System. App!. Microbio!.

11,312-319 (1989)

The Micro£lora Involved in Aerobic-Thermophilic Sludge Stabilization

REINHARD HENSEL l , WILHELM DEMHARTER2 , and REINHOLD HILPERT3

Botanisches Institut der Universitat Miinchen, 8000 Miinchen 19, Federal Republic of Germany

Received September 22, 1988

Summary
The microflora adapted to the habitat of aerobic-thermophilically treated sludge was investigated in a
laboratory-scale fermenter as well as in plants working in practice. As demonstrated by the immuno-
fluorescence technique, the dominant microflora of the laboratory model plant working under steady-state
conditions largely corresponds to that of several thermophilically operated municipial sludge stabilization
plants, indicating that the microflora of the model plant can be regarded as representative of the habitat of
aerobic-thermophilic sludge stabilization in general. As dominant members of this microbial community,
strains of Bacillus thermocloacae were identified; strains of Sphaerobacter thermophilus, Bacillus
sphaericus, Bacillus pallidus and Bacillus stearothermophilus are less abundant.
The influence of the operational parameters of the thermophilic process (e. g. quality of the substrate,
discontinuous or continuous operation mode, hydraulic retention time) on the microflora is discussed.

Key words: Thermophiles - Viable cell counts - Physiologically active microflora - Indirect immuno-
fluorescence - Bacillus - Thermus - Sphaerobacter

Introduction

Sewage sludge produced by mechanical and biological a sufficiently "stabilized" product exhibiting a dry matter
waste-water treatment is either fed back to the production content of about 5-6% and a ratio of biological (BOD)
circuit (e. g. as fertilizer in agriculture) or withdrawn from and chemical oxygen demand (COD) below 0.15.
it, (e. g. stored at special depositories or burned) depend- The sludge stabilization is usually done by anaerobical
ing on its content of toxic compounds. For either purpose digestion at mesophilic temperatures or, with increasing
the sludge must be reduced in volume and in its biochemi- frequencies over the last 10-15 years in the Federal
cal oxygen demand (BOD) for economical transport or Republic of Germany, by an aerobical treatment at ther-
storage. Therefore, the surplus sludge especially from large mophilic temperatures (50-60 Qq. This aerobic-ther-
plants (population equivalent;:;:' 10000) working with a mophilic stabilization uses the self-heating processes of
high sludge load requires an additional treatment yielding bacterial metabolism, thus not requiring any external

1 Present address: Max-Planck-Institut rur Biochemie, 8033 Martinsried, Federal Republic of Germany
2 Present address: Lehrstuhl rur Wassergiitewirtschaft der Technischen Universitat Miinchen, 8046 Garching, Federal Republic of
Germany
3 Present address: Messerschmitt-Bolkow-Blohm GmbH, 8000 Miinchen 80, Federal Republic of Germany
Abbreviations: DSM Deutsche Sammlung von Mikroorganismen und Zellkulturen (Braunschweig, FRG)
PBS phosphate buffered saline
BSA bovine serum albumine
IgG immunoglobuline G
BOD ~.iochemical oxygen demand
COD chemical oxygen demand
S.D. standard deviation
ABTS 2.2' -azinobis (3-ethylbenzthiazolinesulfonic acid)
 The Microflora Involved in Aerobic-Thermophilic Sludge Stabilization 313

