ent 384 (2007) 229–238www.elsevier.com/locate/scitotenv

Science of the Total Environm

Toxicity assessment of heavy metal mixtures by Lemna minor L.

Tea Horvat a,⁎, Željka Vidaković-Cifrek b, Višnja Oreščanin a,Mirta Tkalec b, Branka Pevalek-Kozlina b

a Ruđer Bošković Institute, Laboratory for Radioecology, Bijenička c. 54, HR-10000 Zagreb, Croatiab University of Zagreb, Faculty of Science, Department of Biology, Roosveltov trg 6, HR-10000 Zagreb, Croatia

Received 12 January 2007; received in revised form 28 May 2007; accepted 5 June 2007Available online 5 July 2007

Abstract

The discharge of untreated electroplatingwastewaters directly into the environment is a certain source of heavymetals in surfacewaters.Even though heavy metal discharge is regulated by environmental laws many small-scale electroplating facilities do not apply adequateprotectivemeasures. Electroplatingwastewaters contain large amounts of various heavymetals (the composition depending on the facility)and the pH value often bellow 2. Such pollution diminishes the biodiversity of aquatic ecosystems and also endangers human health.

The aim of our study was to observe/measure the toxic effects induced by a mixture of seven heavy metals on a bioindicator speciesLemna minor L. Since artificial laboratory metal mixtures cannot entirely predict behaviour of metal mixtures nor provide us withinformations relating to the specific conditions in the realistic environment we have used an actual electroplating wastewater sampledischarged from a small electroplating facility. In order to obtain three more samples with the same composition of heavy metals but atdifferent concentrations, the original electroplating wastewater sample has undergone a purification process. The purification process usedwas developed by Oreščanin et al. [Oreščanin V, Mikelić L, Lulić S, Nađ K, Rubčić M, Pavlović G. Purification of electroplatingwastewaters utilizing waste by-product ferrous sulphate and wood fly ash. J Environ Sci Health A 2004; 39 (9): 2437–2446.] in order toremove the heavy metals and adjust the pH value to acceptable values for discharge into the environment. Studies involving plants andmultielemental waters are very rare because of the difficulty in explaining interactions of the combined toxicities. Regardless of thecomplexity in interpretation, Lemna bioassay can be efficiently used to assess combined effects of multimetal samples. Such realisticsamples should not be avoided because they can provide us with a wide range of information which can help explain many differentinteractions of metals on plant growth and metabolism. In this study we have primarily evaluated classical toxicity endpoints (relativegrowth rate, Nfronds/Ncolonies ratio, dry to fresh weight ratio and frond area) and measured guaiacol peroxidase (GPX) activity as earlyindicator of oxidative stress. Also, we have measured metal accumulation in plants treated with waste ash water sample with EDXRFanalysis and have used toxic unit (TU) approach to predict whichmetal will contribute themost to the general toxicity of the tested samples.© 2007 Elsevier B.V. All rights reserved.

Keywords: Heavy metals; Lemna minor; Ecotoxicity; Metal accumulation; GPX stress marker; Toxic units

1. Introduction

Heavy metals are normally present at low concentra-tions in freshwaters (Le Faucheur et al., 2006) but due to

⁎ Corresponding author. Aleja pomoraca 9, HR-10020 Zagreb,Croatia. Tel.: +385 16526808; fax: +385 16551712.

E-mail address: t.horvat@hi.htnet.hr (T. Horvat).

0048-9697/$ - see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.scitotenv.2007.06.007

the anthropogenic activities they are also the most com-mon non biodegradable pollutant reported at elevatedconcentrations in a majority of parts of the world (Mallickand Mohn, 2003). A certain source of heavy metals is thedischarge of untreated electroplating wastewaters directlyinto the environment. Even though the heavy metal dis-charge is regulated by environmental laws many small-scale electroplating facilities do not apply adequate

230 T. Horvat et al. / Science of the Total Environment 384 (2007) 229–238

protective measures. Electroplating wastewaters containlarge amounts of various heavy metals (the compositiondepending on the facility) and the pH value often bellow 2.These polluted waters may then connect through thehydrographic system and generate a diffuse pollution thatendangers plants and animals of the contaminated eco-systems (Teisseire et al., 1998). Such pollution diminishesthe biodiversity of aquatic ecosystems and also endangershuman health both through the use of polluted surfacewaters for domestic purposes and the consumption of anyeatable species living in them either directly or throughfood-chain magnification (Khan et al., 2000; Miretzkyet al., 2004).

