Improvement of pesticide mineralization in on-farm biopurification systems by bioaugmentation with pesticide-primed soil

Page 1

R E S E A R C H A R T I C L E

Improvementof pesticidemineralization inon-farmbiopuri¢cationsystemsbybioaugmentationwith pesticide-primed soilKristel Sniegowski1, Karolien Bers1, Kris Van Goetem1, Jaak Ryckeboer1, Peter Jaeken2, PieterSpanoghe3 & Dirk Springael1

1Division of Soil and Water management, Katholieke Universiteit Leuven, Heverlee, Belgium; 2Ecology Department, Proefcentrum Fruitteelt vzw,

Sint-Truiden, Belgium; and 3Laboratory of Crop Protection Chemistry, Ghent University, Ghent, Belgium

Correspondence: Kristel Sniegowski,

Division of Soil and Water management,

Katholieke Universiteit Leuven, Kasteelpark

Arenberg 20, 3001 Heverlee, Belgium.

Tel.: 132 16 32 16 04; fax: 132 16 19 97;

e-mail: [email protected]

Present address: Peter Jaeken, Phytofar, A.

Reyerslaan 80, 1030 Brussels, Belgium.

Received 28 June 2010; revised 3 December

2010; accepted 3 December 2010.

Final version published online 13 January 2011.

DOI:10.1111/j.1574-6941.2010.01031.x

Editor: Kornelia Smalla

Keywords

biopurification system; pesticide-primed soil;

bioaugmentation; linuron mineralization;

Variovorax.

Abstract

Microcosms were used to examine whether pesticide-primed soils could be

preferentially used over nonprimed soils for bioaugmentation of on-farm biopuri-

fication systems (BPS) to improve pesticide mineralization. Microcosms contain-

ing a mixture of peat, straw and either linuron-primed soil or nonprimed soil were

irrigated with clean or linuron-contaminated water. The lag time of linuron

mineralization, recorded for microcosm samples, was indicative of the dynamics of

the linuron-mineralizing biomass in the system. Bioaugmentation with linuron-

primed soil immediately resulted in the establishment of a linuron-mineralizing

capacity, which increased in size when fed with the pesticide. Also, microcosms

containing nonprimed soil developed a linuron-mineralizing population, but after

extended linuron feeding. Additional experiments showed that linuron-miner-

alization only developed with some nonprimed soils. Concomitant with the

increase in linuron degradation capacity, targeted PCR-denaturing gradient gel

electrophoresis showed the proliferation of a Variovorax phylotype related to the

linuron-degrading Variovorax sp. SRS16 in microcosms containing linuron-

primed soil, suggesting the involvement of Variovorax in linuron degradation.

The correlation between the appearance of specific Variovorax phylotypes and

linuron mineralization capacity was less clear in microcosms containing non-

primed soil. The data indicate that supplementation of pesticide-primed soil

results in the establishment of pesticide-mineralizing populations in a BPS matrix

with more certainty and more rapidly than the addition of nonprimed soil.

Introduction

Since 1940, pesticides are intensively used worldwide. An

important environmental issue of pesticide use is the pollu-

tion of ground and surface water as a result of either diffuse

(run-off, percolation and spray drift) or point contamina-

tion (direct losses through spillage and leakages). Recent

studies showed that direct losses account for 40–90% of the

surface water pollution (De Wilde et al., 2007). To minimize

direct pesticide losses, the installation of biopurification

systems (BPS) to treat pesticide-contaminated wastewater

on the farm yard has been proposed (Torstensson & del Pilar

Castillo, 1997; Vischetti et al., 2004; De Wilde et al., 2007).

In on-farm BPS, the contaminated water is conducted over

a solid matrix, called a biomix, which is composed of a

mixture of various materials, for example straw, peat and

soil, in which biodegradation and sorption result in pesti-

cide removal. BPS are considered a simple, low-cost, prac-

tical and labor-extensive approach for farmers to treat

pesticide-contaminated wastewater on the farm. Despite

the high pesticide removal percentage observed in BPS

(Fogg et al., 2003a, b, 2004; Pigeon et al., 2005), degradation

remains poor for some pesticides (Fogg et al., 2003a, 2004).

Moreover, a rapid complete degradation is advised to avoid

possible toxicity effects of accumulated contaminants (Hen-

riksen et al., 2003), aging (Johannesen et al., 2003; Zhao

et al., 2003) and the occurrence of mobile and toxic

metabolites (Coppola et al., 2007). In addition, degradation

during start-up of the system is poor (Fogg et al., 2004)

because the appropriate microorganisms need to proliferate

FEMS Microbiol Ecol 76 (2011) 64–73c� 2011 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

MIC

ROBI

OLO

GY

EC

OLO

GY

Page 2

in the biomix before maximum degradation rates are

obtained. To ensure rapid and complete degradation of the

pesticides in BPS, bioaugmentation of the biomix with

microorganisms containing catabolic pathways enabling

complete mineralization of pesticides is suggested. Pesti-

cide-mineralizing microorganisms can often be found in

soils or other ecosystems with a long history of pesticide

contamination, designated as pesticide-primed materials.

The organisms can be inoculated either as (formulated)

cultured strains or along with the pesticide-primed material

(e.g. soil) in which they have developed. In comparison with

pure cultures, the latter is expected to contain (1) a larger

gene pool and higher diversity in microorganisms that

contribute to pesticide degradation and (2) populations

better adapted to in situ conditions. In addition, this

approach does not require the isolation of the appropriate

organisms. This is important since, to date, pure strains able

to mineralize a pesticide have been reported only for a few

pesticide compounds. Moreover, bioaugmentation using

lab-cultured pollutant-degrading isolates often had limited

success (Chatterjee et al., 1982; Grigg et al., 1997; Struthers

et al., 1998; Topp, 2001; Mertens et al., 2006; Moran et al.,

2006; Singh et al., 2006; Bazot & Lebeau, 2008). On the other

hand, the few reports on bioaugmentation of contaminated

matrices with pollutant-primed materials showed promising

results (Barbeau et al., 1997; Runes et al., 2001; Grundmann

et al., 2007). Despite the apparent high potential of applying

primed materials for bioaugmentation of contaminated

ecosystems, little research has been performed on this topic.

