Science of the Total Environment 408 (2010) 3162–3168

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r.com/ locate /sc i totenv

Effect of electrokinetic remediation on indigenous microbial activity and communitywithin diesel contaminated soil

Seong-Hye Kim a, Hyo-Yeol Han b, You-Jin Lee c, Chul Woong Kim a, Ji-Won Yang a,⁎a Nano Environmental Engineering Lab, Dept. of Chemical & Biomolecular Engineering, KAIST, 335 Gwahangno, 373-1 Guseong-dong, Yuseong-gu, Daejeon 305-701, Republic of Koreab Daedeok Research Institute Honam Petrochemical Corp., 24-1, Jang-dong, Yuseong-gu, Daejeon, 305-726, Republic of Koreac Power Facility-IT Research Center, KERI, 70 Boolmosangil, Changwon, 641-120, Republic of Korea

⁎ Corresponding author. Tel.: +82 42 350 3964; fax:E-mail address: jwyang@kaist.ac.kr (J.-W. Yang).

0048-9697/$ – see front matter © 2010 Elsevier B.V. Adoi:10.1016/j.scitotenv.2010.03.038

a b s t r a c t

a r t i c l e i n f o

Article history:Received 11 November 2009Received in revised form 17 March 2010Accepted 23 March 2010Available online 8 May 2010

Keywords:Electrokinetics (EK)Total petroleum hydrocarbon (TPH)DieselDGGEShannon–Weaver indexMicrobial community

Electrokinetic remediation has been successfully used to remove organic contaminants and heavy metalswithin soil. The electrokinetic process changes basic soil properties, but little is known about the impact ofthis remediation technology on indigenous soil microbial activities. This study reports on the effects ofelectrokinetic remediation on indigenous microbial activity and community within diesel contaminated soil.The main removal mechanism of diesel was electroosmosis and most of the bacteria were transported byelectroosmosis. After 25 days of electrokinetic remediation (0.63 mA cm−2), soil pH developed from pH 3.5near the anode to pH 10.8 near the cathode. The soil pH change by electrokinetics reduced microbial cellnumber and microbial diversity. Especially the number of culturable bacteria decreased significantly andonly Bacillus and strains in Bacillales were found as culturable bacteria. The use of EDTA as an electrolyteseemed to have detrimental effects on the soil microbial activity, particularly in the soil near the cathode. Onthe other hand, the soil dehydrogenase activity was enhanced close to the anode and the analysis ofmicrobial community structure showed the increase of several microbial populations after electrokinetics. Itis thought that the main causes of changes in microbial activities were soil pH and direct electric current. Theresults described here suggest that the application of electrokinetics can be a promising soil remediationtechnology if soil parameters, electric current, and electrolyte are suitably controlled based on theunderstanding of interaction between electrokinetics, contaminants, and indigenous microbial community.

+82 42 350 3910.

ll rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

Soil is an important habitat for living organisms. Soil pollution is anincreasing environmental problem and technologies to remediatecontaminated soil have been greatly developed. Electrokinetics (EK) isone of the recently developed remediation technologies. Electroki-netics can provide the degradation of soil contaminants and theremoval of metals by direct movement of pollutants, in response tothe direct electric current (Maini et al., 2000). This process has beendemonstrated to be successful with wide ranges of pollutants such asorganic chemicals and heavymetals, and has emerged as one practicalengineering technique for remediation of contaminated soils (Kimet al., 2009; Ko, 1999; Maini et al., 2000). The main advantages ofelectrokinetic remediation are that this process can be performed insitu and is particularly effective for soil with low permeability. Inrecent years, there has been increasing interest in applying electro-kinetics to the site contaminated with hydrophobic organic com-

pound (HOC) that has low volatility, low mobility, low solubility, andlow degradability (Park et al., 2007; Saichek and Reddy, 2003).

As mentioned above, the underlying mechanism of electrokineticsis the introduction of electric current into soil. The introduced electriccurrent in electrokinetics leads to the migration of contaminants viaelectroosmosis, electromigration, and electrophoresis; the processproduces hydrogen ions at the anode and hydroxyl ions at thecathode, resulting in a pH gradient. These phenomena cause changesin the soil condition (Acar and Alshawabkeh, 1993). Electromigrationand electrophoresis result in the movement of ions, ion complex, andcharged particles including microorganisms toward the electrode ofthe opposite charge. Electroosmosis results in the movement of soilmoisture toward the cathode. Chen et al. (2006) observed that theavailable nitrogen, phosphorus and potassium in soil were changedafter EK remediation.

The changes in soil condition such as soil pH, temperature (canincrease by 1–3 °C as a result of EK process), and bioavailability, andthe application of electric current have direct effects on soil microbialactivity. Several studies made efforts to combine EK process andbioremediation (EK-bioremediation) to enhance the transport ofbacteria or nutrients for effective biodegradation (DeFlaun andCondee, 1997; Schmidt et al., 2007; Shi et al., 2008a; Wick et al.,

Fig. 1. A schematic view of the lab-scale electrokinetic reactor.

