FEMS Microbiology Ecology 50 (2004) 123–132
www.fems-microbiology.org
Variations in microcolony strength of probe-defined bacteria
in activated sludge flocs
Morten M. Klausen 1, Trine R. Thomsen, Jeppe L. Nielsen, Lene H. Mikkelsen,
Per H. Nielsen *
Section of Environmental Engineering, Department of Life Sciences, Aalborg University, Sohngaardsholmsvej 57, DK-9000 Aalborg, Denmark
Received 16 December 2003; received in revised form 5 May 2004; accepted 14 June 2004
First published online 25 June 2004
Abstract
The strength of activated sludge flocs is important for the flocculation, settling and dewatering properties of activated sludge and
thus the performance of wastewater treatment plants. Little is known about how different bacteria affect the floc properties, so in this
study it was investigated whether the strength and other characteristics of large microcolonies within activated sludge flocs from a
full-scale nutrient removal plant varied significantly between different phylogenetic groups of bacteria. The investigation was carried
out by using a shear method for deflocculation of activated sludge flocs, combined with different chemical manipulations under
defined conditions. The identification and quantification of the microcolony-forming bacteria were conducted with group-specific
gene probes and fluorescence in situ hybridization. The focus was on the microcolonies and not on the entire sludge flocs. In general,
the results showed large difference in the strength and colloid-chemical properties of the different probe-defined microcolonies. By
applying extensive shear to the system, less than 12% of the microcolony biovolume of the Beta-, Gamma- and Deltaproteobacteria
and Actinobacteria could be disrupted, thus forming strong microcolonies. Alphaproteobacteria and Firmicutes formed weaker
microcolonies (42–61% could be disrupted by shear). For most groups, several intermolecular forces determined the strength of the
microcolonies: hydrophobic interactions, cross-linking by multivalent cations and perhaps entanglements of extracellular polymeric
substances. However, the dominant force varied between the various phylogenetic groups. The large difference between the different
phylogenetic groups indicated that only a few species were present within each group, rather than many different bacterial species
within each phylogenetic group had similar floc properties.
Ó 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.
Keywords: Activated sludge flocs; Deflocculation; Extracellular polymeric substances; Microcolony strength; Fluorescence in situ hybridization;
Phylogenetic groups of bacteria
1. Introduction
The success of the activated sludge process for
wastewater treatment is highly dependent on efficient
solid–liquid separation processes. The separation effi*
Corresponding author. Tel.: +45-963-58493/963-58535; fax: +45963-50558.
E-mail address: [email protected] (P.H. Nielsen).
1
Present address: DHI Water & Environment, Wastewater &
Process Technology Department, Agern Alle 11, DK-2970 Hørsholm,
Denmark.
ciency is largely affected by the sludge structure in terms
of particle size, shape, density, porosity and, in particular, by the fraction of small particles in the wastewater
treated [1,2]. Reduced floc stability and strength can
lead to deflocculation and increasing numbers of small
flocs and free bacteria in the wastewater treated, when
the flocs are exposed to hydrodynamic shear forces.
Besides leading to a deterioration of the effluent quality,
small particles also cause decreased dewaterability of the
sludge [3,4].
The activated sludge floc consists of inorganic and
organic particles produced in the floc or adsorbed from
0168-6496/$22.00 Ó 2004 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.femsec.2004.06.005
124
M.M. Klausen et al. / FEMS Microbiology Ecology 50 (2004) 123–132
the wastewater, bacterial microcolonies, and individual
bacteria, all embedded in a three-dimensional gel matrix of flocculated extracellular polymeric substances
(EPS), where filamentous microorganisms may form a
backbone for the structural integrity [4–7]. The
strength of activated sludge flocs is, in the same way as
that of other biological aggregates, dependent on the
interparticle forces between the different floc constituents. Common colloid-chemical interactions are assumed to be involved in the binding of the various
entities, such as DLVO-type interactions [8–10],
bridging of EPS by means of divalent [11–13] and trivalent cations [6], and hydrophobic interactions [13–
15].
The intermolecular forces can be altered by change in
ionic strength and composition, pH or other chemical
means. Recently, we have reported that also an active
aerobic metabolism of the bacteria is essential to
maintain strong aggregates on a short-term basis [16,17].
By adding certain substrates, stronger flocs were formed,
while other substrates yielded weaker flocs, indicating
the presence of several groups of bacteria with different
nutritional and floc-forming properties. In other words,
biological aggregates cannot be regarded simply as inert
organic aggregates obeying the general colloid-chemical
laws, because the biological mechanisms interact with
these in a dominating way.
Only very limited information is presently available
on the identity and properties of most floc-forming microorganisms in activated sludge. It is known that many,
if not most, bacteria within the flocs form microcolonies
[18–20] and also that some of these may be very strong
[21]. Thus, the strength of flocs will be associated partly
with the presence of weak or strong microcolonies of
certain bacterial species within the flocs and partly with
the properties of the other components, such as individual cells. Approximately 10% of the suspended solids
(SS) concentration can be eroded away by severe shear
forces, while even more can be disintegrated by adding
various chemicals, e.g., sulfide and detergents [22]. Thus,
further knowledge about important physico-chemical
and biological properties governing the strength of activated sludge flocs may be obtained by investigating the
strength and properties of microcolonies of different
species or groups of bacteria within the sludge flocs.