Moreover, no reports exist of using pesticide-primed mate-

rials for bioaugmenting BPS despite the low associated cost

of such an approach for farmers. However, BPS contain a

complex matrix prepared of components from different

ecosystems harboring different microbial communities and

provide a high nutrient content with multiple substrates for

microbial growth. It has to be examined whether pesticide-

mineralizing populations introduced by inoculating pesti-

cide-primed materials can compete and proliferate in such a

complex biotechnological matrix.

Therefore, in this study, the dynamics of the pesticide

mineralization capacity of lab-scale BPS microcosms (BM),

containing a mixture of soil, straw and peat, were compared

when inoculated with either a pesticide-primed soil posses-

sing a pesticide mineralization capacity or a nonprimed soil

without apparent pesticide mineralization capacity and

irrigated or not with a pesticide-containing solution. The

phenylurea herbicide linuron was used as model pesticide

and soil originating from an agricultural field that had been

treated annually with linuron and that contains linuron-

mineralizing organisms (Breugelmans et al., 2007) was used

as the model pesticide-primed material. The linuron miner-

alization capacity of the BMs was monitored by means of14C-linuron mineralization assays. Because linuron-degrad-

ing isolates originating from linuron-treated soils, including

the linuron-primed soil used in this study, almost exclu-

sively belong to the genus Variovorax (Dejonghe et al., 2003;

Sorensen et al., 2005; Breugelmans et al., 2007), the number

of Variovorax and composition of the Variovorax commu-

nity within the BMs was monitored by means of targeted

molecular techniques.

Materials and methods

Pesticides used

Linuron [3-(3,4-dichlorophenyl)-1-methoxy-1-methyl urea]

(purity, 99.5%) was purchased from Sigma Aldrich (Belgium).

[phenyl-U-14C] linuron (16.93 mCi mmol�1, radiochemical

purity 4 95%) was obtained from Izotop, Hungary.

BM set-ups

BM were set up in glass cylinders (height 10 cm; diameter

4 cm) containing a glass filter positioned at 8 cm depth and

filled with the appropriate mixture of soil, peat and straw

(Table 1). The linuron-primed soil (soil L) was sampled

from the A-horizon of a potato field in Halen, Belgium, in

April 2005. The field had been treated for several years with

linuron and was shown to contain a linuron-mineralizing

microbial community (Breugelmans et al., 2007). Non-linuron-

primed soils were obtained from seven different locations

representing four different ecosystems, i.e. garden (soils G1 and

Table 1. Overview of the different BM set-ups operated in this study

Set-up� Origin soil

Linuron-

primed soil

Linuron

treatment

Moisture

content

(w/w%) pH (� SD)

Experiment 1: mixture: straw (25 vol%); peat (25 vol%); soil (50 vol%)

L� Agriculture 1 � 56.12 6.20 (� 0.12)

L1 Agriculture 1 1 60.52 6.20 (� 0.12)

C� Construction

site

� � 48.00 6.05 (� 0.22)

C1 Construction

site

� 1 48.14 6.05 (� 0.22)

Experiment 2: mixture: straw (37.5 vol%); peat (37.5 vol%);

soil (25 vol%)

O1w

No soil � 1 217.78 5.11 (� 0.10)

L1 Agriculture 1 1 89.52 5.44 (� 0.26)

C1 Construction

site

� 1 71.03 5.49 (� 0.21)

A11 Agriculture � 1 42.67 5.56 (� 0.55)

A21 Agriculture � 1 34.64 5.09 (� 0.15)

F11 Forest � 1 82.33 5.61 (� 0.10)

F21 Forest � 1 57.54 4.87 (� 0.11)

G11 Garden � 1 74.19 5.43 (� 0.05)

G21 Garden � 1 42.25 5.38 (� 0.30)

�1 and � indicate treatment with and without linuron, respectively.wStraw: 50 vol%, peat: 50 vol%, soil: 0 vol%.

FEMS Microbiol Ecol 76 (2011) 64–73 c� 2011 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

65Pesticide-primed soil to improve on-farm biopurification

Page 3

G2), agriculture field (soils A1 and A2), forest (soils F1 and F2)

and a construction site (soil C). To the best of our knowledge,

these soils were never treated with linuron and, in addition, did

not show any capacity to mineralize linuron for a period of 70

days in 14C-linuron mineralization assays (data not shown). All

soils were top soils sampled from the upper 20 cm. The soils

were stored at 4 1C in the dark and sieved before use. After

filling, the BMs were positioned on a dish to collect drainage

water and placed in a glass jar closed with an air-open lid to

avoid desiccation and contamination. No leaching of linuron

was observed.

Two BM experiments were set up as outlined in Table 1,

which provides an overview of the compositions, moisture

contents and pHs of the biomix in the various BM set-ups. The

moisture content, based on the weight of a sample (� 0.500 g)

taken from the upper layer of the BMs before and after 2 days

incubation at 60 1C, was determined at each sampling time.

The pH was measured at the start and the end of both

experiments in a 0.01 M CaCl2 extract (soil : liquid ratio 1 : 5).

The pHs determined at the start of the experiments did not

differ significantly within each set-up. In addition, neither pH

nor moisture content changed significantly during the incuba-

tion period (data not shown).

In the first experiment, BMs contained a mixture of

25 vol% cut straw (� 0.5 cm2), 25 vol% peat and 50 vol% soil.

The physico-chemical characteristics of the used substrata are

shown in Table 2. BMs of set-ups L� and L1 were inoculated

with the linuron-primed soil L and packed to a density of

0.94 g cm�3, while BMs of set-ups C� and C1 were inoculated

with nonprimed soil C packed to a density of 0.78 g cm�3.

BMs of set-ups C� and L�were irrigated with sterile tap water,

while BMs of set-ups C1 and L1 received sterile tap water

containing 60 mg L�1 linuron. Both solutions were manually

spread evenly over the surface of the matrix using a 1-mL

pipette during 12 weeks. Each week, the solutions were applied

on Monday (1 mL), Wednesday (1 mL) and Friday (1.5 mL),

resulting in an average added volume of 3.18 L m�3 day�1.

Each set-up included triplicate BMs. All BMs were incubated

in the dark at 25 1C. The upper 1 cm of the matrix in the BMs

was mixed with a sterile spatula, before taking samples to

examine the linuron mineralization capacity. The samples

were taken exactly 0, 2, 4, 8 and 12 weeks after starting the

treatments. Additional samples were taken at weeks 0, 2 and

12 for analysis of the Variovorax community.