3163S.-H. Kim et al. / Science of the Total Environment 408 (2010) 3162–3168

2004). In these studies, the movement of bacteria and nutrient wasobserved but no data were available on the influence of electriccurrent on soil microorganism. This is a serious flaw, since in situremediation technology significantly depends on soil microbialcommunity and activity. To date, only a few studies have examinedthe effect of EK on soil microbial communities (Lear et al., 2007) andmicrobial enzyme activity (Wang et al., 2009). However, in theseprevious studies, the direct effect of EK on soil microbial activity couldnot be ascertained and no study has been conducted to analyze thechanges in microbial community during EK remediation. Thus, little isknown about the effect of the EK remediation on the indigenous soilmicrobial ecosystem, and further study about the interaction betweenEK process and soil microbial community is needed to enhance thecombination effect of in situ EK remediation and biodegradation.

The aim of this study was to investigate the impact of electroki-netics on soil microbial activity and community. Changes in soilindigenousmicrobial activity and community structurewere analyzedafter EK remediation of diesel contaminated soil. The interactionsbetween soil condition changes due to EK process, and indigenousmicroorganisms were determined by various analytical methods andmolecular biology tools.

2. Materials and methods

2.1. Soil electrokinetics reactor and sampling

Soil was taken from a gas station in Busan, Korea and the propertiesare provided in Table 1. The soil was air-dried and sieved in a 2 mmmesh and remoistened to a water content of 19.8% by adding de-ionized water. Soil was stored in the dark at 4 °C until it was used forthe experiment.

For the EK process, soil was packed into an acrylic soil reactor(20 cm×4 cm×4 cm, Fig. 1) and pressurized. The anode chamber wasfilled with 0.005 M ethylenediaminetetra acetic acid (EDTA, pH 11.6)and the cathode chamber was filled with de-ionized water. Graphiteelectrode (DSA®, DOES., Co, Ltd, Korea) were used and directelectrical current of 0.63 mA cm−2 was applied by PowerPac 200(Bio-Rad, Hercules (CA), USA). Another soil reactor was similarlyconstructedwithout application of electrical current as a control. Afterthe process was completed, the soil in the reactor was divided into sixsections for analysis. Soil pH and moisture content were determinedat the final sampling time. Soil pH was measured using a pH probe(Thermo Fisher Scientific Inc., Waltham (MA), USA) and soil moisturewas calculated by drying soil at 105 °C to a constant dry weight.

In order to separate the effects of soil pH and electric current,additional control experiments were performed by varying the soil pHin the range of changed soil pH after EK process (pH 3.0–11.0). Batchstudies were conducted with 100 g of soil for 10 days and initial soilpH was adjusted with 1 N of HCl or NaOH.

Table 1Initial physical and chemical characteristics of soil.

Soil property

Sand (%) 90Silt (%) 3.7Clay (%) 6.3Water content (%, w/w) 2.37pH (H2O, 1:1) 7.07TOC (%) 3.07Total N (mg/kg) 67.88Total available P (mg/kg) 38.5CEC (cmolc/kg) 3.02Ca (cmolc/kg) 2.62Mg (cmolc/kg) 0.15K (cmolc/kg) 0.11Na (cmolc/kg) 0.04TPH (mg/kg) 6800

2.2. TPH analysis

Total petroleum hydrocarbons (TPH) from 3 g of soil wereextracted with n-hexane in a rotary shaker (25 °C, 40 rpm, 24 h).Hexane extracts from each sample were evaporated to constantweight and injected into a gas chromatograph (6890N Network GCSystem, Agilent Technologies, USA) equipped with a flame ionizationdetector (FID). The HP-5 capillary column was used with heliumcarrier gas at a flow rate of 1 ml min−1, hydrogen gas at 40 ml min−1

and air at 400 ml min−1. The temperature program used was50 °C2 min−1, 8 °C min−1 and 320 °C10 min−1.

2.3. Soil microbial number

Soil microbial cell numbers were measured by viable cell countingmethod and real-time PCR of 16S rRNA gene to enumerate bothculturable and total bacteria. For viable counting, 1 g soil was dilutedwith 9 ml of sterile distilled water. After 10-fold serial dilution, 30 μlwas spread onto Difco™ PCA (Becton, Dickinson and Company,France) medium and incubated at 30 °C. Colonies on the plates wereidentified by colony PCR of 16S rRNA gene using primer set of 9f (5′-GAGTTTGATCCTGGCTCAG-3′) and 1512r (5′-ACGGCTACCTTGTTAC-GACTT-3′). Total soil microbial genomic DNAwas extracted from 0.5 gsoil (dry weight) using FastDNA® Spin for Soil Kit (MP Biomedicals,LCC., Ohio, USA). Real-time PCR was performed using CFX96 Real-Time PCR Detection System (Bio-Rad, Hercules (CA), USA) and PrimeQ-Master Mix (Genetbio, Korea) with SYBR green. The standard curvefor total bacterial 16S rRNA genes was made using genomic DNAextracted from E. coli. 16S rRNA geneswere amplified using primer setof 9f and 536r (5′-GTATTACCGCGGCTGCTG-3′). The real-time PCRconditions were 5 min at 95 °C, 40cycles at 95 °C for 1 min, 55 °C for1 min, and 72 °C for 1 min.

2.4. Soil dehydrogenase assay

Soil dehydrogenase activity was estimated by monitoring the rateof reduction of 2,3,5-triphenyl tetrazolium chloride (INT) to iodoni-trotetrazolium formazan (INTF) and calculated as μgINTFg−1 soil asdescribed by Mathew and Obbard, (2001) with some modifications.5 g of soil wasmixedwith 2.5 ml of distilledwater and 1 ml of Tris/INTand incubated in an orbital incubator (100 rpm) at 40 °C for 2 h in thedark. After adding 10 ml of extraction solution (dimethylformamide:ethanol in a 1:1), INTF was measured on a spectrophotometer at464 nm in the dark.