The aim of this study was to investigate whether the
strength of microcolonies within activated sludge flocs
from a full-scale nutrient removal plant varied between
different phylogenetic groups of bacteria and also to
evaluate the changes in microcolony strength when their
stability was affected by biological or physico-chemical
means. The investigation was carried out by combining
a shear test for deflocculation of activated sludge flocs
under defined conditions with identification and quantification of the microcolony-forming bacteria by fluorescence in situ hybridization (FISH).
2. Materials and methods
2.1. Activated sludge
The experiments were carried out with activated
sludge from the aeration tank of Aalborg East wastewater treatment plant (WWTP) (Aalborg, Denmark),
which is an advanced plant performing biological nitrogen and phosphorus removal as well as chemical
phosphorus removal by FeSO4 addition. The plant is
designed for 100,000 population equivalents and mainly
receives domestic wastewater. The sludge age (sludge
residence time) is 25–30 days. The activated sludge was
transported to the laboratory within 30 min from sampling. The experiments were started immediately upon
arrival at the laboratory. The total SS concentration in
the experiment was kept to 6.6 g SS/l with an organic
fraction of 65–70%. Floc and filament characteristics
were described according to Eikelboom and van Buijsen
[23].
2.2. Shear experiments
In all experiments, a baffled shear reactor, consisting
of a 105-mm diameter cylinder with four vertical baffles,
each 11 mm deep was used. A single bladed paddle of
width 5.0 cm and height 1.2 cm was placed centered
both vertically and horizontally in the reactor. A 600-ml
sludge sample was sheared for 2 h at 1560 rpm, corresponding to a turbulent shear rate G ¼ 2200 s1 . According to the model developed by Mikkelsen and
Keiding [22,24], this shear rate and 2 h duration at room
temperature would lead to a mass fraction of approximately 5–15% being released from the activated sludge.
The shear rate of 2200 s1 is very high but necessary in
order to break up the stronger microcolonies, while
other floc components eroded away also at lower shear
rates. The paddle speed was provided by a Heidolph
electronic mixer and controlled by tachometer measurements. The reactors could be covered with airtight
lids and flushing with nitrogen when anaerobic conditions were necessary.
2.3. Biological manipulations
Three different biological manipulations were applied
during shear experiments. One reactor was kept aerobic
at room temperature to ensure aerobic biological activity. The second reactor was kept at approximately 4
°C by placing the reactor in an ice bath during the shear
experiment (and 1 h before) to lower the aerobic biological activity, thereby causing reduced floc stability
and increased deflocculation [17]. In the third reactor,
anaerobic conditions were achieved by bubbling of oxygen-free nitrogen gas through the sludge for 1 h prior
to and during the deflocculation experiment, thereby
M.M. Klausen et al. / FEMS Microbiology Ecology 50 (2004) 123–132
inhibiting the aerobic biological activity and stimulating
anaerobic processes [16].
2.4. Manipulation of surface properties
Hydrophobic properties were manipulated by addition of the anionic detergent sodium dodecylsulfate
(SDS) to a final concentration of 3.5 mM prior to the
experiment to increase the repulsion between hydrophobic entities in the EPS matrix. The surface charge of
floc constituents was altered by adjusting the sludge pH
to 3 or 9 by addition of concentrated HCl or 1 M
NaOH, respectively. There was some buffer capacity in
the sludge; therefore, pH values were controlled and
adjusted regularly during the experiments. The pH was
7.2 in the untreated activated sludge.
The surface charge was also manipulated by affecting
the bridging of EPS by multivalent cations. In one reactor, sulfide was added to a final concentration of 3
mM under anaerobic conditions. This causes a reduction and precipitation of the strong bridging cation Fe3þ
as black FeS, thereby reducing the floc stability even
further, compared to the anaerobic conditions alone [6].
In a second reactor, the chelating agent ethylenediaminetetraacetic acid (EDTA) was added to a final
concentration of 6 mM to remove Ca2þ and Mg2þ from
the sludge matrix.
2.5. Preparation of samples for FISH
At the end of each experiment, 2 4 ml of sludge
were sampled and centrifuged for 2 min at 2200 rpm
(1300 g) to separate the floc fraction from the deflocculated material. The supernatant was checked and
found to be devoid of flocs/microcolonies larger than 6
lm. The supernatant was removed and the pellet was
washed 3 times with deionised water to remove loosely
attached matter. The pellets were fixed in 4% paraformaldehyde (PFA) for Gram-negative cells or 50% ethanol for Gram-positive cells. FISH was carried out as
described by Amann et al. [25].
To make sure that changes in microcolony size and
composition were due to the applied treatments and not
the fixation procedure, an untreated sample was fixed
according to the procedure described above, and another sample was prepared by embedding untreated
sludge in 0.5% agarose gel prior to fixation. PFA (4%)
was placed on top of the embedded sample, and the cells
were fixed for 3 h at 4 °C and then rinsed by buffer. To
reveal any changes in microcolony volume caused by the
fixation procedure, microcolony composition of a bacterial group with weak microcolonies according to the
microcolony biovolume reduction (MBR) results was
examined. No effects of the fixation procedure could be
observed.