In aquatic ecosystems aquatic plants have an importantrole in the food chain since they are primary producers andregulators of oxygen level and at the same time play asignificant role in the biogeochemical cycling of theelements (Sinha et al., 2005). The aim of our study was toobserve/measure the toxic effects induced by the mixtureof seven heavy metals on a bioindicator species Lemnaminor. We have chosen Lemna for three reasons: aquaticplants are the first link in relation to metal contentsof aquatic environments (Singh et al., 2006), it is a bio-indicator of the ecological relevance for the detection andmonitoring ofmetal pollution (Garnczarska andRatajczak,2000) and has a known ability to accumulate heavy metals(Rahmani and Sternberg, 1999; Singh et al., 2000; Axtellet al., 2003; Miretzky et al., 2004). Artificial laboratorymetal mixtures cannot entirely predict the behaviour ofmetalmixtures nor provide uswith informations relating tothe specific conditions in the realistic environment. Toestablish combined effects of metal mixture sample wehave used an actual electroplating wastewater sampledischarged from a small electroplating facility. In order toobtain three more samples with the same composition ofheavy metals but at different concentrations, the originalelectroplating wastewater sample has undergone a purifica-tion process. The purification process used was developedby Oreščanin et al. (2004) in order to remove heavy metalsand adjust the pH value to acceptable values for dischargeinto the environment. Studies involving plants and multi-elemental waters are very rare because of the difficulty inexplaining interactions of the combined toxicities. Existingstudies deal mainly with the capacity of aquatic plants toremove metals from water by means of accumulation.

The hypothesis tested in this study is that regardless ofcomplexity in interpretation, Lemna bioassay can beefficiently used to assess combined effects of multimetalsamples. Such realistic samples should not be avoidedbecause they can provide us with a wide range of infor-mation which can help explain many different interactionsof metals on plant growth and metabolism.

Although many toxicological studies mainly usebiomarkers as early stress indicators (Härtling and Schultz,1998) Da Rosa Corrêa et al. (2006) emphasized, that theunderstanding of exposure assessment also has to be linkedwith visible damage (i.e. the change in the number offronds, colony number and organisation, a decrease orincrease in relative surface area and weight). This is thereason why we have primarily evaluated classical toxicityendpoints. Furthermore, under such adverse environmentalconditions as the presence of heavy metals, active oxygenspecies are formed in plants (Schützendübel and Polle,2002). Plant cells posses an efficient defence system forreducing harmful effects of oxidative stress (Sergiev et al.,2000; Kapchina-Toteva et al., 2004). Among defencesystem enzymes, peroxidases have low substrate specificitybut can be employed as a sensitive and accurate stressmarker for early detection of stressful conditions (Lepedušet al., 2004) such is sublethal metal toxicity (Radotic et al.,2000; Sinha et al., 2005). We have evaluated L. minor'sresponse to mixed metal induced oxidative stress throughguaiacol peroxidase activity. Additionally, in the case ofmixed metal toxicity, combined effects of multimetal pol-lution could be analyzed by the use of toxic unit approach(TU). This approach is mostly used to test the responseaddition model for the chemical mixtures (Pape-LindstromandLydy, 1997; Van der Geest et al., 2000;An et al., 2004).Toxic unit approach has limitations, but is useful as ameansof comparing the likely relative toxicity induced by heavymetals. In our study we have limited the use of TUapproach to predict which metal will contribute the most tothe general toxicity of the tested samples.

2. Materials and methods

2.1. Sample preparation

The first sample is original electroplating wastewater(EWW) sampled from the waterway near a smallelectroplating facility. The portion of EWW samplewas purified by adding ferrous–sulphate in order toreduce carcinogenic and mobile Cr (VI) into less toxicand less mobile Cr (III), and adding of wood fly ashrinsed three times with distilled water with solid/liquidratio 1:100 to neutralise and remove other present heavymetals (Oreščanin et al. (2004). Briefly, ferrous–sul-phate was added into electroplating wastewater andmixed. Then wood fly ash, which serves both as aneutralising agent as well as a coagulant, was added intothe solution until the pH 8 was reached. After the pHadjustment the mixture was mixed again and allowed tosettle. The end products of this method are purifiedwater (pH 8, suitable for further technical use or

231T. Horvat et al. / Science of the Total Environment 384 (2007) 229–238

discharge into the environment) and remaining wasteash. Remaining waste ash is suggested for depositing ondomestic waste repositories or alternatively, for regen-eration of the heavy metals. Our other samplesoriginating from the above described procedure arepurified wastewater (PWW), the water extract of rinsedfly ash (FA) and the water extract of waste ash (WA)remained after the purification of electroplating waste-water with ferrous sulphate and rinsed fly ash. Therinsing of fly ash before using has been found necessaryin order to remove the excess of alkali. Water extractablefractions of fly and waste ash have been preparedaccording to the protocol for the standardized Germanleaching test DIN 38414-S4. All samples have beenpreconcentrated with ammoniumpyroloidinedi-thiocar-bamate (w/v 1%), prepared as thin targets and analyzedfor elemental analysis by energy dispersive X-rayfluorescence (EDXRF) (Oreščanin et al., 2004). Ele-mental concentrations measured in the samples as wellas maximum concentrations allowed in wastewaters andwater extracts suitable for direct discharge into theenvironment are given in Table 1.

2.2. L. minor and classical ecotoxicity parameters

L. minor has been taken from the laboratory stockculture, grown from the plants originally collected in theBotanical Garden of the Department of Biology, Facultyof Science, University of Zagreb. Originally collectedplants have been sterilised according to Krajnčič andDevidé (1980) and maintained on Pirson–Seidel'snutrient solution (Pirson and Seidel, 1950). The stockculture was grown at 24±2 °C with 16/8 light/dark cycleand a photon flux density of 40–50 μmol/m2/s.