2.5. Soil microbial community analysis

Microbial communities were examined by DGGE analysis of PCR-amplified 16S rRNA gene fragments from each soil as described byMuyzer et al. (1993). The 16S rRNA genes for DGGE were amplifiedusing primers 341f (5′-CCTACGGGAGGCAGCAG-3′) with attached GC-clamp and 536r (5′-GTATTACCGCGGCTGCTG-3′). The PCR productswere separated as follows: 8% polyacrylamide gels and denaturinggradient from 35% to 60% were used; gels were electrophoresed in 1×

Fig. 2. Soil pH (A) and moisture content (%) (B) after EK remediation.

3164 S.-H. Kim et al. / Science of the Total Environment 408 (2010) 3162–3168

TAE buffer at 60 °C and 60 V for 15 h using Bio-Rad DCode UniversalDetection Mutation system (Bio-Rad, Hercules, CA, USA).

Similarities between PCR-DGGE fingerprints were calculated usingthe Dice coefficient; unweighted pair-group method using arithmeticaverages (UPGMA) was used for cluster analysis. For prominent bandanalysis in DGGE, several bands from DGGE profiles were cut off andsequenced according to the following protocol: bands of interest inthe DGGE gel were excised with a cutter blade under UV illuminationand placed in 100 μl of distilled water. The DNA was extracted fromthe cut-out gel after 3h incubation at room temperature and thesupernatant was concentrated to 10 μl by ethanol precipitationmethod. Using 5 μl of the concentrated supernatant as the templateDNA, PCR was performed with primers 341f–536r. After purification,PCR products were cloned and sequenced. The closest matches of thesequences were identified through BLAST search in GenBankdatabase. The Shannon–Weaver index which is one of the diversityindices was used to measure soil microbial diversity (Shannon, 1997).It was calculated according to following equation.

H′ = −∑pi ln pi

H′ Shannon–Weaver indexpi (intensity of band i)/(total band intensity).

2.6. Data analysis

Each sample was analyzed in triplicate. Means and standarddeviation were calculated and plotted using SigmaPlot 10.0 (SystatSoftware Inc. USA).

3. Results

3.1. Influence of EK remediation on soil properties

Fig. 2 shows the change of soil pH and moisture content after25 days of EK remediation. The pH profile of the soil sampled fromdifferent column sections developed across the length of theelectrokinetic reactor from pH 3.5 near the anode to pH 10.8 nearthe cathode. The soil moisture contents were higher in EK remedi-ation than in control due to the supply of electrolyte. However, EKremediation showed the reduced soil moisture content compared tothe original soil, and the soil close to the cathode had higher moisturecontent than other soil, indicating the influence of electroosmosis.

In the hydrocarbon degradation analysis, the highest dieselremoval efficiency was observed in the soil near the anode in the EKremediation and over 60% of the light hydrocarbons (C10–C16) weredegraded near the anode (Table 2). The soil in themiddle and near thecathode showed high residual diesel concentration, indicating thatthe diesel was removed by electroosmosis. Preferential degradation ofn-alkanes with respect to isoprenoid compounds was indicated bydecrease in the n-C17/Pr (pristane) and n-C18/Ph (phytane) ratios. Thecalculated index values were reversely correlated with biodegrada-tion efficiencies; the values were 0.54 and 0.71 in the soil near theanode, and 1.26 and 1.19 in the soil near the cathode, respectively.

3.2. Microbial enumeration after EK remediation

Both culturable bacterial number and total bacterial number weremeasured using viable cell counting and real-time PCR. At the end of theEK remediation, the culturable bacterial number decreased significantly(data not shown). The highest number of culturable bacteria in the EKapplied soil was only 6105 CFU g−1 soil (in the middle of the reactor),whereas control showed about 2.5×107CFU g−1 soil. The culturablebacterial numbers near the anode and the cathode were 229 CFU g−1

soil and 48 CFU g−1 soil, respectively.

The 16S rRNA gene PCR of colonies in the EK soil revealed thatmost of the colonies were identified as belonging to Bacillus,Paenibacillus, and Brevibacillus with over 98% of sequence similarities.In the control, the most abundant culturable bacteria were Rhodo-coccus, Pseudomonas, and Streptomyces.

In the results from real-time PCR, the 16S rRNA gene copy numberswere almost constant within the control. However, the results fromEK remediation soil indicated significantly decreased 16S rRNA genecopy numbers compared to the control (Fig. 3A). The 16S rRNA genequantification by real-time PCR in EK remediationwas correlatedwiththe residual diesel concentration, indicating that the transport ofbacteria was affected by electroosmosis. The highest gene copynumbers were observed in the middle; the soil near the electrodesshowed low gene copy numbers. In general, the copy numbers of the16S rRNA gene in EK soils were in the order of 104–5copiesg−1 soil,whereas those of the control were about 108copiesg−1 soil.

3.3. Change of soil enzyme activity after EK remediation

The dehydrogenase activity in the soil was used to measure soilmicrobial enzyme activity (Fig. 3B). Before the experiment, theoriginal soil showed an enzyme activity of 12.63μgINTFg−1 soil,which was a slightly higher value than that of the control after theexperiment. In EK remediation, big differences in soil dehydrogenase

Table 2Degradation efficiency (%) of TPH fraction after EK remediation. a,bIndices of biodegradation of hydrocarbons which indicate preferential degradation of n-alkanes with respect toisoprenoid compounds; Pr, pristane; Ph, phytane.