125
2.6. Oligonucleotide probes and fluorescence in situ
hybridization
A range of 10 different general oligonucleotide probes
targeting different phylogenetic groups of bacteria
within the Bacteria domain were used to screen for numerically important microcolony-forming bacteria
within the sludge. The probes used targeted: Alphaproteobacteria (ALF968), Betaproteobacteria (BET42a),
Beta-one and Beta-two subgroups within Betaproteobacteria (BONE23a and BTWO23a), Gammaproteobacteria (GAM42a), Deltaproteobacteria (SRB385), many
Bacteroidetes (CF319a), most Actinobacteria (HGC69a),
many Firmicutes (LGC354b), and the Bacterial probes
(EUB-I, -II, -III mix).More details on these probes can
be found in the probeBase database [26]. Oligonucleotides were 50 -labelled with 5(6)-carboxyfluoresceinN-hydroxy-succinimide ester (FLUOS) or with sulfoindocyanine dyes (Cy3) (Thermo Hybaid, Germany)
2.7. Quantification of microcolonies
The numbers and average diameters of various FISHlabeled microcolonies were measured in 40 activated
sludge flocs using a Zeiss Axioskop 2 Plus epifluorescence microscope equipped with a 100 oil objective.
The average diameter of microcolonies was measured by
categorizing them manually in size intervals of 6 lm (6–
12, 12–18, 18–24, 24–30, 30–36, 36–42 lm). Sizes from 0
to 6 lm were excluded because they were mostly individual cells or very small microcolonies. A microcolony
size distribution could be calculated from the data
measured. By assuming spherical microcolonies and an
average microcolony diameter corresponding to the
middle of each size interval, the microcolony biovolume
of the bacterial groups before and after treatment could
be estimated. The relative strength of the microcolonies
was then evaluated by estimating the percentage MBR
caused by a given treatment. By comparing different
bacterial groups, it was possible to identify groups
forming the strongest or weakest microcolonies, respectively. By comparing changes in MBR for a specific
probe-defined group when different treatments were
applied, it was possible to find out which factors that are
of importance for the strength of microcolonies of that
particular group. The quantification method is not as
precise as for example use of confocal laser scanning
microscopy, e.g., [28], but much faster and sufficiently
precise to see major changes in microcolony biovolumes,
as was the objective of this study.
The number of flocs for quantification was chosen
based on a percentile bootstrap analysis [27]. The
method was tested on data set of 40, 60 and 80 flocs
(data not shown). Forty flocs were selected for the study
as the estimated average microcolony biovolume and
95% confidence interval did not differ significantly from
126
M.M. Klausen et al. / FEMS Microbiology Ecology 50 (2004) 123–132
2.8. Analytical methods
Suspended solids (SS) and volatile suspended solids
(VSS) were analyzed according to APHA [29]. The absorbance at 650 nm (ABS650) of the supernatant was
measured on a spectrophotometer after 2.0 min of centrifugation at 1300g. The total content of iron (Fe) and
calcium (Ca) in the sludge was analyzed after extraction
by HNO3 . Sludge (2 ml) was added to 8 ml of 0.5 M
HNO3 and extracted overnight. The extracted samples
were centrifuged for 8 min at 5400g and the supernatant
filtered through a 0.45-lm filter. The extract was analyzed for Fe and Ca using a Perkin–Elmer Analyst 100
atomic absorption spectrophotometer.
3. Results
The activated sludge flocs investigated could be
characterized as strong flocs with good flocculation and
settling properties with a sludge volume index (SVI) of
approximately 90–100 ml/g. The number of filamentous
bacteria was low (Filament index of 1–2) with Eikelbooms Type 0041 and Microthrix parvicella as the
dominant types. The community structure based on the
microcolony biovolume (for colonies with a diameter
above 6 lm) of the different bacterial groups was
evaluated over a one-month period by FISH in the
WWTP investigated and it showed a stable composi-
Microcolony Biovolume [µm3/floc]
the ones obtained for 80 flocs. In this investigation, the
percentile bootstrap method was applied to overcome
the fact that the measured size distributions from which
the microcolony biovolume are estimated do not follow
a normal distribution and further to reduce time-consuming counting of a large number of flocs to estimate
the confidence intervals of the size distributions and thus
the microcolony biovolume. All results are shown as
average with 95% confidence intervals unless stated
otherwise.
5000
4000
3000
2000
1000
0
Alp
ha
Be
ta
BO
NE
BT
WO
Ga
De
lta
mm
a
Ac
Fir
Ba
tin
mi
cte
cu
ob
roi
tes
ac
de
ter
s
ia
Fig. 1. Biovolume of microcolonies with a diameter above 6 lm of the
dominant bacterial groups in the activated sludge sampled on two days
within a month. Alpha: Alphaproteobacteria, Beta: Betaproteobacteria,
BONE and BTWO: Beta-one and Beta-two subgroups within Betaproteobacteria, Gamma: Gammaproteobacteria, Delta: Deltaproteobacteria.
tion (Fig. 1). The numerically most important microcolony formers belonged to the Beta-, Alpha-, and
Deltaproteobacteria. In each floc the number of microcolonies of the different groups above 6 lm in diameter varied between 1 and 3 (Table 1). An average
of 15 5 microcolonies was present in one activated
sludge floc ðn ¼ 40Þ. The average diameter of microcolonies (above 6 lm) in an activated sludge floc was
found to be 12 0.6 lm.
The strength of the various probe-defined microcolonies was investigated by changing factors affecting the
microbial activity and the surface chemistry of the flocs.