In the second experiment, BMs contained eight different

mixtures consisting of 37.5 vol% cut straw, 37.5 vol% peat and

25 vol% of soil and were operated in triplicate. The straw and

peat used were the same as those used in the first experiment.

BMs of set-up L1 were inoculated with the linuron-primed

soil, while BMs of the other set-ups were inoculated with

either one of the seven nonprimed soils (G1, G2, A1, A2, F1, F2,

and C). As a control, a ninth set-up was included in which the

BMs contained a mixture of 50 vol% straw and 50 vol% peat

only. Each of these BM set-ups received a linuron-containing

solution for a period of 20 weeks as described above. Samples

were taken after 0, 15 and 20 weeks of incubation at 25 1C for

determination of the 14C-linuron mineralization capacity as

described for the first set-up. Additional samples were taken at

weeks 0 and 15 for analysis of the Variovorax community.

14C-linuron mineralization assays14C-linuron mineralization assays were performed as de-

scribed by Breugelmans et al. (2007). The matrix sample

(200 mg) was added to a volume of 5 mL MMN minimal

medium (Breugelmans et al., 2007), pH 5.8, containing both

unlabeled (20 mg L�1) and 14C-labeled (31mg L�1) linuron

as the only carbon and nitrogen source (a final radioactivity

of 213 Bq mL�1) in 20-mL pyrex tubes containing NaOH

traps. During incubation at 20 1C on a rotary shaker at

150 r.p.m., the amount of 14CO2 produced with reference to

the initial added amount of 14C-linuron was measured and

cumulative mineralization curves were established. As a

negative control, 14CO2 production of the medium without

inoculation of a biomix sample was monitored. The lag

phase, defined as the period between initiating the miner-

alization assay and start of mineralization, was calculated as

the intersection of the x-axis with the linear regression line

between two successive points of the mineralization curve,

where the amount of 14CO2 showed the largest increase. The

slope of the linear regression line of maximum mineraliza-

tion determined the maximum mineralization rate (Broos

et al., 2005).

Most probable number (MPN) mineralizationmethod

An estimation of the size of the active linuron-mineralizing

biomass in samples of the BM matrix was performed using

Table 2. Physicochemical characteristics of the substrata used in the biomix of the BMs

Substratum Moisture content % (w/w) pH Specific density (g cm�3) Total C (%) Total N (%) C/N (%)

Soil C 2.83 6.9 2.81 0.95 0.08 12.1

Soil L 11.53 4.9 ND 0.80 0.06 13.3

Straw 8.40 6.6 1.56 43.25 0.46 95.1

Peat 49.67 6.4 1.58 43.97 0.93 47.1

ND, not determined.

FEMS Microbiol Ecol 76 (2011) 64–73c� 2011 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

66 K. Sniegowski et al.

Page 4

an MPN approach. A sample (10 g, wet weight) taken from

the biomix of the BM was added to 25 mL MgSO4 (10�2 M)

and incubated overnight on a shaker to remove the cells

from the matrix. After settling of the matrix material for 2 h,

a decimal serial dilution was made with 2 mL of the aqueous

extract. Aliquots of 0.5 mL of these dilutions were used as

inoculum for 14C-linuron mineralization assays in triplicate,

operated as described above. Positive tubes for MPN calcu-

lation were those vials where 4 10% total 14CO2 was

produced within 60 days. MPN calculations were performed

using the computer-assisted method developed by Briones

& Reichardt (1999) and the MPN number was expressed as

active mineralizing units per gram dry weight (dw) of the

sample (AMU g�1 dw).

Molecular techniques

Total DNA was extracted from 400-mg samples of the

biomix as described by Uyttebroek et al. (2006). The copy

number of the Variovorax 16S rRNA gene was determined by

real-time PCR, performed in a Rotor Gene (RG 3000)

apparatus from Westburg using the Variovorax-specific 16S

rRNA gene primers VarF and VarR. Details of the method

are described in Bers et al. (2011). Real-time PCR deter-

mined the bacterial 16S rRNA gene copy number as

described by Haest et al. (2010) using primers Eub341F and

Eub534R (Muyzer et al., 1993). The detection limit of both

methods was 105 copies g�1 biomix dw.

Denaturing gradient gel electrophoresis (DGGE) finger-

prints of the Variovorax community were performed using

16S rRNA gene fragments amplified with primers VarF-GC

and VarR in either a single or a double PCR as described by

Bers et al. (2011) on a polyacrylamide gel (10%) with a

denaturating gradient from 45% to 75% as described by

Uyttebroek et al. (2006). The 16S rRNA gene fragments in

the gel were mobilized through an electric field set at 120 V

for 15 h. On each gel, a reference DGGE migration Variovorax

marker sample (ref-Var) was loaded, containing 16S rRNA

gene fragments PCR amplified with primers GC-VarF and

VarR from Variovorax sp. SRS16 (Sorensen et al., 2005),

Variovorax sp. DSM66 and Variovorax sp. WDL1 (Dejonghe

et al., 2003) representing three major phylotypes within the

Variovorax genus (Breugelmans et al., 2007). In addition,

reference DGGE migration marker samples (ref-nonVar)

containing 16S rRNA gene fragments with different GC

content from various bacteria (not belonging to the Variovor-

ax genus) amplified with bacterial primers 63F and 518R (El

Fantroussi et al., 1999) were loaded on each gel. The detection

limit of the Variovorax-specific PCR-DGGE method was

approximately 105 copies g�1 biomix dw. The significance of

differences between averages of the 16S rRNA gene copy

numbers of triplicate BM samples were analyzed by ANOVA

(Po 0.05).

Results

Linuron mineralization capacity of BMbioaugmented with linuron-primed soil andnonprimed soil

In the first experiment, bioaugmentation of BPS with linuron-

primed soil L and nonprimed soil C was compared. Initially,

only samples from BMs inoculated with linuron-primed soil

L, i.e. BMs of set-ups L� and L1, showed linuron mineraliza-

tion with a lag time of approximately 9.9� 0.4 days and a

maximum mineralization rate of 14.2� 1.5% day�1. After 2

weeks of linuron treatment, the lag time of linuron miner-

alization recorded for samples of BMs of set-up L1 was

reduced to 2.6� 1.4 days (Fig. 1). At week 12, the lag time

further decreased to 0.46� 0.56 days. MPN counting showed

that the linuron-mineralizing population in BMs of set up L1

increased from 4.5� 102 to 1.8� 104 AMUg�1 during the 12

weeks of linuron treatment. The lag time of the mineralization

curves performed with the dilution series increased with

higher dilution. Therefore, it was concluded that the reduction

of the lag time observed in the BMs of set-up L1 corre-

sponded to an increase in the linuron-mineralizing biomass

(Sniegowski et al., 2009). With the samples taken from BMs

of set-up L�, the lag time initially decreased to 6.1� 0.1 days

at week 2, but again increased to 8.9� 0.9 days at week 12.