TPHfraction

Normalized distance from anode Control

0 0.2 0.4 0.6 0.8 1

C10–C16 72.8±0.24 64.7±0.40 41.1±0.31 37.8±0.47 42.8±0.11 35.7±0.29 31.1±0.11C16–C20 57.2±0.13 51.2±0.08 35.4±0.18 39.6±0.24 41.4±0.40 32.4±0.02 35.4±0.27C20–C24 58.6±0.21 53.9±0.24 35.8±0.34 28.4±0.11 43.5±0.12 33.1±0.20 35.8±0.13C24–C28 56.5±0.39 45.4±0.22 31.8±0.15 32.4±0.27 37.4±0.02 25.2±0.17 26.8±0.08n-C17/Pra 0.54±0.05 0.65±0.09 1.12±0.01 1.21±0.02 0.98±0.01 1.26±0.00 1.15±0.06n-C18/Phb 0.71±0.17 0.75±0.03 1.20±0.00 0.95±0.08 1.05±0.03 1.19±0.12 1.05±0.02

3165S.-H. Kim et al. / Science of the Total Environment 408 (2010) 3162–3168

activity between EK applied soil samples were observed. The enzymeactivity near the cathode decreased compared to that of the controland the highest enzyme activity was shown in the middle of thereactor (19.13±0.79μgINTFg−1 soil). Interestingly, soil dehydroge-nase activity near the anode was enhanced after EK remediation,despite decreased bacterial number.

3.4. Effect of pH on soil microbial activities

In the pH adjusted experiments, the biodegradation efficiencies ofhydrocarbons were lower than those of EK applied soil and the control

Fig. 3. Total bacterial 16S rRNA gene copy number (A) and soil dehydrogenase assay(B) after EK remediation.

(soil without electric current) due to the short experimental period.However, the results indicated that the biodegradation of hydro-carbons was active under a neutral or slightly basic condition (datanot shown). At an initial soil pH of 7 and 8, the degradation efficienciesof light hydrocarbons (C10–C16) were 23.2% and 26.8%, respectively,whereas at extreme initial soil pH, the degradation efficiencies were16.0% (pH 3) and 18.9% (pH 11). The highest n-C17/Pr and n-C18/Phratios were observed at extreme initial soil pH. The n-C17/Pr and n-C18/Ph ratios were 1.52±0.02 and 1.60±0.07 at pH 3, and 1.20±0.02and 0.84±0.03 at pH 7, respectively.

The 16S rRNA gene quantification showed the highest gene copynumbers at soil pH 7 (1.11×109copiesg−1 soil) and the largestdecrease was seen at pH 3 (1.75×104copiesg−1 soil). In theidentification of colonies by 16S rRNA gene PCR, Bacillus relatedstrains were predominant in strong acidic and alkaline conditions,whereas Pseudomonas and Rhodococcus were abundant in an initialsoil pH of 5–8.

The results from a soil dehydrogenase assay were correlated withthose obtained fromdiesel degradation and bacterial enumeration. Soil ofinitial pH 8 showed slightly higher soil dehydrogenase activity (16.35±0.24μgINTFg−1 soil) than soil of pH 7 (15.26±0.71μgINTFg−1 soil). Soildehydrogenaseactivities at extremesoil pHwere11.04±0.68μgINTFg−1

soil (pH 3) and 11.45±0.5μgINTFg−1 soil (pH 11), respectively.

3.5. Microbial community analysis

DGGE analysis of the 16S rRNA gene was used to investigatedifferences of microbial communities between the EK impacted soilsand the controls (Fig. 4). The cluster analysis of the DGGE bandpattern showed that the samples grouped mainly according to pHdevelopment, indicating that pH is one of the main properties thataffect microbial community.

A decrease in prominent DGGE bands was observed in EK appliedsoil and soilwith low initial pH (pH3). Bands 1, 3 and 6were found in allsoil samples and bands 6 and 9 gradually became prominent in thesamples near the anode. Bands 4 and 5 were prominent only in C1, C2and soils of initial pH above 7, although thebandswere also consistentlyfound in all soil samples. Severalprominentbands inDGGEprofileswereselected for analysis of microbial communities in the soil samples(Table 3). The majority of the DGGE bands were identified as belongingto Bacillus. Band 3 was identified as Alcanivorax dieselolei strain S16-10(98%),which can assimilate diesel as a sole carbon source. Bands 6 and 9showed 98% and 100% 16S rRNA gene similarity to Chlostridium sp.HPB-1 and Pseudomonas sp. PNP4, respectively.

Analysis of the DGGE patterns with the Shannon–Weaver index ofdiversity (H′) revealed that the microbial communities of the EKimpacted soil had lower diversity than did the control soil. In theoriginal soil, the diversity index (H′) was 3.39 and the diversityindices of the control and soil of initial pH 7 (3.26) were similar to thatof the original. The most reduced microbial diversity was observed inthe soil near the anode in the EK process (2.47) and soil of initial pH 3(2.48).

Fig. 4. Microbial community analysis by PCR-DGGE and clustering. Some prominent DGGE bands indicated by arrows and numbers were identified (Table 3). Numbers (0.0–1.0),normalized distance from anode; C1 and C2, control soil before and after experiment; pH 3–pH 11, initial soil pH in the pH adjusted soil experiment; H′, Shannon–Weaver index ofmicrobial diversity.

