FEMS Microbiology Ecology 49 (2004) 109–120
www.fems-microbiology.org
Chlorobenzene biodegradation under consecutive
aerobic–anaerobic conditions
Gerd U. Balcke
a
a,*
, Lea P. Turunen a, Roland Geyer
Dietmar Schlosser c
b,c
, Dirk. F. Wenderoth d,
Department of Hydrogeology, UFZ Centre for Environmental Research Leipzig-Halle, Theodor-Lieser-Str. 4, D-06120 Halle, Germany
b
Center for Biomarker Analysis, University of Tennessee, Knoxville, TN 37932, USA
c
Groundwater Microbiology Group, UFZ Centre for Environmental Research Leipzig-Halle, D-06120 Halle, Germany
d
Microbial Ecology, German Research Centre for Biotechnology, D-38124 Braunschweig, Germany
Received 21 November 2002; received in revised form 5 June 2003; accepted 28 August 2003
First published online 17 April 2004
Abstract
The biodegradation of monochlorobenzene, the main contaminant in a quaternary aquifer at Bitterfeld, Central Germany, was
studied in microcosm experiments employing either original groundwater or defined mineral media together with the indigenous
microbial community from the polluted site. The impact of consecutive aerobic–anaerobic–aerobic incubations on monochlorobenzene biodegradation, microbial diversity, and pH development was examined. The related changes in microbial community
composition were analyzed by 16S rRNA gene-based single-strand conformation polymorphism (SSCP) fingerprints and sequencing
of dominant bands and by quantitative analysis of bacterial respiratory chain quinones as biomarkers. Under aerobic conditions,
the indigenous microbial community of the groundwater degraded monochlorobenzene mainly via the modified ortho-pathway.
Respiratory chain quinones and SSCP analysis suggested dominance of the genera Acidovorax and Pseudomonas. A shift to anoxic
conditions resulted in monochlorobenzene biotransformation but no dechlorination. The ability to degrade monochlorobenzene
aerobically remained preserved throughout a fortnightly anoxic period at sufficiently high buffer capacity. Acidification, caused by
monochlorobenzene biodegradation, was alkalinity-controlled. At low initial alkalinity a substantial decrease in pH, monochlorobenzene degradation, and total counts of live cells, accompanied by a change of the microbial community composition, was
observed.
Ó 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.
Keywords: Chlorobenzene biodegradation; Dechlorination; Microbial diversity; Alkalinity; Groundwater aeration; Optode principle
1. Introduction
Groundwater aeration strategies make use of the capability of indigenous bacteria to degrade organic pollutants aerobically. Oxygen, as the terminal electron
acceptor, can be provided via the gas phase (in situ air
sparging, in-well aeration), via the liquid phase (hydrogen peroxide amendments), or via the solid phase (oxygen release compounds) [1–4].
*
Corresponding author. Tel.: +49-345-5585208; fax: +49-345558559.
E-mail address: [email protected] (G.U. Balcke).
The maintenance of aerobic conditions within a previously anoxic aquifer is constrained by the low oxygen
solubility. The amount of oxygen that can be dissolved
in groundwater depends on several factors such as hydrostatic pressure, temperature, partial pressure of O2 ,
and salinity. Considering, e.g., the injection of pure
oxygen into a depth of 20 m below the groundwater
table, a theoretical solubility of 150 mg/l dissolved oxygen (DO) is possible at 14 °C. However, besides the
biodegradation of pollutants, biogeochemical processes
such as oxidation of ferrous iron or mineral surfaces
may compete for oxygen [5–7]. Together with a heterogeneous transport of DO, this may result in oxygen
0168-6496/$22.00 Ó 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.femsec.2003.08.014
110
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
gradients and anaerobic micro-niches even if an aquifer
contains DO [8]. Moreover, within biofilms, e.g., attached to aquifer minerals, oxygen gradients may become established [9–11].
Oxygen amendment to a previously anoxic aquifer
may further cause a number of acid forming processes.
Most prominently, the biocatalytic dehalogenation of
chlorinated hydrocarbons may lead to HCl and CO2
production, decreasing the pH of the groundwater at
high pollutant load and limited buffer capacity [12].
Such alterations can affect the composition and the
degradation capacity of the indigenous microbial communities and therefore have to be kept within ranges
non-limiting to the biodegradative process [13].
Bacterial lipids (quinones, glycolipids, glycerophosphatidyl lipids) can provide qualitative and quantitative
biomarkers for key microbial parameters that may be
crucial for biodegradation (e.g., community composition, viable biomass, metabolic activity, and nutritional
status) [14–16]. A greater specificity in the detection of
structural microbial diversity can be obtained by selective molecular genetic methods [17,18]. Lipid biomarker
analysis and molecular genetic approaches should
therefore be considered as complementary methods.
In this study we combined the analysis of membraneassociated respiratory quinones, detected by a sensitive
tandem-mass-spectrometry method, with the singlestrand conformation polymorphism (SSCP) analysis of
amplified parts of 16S rRNA genes to assess the microbial community composition. We focused on effects
of transient environmental conditions on biodegradation of monochlorobenzene (MCB), the main contaminant in a quaternary aquifer at Bitterfeld, Central
Germany. In particular, the impact of shifts from aerobic to anaerobic and thereafter back to aerobic conditions on MCB biodegradation, pH development, and
microbial community composition was studied in microcosm experiments.
2. Materials and methods
2.1. Experimental setup
For microcosm experiments, a modified biochemical
oxygen demand measurement system was used, consisting of 500-ml glass bottles, a PTFE washer, and
OxitopÒ C pressure sensing heads (WTW, Weilheim,
Germany). Each sample contained 250 ml of solution.
The big headspace was chosen to provide the samples
with a large O2 reservoir to guarantee oxygen excess
conditions throughout the experimental phase when
desired. A 25 mm2 piece of an oxygen sensitive optode
foil (POF-PtSt3, Presens, Regensburg, Germany) was
taped onto the inner glass surface 1 cm above the bottom of the microcosms, using a minimal amount of
silicon rubber glue (RS Components, Nothants, UK).
The method to measure oxygen via dynamic fluorescence quenching allows measuring O2 concentrations in
aqueous solution and headspace without opening the
bottles [19,20]. In combination with the total pressure,
O2 partial pressures can be obtained upon exposing the
optode spot to both the headspace and the liquid phase.