BMs from set-up C1 and set-up C� did not show a

linuron-mineralizing capacity initially. However, BMs from

(b)

Time (days)

0

10

20

30

40

50

60

Initial situation2 weeks without linuron supply12 weeks without linuron supply

(a)

Time (days)0 5 10 15 20 250 5 10 15 20 25

% 14

C-l

inu

ron

min

eral

ized

0

10

20

30

40

50

60

Initial situation2 weeks linuron supply12 weeks linuron supply

Fig. 1. Linuron mineralization kinetics recorded

for samples taken from the BMs of (a) set-up L1

and (b) set-up L� in the first experiment. BMs of

set-up L1 were continuously irrigated with tap

water containing linuron, while BMs of set-up L�

received noncontaminated water. The arrow in-

dicates the changes in lag time. The data are

average values with indicated SDs of samples

taken from three replicate BMs.

FEMS Microbiol Ecol 76 (2011) 64–73 c� 2011 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

67Pesticide-primed soil to improve on-farm biopurification

Page 5

set-up C1 clearly acquired the capacity to mineralize linuron

upon continuous irrigation with linuron. This capacity devel-

oped in the three replicate BMs, after different periods of

linuron feeding (Fig. 2). Moreover, for one replicate BM, a

significantly lower mineralization rate than that for the other

two was recorded. In addition, the mineralization curve

recorded with this replicate demonstrated a rather linear

mineralization curve, indicating that mineralization was not

linked to growth. At week 12, the average recorded lag time

was 4.5� 1.8 days. BMs from set-up C� did not develop an

observable linuron mineralization capacity during the 12-

week incubation period (data not shown).

Adaptability of BPS bioaugmented withnonprimed soils

Because BMs of set-up C1 developed a linuron mineraliza-

tion capacity upon the continuous supply of a linuron-

containing solution, it was tested whether inoculation with

other nonprimed soils led to a similar development. There-

fore, a second experiment was performed in which non-

primed soils of different origins were used to bioaugment

BMs. To ensure that the acquisition of a linuron mineraliza-

tion capacity was related to the inoculation of the soil, a set-

up containing control BMs without soil addition was

included. BMs containing linuron-primed soil L and non-

primed soil C were included in the experimental set-up. All

systems were irrigated with the linuron-containing solution.

Because in the first experiment, the development of a linuron-

mineralizing capacity in BMs containing nonprimed soil was

observed only after 12 weeks of linuron irrigation, matrix

samples in the current experiment were taken for linuron

mineralization assays after 15 and 20 weeks of incubation.

BMs of set-ups L1 and C1 behaved identical to those in the

first experiment (data not shown). However, the development

of a linuron mineralization capacity was observed with only

one of the other tested nonprimed soils, i.e. soil A1, and in

only one replicate BM (lag time 4.9 days). After 20 weeks of

linuron treatment, the recorded linuron mineralization curves

of the BMs were similar to those recorded after 15 weeks of

treatment (data not shown).

Size and composition of the Variovoraxcommunity in bioaugmented BPS

In the first experiment, the Variovorax 16S rRNA gene copy

number increased both in BMs of set-up L1 and in BMs of

set-up L� between the start of the experiment and week 12

while the number of bacterial 16S rRNA gene copies

remained stable (Table 3). The increases were, however,

insignificant. In BMs of set-ups C� and C1, the Variovorax

16S rRNA gene copy number decreased after 12 weeks of

incubation, but only the decrease observed in set-up C�was

significant (Table 3). Double PCR-DGGE analysis of the

Variovorax community in BMs of set-ups L�, L1, C� and C1

showed similar fingerprints at week 0 for all set-ups with

bands at positions V1, V2 and V4 (Fig. 3a). Position V1

corresponds to the migration position of the 16S rRNA gene

fragment amplified from the linuron-degrading Variovorax

sp. WDL-1, which belongs to Variovorax phylotype B (Bers

et al., 2011). 16S rRNA gene fragments at position V2

correspond to Variovorax members of phylotype C, which

does not contain Variovorax strains with a linuron degrada-

tion capacity (Bers et al., 2011). The fingerprints of week 2 and

week 12 showed the same bands at positions V1, V2 and V4.

However, in set-ups L� and L1 an additional band appeared at

position V3. This band especially dominated in set-up L1 at

week 12 and as such can be associated with the observed

increased capacity to mineralize linuron. Position V3 corre-

sponds to the migration position of the 16S rRNA gene

fragment amplified from the linuron-degrading Variovorax

sp. SRS16, which is a member of the Variovorax phylotype A

(Bers et al., 2011). A band at position V3 also appeared in the

profiles of set-ups C� and C1 at week 12, but no clear

differences in intensity were found between C� and C1.

Real-time PCR results performed with DNA extracted

from the samples of set-ups L1, C1, A11 and O1, taken at

weeks 0 and 15 in the second experiment, showed that the

Week 0

Time (days)

% 14

C-l

inu

ron

min

eral

ized

0102030405060

Week 2

Time (days)

Week 4

Time (days)

Week 12

Time (days)0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60

BM 1 BM 2 BM 3

Fig. 2. Evolution of the linuron mineralization kinetics recorded with samples taken from the three replicate BMs designated as BM1, BM2 and BM3 of

set-up C1 inoculated with the non-linuron-primed soil in the first experiment.