3166 S.-H. Kim et al. / Science of the Total Environment 408 (2010) 3162–3168

4. Discussion

To our knowledge, this is the first in-depth study aimed at in-vestigating the effect of EK on indigenous soil microbial activity andcommunity structure in the presence of an organic pollutant (diesel).Although EK remediation was applied to remove contaminant fromsoil, the process altered soil parameters, especially soil pH, which iscritical to microbial activity. In the present study, the possiblemechanisms of the impact of electrokinetics on soil microbial activityand community were (1) change of soil properties that are crucialfor microbial metabolism; (2) effect of electric current on microbialactivity.

4.1. Effect of the change in soil properties after EK remediation

Among the various soil parameters, the most apparent changeinduced by EK remediation was observed in soil pH. The soil pH iscrucial for microbial activity and it affects microbial cell membraneintegrity and function, and the bioavailability of nutrients andcontaminants. Many previous studies have revealed that the soil pHhad marked effects on microbial biomass, community structure, andresponse to substrate addition, and that low soil pH decreasedmicrobial diversity and increased Gram-positive microbial commu-nities (Lear et al., 2004; Pietri and Brookes, 2009).

Similarly, in the present study, significant changes in microbialcommunity and activity were observed after EK remediation. The

Table 3Sequence analysis of several prominent bands in DGGE profiles.

Band Related bacterial sequence (accession no.) Similarity (%)

1 Bacillus sp. JZDN43 (DQ659022) 982 Rhodococcus sp. (AJ585370) 993 Alcanivorax dieselolei strain S16-10 (FJ218297) 984 Clostridium sp. HPB-1 (AY862515) 975 Bacillus sp. MN39 (DQ336203) 946 Clostridium sp. HPB-1 (AY862515) 987 Pseudomonas argentinensis strain PA01 (AY691189.2) 998 Pseudomonas sp. HPC 919 (AY956955.1) 939 Pseudomonas sp. PNP4 (DQ282190) 10010 Uncultured bacterium partial 16S rRNA gene, isolate

DGGE band 12 (AM232773)99

11 Bacillus sp. YT0042 (AB362831.1) 9512 Bacillus sp. M8-C23 (FJ763930.1) 9113 Bacillus sp. HZBN43 (EF625229.1) 99

bacterial number, DGGE band patterns, and the Shannon–Weaverindex showed the most reduced values in the soils of extreme pH,including those near the electrodes, and highest values in the soil ofneutral pH. This indicates that the pH changes generated by EKremediation was a leading cause of decreased microbial number anddiversity. However, EK applied soil showedmore apparent decrease inculturable bacteria compared to the soil samples of extreme initial pH.

EDTA is a complexing agent that is often used to enhance thetransport of contaminants in EK remediation (Nogueira et al., 2007;Popov et al., 1999). Because the contaminant (diesel) in this study iswater-immiscible, its removal mechanism by electrokinetics mainlydepends on the electroosmosis. Therefore, it is supposed that the useof EDTA as an electrolyte enhanced the removal efficiency of thecontaminant by increasing the electroosmotic flow. However, itseemed that the high alkalinity and toxicity of EDTA causeddetrimental effects on the soil microbial activity and community,especially in the soil near the cathode where the EDTA wasaccumulated due to the electroosmosis. Epelde et al. (2008) observeddecrease in soil microbial respiration, dehydrogenase activity, andmicrobial diversity in Pb phytoextraction with EDTA. The significantdecrease in culturable bacterial number and soil dehydrogenaseactivity near the cathode after EK process in this study, also suggestthe negative impact of EDTA as described in the previous studies(Epelde et al., 2008; Kos and Lestan, 2003).

It is reported that stress from the growth conditions reduces thetotal bacterial number and is a reason for cells entering a viable butnon-culturable state (Ibekwe and Grieve, 2004; Ohtomo and Saito,2001). Although electric current in EK remediation can changebacterial membrane composition and metabolic activity, manystudies revealed that weak DC treatment has no negative effect onmicrobial viability and activity (Lohner and Tiehm, 2009; Shi et al.,2008c; Tiehm et al., 2009). Thus, it is assumed that the decreasedbacterial culturability was mainly due to the changes of soil pH andadverse effect of EDTA.

4.2. Effect of the direct electric current

Along with the changes in soil pH, the applied direct electriccurrent also caused a detectable impact on soil microbial activity andcommunity in this study. The positive correlation between the resultsfrom real-time PCR and diesel analysis indicated that most of thebacteria were transported by electroosmosis (Shi et al., 2008b; Wicket al., 2004). The enhanced microbial activities by EK remediation

3167S.-H. Kim et al. / Science of the Total Environment 408 (2010) 3162–3168

were detected in the DGGE profiles (bands 6 and 9 in Fig. 4), n-C17/Prand n-C18/Ph ratios (Table 2) and soil dehydrogenase assay (Fig. 3B).