Table 1Elemental composition of liquid samples and maximum allowedconcentrations of these elements in wastes

Element (μg l−1) EWW PWW MAV-WW FA WA MAV-E

Pb 8600 2 200 16 19 1000Cr(VI) 2450 7 100 37 18 100Cr(III) 1140 8 1000 16 n.d. /Mn 8300 61 1000 9 n.d. /Fe 69000 33 2000 120 105 /Ni 1300 5 1000 1 7 1000Cu 3100 11 100 28 15 5000Zn 505000 40 1000 440 85 5000

Abbreviations:MAV-WW — maximum allowed values for waste waters suitable forthe discharge into the environment.MAV-E— water extract for the waste materials suitable for depositionon the domestic waste repository.n.d. — not detected.

For establishing classical ecotoxicity parametersrelative growth rate, Nfronds/Ncolonies ratio, dry tofresh weight ratio and frond area have been monitored.

Uniform, healthy duckweeds with 2–3 fronds percolony have been selected as test specimens and placedin 100 ml Erlenmeyer flasks containing 60 ml ofPirson–Seidel's nutrient solution (control), or Erlen-meyer flasks containing tested samples (EWW, PWW,FA and WA) prepared with the same nutrients andadjusted to pH 4.55. All experiments have lasted15 days and have been prepared in 8 replicates.

The frond number and the number of colonies havebeen monitored on days 0, 3, 5, 7, 9, 11, 13 and 15 bycounting the visible fronds and floating colonies. Daughterfronds have been included as soon as they had emergedfrom the pouch of the mother frond. The range finding testhas been performed with electroplating wastewater(EWW) to find sublethal concentrations. Experiment hasbeen set with 10, 25, 50, 75 and 100% (v/v) concentrationof electroplating wastewater and the control. The plants inconcentration of 100, 75 and 50% (v/v) electroplatingwastewater have become chlorotic and died within 3 daysso 10 and 25% (v/v) electroplating wastewater have beenchosen for further experiments (data not shown).

The relative plant growth (RG) has been calculatedusing the expression (Ensley et al., 1994)

RG ¼ no: of fronds at dayn� no: of fronds at day 0no: of fronds at day 0

where n=3,5,7,9,11,13,15 and expressed as means of8 replicates±standard deviation.

The wet and dry weight was measured at the end ofthe 15 day experiment. Before weighing the plants havebeen placed on the net and rinsed with distilled water 7times, then placed on the filter paper to remove theexcess water. After weighing the fresh weight, the plantshave been oven dried at 108 °C overnight to constantweight. Dry to fresh weight ratio has been calculated foreach set of 8 replicates.

For the measurement of the total frond area the plantshave been harvested on 15th day, dried on the filter paperand scannedwithHP-ScanJet 3400C scannerwith defaultresolution set at 200 d.p.i. Images have been analysed inAdobe Photoshop 7.0. to calculate the frond area. Theresults are expressed as the total frond area on the 15th daydivided by the number of fronds on the 1st day.

2.3. Metal accumulation

Metal accumulation was measured by EDXRFanalysis of Lemna plants treated with a sample of

232 T. Horvat et al. / Science of the Total Environment 384 (2007) 229–238

water extract of waste ash (WA) remained after thepurification of electroplating wastewater with rinsed flyash because this form of ash is suggested to be safelydeposited on domestic waste repositories. Preliminaryleaching tests showed that heavy metals scavengedduring the purification process are firmly bound in re-maining waste ash (Oreščanin et al., 2004). Still wewanted to establish which metals L. minor wouldaccumulate from leachate in the environment.

Eight replicates of the plants treated with remainingwaste ash and 8 corresponding controls have beenharvested after 7 days of treatment. The samples of L.minor have been dried to the constant weight at 108 °Cand prepared for the analysis by desegregations withliquid nitrogen in an agate mortar. The obtained powderedsamples have been pressed into pellets and concentrationsof the elements in targets have been analysed by EDXRF.The samples have been irradiated by X-rays generatedfrom the 109Cd annular source. The spectra have beencollected by Genie-2000 software (Canberra, Meriden,CT USA). The collecting time was 40,000 s. The spectraldata have been analyzed by WinAxil software (Canberra,Meriden, CT USA). The calibration file for spectrumfitting and quantification for plant material have beencreated on the basis of the measurements of the standardreference material Orchard leaves prepared and analyzedin the same way as the L. minor samples. Elementalconcentrations have been calculated with the “Funda-mental parameters” method from the WinFund package.

2.4. Guaiacol peroxidase (GPX) stress marker

As a toxicity stress marker we have measured the levelof the guaiacol peroxidase (GPX) activity expressed as aunit of enzyme per mg fresh weight. For this set of

Fig. 1. Relative growth rate of plant Lemna minor treated with tested samplesas means of eight replicates±standard deviation (SD). An asterisk (⁎) denotePb0.001 of the mean value of the growth rate between the sample and itselectroplating wastewater, PWW-purified wastewater, FA-water extract of fl

experiments about 10 healthy colonies with 2 fronds havebeen taken from the stock culture and placed in 250 mlErlenmeyer flasks containing 100 ml of Pirson-Seidel'snutrient solution (control) or 100 ml of the tested samples,prepared in the same way as for the classical toxicityendpoints. All samples have been prepared in 4 replicates.On the 3rd, 6th and 8th day plant materials have beencollected from each treatment as well as from their re-spective controls.