energy source for heating. The main advantage of the caldolyticus" DSM 405; "B. caldotenax" DSM 406; "B. ther-
aerobic-thermophilic treatment results from a disinfecting modenitrificans" DSM 466; B. firmus DSM 12; B. thermog-
action leading to an inactivation of pathogenic microor- lucosidasius DSM 2524; B. subtilis DSM 10; B. pumilus DSM
ganisms and viruses (Strauch and Wassen, 1976; Strauch, 27; B. megaterium DSM 32; B. licheniformis DSM 13; B.
sphaericus DSM 461, 439, 1867; B. stearothermophilus DSM
1980; Riiprich and Strauch, 1984). 22,297,456,458; Thermus ruber DSM 1279; T. aquaticus DSM
The qualitative and quantitative effect of the ther- 625; "T. thermophilus" DSM 579.
mophilic treatment on the degradation efficiency will be Design and operation parameters of the laboratory scale fer-
discussed elsewhere (Demharter et al., in preparation). menters. The laboratory scale fermenters consisted of cylindrical
Here we focus on the composition of the microflora glass vessels (Biostat V, Braun Melsungen) with a total volume of
adapted to the anthropogenic habitat of the aerobic-ther- 51 (working volume: 2.51) thermostated at 60°C by an external
mophilic sludge stabilization to provide a basis for a heat exchanger. The aeration rate was 1-2Umin. The vapour of
biological understanding of this technological process. the used air was condensed by a cooler. The stirring rate was
Although several isolates have been characterized from varied between 700 and 800 rpm. The fermenters were fed either
continuously with the completely soluble waste water of a yeast
plants operating at high temperatures (von Steldern et al.,
factory or semi continuously with thickened surplus sludge (feed-
1974; Sonnleitner and Fichter, 1983; Ottow et al., 1984), ing interval: 12 h) of a municipal sewerage (MiincheniGrosslap-
no indications could be given by the authors as to how pen), respectively. The characteristics of the two different sub-
representative the isolated strains could be regarded for strates, as well as the operation parameters of both fermenters are
the respective habitat. The strikingly low viable cell counts summarized in Table 1.
as determined by von Steldern et al. (1974) and Ottow et Determination of technological parameters. The dry matter of
al. (1984) at least suggest that the microflora could only be a sample was determined as residue weight after drying the sam-
isolated with a low efficiency. ple for 12 h at 105°C, the organic dry matter as difference in
In our study we therefore estimated the quantitation of weight after glowing the dry residue for 4 h at 550°C. The deter-
mination of the chemical oxygen demand (COD) was performed
the isolation procedure by determining the counts of the
according to Wolf and Nordmann (1977), that of the biochemi-
whole physiologically active population in the respective cal oxygen demand (BOD) using a BOD equipment ("Sap-
biotope. For characterizing the typical micro flora of the romat") supplied by Voith Co., Heidenheim.
habitat of aerobic-thermophilic sludge stabilization, we Determination of the cell counts of bacterial cultures. The
focused first on the bacterial population established in a determination of the total cell counts as well as the counts of the
thermostated laboratory-scale fermenter operating under physiologically active bacteria using the reduction of 2-(p-
steady-state conditions. For analyzing the main represen- iodophenyl-)-3-p-nitrophenyl-5 -phenyltetrazolium chloride were
tatives of the microflora adapted to that habitat, this performed according to Zimmermann et al., (1978). The counts
model system seemed to be more suitable than plants of viable cells were determined after plating the homogenized and
diluted samples on freshly prepared agar plates and incubation at
working in practice, since changes of operation parame-
the temperature of the plant from which the samples were taken.
ters leading to shifts within the micro flora could be In order to prevent drying, the plates were incubated under a
avoided. Applying the indirect immunofluorescence vapour saturated atmosphere (incubation time: 3-4 d). For
technique we then checked as how representative this counting the viable cells in plants fed with sludge the media of the
population can be regarded also for the core population of plates were composed according to Ottow (1974) supplemented,
plants working in practice. however, with 100 mUl aqueous sludge extract (final pH: 8.5). In
the case of the laboratory scale fermenter fed with waste of a
yeast factory, the growth medium was made as previously
described (Hensel et al., 1986). All microbial determinations of
Material and Methods the laboratory plants were done over an observation period of
two months after an adaptation time of at least 4-6 weeks.
Microorganisms. The following strains were obtained from the Preparation of the antisera. A logarithmically grown pure cul-
Deutsche Sammlung von Mikroorganismen und Zellkernen ture of bacteria was harvested by centrifugation, washed three
(Braunschweig, FRG) and grown under the conditions recom- times with sterile phosphate-buffered saline (PBS, pH: 7.5) and
mended in its catalogue of strains (Claus et al., 1983): "Bacillus formalinized (1.5% formaline) at 37°C over night. After washing

Table 1. Operation para-

Fermenter I Fermenter II
meters of the laboratory
fermenters maintained at
Intake: source sludge from municipal sewerage yeast factory waste
60°C
COD (g O2/1) 60.0 20.0
BOD (g 0 211) 21.1 15.2
Dry matter (gil) 39.7 25.6
Organic dry matter (gil) 31.5 17.4
Operation mode semi continuously continuously
Retention time (d) 2.5 1
Load (g organic dry matterll x d 12.6 17.4
O 2 content (% saturation) 80 80