The application of electric current suggests a new approach tostimulate microbial activity. Much attention has been paid toinvestigate the effect of electric current on the bacterial metabolismin recent years. In the presence of electric current, substrate utilizationincreased and microbial metabolism was improved by direct (transferof electrons from electrode to bacteria) and indirect (transfer ofelectrons through hydrolysis of water) stimulation (Thrash andCoates, 2008). Aulenta et al. (2008) observed enhanced dechlorina-tion of trichloroethene (TCE) by bioelectrocatalytic H2 production.Lohner and Tiehm (2009) demonstrated electrolytically stimulatedmicrobial perchloroethene (PCE) reduction and vinyl chloride (VC)oxidation with high process efficiencies. She et al. (2006) foundstimulating effect of a weak DC electric field (10 mA) on cell growth,glucose uptake and dehydrogenase activity, and water electrolysiswas suggested as a probable explanation for the stimulating effect ofDC. Lear et al. (2004) found the increase in the microbial utilization ofthe simpler compound and concluded that this might be a reflection ofbacterial requirement for an immediate source of energy to maintaintheir system.

In this study, soil microbial activity was also stimulated by a weakelectric field. Because decrease of TPH could occur by non-biologicalprocesses, changes in hydrocarbon composition that are indicative ofbiodegradation must be measured. The n-C17/Pr and n-C18/Ph ratioshave been used as good indicators of hydrocarbon biodegradation inwater, groundwater and soil environment (Atlas, 1981; Nievas et al.,2005; Pritchard and Costa, 1991). The indices value decreasedparticularly in the soil near the anode, whereas the most decreasedvalue was observed at neutral condition in pH adjusted soilexperiments. This indicates that the biodegradation activity near theanode was stimulated by EK process, in spite of low soil pH.

Although total bacterial number was lowered by EK remediation,soil dehydrogenase activity was enhanced except the soil near thecathode. Soil dehydrogenase activity is measured as an importantindex of the overall microbial activity (Carcia et al., 1994; Kim et al.,1994). Dehydrogenase enzyme is known to oxidize soil organicmatter by transferring protons and electrons and these processes arepart of respiration pathways of soil microorganisms. Therefore,studies on the activities of dehydrogenase enzyme in the soil canassess, indirectly, the oxygen or redox level of soil. Since decrease inbacterial number was observed in this study, the increased soildehydrogenase activity can be explained by enhanced cell respiration.She et al. (2006) suggested that bacterial cell respiration might havebeen enhanced by the generation of anodic oxygen by electrolysis andseveral studies reported a positive correlation between dehydroge-nase activity and oxygen uptake rate (OUR) (Awong et al., 1985;Caravelli et al., 2006; Kim et al., 1994). Based on the previous studies,it is thought that the electrolysis of water in the anode elevated levelsof dissolved oxygen (Jass et al., 1995) thus, increased delivery ofoxygen by electroosmosis and enhanced oxygen uptake rate (OUR) ofsoil microorganisms. The increase in OUR could have stimulatedbiodegradation of hydrocarbons (Atlas, 1981) and increased soildehydrogenase activity. Moreover, harsh environmental conditioninduced by EK process might have enhanced soil microbial activity tomaintain bacterial survival and cellular system with increasedutilization efficiency in response to the stress. It is also possible thatthe electrophoretic transport of negatively charged mineral particletowards the anode stimulated the metabolism of acidophilic bacteria(Jackman et al., 1999).

4.3. Strategies for the enhanced EK-bioremediation

Several studies have reported stimulation of bioremediation byelectric current. The pH was controlled by electrolyte circulation (Kimet al., 2005), mixing of the anolyte and catholyte (Rabbi et al., 2000),

and buffer solution (Niqui-Arroyo et al., 2006). The effect of directelectric current on microbial activity and viability was investigated(Cang et al., 2009; Luo et al., 2005; Shi et al., 2007; Tiehm et al., 2009).However, when electric current is applied, different bacterialresponses to changes in the physico-chemical properties, bioavail-ability, and toxic electrode-effect can be observed depending on theamperage, treatment period, cell type, and medium (Wick et al.,2007). Therefore for the successful in situ EK-bioremediation,investigating the characteristics of the contaminated site includingindigenous microbial community is indispensable to achieve theoptimal biodegradation activity without any unintended negativeeffect by EK process. Previous studies have revealed that EK processincreased the number of Bacillus and Arthrobacter (Lear et al., 2004)and that bioelectrical reactors enhanced the metabolism of severalstrains in Clostridium, Ralstonia, Pseudomonas, and Brevibacterium(Thrash and Coates, 2008). The results from the current study alsosupport these findings, suggesting that bacteria with the ability totolerate environmental stress can be useful in EK-bioremediation.

Electrode and electrolyte are other important parameters in EK-bioremediation. For instance, Tiehm et al. (2009) observed inhibitioneffects on vinyl chloride (VC) degrading microbial viability afterexposure to the electric currents (14 mA cm−2) and these inhibitioneffects were more prominent when DSA electrodes were used.Separation of electrodes from bacteria by a bipolar membrane didnot cause any inhibition effect. In the present study, 0.005 M of EDTAwas used as an electrolyte. Although EDTA is effective to enhance theremoval of hydrophobic organic compounds (HOCs), it has negativeeffect on soil microbial communities. Other chelating agents such ascitric acid and EDDS (ethylene diamine disuccinate) have beenreported to be effective in EK remediation and proved to be lesstoxic to the soil microbial community (Epelde et al., 2008; Popov et al.,1999). It is also possible to introduce nutrients by electrokinetics withelectrolyte (Schmidt et al., 2007) and the addition of lower alcohol asa co-solvent may enhance the biodegradation activity by promotingco-metabolism of contaminants (Boulding, 1996).