For measuring the GPX activity and the soluble proteincontent, 100 mg of plant material has been homogenised in1.5ml of cold 50mMpotassium phosphate buffer (pH 7.0),containing 1 mM EDTA and 5% (w/v) polyvinylpolypyr-rolidine. The homogenates have been centrifuged at20,000 ×g at 4 °C for 30 min and supernatant has beenused for enzyme assay. The GPX activity has been deter-mined spectrophotometrically at 470 nm by the guaiacoloxidationmethod. The reactionmixture consisted of 50mMpotassium phosphate buffer (pH 7.0) with 18 mM guaiacoland 5 mM H2O2 (Chance and Maehly, 1955). The reactionhas been started by adding 20 μl of supernatant sample to1 ml of the reaction mixture. The results are expressed asμmol of formed tetraguiacol per minute (U).

The soluble protein content had been determinatedby the method of Bradford (Bradford, 1976) usingbovine albumin serum as a standard.

2.5. Statistical analysis

Since classical experimental end points have beenperformed in 8 replicates, the data in each test serieshave been tested for normality and homogeneity ofvariances (Shapiro–Wilk's test). After ANOVA analysisof variance, the differences between the treatments andbetween days of the treatment have been tested with

and their corresponding control during 15 day treatment. Data are givens significant difference Pb0.05 and ⁎⁎⁎denotes significant differencecorresponding control on the same sampling day. C-control, EWW-

y ash, WA-water extract of waste ash.

Fig. 2. Nfr/Ncol ratio recorded in Lemna minor during 15 day treatment. Data are given as means of eight replicates±standard deviation (SD). Anasterisk (⁎) denotes significant difference (Pb0.05) of the mean value between the sample and its corresponding control on the same sampling day.

233T. Horvat et al. / Science of the Total Environment 384 (2007) 229–238

Newman–Keuls test (Pb0.05 and Pb0.001). Data arepresented as mean±1 standard deviation. All the abovementioned tests are included in Statistica 7.0. (StatSoft.Inc., Tulsa, USA) software package.

The GPX activity measures have been performed in 4replicates and also tested for normality and homogeneity asdescribed above. The data obtained have been tested withNewman–Keuls test for analysis of variance (Pb0.05).

2.6. Toxicity assessment of multimetal pollution usingtoxic unit approach

To express concentrations of the metals in our mixedmetal samples we used toxic unit (TU) model which

Fig. 3. Dry to fresh weight ratio in Lemna minor after 15 days of treatment. Dastandard error (SE). An asterisk (⁎) denotes significant difference (Pb0.05) othe 15th day of the treatment.

expresses metal concentrations as fractions of theirmedian effective concentration (EC50). The sum of TUis expressed as (An et al., 2004)

XTUi ¼

Xn

i¼1

ciECxi

where n is the number of metals in the sample, ci ismeasured concentration of the individual metal in thesample, and ECxi is the concentration of ith samplecomponent that cause x% effect. Since in our study wehave not tested single metals, but multimetal samples,for all calculations we used EC50 values taken fromliterature data for 4 days metal toxicity of Ni, Cu, Fe,

ta are given as means of eight replicates±1 standard deviation (SD)±1f the mean value between the sample and its corresponding control on

Fig. 4. Total frond area of treated Lemna plants recorded on the 15th day of treatment. Data are given as means of eight replicates±1 standarddeviation (SD)±1 standard error (SE). Asterisks (⁎⁎⁎) denotes significant difference (Pb0.001) of the mean value between the sample and itscorresponding control on the 15th day of the treatment.

Table 2Elemental composition (μg g−1 dwt) of dried Lemna plants after7 days of treatment with water extract of waste ash (WA) and theircorresponding control

Element Control WA

Pb 1.3±0.06 3.5±0.1Cr 1.0±0.2 0.9±0.2Mn 29±2 39±3Fe 124±1 203±3Ni 0.17±0.06 0.5±0.1Cu 3.6±0.1 3.9±0.2Zn 10.4±0.3 23.5±0.6Dry weight (g) 0.7686 1.1884

Data are given as mean values of eight replicates±1 SD (n=8).

234 T. Horvat et al. / Science of the Total Environment 384 (2007) 229–238

Pb, Zn, Mn and Cr (VI) (Wang, 1986). Using toxicunits we calculated the proportion of the toxicity ofindividual metal to the total toxicity of mixed metalsamples.

3. Results

3.1. Classical ecotoxicity parameters

3.1.1. Growth rateThe Newman–Keuls test has shown statistically

significant difference (Pb0.05) on the 3rd day of thetreatment (Fig. 1) between the samples of 25% (v/v)electroplating wastewater (EWW), 10% (v/v) electro-plating wastewater and water extract of the waste ash(WA) compared to the corresponding control. Samplesof purified water (PWW) and fly ash (FA) showed nosignificant difference in respect to the control on the 3rdday. On the 5th and 7th day the plants treated with 25%(v/v) electroplating wastewater, 10% (v/v) electroplat-ing wastewater have shown statistically significantdifference at Pb0.001 and water extract of the wasteash has shown statistically significant difference atPb0.05 in the growth rate opposed to the correspondingcontrol. From the 9th until 15th day the growth rate hasbeen significantly different (Pb0.001) from the controlonly in the plants treated with 25% (v/v) electroplatingwastewater and 10% (v/v) electroplating wastewater.