The O2 detection limit of this equipment was determined
to be 10 lg/l. In control experiments it was assured
that neither optode nor glue caused measurable MCB
sorption.
2.2. Media and inoculation
Groundwater originated from a quaternary aquifer
(infiltration well 28 m below sea level) at the pilot test
site SAFIRA at Bitterfeld [21]. Alkalinity, DO, pH, and
redox potential were measured immediately at the site.
Groundwater was originally contaminated with MCB,
1,2-dichlorobenzene, 1,4-dichlorobenzene, and benzene
at 21.9, 0.05, 0.27, and 0.09 mg/l, respectively. Microcosms were supplemented with either 250 ml of original
groundwater or artificial mineral media (MM) as depicted in Table 1. To investigate the effect of buffer capacity on acidification caused by biodegradation of
MCB, the initial alkalinity was varied for MM (further
on referred to as MML ¼ mineral medium with low,
MMM ¼ mineral medium
with
medium, and
MMH ¼ mineral medium with high hydrogen carbonate
starting concentration) (Table 1). The MM solutions
were sterilized by filtration through 0.2 lm polycarbonate filters (NucleporeÒ , Whatman, Maidstone, UK)
before inoculation.
Original groundwater contained the indigenous bacterial community (2.8 104 cells/ml) whereas in MM
microcosms 15 ml were replaced by an inoculum of an
aerobic groundwater enrichment culture to give a final
cell concentration of 6.3 106 cells/ml. The inoculum
was prepared in the following manner. Five hundred
milliliters glass bottles containing 250 ml of groundwater were aerated thoroughly, spiked with MCB at 100
mg/l, sealed, and incubated for 7 days at 14 °C in the
dark under shaking (150 rpm). The rather high concentration was chosen to bring out the potential chloride
formation at a background of 460 mg/l chloride in the
groundwater. Furthermore, it seemed reasonable to
work at this level for concentrations up to 55 mg/l MCB
were determined previously in the groundwater.
2.3. Cultivation regime
The aerobic–anaerobic–aerobic consecutive experiments were conducted according to the regime illustrated in Fig. 1 (see arrows). At the beginning (except
microcosms containing original groundwater since it
þ
naturally contained PO3
4 and NH4 , Table 1) and before
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
111
Table 1
Composition of groundwater and artificial mineral salt media
Cl
Br
SO2
4
PO3
4
Groundwater (mg/l)
Mineral salt media (MM) (mg/l)
460
0
800
15
23c
1200
800
20
MML
HCO
3
NO
3
NO2
þ
Na
Kþ
NHþ
4
Ca2þ
Mg2þ
293
not detectable
not detectable
180
180
20
300
40
pH
Redox potential
DO
MCB
1,2-Dichlorobenzene
1,4-Dichlorobenzene
Benzene
6.7
)5 mVa /420 mVb
0.8a /9.74b
21.90a
0.05a
0.27a
0.09a
b
220
80
170
50
300
40
b
MMM
b
MMH
450
620b
170
250
+Trace-element solution: Zn2þ 0.5; Cu2þ 0.1; Mn2þ 0.5; Fe2þ 0.5
6.7
420 mVb
9.98b
9.69b
9.45b
–
No NO
3 and NO2 detectable.
Before areation with air/CO2 .
b
After areation with air/CO2 .
c
Due to inoculation with pre-incubated groundwater.
a
starting of each subsequent batch (now including also
groundwater microcosms), microcosms were suppleand NHþ
mented with PO3
4
4 at 20 and 50 mg/l, respectively, to avoid nitrogen and phosphorus limitation.
Thereafter, they were vigorously purged for 15 min with
mixtures of sterile air at aerobic batches (or nitrogen at
the anaerobic phase) and CO2 in order to remove remaining MCB from the previous batch (or the groundwater) and to set aerobic (or anoxic) conditions. During
this, the CO2 partial pressure was increasingly enhanced
to adjust exactly a pH of 6.7 (as measured in the original
groundwater at site) at the beginning of the first aerobic
batch and later to the pH value obtained at the end of
the previous batch, respectively, for subsequent treatments. In groundwater microcosms, this procedure retained the original alkalinity even if the total hydrostatic
pressure decreased from 2.3 atm in the well to 1 atm
under laboratory conditions. While pulling up the gas
frit, used for purging, MCB was added each time at 100
mg/l and microcosms were closed immediately. Microcosms were incubated at 14 °C in the dark under shaking
(150 rpm). Optode control measurements (see below)
revealed O2 concentrations as low as 20 lg/l when nitrogen served as flushing gas, thus ensuring that less
then 2.1% of the added MCB could be transformed to
the corresponding catechol. The oxygen levels within the
microcosms were controlled repeatedly during the
course of the incubation in order to exclude the possibility of small oxygen fluxes, in particular, throughout
the anaerobic period. Previous stirring tests over 10 days
with nitrogen flushed water also ensured the tightness of
the setup.
In an additional set of experiments attempted to follow
MCB biodegradation under defined oxygen-limiting
conditions, groundwater microcosms were fully aerobically preincubated for 7 days as described above to enrich
MCB-degrading bacteria. Thereafter, microcosms were
þ
amended with PO3
4 and NH4 as already mentioned and
purged with a sterile mixture of oxygen, argon, and carbon dioxide to adjust the initial oxygen partial pressure at
a molar O2 :MCB ratio of 2.5 (related to the MCB newly
added at 100 mg/l). After this, microcosms were incubated
for 9 days as already described.
Finally, control experiments using MMM were carried out in order to prove the MCB degradation when
MCB was re-spiked to the anoxic culture. At this set of
experiments three subsequent anaerobic incubations of
14 days each followed the two aerobic growth periods.
The batches were incubates as triplicates. Otherwise, the
treatment remained the same as above.
2.4. Analytical procedures
Oxygen concentrations were measured twice a day by
holding an optical fiber combined to a fiber optic device
(FIBOX2, Presens, Germany) onto the optode spot. The
calculated oxygen demand was not corrected for endogenous carbon degradation (biomass decomposition).
112
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
Fig. 1. Experimental regime for consecutive aerobic–anaerobic–aerobic incubations of microcosms. The arrows symbolize the partial
pressure adjustment and the injection of 100 mg/l MCB at the beginning of each incubation. (Please note that the data only depict endpoint analyses of sequential batches. The straight lines solely resemble
that the concerning samples are associated but have no kinetic
meaning.) Removal of MCB (a), oxygen consumption (b), chloride
formation (c), and pH development (d) during consecutive aerobic–
anaerobic–aerobic incubations of microcosms containing original
groundwater (d), MMH (r), MMM (.), and MML (s). Symbols
represent an average of five microcosms running in parallel (±SD).