FEMS Microbiol Ecol 76 (2011) 64–73c� 2011 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

68 K. Sniegowski et al.

Page 6

number of bacterial and Variovorax 16S rRNA gene copies

were initially not significantly different between the different

set-ups (Table 3). The number of Variovorax 16S rRNA gene

copies in BMs of set-ups A11, C1 and L1 increased after 15

weeks of linuron treatment. However, the increase was

significant only for set-ups C1 and L1. In set-up A11, no

significant increase was recorded in the Variovorax 16S

rRNA gene copy number in the BM replicate that had

developed a linuron mineralization capacity compared

with the other two replicate BMs (data not shown). The

Variovorax population in BMs of set-up O1 remained

constant in size, while the total number of bacteria increased

six times. Variovorax 16S rRNA gene double PCR-DGGE

fingerprints in samples taken from BMs of set-ups L1, C1,

A11 and O1 of the second bioaugmentation experiment are

shown in Fig. 3b. DGGE fingerprints of all set-ups at week 0

displayed profiles similar to those observed at week 0 in

experiment 1, but with additional bands such as at position

V3. This also accounted for the set-up O1 where no soil was

added, indicating that the observed Variovorax populations

originated from the biomix substrata peat or straw. As in the

Variovorax profiles of set-up L1 in the first experiment, an

additional band at position V3 became clearly dominant in

BMs of set-up L1. Interestingly, at week 15, in set-up A11,

only the replica BM showing enhanced linuron-mineralizing

capacity displayed a dominant band at position V1 in the

Variovorax DGGE profiles, although this was clearly ob-

served only after a single PCR approach (Fig. 3b lower gel),

which associates this band with the observed linuron

mineralization activity. Position V1 corresponds to the

Variovorax phylotype B, which, as phylotype A, includes

predominantly linuron-degrading Variovorax strains. The

same band also appeared in all profiles of set-up C1, but was

also observed at week 0.

Discussion

This study demonstrates for the first time the potential of

bioaugmenting BPS by means of pesticide-primed soil.

Although this bioaugmentation approach has been pre-

viously applied with success in polluted soil and wetlands,

it has never been tested in a complex biotechnological

matrix such as the biomix of a BPS with 8 and 2.5 times

higher total carbon and nitrogen percentages than an

average soil biotope, and sustaining a high biomass and rich

biodiversity. Moreover, in contrast with previous studies, a

comparison was made between systems bioaugmented with

pesticide-primed soil and nonprimed soil. Clear differences

in initial linuron mineralization capacity and the time to

establish a maximum mineralization capacity were observed

between the linuron-primed soil and any nonprimed con-

trol soil. The nonprimed perfect control would have been a

non-linuron-treated soil originating from the same field siteTab

le3.

Num

ber

of

bac

terial

and

Var

iovo

rax

16S

rRN

Agen

eco

pie

sin

BM

sof

exper

imen

t1

(set

-ups

L�,L1

,C�

and

C1

)at

wee

k0

and

afte

r2

and

12

wee

ksof

irrigat

ion

with

tap

wat

erw

ith

or

without

linuro

nan

din

BM

sof

exper

imen

t2

(set

-ups

A11

,L1

,O

1an

dC

1)at

wee

k0

and

afte

r15

wee

ksof

irrigat

ion

with

linuro

n

Exper

imen

t1

Set-

up

L�Se

t-up

L1Se

t-up

C�

Set-

up

C1

Var

iovo

rax

Bac

teria

Var

./bac

t.(%

)V

ario

vora

xBac

teria

Var

./bac

t.(%

)V

ario

vora

xBac

teria

Var

./bac

t.(%

)V

ario

vora

xBac

teria

Var

./bac

t.(%

)

Wee

k0

3.4�

1.4�

10

72.6�

0,1�

10

10

0.1

3�

0.0

66.7�

4.6�

10

71.6�

0.8�

10

10

0.4

2�

0.3

55.4�

3.8�

10

73.3�

1.4�

10

91.6

6�

1.1

14.1�

3.8�

10

74.9�

1.4�

10

90.8

4�

1.1

1

Wee

k2

8.7�

3.4�

10

73.7�

4.6�

10

10

0.2

3�

0.1

07.5�

1.1�

10

72.8�

0.3�

10

10

0.2

7�

0.0

42.1�

0.4�

10

74.9�

0.5�

10

90.4

3�

0.0

82.3�

2.4�

10

74.1�

1.9�

10

90.5

5�

0.0

6

Wee

k12

1.2�

0.7�

10

83.7�

3.5�

10

10

0.3

1�

0.2

23.1�

1.4�

10

83.1�

0.1�

10

10

0.7

2�

0.1

61.3�

1.1�

10

62.5�

1.1�

10

80.5

4�

0.4

99.0�

0.8�

10

63.3�

3.0�

10

90.2

8�

0.3

7

Set-

up

A11

Set-

up

L1Se

t-up

O1

Set-

up

C1

Exper

imen

t2

Var

iovo

rax

Bac

teria

Var

./bac

t.(%

)V

ario

vora

xBac

teria

Var

./bac

t.(%

)V

ario

vora

xBac

teria

Var

./bac

t.(%

)V

ario

vora

xBac

teria

Var

./bac

t.(%

)

Wee

k0

4.5�

1.4�

10

74.4�

0.3�

10

81.0

6�

0.4

23.0�

1.5�

10

72.8�

0.5�

10

91.1

0�

0.5

71.7�

0.4�

10

81.8�

0.2�

10

10

0.1

9�

0.0

41.7�

0.8�

10

75.8�

5.7�

10

80.6

9�

0.7

6

Wee

k15

1.7�

0.8�

10

83.4�

3.3�

10

93.0

6�

2.0

01.0�

0.5�

10

92.2�

1.2�

10

10

5.0

4�

1.7

01.2�

0.3�

10

81.1�

1.1�

10

11

0.0

05�

0.0

51.2�

4.3�

10

88.0�

8.3�

10

93.3

1�

3.2

7

The

dat

aar

eav

erag

eva

lues

with

indic

ated

SDs.

FEMS Microbiol Ecol 76 (2011) 64–73 c� 2011 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

69Pesticide-primed soil to improve on-farm biopurification

Page 7

as the linuron-primed soil but such a soil was not available.

Nevertheless, the effect of linuron priming is accentuated in

the second experiment where only two of the tested non-

primed soils resulted into the development of a linuron

mineralization capacity.