3.1.2. Nfr/Ncol ratioOn the 1st day there have not been any statistically

significant differences between the tested samples andthe corresponding control. On the 3rd day Nfr/Ncol ratiohas shown statistically significant difference (Pb0.05)between 10% (v/v) electroplating wastewater and 25%(v/v) electroplating wastewater compared to the controlplants (Fig. 2). This has repeated on the 5th, 7th, 9th,11th, 13th and 15th day of the treatment, indicating thatonly the electroplating wastewater samples exhibitsignificant change in Nfr/Ncol ratio compared to thecontrol and other tested samples. Avisual observation offronds grown in both electroplating wastewater samples

Fig. 5. Activity of guaiacol peroxidase measured on the 3rd, 6th and 8th day of the treatment. Data are given as means of four replicates±1 standarddeviation (SD). Significant difference is set as Pb0.05. An asterisk (⁎) denotes significant difference of the mean value of the guaiacol peroxidaseactivity between the sample and its corresponding control on the same sampling day.

Fig. 6. Proportion of single metal toxicity expressed as toxic unit (TU) tothe joint toxicity expressed as sum of toxic units ΣTU for each sample.

235T. Horvat et al. / Science of the Total Environment 384 (2007) 229–238

has shown that the fronds had been considerably smallerthan both control plants and plants in other treatments.There have also been fewer colonies, mostly withbleached or chlorotic fronds.

3.1.3. Dry to fresh weight ratioDry to fresh weight ratio measured at the end of the

15 day treatment has shown significant difference(Pb0.05) in the samples of 10% (v/v) electroplatingwastewater and 25% (v/v) electroplating wastewater(Fig. 3). The samples of purified water and waterextracts of both fly and waste ash have shown nostatistically significant difference in dry to fresh weightratio compared to the control plants.

3.1.4. Total frond areaThe fronds of the plants grown in the 10% (v/v) and

25% (v/v) electroplating wastewater have been, aspreviously mentioned, mostly bleached or chlorotic,therefore there have been few green, healthy frondswhat resulted in significant difference (Pb0.001) in thetotal frond area from the corresponding control (Fig. 4.).The plants grown in the purified water samples have notbeen significantly different from the control plants.Plants treated with water extracts of fly and waste ashhave also not been significantly different from the totalfrond area of the control plants.

3.2. Metal accumulation

The concentrations of the elements measured in thecontrol and water extract of the waste ash (WA) treatedLemna plants are given in the Table 2. An increase in theaccumulation of Pb, Mn, Ni, Zn and especially of Fe isnoticed. This complies well with the known ability ofLemna species to accumulate heavy metals.

3.3. GPX activity and soluble protein content

The measurements of the activity of guaiacolperoxidase are given in the Fig. 5. On the 3rd day ofthe treatment only 25% (v/v) electroplating wastewaterinduced statistically significant increase in the peroxi-dase activity indicating oxidative stress induced by thepresence of heavy metals. On the same day a significantincrease of soluble protein content has been recorded inall tested samples (the data not shown).

On the 6th day of the treatment the guaiacol peroxidaseactivity in 25% (v/v) electroplating wastewater (EWW)sample has decreased, so there has not been anystatistically significant difference between the sampleand the control nor the control and other treatments. Bythe 6th day the protein contents in all treatments have alsodecreased to normal (corresponding with the control). Onthe 8th day there has again been an increase in theguaiacol peroxidase activity in the 25% (v/v) electroplat-ing wastewater treatment indicating the detoxification ofH2O2 produced under prolonged oxidative stress. On the

236 T. Horvat et al. / Science of the Total Environment 384 (2007) 229–238

same day though, there has been no record of higher levelof soluble protein content in any of the treatments.

3.4. Toxicity assessment of multimetal pollution usingtoxic unit approach

As we have stated earlier, we have not tested singlemetals butmixedmetal samples, sowe have used literaturedata for EC50 inhibition concentrations of Ni, Cu, Fe, Pb,Zn, Mn and Cr (VI) during 4 day treatment (Wang, 1986).EC50 inhibition value for Cr(III) was not available, butsince Cr(III) is considered to be 1000 times less toxic thanCr(VI) (Peer et al., 2005)we calculatedTUwithoutCr(III).During longer exposure time EC50 values decreases(Wang, 1986), so results given in Fig. 6. were used only toestimate the proportion of single metal toxicity to the jointtoxicity of the sample. We assumed that actual metaltoxicities at the end of the 14 day treatment had the samedistribution even though EC50 values were presumablylower. Fig. 6. shows that in electroplating wastewatersamples (10 and 25% v/v EWW) the majority of pollutionis caused by Zn2+ (66.2%) and Fe2+ (24.45% related to thesum of toxic unit of all metals in EWW samples). Inpurified water sample proportion of Zn concentration(expressed as toxic units) considerably decreased to10.98% Zn, while Fe concentration (expressed as pro-portion to the sum of all TU of the sample) remained24.45% even though actual concentration of Fe decreased.In the water extract sample of wood fly ash (FA) largeamount of the observed effect was contributed to Zn(35.08%),whilewater extract ofwaste ash (WA) containedonly 12.34% Zn2+, indicating strong binding of metal tothe waste ash. In this sample (WA) considerable effect alsois ascribed to 41.18% Fe2+ and 22.57% Ni2+ in relation tothe sum of toxic units of all metals in the sample.