The pressure sensors stored pressure data automatically in intervals of 20 min. At the end of each phase all
pressure data were retrieved from the OxiTopÒ C pressure sensor heads using an infrared coupled data logger
(OC110, WTW, Weilheim, Germany).
MCB concentrations were analyzed by gas chromatography (GC) on a Varian 3400 CX gas chromatograph (Darmstadt, Germany), equipped with an
HP-5MS column (Hewlett–Packard, Waldbronn, Germany) and a flame ionization detector (FID). Helium
served as carrier gas. Samples were 20-fold diluted in
headspace GC vials and volatile compounds were allowed to adsorb onto solid micro extraction phase out
of the headspace (85 lm polyacrylate coating, Supleco,
Taufkirchen, Germany) for 8 min.
Polar degradation metabolites were assessed by high
performance liquid chromatography (HPLC), employing a Merck–Hitachi HPLC system (Darmstadt, Germany) consisting of an L-7455 diode array detector, an
L-7120 gradient pump, an L-7200 auto-injector, and a
column oven. Metabolites were separated on an OH18HY ion exclusion chromatography column (Merck,
Darmstadt, Germany), using 5 mM H2 SO4 as an eluent
at a flow rate of 0.75 ml/min. The column oven temperature was set to 85 °C.
Chloride concentrations and those of other ions were
quantified using a DX-120 ion chromatography system
(Dionex GmbH, Idstein, Germany), equipped with a
conductivity detector and an IonPack AS 14-4-mm
column. The eluent was a bicarbonate buffer composed
2
of 1 mM HCO
3 and 3.5 mM CO3 .
concentrations
were
spectrophotometrically
NHþ
4
determined with the help of the MicroquantÒ 14750
system (Merck, Darmstadt, Germany).
Cells were enumerated by fluorescence staining, using
the V-7023 staining kit (Molecular Probes Europe BV,
Leiden, The Netherlands) and following the protocol of
the manufacturer. The kit allows discrimination of dead
and live bacteria and relies on a selective membrane
integrity-based stain penetration into the cells, where
cells with intact membranes are stained with the nucleic
acid stain 40 ,6-diamidiono-2-phenylindole (DAPI) and
cells with damaged membranes are stained with the
SYTOXÒ Green nucleic acid stain. Before staining,
samples were subjected to low-energy ultra-sound for 3
min to disrupt biomass clusters that may affect cell
counting, which has been found not to bias the viability
of the bacteria. For cell counting, sample aliquotes were
filtered through 0.2 lm black polycarbonate filters
(NucleporeÒ , Whatman, Maidstone, UK). Counting
was carried out on a Zeiss Axioskop fluorescence microscope (Carl Zeiss, Oberkochen, Germany), using a
400-fold magnification and a 10 10 mm ocular grid.
For SSCP analysis, 5-ml samples were filtered
through polycarbonate membrane filters (0.2 lm pore
size, NucleporeÒ , Whatman, Maidstone, UK). The fil-
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
ters were folded using sterile tweezers, wrapped into
sterile aluminum foil, and frozen at )24 °C instantly.
For DNA extraction, frozen filters were cut into small
pieces, transferred to the first tube of the BIO 101 Fast
DNA kit for soil (QBiogene-ALEXIS GmbH,
Gr€
unberg, Germany), and further processed according
to the manufacturer’s protocol. Sequences of the two
primers used for the amplification of bacterial 16S
rDNA and their positions in the E. coli 16S rRNA gene
were forward primer Com1 (50 CAGCAGCCGCGGTAATAC30 , positions 519–536) and reverse
primer Com2-Ph (50 CCGTCAATTCCTTTGAGTTT30 ,
positions 907–926) [23], the latter containing a 50 -terminal phosphate group. Each polymerase chain reaction (PCR) was performed in a total volume of 50 ll in
micro reaction tubes (0.1 ml volume, Eppendorf,
Hamburg, Germany), containing 1 PCR buffer with
1.5 mM MgCl2 , desoxynucleoside triphosphate solution (200 mM each dATP, dCTP, dGTP and dTTP),
primers Com1 and Com2-Ph (0.5 lM each), 5 ng DNA
as template, and DNA polymerase (0.5 U; HotStar
Taq, Qiagen, Germany). Thermocycling was conducted
in an Eppendorf Mastercycler (Hamburg, Germany),
starting with an initial denaturation for 15 min at 95 °C.
A total of 30 cycles were performed, each including 90 s
at 94 °C, 40 s at 50 °C, and 40 s at 72 °C, and a final
elongation for 10 min at 72 °C. Purity and size of
PCR products were analyzed by agarose gel electrophoresis (1.5% agarose, 1 TBE, pH 8.0, running buffer) and ethidium bromide staining [22]. The whole
PCR product was used for SSCP analysis. Single strand
preparation gel electrophoresis and silver staining of
the polyacrylamid gel were done according to Schwieger and Tebbe [23]. For image analysis the gels were
digitized to create tif-files. Cluster analysis of 16S rDNA
fingerprints originating from one gel was performed
using the software package GelCompare II (Applied
Maths, Kortrijk, Belgium). Background was subtracted using rolling circle correction (circle diameter,
30) and lanes were normalized. Only bands with an intensity of at least 2% of the total intensity per lane were
considered for statistical analysis. Dendrograms were
constructed based on gel-scans of each fingerprint applying the unweighted pair group method using arithmetic averages (UPGMA) as cluster algorithm and the
Pearson coefficient that includes position and intensity
of single bands. Products identified in silver-stained
polyacrylamide gels were excised with sterile scalpels
and transferred to an Eppendorf tube and mixed with 50
ll elution buffer (10 mM Tris–HCl, 50 mM KCl, 1.5
mM MgCl2 , 0.1% Triton X-100, pH 9.0). After incubation for 20 min at 95 °C, 5–20 ll of the band solution
was applied to a PCR containing Com primers under
the conditions described above. PCR products were
checked for size and purity by agarose gel electrophoresis as already described. The re-amplified PCR-prod-
113
ucts were then extracted from the agarose gel (1.5%,
TAE-buffer), purified using the Qiaquick MinEluteGel
Purification Kit (Qiagen, Hilden, Germany) and sequenced with the ABI PRISM BigDye Terminator Cycle
Sequencing Ready Reaction Kit including one Com
primer (Applied Biosystems, Foster City, CA, USA).