In contrast to nonprimed soil, bioaugmentation with

linuron-primed soil immediately enhanced the linuron

mineralization capacity of the BMs. Moreover, feeding with

linuron further increased the linuron-mineralizing capacity

in the matrix. MPN counting demonstrated that the ob-

served enhancement in linuron mineralization capacity

could be related to an increase in size of the linuron-

mineralizing population. Therefore, it can be concluded

that despite the complex background, the high biomass and

high biodiversity, the community or at least the linuron-

degrading fraction in the added soil could proliferate in the

biomix in case fed with linuron. Various studies, mostly

performed with suspended cultures inoculated with pure

bacterial strains, reported on the effect of a highly degrad-

able carbon and nitrogen content on mineralization/degra-

dation of pesticides or other pollutants. Positive effects of

additional C and/or N sources on pesticide degradation

were reported by Cullington & Walker (1999), Aslan &

Turkman (2005) and Fogg et al. (2003b), while Breugelmans

et al. (2010) noticed negative effects on degradation of

linuron. On the other hand, the maintenance of the linuron-

mineralizing capacity in BMs of set-up L� throughout the

experiment indicate that the linuron-degrading population

present in the primed soil is sufficiently competitive in its

new biotope and can establish even without apparent

selective conditions. Similarly, Johnsen et al. (2007) reported

the long-term survival of a polyaromatic hydrocarbon

(PAH)-degrading community present in inoculum PAH-

primed soil, in soils without PAH pollution.

BMs of set-up C1, inoculated with nonprimed soil C,

started to mineralize linuron after an extended period of

linuron supply. A similar result was obtained with one soil of

the other tested non-linuron-primed soils. Development of

a linuron mineralization capacity in those BMs occured only

when the pesticide was supplied, showing the selective

nature of the process. Because the set-ups showing this

outcome had the maximum moisture contents and because

the pH did not significantly differ between set-ups, it is very

unlikely that such differences in environmental conditions

Week 15Week 0

V4V3

V1

V2

V4V3

V1

V2

A1+A1

+ C+ O+L+ O+L+C+RVR R– – – – – – + + + – – – + – – + + + + + + – – –

A1+

–––+++++–+––++–––– ––

Week 0

A1+ C+ O+

Week 15

L+ O+L+C+

V4V3

V1

V2

V4V3

V1

V2

RVR R

Double PCR

R

Week 12Week 2Week 0

C– C+ L– L+ C– C+ L– L+ C– C+ L– L+R RVR RV

V4V3V2

V3V2

V1

V4

V1

(a)

(b)

Single PCR

Double PCR

Fig. 3. 16S rRNA gene-based DGGE fingerprints of the Variovorax community obtained from (a) BM samples taken at weeks 0, 2 and 12 from BMs of

set-ups C�, C1, L� and L1 in the first experiment and (b) BM samples taken at weeks 0 and 15 from set-ups A11, C1, L1 and O1 in the second

experiment. (a) Only DGGE profiles obtained with 16S rRNA gene fragments produced by double PCR are shown. (b) DGGE profiles obtained with 16S

rRNA gene fragments produced by double PCR (upper gel) and single PCR (lower gel). DGGE fingerprints of reference DGGE migration markers ref-

nonVar and ref-Var are indicated with ‘R’ and ‘V’, respectively. Band V3 corresponds to phylotype A, band V2 with phylotype C and band V1 with

phylotype B.

FEMS Microbiol Ecol 76 (2011) 64–73c� 2011 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

70 K. Sniegowski et al.

Page 8

were responsible for the observed differences in linuron

mineralization development. Instead, it can be hypothesized

that either the microbial community in those BMs geneti-

cally adapted to degrade linuron or a linuron-mineralizing

bacterial population initially present in the soil at undetec-

tably low numbers proliferated when linuron was added.

Interestingly, different replicate BMs developed a linuron-

mineralizing capacity either after different treatment periods

(in case of bioaugmentation with soil C) or only in one BM

replicate (in case of bioaugmentation with soil A1). This can be

explained by the occurrence of genetic adaptation at different

time points or in one replicate, or by the fact that low initial

numbers of mineralizing biomass were unevenly distributed in

the soil sample used for inoculation. A similar observation was

reported by Cullington & Walker (1999) with diuron.

The number of Variovorax in BMs containing linuron-

primed soil L and receiving linuron (set-up L1) increased

concomitantly with the increase in linuron mineralization

capacity, but this was statistically significant only in BMs

operated in the second experiment. The increase in Vario-

vorax 16S rRNA gene copy number was less pronounced

compared with the 100-fold increase observed in soil L when

it was supplied linuron on a long-term base (Bers et al.,

2011). The reason for the observed difference in prolifera-

tion of Variovorax between set-ups L1 in the first and second

experiment might be due to the longer incubation period (3

weeks) and/or due to the higher amount of nutrients

because of the higher fraction of straw and peat (only

25 vol% soil inoculum) in the biomix used in the second

experiment. On the other hand, in both experiments,

Variovorax DGGE fingerprinting clearly demonstrated the

proliferation of a particular Variovorax population in BMs

bioaugmented with linuron-primed soil L and fed with

linuron. This Variovorax population belonged to phylotype

A, which includes only linuron-degrading Variovorax strains

such as Variovorax SRS16 (Breugelmans et al., 2007). The

same phylotype A band proliferated in soil L when con-

tinuously fed with a linuron-containing solution (Bers et al.,

2011). The Variovorax 16S rRNA gene PCR products

recovered from these BMs were cloned but only some

randomly chosen clones were sequenced because the asso-

ciation of bands with specific phylotypes was demonstrated

previously (Bers et al., 2011). Nevertheless, BLAST analysis of

these sequences identified them all as Variovorax 16S rRNA

genes (E-value o 10�18) with sequences 100% identical to

those reported by Bers et al. (2011). The enrichment of

phylotype A Variovorax in these BMs was probably not

detected by the Variovorax real-time PCR due to masking

by the large number of Variovorax initially present in the

biomix substrata. These data suggest that linuron-degrading

Variovorax populations endogenous to the primed soil were

successfully transferred to the BM biomix. In addition they

suggest that strains belonging to phylotype A proliferated in

the BM upon linuron supply, thereby enhancing the linuron

degradation capacity. Because those populations only pro-

liferated in case linuron was supplied and because they could

be associated with a phylotype containing only linuron-

degrading Variovorax, it can be suggested that these popula-

tions are involved in linuron mineralization in the L1 set-

up. Conclusive evidence for this link can be provided using

techniques such as DNA/RNA-stable isotope probing (Du-

mont & Murrell, 2005). It cannot be excluded that bacteria

other than Variovorax were also involved in linuron degra-

dation. DGGE analysis of bacterial 16S rRNA genes was

performed but no differences were observed between set-ups

L1 and L� (data not shown).