When these factors were changed an alteration in MBR
could be observed. The results are shown in Figs. 2–4
where they are compared to an untreated sample. The
changes in MBR for the different populations compared
to aerobic shear at room temperature (control) are
summarized in Table 2. A reduction in microcolony
biovolume indicates that the colonies with a diameter
above 6 lm were disrupted by the applied treatment
either to microcolonies smaller than 6 lm (disruption)
or to individual cells (erosion). The applied quantifica-
Table 1
Typical characteristics of microcolonies above 6 lm in diameter from the various bacterial groups (average of two sludge samples with each 40
flocs standard error on the mean)
Bacterial group
Percentage of total
quantified groups (%)
Number of colonies pr. floc
Average diameter (lm)
Alphaproteobacteria
Betaproteobacteria
Beta-one subgroup within Betaproteobacteria
Beta-two subgroup within Betaproteobacteria
Gammaproteobacteria
Deltaproteobacteria
Bacteroidetes
Actinobacteria
Firmicutes
Total/average
13
28
7
21
7
9
4
6
5
100
2 0.5
3 0.8
1 0.5
3 1.0
1 0.3
2 0.6
1 0.5
1 0.3
1 0.7
15 5.0
12 0.4
14 0.5
12 0.7
13 0.5
13 1.1
11 0.4
12 0.5
11 0.6
11 0.6
12 0.6
M.M. Klausen et al. / FEMS Microbiology Ecology 50 (2004) 123–132
100
127
100
A
A
80
80
60
60
40
40
20
20
100
B
80
60
40
20
0
100
C
80
Microcolony Biovolume Reduction [%]
Microcolony Biovolume Reduction [%]
0
0
-20
100
B
80
60
40
20
0
-20
100
C
80
60
60
40
40
20
20
0
0
-20
Alp
ha
Be
ta
BO
N
E
BT
WO
Ga
mm
a
De
lta
Ba
cte
Fir
Ac
mi
tin
cu
ob
rio
tes
ac
de
ter
tes
ia
Fig. 2. Microcolony biovolume reductions (MBR) of dominant bacterial groups by shear treatment at 20 °C (A), at 4 °C (B), and under
anaerobic conditions (C), compared to untreated sample. Alpha: Alphaproteobacteria, Beta: Betaproteobacteria, BONE and BTWO: Betaone and Beta-two subgroups within Betaproteobacteria, Gamma:
Gammaproteobacteria, Delta: Deltaproteobacteria.
tion method cannot distinguish between these phenomena. A significant MBR was not simply a shearing of the
enmeshed microcolony from the main floc because the
applied centrifugation procedure secured that all microcolonies larger than 6 lm were separated together
with the floc remains.
3.1. Effect of microbial activity on the microcolony
strength
Lack of microbial activity is known to cause some
deflocculation of activated sludge flocs and this was
studied using two different conditions during the shear
experiments. One reactor was kept aerobic at room
temperature to ensure aerobic biological activity, while a
reactor at 4 °C significantly lowered the biological activity. An anaerobic reactor inhibited the aerobic activity and, in addition, it may have stimulated the
anaerobic activity.
In the aerobic experiment at room temperature, the
microcolonies affiliated to the Alphaproteobacteria were
Al p
ha
Be
ta
BO
NE
BT
WO
Ga
mm
a
De
lta
Ba
Ac
Fir
tin
mi
cte
ob
cu
rio
tes
ac
de
ter
tes
ia
Fig. 3. Microcolony biovolume reductions (MBR) of dominant bacterial groups by addition of detergent (A), pH 3 (B), and pH 9 (C),
compared to untreated sample. Alpha: Alphaproteobacteria, Beta:
Betaproteobacteria, BONE and BTWO: Beta-one and Beta-two subgroups within Betaproteobacteria, Gamma: Gammaproteobacteria,
Delta: Deltaproteobacteria.
strongly affected by the shear applied with an MBR of
61%. Also the Firmicutes and Bacteroidetes were affected
(42% and 28%, respectively, Fig. 2A and Table 2), while
the other groups were hardly affected by shear.
When the temperature was lowered more bacterial
groups were affected by shear. Bacteria from the Betaproteobacteria and especially the Beta-one subgroup
were reduced in microcolony biovolume, suggesting that
the microcolony strength of these bacteria to some extent depended on short-term microbial activity (Fig. 2B
and Table 2). Also microcolonies from the Bacteroidetes
were further reduced in volume, compared to the
shearing at room temperature (shearing control, Table
2). The other groups were not affected, compared to the
shearing control.
The introduction of anaerobic conditions had a large
effect on the MBR of several groups of bacteria, especially Beta- and Deltaproteobacteria and Firmicutes
(Fig. 2C). This indicated that bacteria from these groups
depended on aerobic conditions to some degree to remain aggregated in the microcolony, or that anaerobic
conditions stimulated their disintegration.
128
M.M. Klausen et al. / FEMS Microbiology Ecology 50 (2004) 123–132
3.2.1. Effects of hydrophobic properties and surface
charge properties
The importance of hydrophobicity was investigated
by adding a detergent in order to weaken the hydrophobic interactions (Fig. 3A and Table 2). The microcolonies of the two bacterial groups, Deltaproteobacteria
and Firmicutes showed a large MBR (73% and 80%)
indicating that hydrophobic properties were important
to keep their microcolony structure intact. The Betaproteobacteria displayed a small but significant increase
in MBR compared to the sheared control, while the
other groups seemed not to be strongly dependent on
hydrophobic interactions.