Sequencing reactions were analyzed on an Applied
Biosystems 377 genetic analyzer.
For analysis of respiratory chain quinones, 3-ml
samples were collected in incinerated glass vials and lyophilized at a freeze dryer (Alpha2-4, CHRIST, Germany). A simultaneous measurement of ubiquinones
(UQ) and menaquinones (MK) at a tandem mass spectrometer (API-365, SCIEX, Canada) coupled to a liquid
chromatography (LC) device was utilized. The LC solvent system and the detection of positive fragment ions
of UQs, as m=z ¼ 197, based on electrospray (ES) ionization method, are described elsewhere [24]. The MKs,
as m=z ¼ 187, could not be measured with ES, therefore,
a prototype of an atmospheric pressure photoionization
(APPI) ion source (SCIEX, Canada) was used for the
simultaneous measurement of UQs and MKs. Additionally, that increased the sensitivity of detection
compared to the also applicable atmospheric pressure
chemical ionization (APCI). The newly developed approach expands the detection limits for both quinone
classes to approximately 10 lg/l within a linear detection
range of about three orders of magnitude (D.C. White
and R. Geyer, unpublished data). Due to the increased
sensitivity we could obtain quinone profiles (fmol range)
in samples with at least 5.9 106 cells/ml. Major quinones in Gram-positive and Gram-negative bacteria,
UQ6 to UQ10, MK4, and MK7 to MK9 (number indicates the length of the isoprenoid side chain), were
measured by multiple reaction monitoring and quantified by comparing with authentic standards of UQ6,
UQ10, and MK4 [14].
3. Results
3.1. Biodegradation experiments
Starting with two successive oxic batches, conditions
for selective aerobic growth on MCB were provided.
After 9 days, 100% of the MCB was converted in all
mineral media (Fig. 1(a)), whereas 60% (in relation to
the amount theoretically required for complete mineralization of MCB) of the pertinent oxygen was consumed (Fig. 1(b)). In the groundwater microcosms 60%
of the MCB had been converted at this time. MCB was
not anymore detectable under any treatment condition
after 18 days, whereas the relative oxygen consumption
had increased to values between 75% and 90%. A
following change to anaerobic conditions and additional MCB supplementation resulted in an MCB
114
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
disappearance of approximately 50% in all mineral salt
media and 92% in original groundwater at the end of the
anaerobic batch (day 32) (Fig. 1(a)). During control
experiments either omitting biomass or employing biomass poisoned by adding 2 g/l of NaN3 and 100 mg/l of
HgCl2 , 92% and 103% of MCB, respectively, were recovered after 9 days, thus ruling out a non-biological
reason for the MCB disappearance. After re-establishment of aerobic conditions, MCB was converted again
completely (Fig. 1(a)). Between 32 and 41 almost similar
amounts of oxygen were consumed, as compared to the
aerobic treatment between day 10 and 18 (Fig. 1(b)).
When MCB was added again, the conversion gradually
decreased in MML microcosms (day 50). Groundwater
microcosms and those containing MMH as well as
MMM kept the ability to completely degrade MCB
aerobically.
The chloride formation within each phase was obtained by comparing the cumulative Cl concentrations
of the beginning and the end of an experimental phase.
Within the first aerobic incubation 60% (MM) to 100%
(GW) of the converted MCB were dechlorinated.
During the second aerobic incubation approximately
80–100% of the converted MCB were dechlorinated,
indicating mineralization to a high extent (Fig. 1(c)).
Higher standard deviations observed in groundwater
microcosms arise from the high chloride background,
which is declining the detection quality of the chloride
newly formed. In contrast, only very little chloride formation could be observed during the following anaerobic phase, even though high amounts of MCB were
removed. Interestingly, subsequent re-aeration of
MMM and MMH microcosms at day 32 resulted in a
chloride formation which significantly exceeded the
theoretical amount expected upon complete dechlorination of newly injected MCB. This was also visible in
groundwater microcosms, besides the higher standard
deviations observed for reasons already mentioned. The
chloride missing during the anaerobic treatments thus
could be partially recovered in the following aerobic
batches, e.g., in MMM microcosms the average chloride
formation accounted for 1% and 125% at day 32 (end of
anaerobic treatment) and 39 (end of the following aerobic treatment), respectively, yielding a total of 126%. A
sum of 153% chloride could be expected upon complete
dechlorination, considering that 53% and 100% of MCB
have been removed during the anaerobic and the subsequent aerobic batch, respectively. This is indicative for
the biotransformation of MCB without dechlorination
under anoxic conditions. Less chloride, as compared to
MMM and MMH microcosms, was produced upon reaeration of MML microcosms between day 32 and 41.
Chloride formation was totally inhibited during the last
aerobic incubation, thus confirming the observed decrease in MCB degradation and oxygen demand under
these conditions very well.
During aerobic batches conducted under excess of
oxygen only traces of potential organic metabolites were
detectable by HPLC under any condition. Therefore, an
additional experiment was performed using groundwater under defined oxygen-limiting conditions (initial
molar ratio of O2 :MCB ¼ 2.5) in order to promote the
accumulation of potential organic degradation products
resulting from aerobic biodegradation of MCB. After 9
days of incubation, 100% of the MCB was converted
whereas only 45% of the total chlorine was recovered as
chloride. Two chlorinated metabolites, cis,cis-2-chloromuconate (2-ClMA) and 3-chlorocatechol (3-CC)
were identified by comparing their HPLC retention
times and UV spectra with those of authentic standards.
Related to the MCB concentration initially added, 36%
and 5%, respectively, accounted for 2-ClMA and 3-CC,
respectively. Two further metabolites of low apparent
abundance remained unidentified. Together this led to a
mass balance of 86% under oxygen limitation, suggesting that aerobic biodegradation occurred mainly via the
modified ortho-pathway [29]. At the end of the anaerobic batch, only 5.0% of 2-ClMA and 1.8% of 3-CC (in
relation to the added MCB concentration) were detected. During the anoxic batches, the proven O2 concentrations were below 20 lg/l.
Microcosms containing MMH displayed pH values
within a range of 6.5–6.8 throughout the experiment
(Fig. 1(d)). In groundwater and MMM microcosms the
pH decreased slowly, ending at pH 5.8 and 6.2, respectively, after 50 days. In MML microcosms the pH drop
occurred much faster, finally approaching pH 3.8.