Linuron is degraded through 3,4-dichloroaniline (3,4-

DCA). Several 3,4-DCA-degrading strains have been iso-

lated previously from soil L in co-culture with linuron-

degrading Variovorax strains and it was suggested that the

former proliferated by growth on 3,4-DCA leaking from the

linuron-degrading Variovorax strains (Breugelmans et al.,

2007). Some of these 3,4-DCA-degrading strains belonged

to Variovorax phylotype C. The band associated with

Variovorax phylotype C in the Variovorax DGGE profiles

migrates at position V2, but its occurrence in BMs of set-up

L1 could not be correlated with the appearance of the

linuron mineralization capacity. An explanation is that the

release of 3,4-DCA was limited during linuron degradation

in the biomix, hindering the proliferation of 3,4-DCA-

degrading bacteria or that, alternatively, 3,4-DCA degrada-

tion in the biomix was performed by other genera such as

Comamonas as observed in other linuron-mineralizing

bacteria consortia (Dejonghe et al., 2003).

No relationship was found between the linuron-miner-

alization capacity dynamics and Variovorax community

dynamics (neither in size nor in structure) in BMs of set-

up C1 in the first experiment. On the other hand, in

experiment 2, a Variovorax population related to phylotype

B (at position V1) appeared in the DGGE profile concomi-

tantly with the appearance of a linuron mineralization

capacity. Moreover, the Variovorax number increased con-

comitantly with the linuron mineralization capacity. Phylo-

type B includes the linuron-degrading Variovorax sp.

WDL1. However, in experiment 2, no set-up was included

without linuron feed and therefore the appearance of this

phylotype B cannot be linked to the observed increased

linuron mineralization capacity. Moreover, the presence of a

band at position V1 was not correlated with an increase in

linuron mineralization capacity in the first experiment. On

the other hand, based on the single PCR-DGGE results, the

proliferation of a Variovorax phylotype B and hence a

WDL1-related Variovorax population could be clearly asso-

ciated with the increased linuron mineralization capacity in

BM1 of set-up A11 because neither of the other two BM

replicas in this set-up showed enhanced linuron

FEMS Microbiol Ecol 76 (2011) 64–73 c� 2011 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

71Pesticide-primed soil to improve on-farm biopurification

Page 9

mineralization capacity nor the DGGE band associated with

Variovorax phylotype B. However, this association was less

clear in the DGGE profiles obtained with the product of the

double-PCR approach. Because a WDL1-related population

was also detected in BMs without inoculum and in BMs

with nonprimed soil A1 (as shown in the double PCR-

DGGE approach) that did not develop a linuron-degrading

capacity, it can be suggested that different WDL1-related

populations were initially present in the straw and peat mix

on the one hand and in the soil inoculum on the other.

In conclusion, the results suggest that it is preferable to

use pesticide-primed soils containing endogenous pesticide-

mineralizing microorganisms over nonprimed soil for

bioaugmentation of on-farm BPS for several reasons. First,

although microorganisms in soils without previous pesti-

cide treatment can apparently develop a capacity to miner-

alize the pesticide used, this is not always the case. Second,

the initiation of mineralization of the target pesticide, in

case nonprimed soil is used, takes much more time com-

pared with a system inoculated with pesticide-primed soil.

This is important because mineralization should start as fast

as possible in a BPS to minimize leaching of pesticides

during the start-up period of the BPS. Furthermore, the

results show that Variovorax populations originating from

the added soil contribute to linuron degradation in the BPS

at least in the systems containing linuron-primed soil.

Acknowledgements

This research was supported by IWT-Vlaanderen Strategic

Basic Research project 73352 and IWT-Vlaanderen Agricul-

tural Research project LBO 040272.

References

Aslan S & Turkman A (2005) Combined biological removal of

nitrate and pesticides using wheat straw as substrates. Process

Biochem 40: 934–943.

Barbeau C, Deschenes L, Karamanev D, Comeau Y & Samson R

(1997) Bioremediation of pentachlorophenol-contaminated

soil by bioaugmentation using activated soil. Appl Microbiol

Biot 48: 745–752.

Bazot S & Lebeau T (2008) Simultaneous mineralization of

glyphosate and diuron by a consortium of three bacteria as

free- and/or immobilized-cells formulations. Appl Microbiol

Biot 77: 1351–1358.

Bers K, Sniegowski K, Breugelmans P, Hendrickxs L, Mot RD &

Springael D (2011) A molecular toolbox to estimate the

number and diversity of Variovorax in the environment:

application in soils treated with the phenylurea herbicide

linuron. FEMS Microbiol Ecol DOI: 10.1111/j.1574-

6941.2010.01028.x.

Breugelmans P, D’huys PJ, De Mot R & Springael D (2007)

Characterization of novel linuron-mineralizing bacterial

consortia enriched from long-term linuron-treated

agricultural soils. FEMS Microbiol Ecol 62: 374–385.

Breugelmans P, Horemans B, Hofkens J & Springael D (2010)

Response of the structure and activity of a linuron-degrading

triple-species biofilm to mixed substrate feeds. Res Microbiol

161: 660–666.

Briones AM & Reichardt W (1999) Estimating microbial

population counts by ‘most probable number’ using Microsoft

Excels. J Microbiol Meth 35: 157–161.

Broos K, Mertens J & Smolders E (2005) Toxicity of heavy metals

in soil assessed with various soil microbial and plant growth

assays: a comparative study. Environ Toxicol Chem 24:

634–640.

Chatterjee DK, Kilbane JJ & Chakrabarty AM (1982)

Biodegradation of 2,4,5-trichlorophenoxyacetic acid in soil by

a pure culture of Pseudomonas cepacia. Appl Environ Microb

44: 514–516.

Coppola L, Castillo MD, Monaci E & Vischetti C (2007)

Adaptation of the biobed composition for chlorpyrifos

degradation to Southern Europe conditions. J Agr Food Chem

55: 396–401.

Cullington JE & Walker A (1999) Rapid biodegradation of diuron

and other phenylurea herbicides by a soil bacterium. Soil Biol

Biochem 31: 677–686.

Dejonghe W, Berteloot E, Goris J, Boon N, Crul K, Maertens S,

Hofte M, De Vos P, Verstraete W & Top EM (2003) Synergistic

degradation of linuron by a bacterial consortium and isolation

of a single linuron-degrading Variovorax strain. Appl Environ

Microb 69: 1532–1541.