Changes caused by surface charge were investigated
by adjusting the sludge pH to 3 to reduce the surface
charge from carboxylic groups in the sludge matrix.
However, low pH also increases the solubility of cationic
species (Fe3þ , Ca2þ , Mg2þ ) and weakens the charge
neutralization and bridging effects of these ions. Adjusting the sludge pH to 9 increased the surface charge
and thus the repulsion between the charged entities. The
impact of these treatments on the microcolony biovolume is presented in Figs. 3B and C and Table 2.
Compared to the shear control, low pH (pH 3) significantly reduced the strength of the microcolonies of
three bacterial groups: Deltaproteobacteria, Bacteroidetes, and Firmicutes. The other groups were almost
unaffected by the lowered pH except for the Alphaproteobacteria. These were positively affected by the low
pH, indicating strengthening of the intermolecular forces within these microcolonies. High pH (pH 9) had
some effect on the microcolonies from Betabacteria,
while the other bacterial groups were hardly affected.
100
Microcolony Biovolume Reduction [%]
80
60
40
20
0
-20
100
80
60
40
20
0
-20
Alp
ha
Be
ta
BO
NE
BT
W
O
Ga
mm
a
De
lta
Ba
Fir
Ac
tin
mi
cte
ob
cu
rio
tes
ac
de
t
e ri
tes
a
Fig. 4. Microcolony biovolume reductions (MBR) of dominant bacterial groups by addition of EDTA (A) and sulfide (B), compared to
untreated sample. Alpha: Alphaproteobacteria, Beta: Betaproteobacteria, BONE and BTWO: Beta-one and Beta-two subgroups within
Betaproteobacteria, Gamma: Gammaproteobacteria, Delta: Deltaproteobacteria.
3.2. Effects of surface properties and cations complexation
on floc strength
A series of experiments manipulating the colloidal
chemistry of the sludge flocs were conducted to see
whether the different bacterial groups exhibited different
physico-chemical properties. The first three experiments
targeted mainly the surface properties of the EPS components, while the last two primarily targeted the cations binding the EPS components together (and thus
indirectly the surface properties).
3.2.2. Effects of surface charge and cations
The importance of multivalent cations was investigated in two experiments (Fig. 4A). Removal of Ca2þ
and Mg2þ and perhaps some Fe3þ was conducted by the
Table 2
Summary of microcolony biovolume reduction (MBR) for the different probe-defined bacterial groups after different treatments
Phylogenetic Group
Shear control MBR (%)
Alphaproteobacteria
Betaproteobacteria
Beta-one subgroup
Beta-two subgroup
Gammaproteobacteria
Deltaproteobacteria
Bacteroidetes
Actinobacteria
Firmicutes
61
10
0
5
8
0
28
0
42
4°C
+
+
Anaerobic
+
+
SDS
pH 3
pH 9
EDTA
)
)
)
+
+
+
+
+
+
+
++
+
+
+
++
+
+
Sulfide
+
++
+
++
+
The results are compared to aerobic shear at room temperature (control, second column). Change in MBR compared to shear control: + or )
shows that the treatment is significantly different from the control (no overlap of the 95% confidence interval).
): less reduction, )15 to )60%
+: more reduction, 15–30%
++: more reduction, 30–60%
No sign: no significant change.
M.M. Klausen et al. / FEMS Microbiology Ecology 50 (2004) 123–132
chelating agent EDTA, while specific removal of ferric
iron was performed by adding H2 S under anaerobic
conditions to form FeS.
When Ca2þ and Mg2þ were removed by EDTA, a
large effect was observed on microcolony biovolume of
Deltaproteobacteria. Also Gammaproteobacteria and
Bacteroidetes were significantly affected, compared to
the shear control (Table 2), suggesting that cross-linkage
by cations was important to the strength of these colonies. The microcolonies of Betaproteobacteria and Actinobacteria were almost unaffected by the treatment.
By adding sulfide under anaerobic conditions
(Fig. 4B), the most pronounced effect was seen on the
microcolonies of the Deltaproteobacteria (Table 2). They
disintegrated more than in the anaerobic experiment
without sulfide adition (66% and 43%, respectively)
(Fig. 2) showing that specific removal of iron weakened
these microcolonies. Betaproteobacteria (mainly Betatwo) and Firmicutes showed an increase in MBR that
was comparable to the MBR under anaerobic conditions, indicating that the removal of iron was not able to
further disrupt these microcolonies.
129
4. Discussion
was not investigated in detail, but for example the
Betaproteobacteria actually seemed to consist of relatively few dominant species belonging to the genus
Nitrosomonas, Azoarcus (Beta-one), Rhodocyclus,
Thauera (Beta-two), and other unidentified bacteria
(unpublished results).
The average size of the identified microcolonies in
the flocs of 12 lm is in good agreement with results
obtained by Snidaro et al. [32]. The floc size distribution was not measured, but a typical floc was 50–75 lm
in diameter (data not shown). The average number of
15 microcolonies underestimates the total number because many microcolonies were smaller than 6 lm and
because the applied group-specific gene probes did not
target all microcolonies. The amount of bacteria present as individual cells or microcolonies less than 6 lm
were not directly quantified, but estimated using the
Bacteria probe mix to be less than 20% of all bacteria
in the sludge investigated. This shows that the large
microcolonies made up the dominant part of the biovolume of the flocs, which was also shown by Li and
Barbusinski [19,20]. It also shows that relatively few
bacterial species or groups made up the dominant part
of the individual flocs, thus being important for the floc
properties.