4. Discussion

As shown in the Results, both 10% (v/v) and 25% (v/v)electroplatingwastewater (EWW) strongly suppressed plantgrowth, which was expected because these samples containconsiderable amounts of potentially toxic metals. Toxic unit(TU) approach to toxicity assessment of the tested samplesshowed that this effect could be mainly contributed to highamounts of zinc and ferrous (see Fig. 6). Themain limitationof the TU approach in our study relates to the fact that thereare numerous and undefined interactions induced bysynergisms or antagonisms of heavy metals present. Eventhough toxic unit approach allows us to predict these inter-actions, we have used it only to express percentage ofindividual heavy metal toxicity in relation to the sum oftoxicity of all heavy metals present in the samples.

After the application of the electroplating wastewatertreatment method (Oreščanin et al., 2004) the obtainedpurified water sample has shown no toxicity effects inany of the observed parameters. The same is found forthe water extract of rinsed fly ash before the treatment.Both samples contain metals in low concentrations andin different proportions than the electroplating waste-water samples (see Fig. 6).

Water extract of the waste ash (WA) has showed slighttoxicity on the growth rate during the first seven days of thetreatment (Pb0.05 opposed to Pb0.001 for electroplatingwastewater samples). Levels of all metals in water extract ofthe waste ash sample are lower than in the i.e. water extractof the fly ash sample which has not induce any toxicity at all(see Table 1). The obtained results suggest L. minor's spe-cific response to the combination of metal concentrationspresent in that sample. Since tolerant or toxic responsehighly depends on the proportion of substances in the me-dium, this proportion of metals induced plant growth inhi-bition during the first seven days of the treatment, but duringthe next seven days plants detoxifying mechanisms havemanaged to overcome it. Among these, observed effect ismainly contributed to 41.18% Fe2+, 22.57% Ni2+, 19.75%Cu2+ and 12.34% Zn2+ in the total sum of TUs of waste ashwater extract. Both Ni and Zn are found to have genotoxicactivity via oxidative pathways and this genotoxic activityinfluences cell division and hence plant growth (Rout andDas, 2003). The fronds grown in the water extract of wasteash sampleswere all healthywith normally formed colonies.Water extract of the waste ash (WA) sample has not inducedany toxic effect in other evaluated parameters (especiallythere were no GPX stress marker results, no increase in dryto fresh weight ratio nor total frond area).

Heavy metals in both electroplating wastewater sam-ples have also caused the formation of small, usually singleand mostly chlorotic fronds resulting in plants beingsignificantly smaller than in other samples. Both growthinhibition and chlorosis are zinc toxicity effects (Rout andDas, 2003) due to 66.2% Zn2+ in the total toxicity of theelectroplating wastewater samples expressed as TU.

The interactions in our multimetalic samples consider-ably limit possibility to explain the results for singlemetals, but TU based prediction of major metal con-tributors to total toxicity of the sample increased certaintyof the interpretation. Zn2+ is an essential microelement thatis indispensable for normal plant growth at low concentra-tions but in higher concentrations it is toxic and inducesoxidative damage (Erdei et al., 2002). The tested samplesof 10% and 25% (v/v) electroplating wastewater contained50.500 and 126.250 μg l−1 of Zn2+ respectively, whichcontributed to the sample's toxicity. Considerable amountsof Fe2+ and Pb2+ are also present, but in terms of TU based

237T. Horvat et al. / Science of the Total Environment 384 (2007) 229–238

prediction of toxicity Cu2+ (3.69%) and Ni2+ (3.78%) aremore significant than Pb. The samples also contain Cr6+

and Cr 3+ in smaller amounts. Cr(III) is known to be lessmobile and less toxic than Cr(VI) and is mainly bonded toorganic matter both in soil and aquatic environments(Beckquer et al., 2003; Shanker et al., 2005).

Chromium (VI) is hazardous heavymetal which causesmembrane damage, ultrastructural changes in the orga-nelles, disrupted metabolic activities, growth inhibition aswell as oxidative damage to lipids, proteins and nucleicacids (Sinha et al., 2005). Dirilgen and Doğan (2002)reported a difference in the accumulation of Cr6+ and Cr3+

in L. minor. Their results show that Lemna accumulatedlarger quantities of Cr6+ which caused smaller growth andsmaller quantities of Cr3+ which resulted in larger Lemnagrowth. We have not been able to differentiate effects ofCr6+ from Cr3+ but there is a possibility that the differencein the absorption of one or the other metal species causedsome of the effects observed.