During the anaerobic phase a slight pH increase could
be observed, most obvious in samples of low initial alkalinity (Fig. 1(d)). The redox potential ranged from
+350 to +500 mV (vs. normal hydrogen electrode) in all
microcosms (data not shown) during aerobic treatments.
In anaerobic batches values between +235 and +340 mV
were observed, with the lowest redox potentials measured in groundwater microcosms.
Cell numbers are depicted as cumulative plots and are
further subdivided into live and dead cells (Fig. 2).
Starting from an initial cell density of 2.4 104 cells/ml
for groundwater and 3.3 106 cells/ml for all MM, an
increase in cell numbers by three orders of magnitude in
groundwater and by 1.3–1.6 orders of magnitude in MM
microcosms was observed within the first 9 incubation
days. During the second aerobic incubation overall cell
numbers increased further on to give maximum concentrations of around 108 cells/ml. In the majority of the
samples living bacteria represented the main fraction of
the cells. A converse composition of living and dead
bacteria was found after the shift to anaerobic conditions (measured at day 41). The fraction of living bacteria accounted only for 3–10% in the MM experiments
and for 27% in the groundwater microcosms. During
following aerobic phases (day 32 to 50) the amount of
1.3
1.2
1.2
1.9
5.2
(36.4)
(0.4)
(0.0)
(4.9)
(10.2)
11.9
1.7
0.1
34.7
73.4
0 (0)
0 (0)
0 (0)
1.1 (0.1)
2.5 (0.3)
(11.2)
(1.1)
(0.7)
(2.1)
(4.0)
3.7
5.2
3.4
14.8
28.7
(30.9)
(9.0)
(0.8)
(0.8)
(1.0)
1
2
3
4
5
(9)
(18)
(32)
(41)
(50)
0 (0)
10.0 (2.1)
12.8 (2.6)
15.4 (2.7)
14.8 (2.0)
0 (0)
31.2 (6.5)
40.6 (8.2)
38.2 (5.4)
32.9 (4.6)
2.6 (7.8)
308.0 (63.7)
382.4 (77.5)
416.9 (58.8)
396.4 (54.9)
3.1
82.6
49.1
180.0
160.8
(9.5)
(17.1)
(10.0)
(25.4)
(22.2)
10.1
43.3
3.7
5.4
7.5
MK8
MK7
MK4
UQ10
UQ9
UQ8
UQ7
UQ6
Quinones lM (% of total within batch)
Batch no.
(days)
living bacteria increased again in highly buffered media
(MMH). In groundwater and MMM counts for living
cells increased first, declined again with further acid
formation, and did not recover when acidification
emerged early (MML). The amount of live cells appeared to diminish with decreasing pH values. The
fraction of dead cells remained constant from the anaerobic phase onwards, indicating persistence of nucleic
acids in dead cells as evidenced by stainability upon
employment of the nucleic acid stain SYTOXÒ Green.
Finally, in control experiments including three subsequent anaerobic batches of 14 days each following
aerobic growth MCB was still removed to a high extent
(see Table 4). Repeated spiking under oxygen exclusion
Table 2
Concentrations of ubiquinones (UQ) and menaquinones (MK) extracted from MML-containing microcosms at the end of consecutive batches
Fig. 2. Concentrations of viable (black bars) and dead (grey bars) cells
during consecutive incubations of microcosms containing either original groundwater or MM at variations as indicated. The overall
column heights display the sum of live and dead bacteria. Microcosms
were analyzed at the end of each batch. The cell densities at the beginning of the experiment were 2.4 104 (original groundwater) and
3.3 106 cells/ml (all MM incubations). The bars reflect average values
of five replicates.
(4.1)
(0.2)
(0.2)
(0.3)
(0.7)
115
MK9
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
116
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
Table 3
Identification of bacterial taxa related to numbered bands in Fig. 3(a) and (b) as determined by 16S rDNA sequence analysis
Band
Closest relative in GenBank database
Accession of closest
relative
% similarity
Typical main quinone
profile of genera
1
2
3
Acidovorax sp.
Acidovorax sp. UFZ
Stenotrophomonas maltophilia (syn. Xanthomonas
maltophilia)
Pseudomonas fluorescence
Pseudomonas sp. uncultured
Nocardioides simplex (syn. Arthrobacter simplex)
Azospirillum sp.
AF078767
AF235013
AJ131784
99
99
98
UQ8 [25]
UQ8 [25]
UQ8 [26]
AJ308307
AJ295645
U81990
AF112477
99
97
97
98
UQ9 [26]
UQ9 [26]
MK-8(H4) [27]
UQ10 [28]
4
5
6
7
Only sequences with a content of not more than 5% unknown nucleotides per sequence were subjected to the database.
Table 4
Repeated anaerobic spiking of 100 mg/l MCB to an aerobically grown consortium
Batch no. (days)
MCB removal % of
theoretical
Chloride formation % of
theoretical
Cell number (alive/dead) [cells/ml] and observed
cell morphology
4 (32)
48.8
7.2
9.4E08/3.1E07
Small thick rods, almost like pure culture
5 (46)
24.9
Beyond detection limit
1.6E08/8.8E08
Small thick rods, accompanied by some rods and rod
chains
6 (60)
27.0
8.2
2.6E09/9.5E08
Very small rods, accompanied by thread-shaped cells
still resulted in approximately 25% MCB removal after
the third subsequent anaerobic incubation. Chloride
formation during the anaerobic batches was again only
minimal.
The cell numbers, dead/alive ratio, and cell morphology changed intensely during the course of the three
anaerobic batches. While during the first anaerobic
batch only three percent of the counted cells showed
damages in their membrane integrity, the majority of the
cells died during the second incubation. The share of live
cells during this period declined from nearly 1E09–
1.6E08 cells/ml. During the last anaerobic phase the
cultures recovered, and increases in cell numbers from
1.9E09 to 3.2E09 cells/ml could be observed despite the
absence of oxygen.
3.2. Quinone profiles
Ubiquinones with 6–10 and menaquinones with 4 and
7–9 isoprenoid units, common respiratory quinones of
Gram-positive and Gram-negative bacteria [14], were
assessed for microcosms containing MML. This medium was chosen since: (i) during the first aerobic batches MCB biodegradation parameters were essentially
comparable to those of MMM, MMH, and groundwater, it may therefore be considered as representative for
the early stages of the experiment, and (ii) later on MCB
biodegradation underwent the most significant changes,
as compared to MMH, MMM, and original ground-
water, suggesting that the microbial community might
have been altered accordingly.