4.1. Overall community composition and microcolony
characteristics
4.2. Characterization of microcolony strength and properties
The dominating floc-forming bacteria in activated
sludge flocs in various industrial [28,30] and municipal
treatments plants [31] are almost always reported to
belong to the Betaproteobacteria. Also Gamma- and
Alphaproteobacteria, Actinobacteria and Firmicutes are
typically present as well, but in lower numbers. In the
nutrient removal plant investigated in this study, the
dominating large microcolonies also belonged to these
phylogenetic groups, and the relative composition was
very stable over the time period of 1–2 months in
agreement with the findings of Lee et al. [31]. Thus, the
results obtained in this study can be regarded as typical
for this plant over longer time periods, but whether it
can be extrapolated also to other plants requires further
studies of the exact population composition on different
plants.
The overall results showed that the large microcolonies of the different bacterial populations had substantial variations in response to the treatments applied,
such as shear, anaerobiosis and manipulation with the
colloid-chemical properties. This reflects that each 16S
rRNA gene probe defined group had specific microcolony-forming properties and thus different floc-forming
properties. This large difference is remarkable, since the
applied gene probes are not species-specific, but only
group-specific and thus indicates that only a few species
were dominating within each probe-defined group. This
The resistance to shear of microcolonies of the different bacterial groups showed large differences. The
strong shearing caused a disruption of microcolonies
from the Alphaproteobacteria and to some extent the
Firmicutes and Bacteroidetes, showing that bacteria
within these groups were kept together by relatively
weak forces. The other bacteria from the Beta-, Gamma-,
and Deltaproteobacteria and Actinobacteria were all very
resistant to shear. As the focus of this study is on the
large microcolonies, there may be other species or groups
present as individual cells or small microcolonies that
respond differently to shear or chemical modifications.
These are of course also important to the overall floc
properties, but were not dealt with in this study.
It is well known that cooling to 4 °C or anaerobic
conditions can cause activated sludge flocs to deflocculate when shear is applied, e.g., [16]. This was also observed here. Cooling caused two of the bacterial groups
to be further reduced in microcolony size, compared to
the sheared samples at room temperature, the Beta-one
group and the Bacteroidetes, so it seems that the direct
microbial activity, e.g., the formation of EPS-components [16], is required to keep the strength of the microcolonies intact. Deflocculation under anaerobic
conditions was dominated mainly by the Deltaproteobacteria, to which many sulfate reducers and some iron
reducers belong [33], and the Firmicutes, to which, e.g.,
130
M.M. Klausen et al. / FEMS Microbiology Ecology 50 (2004) 123–132
the genus Clostridium belongs. The observation that
Deltaproteobacteria to some extent deflocculated under
anaerobic conditions but not when cooled to 4 °C
indicate that their anaerobic activity promoted a
deflocculation.
In order to understand more about the important
colloid-chemical properties of the different probedefined bacterial groups, various chemical manipulations were carried out. These affect all floc components
such as adsorbed matter and the EPS ‘‘cloud’’ around the
various microcolonies [5,34] as well as the individual
microcolonies within the floc. We have focused only on
the large microcolonies in this study, so the results suggest which forces could be most important to maintain
and keep together the dominant microcolonies from a
certain probe-defined population. The experiments do
not give a definitive explanation for two reasons: each
probe-defined population may consist of several species
with different properties, and, most likely, several forces
are involved in the binding of a certain microcolony, so
the result will always be mixed.
Hydrophobic interactions have been suggested as one
of the important forces in adhesion of cells to surfaces
and also of importance in activated sludge [13,15]. It has
also been shown that variations among the hydrophobicity of various structures exist (e.g., different microcolonies) in activated sludge flocs [14], which is also
confirmed in this study. Mainly the colony-strength of
Deltaproteobacteria and Firmicutes seemed to depend on
hydrophobic interactions. The Gram-positive Actinobacteria were not affected by the added detergent despite
the fact that many Gram-positive bacteria are often
assumed to be relatively hydrophobic, e.g., the Nocardioforms [35].
Electrostatic interactions and cross-linkage by
multivalent cations of the EPS components on the cell
surfaces have been considered as the most important
forces in keeping the activated sludge floc intact
[11,17,36]. In our experiments we could also see that
these interactions were of importance to several types
of microcolonies. In general, increasing the surface
charge by increasing pH slightly reduced the strength
of several types of microcolonies, but the most pronounced effect was observed when the divalent cations
were removed from the floc matrix by complexation
or change in pH. When Ca2þ and Mg2þ were removed
by complexation, the Delta- and Gammaproteobacteria
and the Bacteroidetes were weakened, indicating that
their EPS components were cross-linked by these divalent cations. The Deltaproteobacteria were more affected by the removal of ferric iron by sulfide,
confirming the importance of cross-linking cations in
their EPS matrix. Thus, the colony-strength of Deltaproteobacteria seemed to depend on both hydrophobic interactions and cross-linkage with cations.
Firmicutes were affected by ferric iron and not EDTA,
indicating that Fe3þ was more important than the
divalent cations. Bacteroidetes were not affected at all
by sulfide, indicating that Fe3þ was less important
than divalent cations.