Dry to fresh weight ratio results also show a significantincrease only in the 10% and 25% (v/v) electroplatingwastewater treatments. That is understandable consider-ing that these samples contain larger amounts of metalsthan other samples, and Lemna's ability to accumulatemetals is well known (Rahmani and Sternberg, 1999;Singh et al., 2000; Axtell et al., 2003; Miretzky et al.,2004). These two samples also significantly inhibit plantgrowth. Garnczarska and Ratajczak (2000) have noticedthat when the growth of L. minor was inhibited by Pb2+

ions that the dry to fresh ratio increased with increase ofthe Pb2+concentration in the medium indicating probablestarch grain accumulation. Our results show similar effectbut TU based toxicity assessment of electroplatingwastewater sample showed only 1.4% Pb2+ in the jointtoxicity. This could be explained by accumulative effectadded to high percentage of Zn2+ and presence of Ni2+

which is known genotoxin.Except for the above mentioned metals, the electro-

plating wastewater samples also contain Cu2+ ions.Copper is an essential micronutrient and it constituentsseveral enzymes but in only slightly increased concen-trations than required for plant life, it becomes toxic, andin large quantities interacts with photosynthesis andrespiratory processes, protein synthesis and the devel-opment of organelles (Frankart et al., 2002). They havealso found that concentrations lower than 400 μg Cu2+

affect L. minor's photosynthetic capacity. For compar-ison, the 10% (v/v) electroplating wastewater and 25%(v/v) electroplating wastewater samples contain 310 and775 μg Cu2+, respectively.

Interestingly, even though it is an early stress indicator,there has not been any increase in guaiacol peroxidase

(GPX) activity in the 10% (v/v) electroplating wastewatersample. Observed increase in the guaiacol peroxidaseactivity in the 25% (v/v) electroplating wastewater samplerecorded on the 3rd day could have been causedmainly byzinc but also by interaction of any of the above mentionedmetals. The plant protectivemechanisms have successfullyreduced oxidative stress by the 6th day, but a prolongedexposure to 8th day, induced stress response again.

The content of soluble proteins has been significantlyincreased in all treated plants when sampled on the 3rdday of the experiment, which corresponds to the signif-icant inhibition of plant growth observed also on the 3rdday of the treatment in all except the purified water(PWW) treatment. Toxic unit approach to prediction ofheavymetal toxicity shows that this sample has the lowestlevels of Zn2+, Pb2+ and Cr6+ of all tested samples. Theexplanation for the increase of soluble protein content inother treatments might lie in the adjustment phase of theplants needed to recover from planting and for adaptationto metal containing medium. However, by the 6th daylevels of proteins have decreased and remained relativelyconstant until the 8th day.

5. Conclusions

In nature there is a number of interactions betweenmetals, some of them have been identified by researchers,some still not, but despite the difficulties in the expla-nation arising from the multielemental composition, sucha composition realistically shows the situation on thepotential environmental site. In this study we tried toassess which metals are major contributors in observedtoxicity parameters with toxic unit approach. The ob-tained results have shown that Lemna bioassay could bemore often used in multimetalic toxicity studies, particu-larly in the attempt to explain many observed effects onplant growth and metabolism in the case of recreation ofmore lifelike environment. Toxic unit approach in assess-ment of toxicity of individual metals from the mixedmetals samples has proven to be useful in understandingmajor contributions to adverse effects.

Acknowledgments

The authors wish to thank Mrs. Iva Kušanić forassistance with preparation of the manuscript.

References

An YJ, Kim YM, Kwon TI, Jeong SW. Combined effect of copper,cadmium, and lead upon Cucumis sativus growth and bioaccu-mulation. Sci Total Environ 2004;326:85–93.

238 T. Horvat et al. / Science of the Total Environment 384 (2007) 229–238

Axtell NR, Sternberg SPK, Claussen K. Lead and nickel removal usingMicrospora and Lemna minor. Bioresour Technol 2003;89:41–8.

Beckquer T, Quentin C, Siscot M, Baudot JP. Chromium availability inultramofic soils from New Caledonia. Sci Total Environ 2003;301:251–61.

Bradford MM. A rapid and sensitive method for the quantization ofmicrogram quantities of protein utilising the principle of protein–dye binding. Anal Biochem 1976;72:248–54.

Chance B, Maehly AC. Assay of catalases and peroxidases. MethodsEnzymol 1955;2:764–75.

Da Rosa Corrêa AX, Rörig LR, Verdinelli MA, Cotelle S, Férard JF,Radetski CM. Cadmium phytotoxicity: quantitative sensitivityrelationships between classical endpoints and antioxidative enzymebiomarkers. Sci Total Environ 2006;357:120–7.

Dirilgen N, Doğan F. Speciation of chromium in the presence ofcopper and zinc and their combined toxicity. Ecotoxicol EnvironSaf 2002;53:397–403.

Ensley HE, Barber JT, Polito MA, Oliver AI. Toxicity and metabolismof 2,4-dichlorophenol by the aquatic angiosperm Lemna gibba.Environ Toxicol Chem 1994;13(2):325–31.

Erdei S, Hegedüs A, Hauptmann G, Szalai J, Horváth G. Heavy metalinduced physiological changes in the antioxidative response system.Acta Biol Szeged 2002;46(3–4):89–90.

Frankart C, Eullaffroy P,Vernet G. Photosynthetic responses ofLemnaminorexposed to xenobiotics, copper, and their combinations. EcotoxicolEnviron Saf 2002;53:439–45.