The relative and absolute concentrations of bacterial
quinones detected at the end of each batch are listed in
Table 2. While quinone profiles obtained at day 9 (first
aerobic incubation) still resembled the profile of the
initial groundwater bacterial community (dominance of
MK4 and 8 and UQ8, 9, and 10), ongoing aerobic
conditions (day 10 to 18) led to an increase in concentrations of UQs with 7–10 isoprenoid units and resulted
in a unequivocal dominance of UQ8, whereas the MK8
concentration strongly decreased. The dominance of
UQs within the profile is characteristic to aerobic Gramnegative bacterial communities [14]. The quinone profile
obtained at the end of the anaerobic incubation (day 19
to 32) was clearly still dominated by UQ8, followed by
UQs9 and 7. The concentration of UQ10 and, to a lesser
extend, also that of UQ9 declined. Summarizing, the
anaerobic treatment did not qualitatively alter the quinone profile observed after the previous aerobic treatment. After the first re-aeration period (day 32 to 41)
UQ9-, MK4-, and MK8-containing bacteria again
gained more influence whilst UQ8 remained essentially
constant in terms of its absolute concentration. After the
second re-aeration period (day 41 to 50) the absolute
concentrations of UQ9 and 8 slightly decreased were
MK8 and 4 concentrations clearly were increased, suggesting an ongoing shift of the microbial community
composition (Table 2).
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
117
Fig. 3. (a) Cluster analysis and SSCP community profiles obtained from 16S rDNA PCR amplicons during consecutive incubations of a representative microcosm containing MML. Samples for analysis were taken at the end of each batch. The band assignment is given in Table 3. Visual
evaluation of band intensities suggests dominance of Acidovorax sp. during early stages of the experiment (day 18 and 32). Intermediate anoxic
conditions showed no remarkable impact on the microbial diversity (day 32). Acidification in later phases (day 41 and 50) led to the detection of
additional bands, thus suggesting promotion of growth of different bacterial species. (b) Cluster analysis of five independent microcosms run in
parallel and sampled at the end of the experiment (day 50) demonstrate a good reproducibility of the obtained data.
3.3. SSCP analysis and sequencing
SSCP fingerprints of the bacterial community were
generated from microcosms containing MML for the
reasons already mentioned. Fig. 3(a) shows a representative cluster analysis of SSCP community profiles of a
MML microcosm over time. The resulting dendrogram
suggests two main groups. A high similarity (98%) was
found within the first group (day 18 and 32) between the
samples from the second aerobic growth phase and the
anaerobic phase. The samples of aerobic growth under
increasingly acidifying conditions (day 41 and day 50)
grouped together with a similarity of 83%. High overall
similarities between all experimental phases (79%) and
an unusually small number of intensive bands reflect
strong selection conditions. Fig. 3(b) illustrates the
similarities between five independent microcosms performed in parallel; samples were taken at the end of the
experiment. All SSCP community profiles clustered together with a similarity of at least 78% reflecting high
reproducibility of the experimental conditions even after
five subsequent incubations.
Sequence analysis of single bands excised from the
SSCP fingerprints was used to identify single members
of the microbial communities, and to give more detailed
phylogenetic information about the most dominant
bacteria. From each fingerprint the main bands were
excised and the respective fragments comprising the V4
and V5 regions were sequenced. Table 3 shows the
phylogenetically closest relative found in the GenBank
database. The band positions are marked in Fig. 3(a)
and (b). Bands from different fingerprints but from same
relative position in the respective fingerprint exhibited
identical sequences as confirmed by sequence analysis
(Fig. 3(a) and (b); Table 3). Two of the sequences found
in the SSCP fingerprints revealed the presence of members of the c-subgroup of the proteobacteria, three
bands were related to the b-subgroup [23,25], and in
each case one member was assigned to the a-proteobacteria [21,28] and the Gram-positive bacteria with
high GC-content [14,27]. The appearance of additional
bands with decreasing pH (Fig. 3(a), days 41 and 50)
could reflect the changing environmental conditions.
Genetic data of these bands suggested the growth of
further pseudomonades but are not shown here due to
high uncertainty in nucleotide sequences.
4. Discussion
Periodical groundwater aeration for the stimulation
of indigenous pollutant degrading bacteria was tested in
subsequent batches. Since the oxygen solubility may
constitute an important constraint in the remediation
process, we particularly focused on transient stages between aerobic and anaerobic conditions following upon
an aeration phase. We have designed a set of consecutive
aerobic–anaerobic incubations with groundwater culture consortia from Bitterfeld and chlorobenzene in order to mimic rapid changes in environmental conditions
118
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
such as oxygen availability and alkalinity of the
groundwater. The choice of batch microcosms as the
simplest setup was made to provide reproducible data
within a controlled environment. The drawbacks of
batch approaches as compared to flowthrough column
experiments, e.g., nutrient depletion over time, were
taken into account as far as possible.
4.1. Aerobic–anaerobic changes and biodegradation of
MCB
Under oxic conditions MCB can be completely degraded by the indigenous microbial community within
short periods of time (Fig. 1). In aerobic MCB degradation experiments employing groundwater derived
from a network of wells of the Bitterfeld quaternary
aquifer Dermietzel and Vieth [30] showed that MCB
biodegradation in groundwater samples taken from the
pollutant plume occurred significantly faster, as compared to samples that were taken in a distance from the
plume. This was implied as an adaptation of the microbial community at the site to MCB [30]. Additional
information about the dissolved oxygen concentration
of the aquifer (approximately 0.8 mg/l) makes the disposition of the bacterial community understandable to
grow fast on MCB when enough oxygen is present.
Genetic analyses and respiratory chain quinone profiles
consistently suggest growth promotion of Gram-negative bacteria such as Acidovorax sp. and pseudomonades, when rigorous aerobic conditions were provided. In
good accordance with our results, pure isolates from
Bitterfeld groundwater, belonging to the genera Acidovorax and Pseudomonas, were found to aerobically degrade MCB (C. Vogt, Center for Environmental
Research, personal communication).