One of the gene probe-defined groups, the Betaproteobacteria, responded to several treatments applied, but the MBR was always below 40% stressing
that 60% remained intact. The explanation for this
high strength could be that many bacteria among the
Betaproteobacteria, and particularly the Beta-one
group, relied mainly on EPS entanglements which has
been considered as a possible mechanism of importance to floc strength in activated sludge [37]. Electron
micrographs of for example microcolonies of the
ammonium oxidizer Nitrosomonas also show a very
dense EPS matrix between the individual cells [38],
and it seems likely that entanglements of the EPS
components from several bacteria may take place.
Also Wilen et al. [39] observed that certain groups
were associated with stronger flocs and among these
were some nitrifiers.
The shear rate applied in the experiments was very
high, so the results cannot directly be compared to deflocculation problems at full-scale wastewater treatment
plants. Other floc components than the microcolonies,
e.g., individual cells, erode at much lower shear rates
and they may be of larger importance for the normal
operation. The highest shear rate sludge flocs usually are
exposed for in treatment plants is G ¼ 500 to 1000 s1
during dewatering, where any deflocculation can be very
critical for the outcome of the dewatering process [1,4].
Therefore, more detailed studies on the identity and
behavior of small microcolonies and individual cells and
other floc components that erode at this shear rate
should be carried out. However, in a survey of several
industrial treatment plants we have observed poor
flocculation and high turbidity in plants with a low
number of Betaproteobacteria (unpublished data), so it
is likely that the observed phenomena based on studies
of larger microcolonies are of more general significance.
In summary, the study clearly showed that various
bacterial species can affect the floc structure in completely different ways. This means that the relative
amount of the different bacterial groups both determines
the general structure and the properties of flocs in a
certain treatment plant. The composition also determines the change in stability when the bacteria are exposed to variations in shear, ionic environments,
substrate, and oxygen conditions. Thus, the ability to
link specific floc-forming properties to specific species
opens up very interesting perspectives manipulating the
floc properties to a much greater degree than is possible
today. Examples could be change in operational conditions so that good floc-formers are selected instead of
poor floc-formers or that physico-chemical enrichment
of specific groups (e.g., nitrifiers or polyphosphate-
M.M. Klausen et al. / FEMS Microbiology Ecology 50 (2004) 123–132
accumulating bacteria) can improve recycling of specific
bacteria during sludge management. In future studies,
this requires proper identification of all dominant flocformers and a linkage to their key physiological and
floc-forming properties.
Acknowledgements
The Danish Technical Research Council within the
research program ‘‘Activity and Diversity in Complex
Microbial Systems’’ supported this study. We thank M.
Stevenson and L. Wybrandt for technical assistance.
References
[1] Nielsen, P.H. (2002) Activated sludge – the floc. In: Encyclopedia
of Environmental Microbiology (Bitton, G., Ed.). Wiley, New
York.
[2] Wanner, J. (1994) Activated Sludge Bulking and Foaming
Control. Technomic Publishing Company, Inc., Basel, Switzerland.
[3] Mikkelsen, L.H., Gotfredsen, A.K., Agerbæk, M.L., Nielsen, P.H.
and Keiding, K. (1996) Effects of colloidal stability on clarification
and dewatering of activated sludge. Water Sci. Technol. 34, 449–
457.
[4] Bruus, J.H., Nielsen, P.H. and Keiding, K. (1992) On the stability
of activated sludge flocs with implications to dewatering. Water
Res. 26, 1597–1604.
[5] Keiding, K. and Nielsen, P.H. (1997) Desorption of organic
macromolecules from activated sludge: effect of ionic composition.
Water Res. 31, 1665–1672.
[6] Nielsen, P.H. and Keiding, K. (1998) Disintegration of
activated sludge flocs in the presence of sulfide. Water Res.
32, 313–320.
[7] Sanin, F.D. and Vesilind, P.A. (1999) A comparison of physical
properties of synthethic sludge with activated sludge. Water
Environ. Res. 71, 191–196.
[8] Gregory, J. (1989) Fundamentals of flocculation. Crit. Rev.
Environ. Cont. 19, 185–230.
[9] Hermansson, M. (1999) The DLVO theory in microbial adhesion.
Colloids Surf. B 14, 105–119.
[10] Zita, A. and Hermansson, M. (1994) Effects of ionic strength on
bacterial adhesion and stability of flocs in a wastewater activated
sludge system. Appl. Environ. Microbiol. 60, 3041–3048.
[11] Higgins, M. and Novak, J.T. (1997) Dewatering and settling of
activated sludges: the case for using cation analysis. Water
Environ. Res. 69, 225–232.
[12] Eriksson, L., Steen, I. and Tendaj, M. (1992) Evaluation of sludge
properties at an activated sludge plant. Water Sci. Technol. 25,
251–265.
[13] Urbain, V. and Block, J.C. (1993) Bioflocculation in activated
sludge: An analytical approach. Water Res. 27, 829–838.
[14] Nielsen, J.L., Mikkelsen, L.H. and Nielsen, P.H. (2001) In situ
detection of cell surface hydrophobicity of probe-defined bacteria
in activated sludge. Water Sci. Technol. 43, 97–103.
[15] Zita, A. and Hermansson, M. (1997) Effects of bacterial cell
surface structures and hydrophobicity on attachment to activated
sludge. Appl. Environ. Microbiol. 63, 1168–1170.
[16] Wilen, B.-M., Keiding, K. and Nielsen, P.H. (2000) Anaerobic
deflocculation and aerobic reflocculation of activated sludge.