Garnczarska M, Ratajczak L. Metabolic responses of Lemna minor tolead ions I. Growth, chlorophyll level and activity of fermentativeenzymes. Acta Physiol Plant 2000;22(4):423–7.

Härtling S, Schultz H. Biochemical parameters as biomarkers for theearly recognition of environmental pollution in Scotish pine trees.I. Phenolic compounds. Z Naturforsch 1998;53 c:331–40.

Kapchina-Toteva V, Slavov S, Batchvarova R, Krantev A, Stefanov D,Uzunova A. Stress markers in Chlorsulphuron tolerant transgenictobacco plants. Bulg J Plant Physiol 2004;30(1–2):103–11.

Khan G, Kuek C, Chaudhary T, Fhoo C, Hayes W. Role of mycorrhizaeand phytochelators in heavy metal contaminated land remediation.Chemosphere 2000;41:197–207.

Krajnčič B, Devidé Z. Report on photoperiodic responses in Lemnaceaefrom Slovenia. Ber Geobot Inst 1980;47:75–86.

Le Faucheur S, Schildknecht F, Behra R, Sigg L. Thiols in Scenedesmusvacuolatus upon exposure to metals and metalloids. Aquat Toxicol2006;80:355–61.

Lepeduš H, Cesar V, Krsnik-Rasol M. Guaiacol peroxidases in carrot(Daucus carota L.) root. Food Technol Biotechnol 2004;42(1):33–6.

Mallick N, Mohn FH. Use of chlorophyll fluorescence in metal-stressresearch: a case study with the green microalga Scenedesmus.Ecotoxicol Environ Saf 2003;55:64–9.

Miretzky P, Saralegui A, Fernández Cirelli A. Aquatic macrophytespotential for simultaneous removal of heavy metals (Buens Aires,Argentina). Chemosphere 2004;57:997–1005.

Oreščanin V, Mikelić L, Lulić S, Nađ K, Rubčić M, Pavlović G.Purification of electroplating wastewaters utilizing waste by-productferrous sulphate and wood fly ash. J Environ Sci Health A 2004;39(9):2437–46.

Pape-Lindstrom PA, Lydy MJ. Synergistic toxicity of atrazine andorganophosphate insecticides contravenes the response additionmixture model. Environ Toxicol Chem 1997;16:2415–20.

Peer WA, Baxter IR, Richards EL, Freeman JL, Murphy AS. Phy-toremediation and hyperaccumulator plants. Molecular biology ofmetal homeostasis and detoxification. Topics in Current GeneticsBer-lin: Springer; 2005. p. 299–340.

Pirson A, Seidel F. Zell- und stoffwechselphysiologiche Untersuchun-gen an der Wurzel von Lemna minor unter besonderer Berücksich-tigung von Kalium- und Calciummangel. Planta 1950;38:431–73.

Radotic K, Ducic T, Mutavdzic D. Changes in peroxidase activity andisozymes in spruce needles after exposure do different concentra-tions of cadmium. Environ Exp Bot 2000;44:105–13.

Rahmani GNH, Sternberg SPK. Bioremoval of lead from water usingLemna minor. Bioresour Technol 1999;70:225–30.

Rout GR, Das P. Effect of metal toxicity on plant growth andmetabolism: I. Zinc. Agronomie 2003;23:3–11.

Schützendübel A, Polle A. Plant responses to abiotic stress. Heavymetal-induced oxidative stress and protection by mycorrhization.J Exp Bot 2002;372:1351–65.

Sergiev I, Alexieva V, Yanev S, Karanov E. Effect of atrazine andspermine on free praline and some antioxidants in pea (Pisumsativum L.) plants. Comp Rend Acad Bulg Sci 2000;53(10):63–6.

Shanker AK, Cervantes C, Loza-Tarera H, Avudainayagnam S.Chromium toxicity in plants. Environ Int 2005;31:739–53.

Singh S, Pradhan S, Rai LC. Metal removal from single and multi-metallic systems by different biosorbent materials as evaluated bydifferential pulse anodic stripping voltammetry. Process Biochem2000;36:175–82.

Singh S, Eapan S, D’Souza SF. Cadmium accumulation and itsinfluence on lipid peroxidation and antioxidative system in anaquatic plant, Bacopa monnieri L. Chemosphere 2006;62:233–46.

Sinha S, Saxena R, Singh S. Chromium induced lipid peroxidation in theplants of Pistia stratiotes L.:role of antioxidants and antioxidantenzymes. Chemosphere 2005;58:595–604.

Teisseire, Couderchet M, Vernet G. Toxic responses and catalseactivity of Lemna minor L. Exposed to folpet,copper, and theircombination. Ecotoxicol Environ Saf 1998;40:194–200.

Van der Geest HG, Greve GD, Boivin ME, Kraak MHS, van GestelCAM. Mixture toxicity of copper and diazinon to larvae of themayfly (Ephoron virgo) judging additivity at different effect levels.Environ Toxicol Chem 2000;19:2900–5.

WangW. Toxicity tests of aquatic pollutants by using common duckweed.Environ Pollut B 1986;11:1–14.