5. Conclusion

The results in this study demonstrate the impact of EK remediationon the microbial activity and community structure within dieselcontaminated soil. The EK remediation with the addition of EDTAeffectively removed diesel from the soil by electroosmosis, but EDTAhad detrimental effect on soil microbial activity near the cathode. Themain factors that caused significant changes in microbial activity weresoil pH and direct electric current. It is assumed that the changed soilpH induced by EK remediation decreased soil microbial number anddiversity but the direct electric current increased biodegradation ofhydrocarbons and soil enzyme activity. This implies that electroki-netic remediation could be promising when combined with biore-mediation by increasing microbial activity. To achieve a successfulcombination of electrokinetics and bioremediation, soil parameters,electric current, electrode, and electrolyte should be carefullyconsidered. Thus, the remediation should be conducted to maximizeboth the EK process and biodegradation and to minimize unintendedeffects on the soil condition and soil ecosystem. However, only a fewstudies have investigated the impact of electrokinetics on soilecosystem until now. Therefore, better understanding of interactionsbetween electrokinetics and indigenous microbial community areneeded, and this will enable us to improve the in situ EK remediationtechnology and sustain a healthy environment.

Acknowledgement

This research was supported by a grant from Korea Ministry ofEnvironment as “GAIA project”.

3168 S.-H. Kim et al. / Science of the Total Environment 408 (2010) 3162–3168

References

Acar YB, Alshawabkeh AN. Principles of electrokinetic remediation. Environ Sci Technol1993;13:2638–47.

Atlas RM. Microbial-degradation of petroleum-hydrocarbons — an environmentalperspective. Microbiol Rev 1981;1:180–209.

Aulenta F, Canosa A, Majone M, Panero S, Reale P, Rossetti S. Trichloroethenedechlorination and H2 evolution are alternative biological pathways of electriccharge utilization by a dechlorinating culture in a bioelectrochemical system.Eviron Sci Technol 2008;16:6185–90.

Awong J, Bitton G, Koopman B. Atp, oxygen-uptake rate and Int-dehydrogenase activityof actinomycete foams. Water Res 1985;7:917–21.

Boulding JR. EPA environmental engineering sourcebook. Chelsea, Michigan: Ann ArborPress Inc; 1996. p. 76.

Cang L, Zhou DM,Wang QY,Wu DY. Effects of electrokinetic treatment of a heavy metalcontaminated soil on soil enzyme activities. J Hazard Mater 2009;2–3:1602–7.

Caravelli A, Giannuzzi L, Zaritzky N. Effect of ozone on filamentous bulking in alaboratory scale activated sludge reactor using respirometry and INT-dehydroge-nase activity. J Environ Eng-Asce 2006;9:1001–10.

Carcia C, Hernandez T, Costa F. Microbial activity in soils under mediterraneanenvironmental conditions. Soil Biol Biochem 1994;9:1185–91.

Chen XJ, Shen ZM, Lei YM, Zheng SS, Ju BX, Wang WH. Effects of electrokinetics onbioavailability of soil nutrients. Soil Sci 2006;8:638–47.

DeFlaun MF, Condee CW. Electrokinetic transport of bacteria. J Hazard Mater 1997;1–3:263–77.

Epelde L, Hernandez-Allica J, Becerril JM, Blanco F, Garbisu C. Effects of chelates onplants and soil microbial community: comparison of EDTA and EDDS for leadphytoextraction. Sci Total Environ 2008;1–3:21–8.

Ibekwe AM, Grieve CM. Changes in developing plant microbial community structure asaffected by contaminated water. FEMS Microbiol Ecol 2004;2:239–48.

Jackman SA, Maini G, Sharman AK, Knowles CJ. The effects of direct electric current onthe viability and metabolism of acidophilic bacteria. Enzyme Microb Tech 1999;5–6:316–24.

Jass J, Costerton JW, Lappinscott HM. The effect of electrical currents and tobramycin onPseudomonas-aeroginosa biofilms. J Ind Microbiol 1995;3:234–42.

Kim CW, Koopman B, Bitton G. Int-dehydrogenase activity test for assessing chlorineand hydrogen-peroxide inhibition of filamentous pure cultures and activated-sludge. Water Res 1994;5:1117–21.

Kim SJ, Park JY, Lee YJ, Lee JY, Yang JW. Application of a new electrolyte circulationmethod for the ex situ electrokinetic bioremediation of a laboratory-preparedpentadecane contaminated kaolinite. J Hazard Mater 2005;1–3:171–6.

Kim DH, Ryu BG, Park SW, Seo CI, Baek K. Electrokinetic remediation of Zn and Ni-contaminated soil. J Hazard Mater 2009;1–3:501–5.

Ko S-O. Electrokinetic removal of hydrophobic organic contaminant in fine soils. KoreanSoc Environ Eng 1999;3:231–8.

Kos B, Lestan D. Influence of a biodegradable ([S, S]-EDDS) and nondegradable (EDTA)chelate and hydrogel modified soil water sorption capacity on Pb phytoextractionand leaching. Plant Soil 2003;2:403–11.

Lear G, Harbottle MJ, van der Gast CJ, Jackman SA, Knowles CJ, Sills G, Thompson IP. Theeffect of electrokinetics on soil microbial communities. Soil Biol Biochem 2004;11:1751–60.

Lear G, Harbottle MJ, Sills G, Knowles CJ, Semple KT, Thompson IP. Impact ofelectrokinetic remediation on microbial communities within PCP contaminatedsoil. Environ Pollut 2007;1:139–46.