Water Res. 34, 3933–3942.
131
[17] Wilen, B.-M., Nielsen, J.L., Keiding, K. and Nielsen, P.H. (2000)
Influence of microbial activity on the stability of activated sludge
flocs. Colloids Surf. 18, 145–156.
[18] Daims, H., Nielsen, J.L., Nielsen, P.H., Schleifer, K.-H. and
Wagner, M. (2001) In situ characterization of Nitrospira-like
nitrite-oxidizing bacteria active in wastewater treatment plants.
Appl. Environ. Microbiol. 67, 5273–5284.
[19] Barbusinski, K. and Koscielniak, H. (1995) Influence of substrate
loading intensity on floc size in activated sludge process. Water
Res. 29, 1703–1710.
[20] Li, D. and Ganczarczyk, J. (1991) Size distribution of activated
sludge flocs. Res. J. WPCF 63, 806–815.
[21] Frølund, B., Palmgren, R., Keiding, K. and Nielsen, P.H. (1996)
Extraction of extracellular polymers from activated sludge using a
cation exchange resin. Water Res. 30, 1749–1758.
[22] Mikkelsen, L.H. and Nielsen, P.H. (2001) Quantification of the
bond energy of bacteria attached to activated sludge floc surfaces.
Water Sci. Technol. 43, 67–75.
[23] Eikelboom, D.H. and van Buijsen, H.J.J. (1983) Microscopic
sludge investigation manual. TNO Research Institute for Environmental Hygiejne, Delft, the Netherlands. MG-TNO A 94a.
[24] Mikkelsen, L.H. and Keiding, K. (1999) Equilibrium aspects of
the effects of shear and solids content on aggregate deflocculation.
Adv. Colloid Interf. Sci. 80, 151–182.
[25] Amann, R., Ludwig, W. and Schleifer, K.H. (1995) Phylogenetic
identification and in situ detection of individual microbial cells
without cultivation. Microbiol. Rev. 59, 143–169.
[26] Loy, A., Horn, M. and Wagner, M. (2003) probeBase – an online
resource for rRNA-targeted oligonucleotide probes. Nucleic Acids
Res. 31, 514–516.
[27] Efron, B. (1994) The Jackknife, the Bootstrap and other resampling plans. Society for Industrial and Applied Mathematics
(SIAM), Capital City Press, Vermont, 92 pp.
[28] Schmid, M., Thill, A., Purkhold, U., Walcher, M., Bottero, J.Y.,
Urbain, V., Ginestet, P., Nielsen, P.H., Wuertz, S. and Wagner,
M. (2003) Characterization of activated sludge flocs by confocal
laser scanning microscopy and image analysis. Water Res. 37,
2043–5052.
[29] American Public Health Association (APHA). (1995) Standard Methods for the Examination of Water and Wastewater. 19th Edn. American Public Health Association,
Washington, DC.
[30] Juretschko, S., Loy, A., Lehner, A. and Wagner, M. (2002) The
microbial community composition of a nitrifying-dentrifying
activated sludge from an industrial sewage treatment plant
analyzed by the full-cycle rRNA approach. J. Syst. Appl.
Microbiol. 25, 84–99.
[31] Lee, N., Jansen, J.L.C., Aspegren, H., Dircks, K., Henze, M.,
Nielsen, P.H., Schleifer, K.-H. and Wagner, M. (2002) Population
dynamics and in situ physiology of phosphorus-accumulating
bacteria in wastewater treatment plants with enhanced biological
phosphorus removal operated with and without nitrogen removal.
Water Sci. Technol. 46, 163–170.
[32] Snidaro, D., Zatarian, F., Jorand, F., Bottero, J.-Y., Block, J.-C.
and Manem, J. (1997) Characterization of activated sludge floc
structure. Water Sci. Technol. 36, 313–320.
[33] Nielsen, J.L., Juretschko, S., Wagner, M. and Nielsen, P.H. (2002)
Abundance and phylogenetic affiliation of iron reducers in
activated sludge as assessed by fluorescence in situ hybridization
and microautoradiography. Appl. Environ. Microbiol. 68, 4629–
4636.
[34] Wilen, B.-M., Jin, B. and Lant, P. (2003) The influence of key
chemical constituents in activated sludge on surface and flocculation properties. Water Res. 37, 2127–2139.
[35] Stratton, H., Seviour, B. and Brooks, P. (1998) Activated sludge
foaming: what causes hydrophobicity and can it be manipulated to
control foaming? Water Sci. Technol. 37, 503–509.
132
M.M. Klausen et al. / FEMS Microbiology Ecology 50 (2004) 123–132
[36] Nielsen, P.H. and Keiding, K. (1998) Disintegration of
activated sludge flocs in presence of sulfide. Water Res. 32,
313–320.
[37] Mikkelsen, L.H. and Keiding, K. (2002) The shear sensitivity of
activated sludge: An evaluation of the possibility for a standardized floc strength test. Water Res. 36, 2931–2940.
[38] Wagner, M., Rath, G., Amann, R., Koops, R. and Schleifer, K.H. (1995) In situ identification of ammonia-oxidizing bacteria.
Syst. Appl. Microbiol. 18, 251–264.
[39] Wilen, B.-M., Jin, B. and Lant, P. (2003) Impacts of structural
characteristics on activated sludge floc stability. Water Res. 37,
3632–3645.