With good reproducibility it could be shown that
during the anaerobic phase MCB disappeared to a high
extent while almost no chloride was formed. The missing
chloride could be partially recovered during subsequent
aerobic incubations, suggesting the formation of chlorinated metabolites during the anaerobic phase. As
ensured by optode measurements the oxygen concentration never exceeded 20 lg/l neither immediately after
MCB spiking at the beginning of the anaerobic incubation, nor during the course of the anaerobic incubations. Furthermore, in preliminary experiments under
defined oxygen-limiting conditions 36% of MCB were
converted into 2-ClMA when 2.5 mol of O2 were provided per one mole of MCB. This suggests that during
anaerobic treatments of microcosms traces of contaminating oxygen might have led to the accumulation of 2ClMA. However, the small concentrations of 2-ClMA
and 3-CC detected at the end of the anaerobic period
could not account for the high MCB removal. We presume that MCB was metabolized by the microbial
community to chlorinated metabolites that could be
detected neither by HPLC (polar compounds) nor GC
(volatile, less polar compounds).
Repeated MCB spiking within three subsequent anaerobic batches could proof ongoing MCB removal. The
extent of this removal seemed to decrease over time, but
was still abundant after 42 days at a rate of approximately 25 per 100 ppm MCB spiked. A limited chloride
formation during the first and the last anaerobic batch is
not reflecting the fate of transformed MCB even though
up to 8% of the theoretically expected chloride (nearly
one third of the MCB removed) has formed.
Cell numbers fluctuated during the anaerobic incubations rendering a community shift that is still to be
analyzed.
So far, anaerobic biodegradation of MCB has not
been described [31]. We could show the disappearance of
MCB at transient aerobic–anaerobic conditions. The
repeatability of an anoxic transformation at repeated
anaerobic spiking strengthens this observation. At the
moment, we cannot give a mechanistic explanation yet.
Sorption to biomass or glassware can be excluded for
tests with empoisoned biomass resulted in nearly complete MCB recovery.
We are currently investigating the community shift at
longer anaerobic periods following aerobic growth. Second, we are attempting to trace chlorinated metabolites
that might have formed during anoxic incubations.
The quinone profiles and 16S rRNA gene-based SSCP
fingerprints reveal that the aerobic microbial community
sustained an anoxic phase of 14 days without becoming
overgrown by other bacteria. According to the composition of the original groundwater no alternative electron
acceptors, such as nitrate or ferric iron, were employed in
MM microcosms. A fairly high redox potential and
lacking of H2 S formation do not support the presence
of sulfate reducers during the anoxic incubation of
groundwater microcosms. Secondary fermentation of
dead biomass cannot be excluded. When aerobic conditions were reestablished, again complete MCB degradation and dechlorination was observed in sufficiently
buffered microcosms. We conclude that at least short
periods without oxygen do not cause major changes in the
composition of the microbial community. This result is of
particular interest when cyclic oxygen sparging is considered in subsurface remediation.
4.2. Increasing acidification and biodegradation of MCB
According to the design of the experiments we intended to overburden the alkalinity-controlled buffer
capacity of the microcosms and to relate the changed
pH conditions to the biodegradation of MCB and the
composition of the bacterial community. Poorly buffered media (MML) clearly dropped in pH, resulting
from biological activities, earlier than medium buffered
media (MMM and groundwater). Highly buffered media
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
(MMM) remained within almost neutral pH values.
Furthermore, MCB biodegradation decreased with decreasing pH. This goes along low numbers of living cells.
After re-establishment of aerobic conditions, viable cell
numbers were only recovered when the pH value was
kept within a neutral range. Quinone profiling and
rDNA sequences imply a shift in the microbial community composition towards more MK8- and MK4containing microbes along with a decreasing pH (Figs. 1
and 3; Tables 2 and 3). It cannot not be readily explained whether the community shift is a result of MCB
utilization or pH decrease. The MCB utilization causes a
pH decrease, and both effects are superimposed. However, the quinone profiling showed increasing values in
MK4–9 along with proceeding acidification (Table 2
and Fig. 1). This result would not be expected in case of
MCB utilization. Here the conditions of the second
aerobic batch should be resembling.
The MCB load (500 mg/l per microcosm in total) in
these experiments was certainly to high to allow a valid
transfer of the results to a flowthrough system. Nevertheless, whether biodegradation of an organic pollutant
contributes to a pH decrease or not depends mainly on
the buffer capacity of the aquifer, including mineral
surfaces. At high buffer capacity the biogenic CO2 will
contribute to the overall alkalinity, e.g., by calcite dissolution [12]. By oxygen amendments to a groundwater,
several naturally occurring acid forming (and hence alkalinity decreasing) processes are promoted [7,13]. Once
the groundwater alkalinity is decreased (or is initially
low), the formed CO2 and other acidic biodegradation
products cannot become buffered anymore and may
cause local regions within the aquifer, where biodegradation becomes less effective due to decreased pH conditions. Sub-oxic transitions, subsequent to aerobic
conditions, are particularly critical since bacterial cells
may accumulate metabolites, which can deliver protons
when biodegradation is continued at re-aeration. The
alkalinity development in sandy aquifers with low clay
and chalk content, as the Bitterfeld quaternary aquifer
[12,13], should be therefore observed critically when
groundwater is aerated. The kinetics of such alkalinityinfluencing processes in interaction with the desired
biodegradation needs to be further investigated.
Acknowledgements
We are thankful to W. Reineke (Wuppertal) and C.
Vogt (Leipzig) for the generous gifts of 2-ClMA and 3CC, respectively. Further, we thank U. Lechner (Halle)
for assistance in the quinone assignment and for valuable discussion of the results. L. Turunen’s research stay
at the UFZ-Halle was funded by the EU program
LEONARDO.
119
References
[1] Aelion, C.M. and Kirtland, B.C. (2000) Physical versus biological
hydrocarbon removal during air sparging and soil vapor extraction. Environ. Sci. Technol. 34, 3167–3173.
[2] Rogers, S.W. and Ong, S.K. (2000) Influence of porous media,
airflow rate, and air channel on benzene NAPL removal during air
sparging. Environ. Sci. Technol. 34, 764–770.
[3] Zappi, M., White, K., Hwang, H.M., Bajpai, R. and Qasim, M.
(2000) The fate of hydrogen peroxide as an oxygen source for
bioremediation activities within saturated aquifer systems. J. Air
Waste Manage. 50, 1818–1830.
[4] Barcelona, M.J. and Xie, G.B. (2001) In situ lifetimes and kinetics
of a reductive whey barrier and an oxidative one ORC barrier in
the subsurface. Environ. Sci. Technol. 35, 3378–3385.