De Wilde T, Spanoghe P, Debaer C, Ryckeboer J, Springael D &

Jaeken P (2007) Overview of on-farm bioremediation systems

to reduce the occurrence of point source contamination. Pest

Manag Sci 63: 111–128.

Dumont MG & Murrell JC (2005) Stable isotope probing –

linking microbial identity to function. Nat Rev Microbial 3:

499–504.

El Fantroussi S, Verschuere L, Verstraete W & Top EM (1999)

Effect of phenylurea herbicides on soil microbial communities

estimated by analysis of 16S rRNA gene fingerprints and

community-level physiological profiles. Appl Environ Microb

65: 982–988.

Fogg P, Boxall ABA & Walker A (2003a) Degradation of pesticides

in biobeds: the effect of concentration and pesticide mixtures.

J Agr Food Chem 51: 5344–5349.

Fogg P, Boxall ABA, Walker A & Jukes AA (2003b) Pesticide

degradation in a ‘biobed’ composting substrate. Pest Manag

Sci 59: 527–537.

Fogg P, Boxall ABA, Walker A & Jukes A (2004) Degradation and

leaching potential of pesticides in biobed systems. Pest Manag

Sci 60: 645–654.

Grigg BC, Assaf NA & Turco RF (1997) Removal of atrazine

contamination in soil and liquid systems using

bioaugmentation. Pestic Sci 50: 211–220.

Grundmann S, Fuss R, Schmid M, Laschinger M, Ruth B, Schulin

R, Munch JC & Schroll R (2007) Application of microbial hot

FEMS Microbiol Ecol 76 (2011) 64–73c� 2011 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

72 K. Sniegowski et al.

Page 10

spots enhances pesticide degradation in soils. Chemosphere 68:

511–517.

Haest PJ, Philips J, Springael D & Smolders E (2010) The reactive

transport of trichloroethene is influenced by residence time

and microbial numbers. J Contam Hydrol 119: 89–98.

Henriksen VV, Helweg A, Spliid NH, Felding G & Stenvang L

(2003) Capacity of model biobeds to retain and degrade

mecoprop and isoproturon. Pest Manag Sci 59: 1076–1082.

Johannesen H, Sorensen SR & Aamand J (2003) Mineralization of

soil-aged isoproturon and isoproturon metabolites by

Sphingomonas sp strain SRS2. J Environ Qual 32: 1250–1257.

Johnsen A, Schmidt S, Hybholt T, Henriksen S, Jacobsen C &

Andersen O (2007) Strong impact on the polycyclic aromatic

hydrocarbon (PAH)-degrading community of a PAH-polluted

soil but marginal effect on PAH degradation when priming

with bioremediated soil dominated by mycobacteria. Appl

Environ Microb 73: 1474–1480.

Mertens B, Boon N & Verstraete W (2006) Slow-release

inoculation allows sustained biodegradation of gamma-

hexachlorocyclohexane. Appl Environ Microb 72: 622–627.

Moran AC, Muller A, Manzano M & Gonzalez B (2006) Simazine

treatment history determines a significant herbicide

degradation potential in soils that is not improved by

bioaugmentation with Pseudomonas sp. ADP. J Appl Microbiol

101: 26–35.

Muyzer G, Dewaal EC & Uitterlinden AG (1993) Profiling of

complex microbial populations by denaturing gradient gel-

electrophoresis analysis of polymerase chain reaction-

amplified genes coding for 16S ribosomal-RNA. Appl Environ

Microb 59: 695–700.

Pigeon O, De Vleeschouwer C, Cors F, Weickmans B, De Ruyckel

B, Pussemier L, Debognie P & Culot M (2005) Development of

biofilters to treat the pesticide wastes from spraying

applications. Comm Agr Appl Biol Sci 70: 1003–1012.

Runes HB, Jenkins JJ & Bottomley PJ (2001) Atrazine

degradation by bioaugmented sediment from constructed

wetlands. Appl Microbiol Biot 57: 427–432.

Singh BK, Walker A & Wright DJ (2006) Bioremedial potential of

fenamiphos and chlorpyrifos degrading isolates: influence of

different environmental conditions. Soil Biol Biochem 38:

2682–2693.

Sniegowski K, Mertens J, Diels J, Smolders E & Springael D

(2009) Inverse modeling of pesticide degradation and

pesticide-degrading population size dynamics in a

bioremediation system: parameterizing the Monod model.

Chemosphere 75: 726–731.

Sorensen SR, Rasmussen J, Jacobsen CS, Jacobsen OS, Juhler RK

& Aamand J (2005) Elucidating the key member of a linuron-

mineralizing bacterial community by PCR and reverse

transcription-PCR denaturing gradient gel electrophoresis 16S

rRNA gene fingerprinting and cultivation. Appl Environ

Microb 71: 4144–4148.

Struthers JK, Jayachandran K & Moorman TB (1998)

Biodegradation of atrazine by Agrobacterium radiobacter J14a

and use of this strain in bioremediation of contaminated soil.

Appl Environ Microb 64: 3368–3375.

Topp E (2001) A comparison of three atrazine-degrading bacteria

for soil bioremediation. Biol Fert Soils 33: 529–534.

Torstensson L & del Pilar Castillo M (1997) Use of biobeds in

Sweden to minimize environmental spillages from agricultural

spraying equipment. Pesticide Outlook 8: 24–27.

Uyttebroek M, Breugelmans P, Janssen M et al. (2006) Distribution

of the Mycobacterium community and polycyclic aromatic

hydrocarbons (PAHs) among different size fractions of a

long-term PAH-contaminated soil. Environ Microbiol 8:

836–847.

Vischetti C, Capri E, Trevisan M, Casucci C & Perucci P (2004)

Biomassbed: a biological system to reduce pesticide point

contamination at farm level. Chemosphere 55: 823–828.

Zhao SH, Arthur EL & Coats JR (2003) Influence of microbial

inoculation (Pseudomonas sp strain ADP), the enzyme atrazine

chlorohydrolase, and vegetation on the degradation of

atrazine and metolachlor in soil. J Agr Food Chem 51:

3043–3048.

FEMS Microbiol Ecol 76 (2011) 64–73 c� 2011 Federation of European Microbiological SocietiesPublished by Blackwell Publishing Ltd. All rights reserved

73Pesticide-primed soil to improve on-farm biopurification