37 a 84±44 a 1238±414 b1 7±3 a6 a 87±13 a 1532±538 b 6±2 a

1 a 68±19 a 27±10 a 1±1 a2 b 94±25 a 1494±721 b 4±3 a

75⁎ 622⁎ 5⁎

4 a 61±8 a 41±6 a 3±1 a8 a 71±27 a 3420±4247 b 3±3 a8 a 60±9 a 493±110 b 1±1 a

118 J. Pastor et al. / Science of the Total Environment 378 (2007) 114–118

increased significantly in response to the chelatingagents (Tables 5 and 6) with the amounts of Cuapproaching hyperaccumulation values when EDTAwas added to the soil. However, no significantdifferences emerged between plants growing inEDTA- and DTPA-treated soil, because of the highvariation in the soils variables determined. Thus,standard deviations were high for the variables plantCu, Al and Fe and although these deviations were higherfor C. telephiifolia growing in the EDTA-amended soilthan the DTPA-treated soil, the difference was notstatistically significant.

The nutritional imbalances suffered by these speciesseems to be a common reaction in plants growing underheavy metal stress (Pastor et al., 2003b) and may makethese species less efficient as phytoextractors.

4. Conclusions

The biomarkers examined indicated no severedamage to the plant community, even when doses ofchelating agents were high. The high copper uptake byC. telephiifolia and A. castellana observed suggeststhat these two species could be effectively used forEDTA-enhanced phytoextraction. In addition, bothplants are perennial, not very palatable for herbivoresand may even prevent erosion, as they provide good soilcover. These characteristics are all highly appropriate fora phytoextracting species. Our data also suggest thatDTPA is not as efficient as EDTA at enhancing metaluptake.

Acknowledgement

This study was funded by MEC Projects CTM2005-02165/TECNO and REN2002-02501/TECNO, project1.2-046/2005/3-B from the Spanish Ministry of theEnvironment and an EIADES CAM project.

References

AdilogluA.The effect of EDTAon the extractability of zinc, cadmiumandnickel in vertisol and alluvial soils. Soil Chem 2003;36:1197–200.

B.O.E.. Real Decreto 1310/90 n° 262; 1 noviembre 1990.Encabo C, Cala V, Gutierrez-Maroto A. Evaluación de la dispersión de

metales pesados del entorno de una mina, mediante el método deespeciación secuencial química. Bol Geol Min 1997;108:57–68.

Hernández AJ, Pastor J. Técnicas analíticas para el estudio de lasinteracciones suelo-planta. Henáres Rev Geol 1989;3:67–102.

I.G.M.E. Mapa Geológico de España, E. 1/50000. Memoria y Hoja n°484 (Buitrago de Lozoya). Madrid, 1991, 105 pp.

Li Z, Suman LM. Extractability of zinc, cadmium and nickel in soilsamended with EDTA. Soil Sci 1996;161:226–32.

Madrid F, LiphadzibMS, KirkhambMB. Heavy metal displacement inchelate-irrigated soil during phytoremediation. J Hydrol 2003;272:107–19.

Monturiol F, Alcala del Olmo L. Mapa de asociaciones de suelos de laComunidad de Madrid. CAM. Consejería de Agricultura yCooperación y CSIC, Madrid, 1990,71 pp.

Pastor J, Gutierrez-Maroto A, Hernández AJ. Biomarcadores a nivel deuna comunidad de pasto y de una población herbácea forrajera parasuelos contaminados por cobre. Anal Biol 2003a;25:103–8.

Pastor J, Hernández AJ, Prieto N, Fernández-Pascual M. Accumulat-ing behaviour of Lupinus albus L. growing in a normal and adecalcified calcic luvisol polluted with Zn. J Plant Physiol2003b;160:1457–65.