[5] Warren, L.A. and Haack, E.A. (2001) Biological controls on metal
behavior in freshwater environments. Earth-Sci. Rev. 44, 261–320.
[6] Søgaard, E.G., Remgigijus, A., Abraham-Peskir, J. and Koch,
B.C. (2001) Conditions for biological precipitation of iron by
Gallionella ferruginea in a slightly polluted groundwater. Appl.
Geochem. 16, 1129–1137.
[7] Ehrlich, H.L. (1998) Geomicrobiology: its significance for geology. Earth-Sci. Rev. 45, 45–60.
[8] Fry, V.A., Selker, J.S. and Gorelick, S.M. (1997) Experimental
investigations for trapping oxygen gas in saturated porous media
for in situ bioremediation. Wat. Res. 12, 2687–2696.
[9] Klimant, I., Kuehl, M., Holst, G. and Glud, R.N. (1996) Optical
measurement of oxygen and temperature in microscale: strategies
and biological application. Sensors and Actuators B, Article 2176.
[10] Schramm, A., De Beer, D., Gieseke, A. and Amann, R. (2000)
Microenvironments and distribution of nitrifying bacteria in a
membrane-bound biofilm. Environ. Microbiol. 2, 680–686.
[11] Amann, R. and K€
uhl, M. (1998) In situ methods for assessment of
microorganisms and their activities. Curr. Microbiol. 1, 352–358.
[12] Kelly, W.R., Herman, J.S. and Mills, A.L. (1997) The geochemical
effects of benzene, toluene, and xylene (BTX) biodegradation.
Appl. Geochem. 12, 291–303.
[13] Pfiffner, S.M., Palumbo, A.V., Gibson, T., Ringelberg, D.B and
McCarthy, J.F. (1997) Relating ground water and sediment
chemistry to microbial characterization at a BTEX-contaminated
site. Appl. Biochem. Biotechnol. 63, 775–788.
[14] Collins, M.D. and Jones, D. (1981) Distribution of isoprenoid
quinone structural types in bacteria and their taxonomic implications. Microbiol. Rev. 45, 316–354.
[15] Boschker, H.T.S. and Middelburg, J.J. (2002) Stable isotopes and
biomarkers in microbial ecology. FEMS Microbiol. Ecol. 40, 85–95.
[16] White, D.C. and Ringelberg, D.B. (1997) Utility of signature lipid
biomarker analysis in determining in situ viable biomass, community structure, and nutritional/physiological status of the deep
subsurface microbiota. In: The Microbiology of the Terrestrial
Subsurface (Amy, P.S. and Haldeman, D.L., Eds.), pp. 117–134.
CRC Press, Boca Raton, FL.
[17] Weinbauer, M.G., Fritz, I., Wenderoth, D.F. and H€
ofle, M.G.
(2002) Simultaneous extraction from bacterioplankton of total
RNA and DNA suitable for quantitative structure and function
analysis. Appl. Environ. Microbiol. 68, 1082–1087.
[18] Amann, R., Fuchs, B.M. and Behrens, S. (2001) The identification
of microorganisms by fluorescence in situ hybridisation. Curr.
Opin. Biotechnol. 12, 231–236.
[19] Woodrow, P.T. and Duke, S.R. (2001) Laser-induced fluorescence
studies of oxygen transfer across unsheared flat and wavy air–
water interfaces. Ind. Eng. Chem. Res. 40, 1985–1995.
[20] Liebsch, G., Klimant, I., Frank, B., Holst, G. and Wolfbeis, O.
(2000) Luminescence lifetime imaging of oxygen, pH, and carbon
dioxide distribution using optical sensors. Appl. Spectrosc. 54,
548–559.
120
G.U. Balcke et al. / FEMS Microbiology Ecology 49 (2004) 109–120
[21] Weiss, H., Schirmer, M. and Merkel, P. (2001) Development of
novel reactive barrier technologies at the SAFIRA test site,
Bitterfeld. In: BAM Workshop In situ Reactive Barriers Versus
Pump-and-treat Methods for Groundwater Remediation, Berlin,
Germany, 18./19.10.2001.
[22] Sambrook, J., Fritsch, E.E. and Maniatis, T. (1989) Molecular
Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, NY.
[23] Schwieger, F. and Tebbe, C.C. (1998) A new approach to utilize
PCR-single-strand-conformation polymorphism for 16S rRNA
gene-based microbial community analysis. Appl. Environ. Microbiol. 64, 4870–4876.
[24] Lytle, C.A., Gan, Y.-D.M., Salone, K. and White, D.C. (2001)
Sensitive characterization of microbial ubiquinones from biofilms
by electrospray/mass spectrometry. Environ. Microbiol. 3, 265–272.
[25] Willems, A., de Ley, J., Gillis, M. and Kersters, K. (1991)
Comamonadacae, a new family encompassing the Acidovorans
rRNA complex, including Variovorax paradoxus gen. Nov.,
comb. Nov, for Alcaligenes paradoxus. Int. J. Syst. Bacteriol. 41,
445–450.
[26] Auling, G. (1993) Pseudomonades. In: Biotechnology: Biological
Fundamentals (Sahm, H., Ed.), Vol. 1, pp. 402–429. VCH,
Weinheim, Germany.
[27] O’Donnell, A.G., Goodfellow, M. and Minnikin, D.E. (1982)
Lipids in the classification of Nocardioides: reclassification of
Arthrobacter simplex (Jensen) lochead in the genus Nocardioides
(Prauser) emend. O’Donnell et al. as Nocardioides simplex comb.
nov. Arch. Microbiol. 133, 323–329.
[28] Oyaizu-Masuchi, Y. and Komagata, K. (1988) Isolation of freeliving nitrogen-fixing bacteria from the rhizosphere of rice. J. Gen.
Appl. Microbiol. 34, 127–164.
[29] Reineke, W. and Knackmus, H.-J. (1984) Microbial metabolism
of haloaromatics: isolation and properties of a chlorobenzenedegrading bacterium. Appl. Environ. Microbiol. 47, 395–
402.
[30] Dermietzel, D. and Vieth, A. (2002) Chloroaromatics in groundwater: chances of bioremediation. Environ. Geol. 41, 683–
689.
[31] Reineke, W. (2001) The Handbook of Environmental Chemistry,
Vol. 2 Part K, Biodegradation and Persistance. Springer, Berlin.