Lohner ST, Tiehm A. Application of electrolysis to stimulate microbial reductive PCEdechlorination and oxidative VC biodegradation. Environ Sci Technol 2009;18:7098–104.

Luo QS, Wang H, Zhang XH, Qian Y. Effect of direct electric current on the cell surfaceproperties of phenol-degrading bacteria. Appl Environ Microb 2005;1:423–7.

Maini G, Sharman AK, Knowles CJ, Sunderland G, Jackman SA. Electrokineticremediation of metals and organics from historically contaminated soil. J ChemTechnol Biot 2000;8:657–64.

Mathew M, Obbard JP. Optimisation of the dehydrogenase assay for measurement ofindigenous microbial activity in beach sediments contaminated with petroleum.Biotechnol Lett 2001;3:227–30.

Muyzer M, Dewaal EC, Uitterlinden AG. Profilling of complex microbial populations bydenaturing gradient gel electrophoresis analysis of polymerase chain reactionamplified genes coding for 16s ribosomal RNA. Appl EnvironMicrob 1993;3:695–700.

Nievas ML, Commendatore MG, Esteves JL, Bucala V. Effect of pH modification on bilgewaste biodegradation by a native. Int Biodeter Biodegr 2005;3:151–7.

Niqui-Arroyo JL, Bueno-Montes M, Posada-Baquero R, Ortega-Calvo JJ. Electrokineticenhancement of phenanthrene biodegradation in creosote-polluted clay soil.Environ Pollut 2006;2:326–32.

Nogueira MG, Pazos M, Sanroman MA, Cameselle C. Improving on electrokineticremediation in spiked Mn kaolinite by addition of complexing agents. ElectrochimActa 2007;10:3349–54.

Ohtomo R, Saito M. Increase in the culturable cell number of Escherichia coli duringrecovery from saline stress: possible implication for resuscitation from the VBNCstate. Microb Ecol 2001;2:208–14.

Park JY, Lee HH, Kim SJ, Lee YJ, Yang JW. Surfactant-enhanced electrokinetic removal ofphenanthrene from kaolinite. J Hazard Mater 2007;1–2:230–6.

Pietri JCA, Brookes PC. Substrate inputs and pH as factors controlling microbial biomass,activity and community structure in an arable soil. Soil Biol Biochem2009;7:1396–405.

Popov K, Yachmenev V, Kolosov A, Shabanova N. Effect of soil electroosmotic flowenhancement by chelating reagents. Colloid Surf A 1999;2:135–40.

Pritchard PH, Costa CF. Epas Alaska oil-spill bioremediation project. Environ Sci Technol1991;3:372–9.

Rabbi MF, Clark B, Gale RJ, Ozsu-Acar E, Pardue J, Jackson A. In situ TCE bioremediationstudy using electrokinetic cometabolite injection. Waste Manage 2000;4:279–86.

Saichek RE, Reddy KR. Effect of pH control at the anode for the electrokinetic removal ofphenanthrene from kaolin soil. Chemosphere 2003;4:273–87.

Schmidt CAB, Barbosa MC, de Almeida MDS. A laboratory feasibility study onelectrokinetic injection of nutrients on an organic, tropical, clayey soil. J HazardMater 2007;3:655–61.

Shannon CE. The mathematical theory of communication (Reprinted). MD Comput1997;4:306–17.

She P, Bo S, Xing XH, van Loosdrecht M, Liu Z. Electrolytic stimulation of bacteriaEnterobacter dissolvens by a direct current. Biochem Eng J 2006;1:23–9.

Shi L, Wick LY, Gunther S, Harms H, Muller S. Flow cytometry analysis of the viability ofpollutant degrading bacteria exposed to weak electric fields. Cytom Part A 2007;1:52–3.

Shi L, Muller S, Harms H, Wick LY. Effect of electrokinetic transport on the vulnerabilityof PAH-degrading bacteria in a model aquifer. Environ Geochem Health 2008a;2:177–82.

Shi L, Muller S, Harms H, Wick LY. Factors influencing the electrokinetic dispersion ofPAH-degrading bacteria in a laboratory model aquifer. Appl Microbiol Biot2008b;3:507–15.

Shi L, Muller S, Loffhagen N, Harms H, Wick LY. Activity and viability of polycyclicaromatic hydrocarbon-degrading Sphingomonas sp. LB126 in a DC-electrical fieldtypical for electrobioremediation measures. Microb Biotechnol 2008c;1:53–61.

Thrash JC, Coates JD. Review: direct and indirect electrical stimulation of microbialmetabolism. Environ Sci Technol 2008;11:3921–31.

Tiehm A, Lohner ST, Augenstein T. Effects of direct electric current and electrodereactions on vinyl chloride degrading microorganisms. Electrochim Acta 2009;12:3453–9.

Wang QY, Zhou DM, Cang L, Li LZ, Wang P. Solid/solution Cu fractionations/speciation ofa Cu contaminated soil after pilot-scale electrokinetic remediation and theirrelationships with soil microbial and enzyme activities. Environ Pollut 2009;8–9:2203–8.

Wick LY, Mattle PA, Wattiau P, Harms H. Electrokinetic transport of PAH-degradingbacteria in model aquifers and soil. Environ Sci Technol 2004;17:4596–602.

Wick LY, Shi L, Harms H. Electro-bioremediation of hydrophobic organic soil-contaminants: a review of fundamental interactions. Electrochim Acta 2007;10:3441–8.