22 System. App!. Microbio!. Vo!' 1113

314 R. Hensel, W. Demharter, and R. Hilpert

the cells five times, the cell density was adjusted to an O.D'S7S = drop of saturated aqueous fructose solution and covered with a
1.0 with PBS. With this suspension rabbits were immunized by cover slide. The fluorescence-labelled bacteria were counted by
the following schedule: At three following days, 0.1, 0.2 and 0.3 epifluorescence microscopy using an ocular net square.
ml of the cell suspension were injected into the ear veins of the In order to check the quantitation of this labelling method in
rabbits, after one week the injection was repeated with 0.4, 0.6, such heterogenous substrate like sludge, a pure culture of B.
0.8 and finally with 0.8, 1.0, 1.0 m!. Four weeks later the rabbits stearothermophilus strain H 11 (cell counts: 2.8 X 109 cells/ml)
were bled. After removing the erythrocytes by clotting and cen- was diluted in different ratios with raw sludge labelled with an
trifugation, the immunoglobulines were enriched by ammonium antiserum against these cells and counted by epifluorescence rnic-
sulfate fractionation (precipitation at 37% ammonium sulfate roscopy. Within the experimental error « 20%) the mean value
saturation), redissolved in PBS, extensively dialyzed against the of the counted cells corresponded to the expected value.
same buffer and frozen at - 20°C. Physiological tests, cell wall analyses and molecular genetic
Determination of the antibody titer. The titer of the antibody analyses were performed as previously described (Hensel et a!.
preparations was determined by an immuno-enzymatic assay 1986).
using peroxidase conjugated anti-rabbit IgG antibodies from
goat: 0.2 ml of a bacteria suspension (homologous antigen,
O.D'S78 = 1.0) was incubated in Eppendorf centrifugation tubes Results
each with 0.1 ml of a dilution series of the antibody preparation
(geometrical dilution series in PBS) for 1 h at room temperature. The thermophilic microflora of the laboratory scale fer-
After centrifugation the supernatant was discarded. The pellet menters
was washed twice and then incubated with 100 fll anti-rabbit IgG
antibodies from goat (Serva, Heidelberg; 1 : 1000 dilution in PBS) Counts of the total, physiologically active, and viable
for 1 h at room temperature. After two washes with 0.75 ml PBS cells. In order to determine the composition of the aerobic-
the pellet was suspended in 0.1 ml from a freshly prepared solu- thermophilic micro flora adapted to sludge or to heavily
tion composed of 19.8 ml of 0.1 molJl sodium acetate pH 5.0, 0.2 loaded waste water independently of the fluctuations of
ml of ABTS (Sigma) stock solution (0.1 molll ABTS in 0.1 molll influent, aeration rate and temperature, we used two
sodium acetate pH 5.0) and 0.02 ul H 2 0 2 (30%). After 30 min laboratory scale fermenters as model plants maintained at
incubation in the dark 1 ml H 2 0 was added; the suspension was 60°C and operated under steady-state conditions. As
centrifuged and the extinction of the supernatant was measured intake we employed secondary sludge from the municipal
at 405 nm. The antibody solutions were diluted to the lowest
concentration yielding maximal extinction, frozen and stored at
sewerage of Miinchen-Grosslappen (influent of fermenter
-80°C. I) and - to demonstrate the influence of the substrate qual-
ImmunofluorescenclL technique. Sludge samples were pretre- ity on the microfloral composition - heavily loaded waste
ated by quick centrifugation (1 sec) to remove solids. Samples of water from a yeast factory as model substrate for indust-
0.05-0.1 ml were pipetted into 1 ml-Eppendorf centrifugation rial wastes (influent of fermenter II). Due to the
tubes and centrifuged for 3 min in an Eppendorf centrifuge. The heterogenenous composition of the sludge, this substrate
supernatant was discarded and 0.03 ml of antibody solution were could only be supplied semicontinuously (twice within 24
added to the pellet. After mixing, the suspension was incubated h), whereas the soluble waste from the yeast factory was
for 20 min on ice and afterwards washed twice with a 0.75% (w/ fed continuously.
v) bovine serum albumine (BSA) solution in PBS. Then 0.03 mlof
Due to the semicontinuous operation the counts of the
a 7.5% BSA solution was added, mixed and incubated again for
20 min on ice. After centrifugation and digcarding of the super- physiologically active cells in fermenter I showed a slight
natant, 0.03 ml of anti-rabbit IgG (raised in goat) conjugated increase within 3-4 h after feeding, followed by a slow
with fluorescein isothiocyanate (FITC, Serva, Heidelberg; 1 : 50 decrease back to the original level. Samples for all kinds of
dilution in 7.5% BSA solution) were added; the suspension was cell counts and for isolating strains were taken from this
mixed and incubated again for 20 min in the cold followed by a fermenter at the time of maximal physiological activity
two-fold wash in 0.75% BSA. whereas in the case of fermenter II sampling was done
For a qualitative assay the pellet was suspended in 0.05 ml irregularly over the time.
60% (v/v) glycerol in PBS and examined directly using an epi- As shown in Table 2, fermenter I exhibited eight-fold
fluorescence microscope (excitation filter: G 436, objective: Neo-
fluar 100 x 1.3 oil, Zeiss, Oberkochen). For a quantitative deter-
higher counts of physiologically active cells than fermenter
mination the pellet was resuspended in a defined volume of II, indicating a higher metabolic rate in this fermenter. The
0.75% BSA solution, diluted in the same solution and passed strikingly high counts of total cells, however, are due to
through a cellulose filter (pore size: 0.2 flm, 25 mm in diameter). the large portion of the physiologically inactive biomass in
The filter was then placed on a microscope slide wetted with a the raw sludge serving as substrate.

Table 2. Total, active and viable cell

Fermenter I Fermenter II
counts of the laboratory fermenters
fed with municipal sludge fed with yeast factory wastes

Total celis/mJl 15.5 X 10 9 ± 3.6 x 109 1.4 x 109. ± 0.4 x 10 9

physiologically
active cells/mJl 8.2 X 10 9 ± 2.3 x 10 9 1.1 X 10 9 ± 0.2 x 10 9
viable cells/mP "' 5.0 X 10 9 ± 1.1 x 10 9 0.5 X 10 9 ± 0.1 x 10 9

1 mean value ± S.D.; n = 5

 The Microflora Involved in Aerobic-Thermophilic Sludge Stabilization 315

After varying media compositions and culture condi- species (cross-linking of the peptide subunits by the dipep-
tions, 60% (fermenter I) or 50% (fermenter II) of the tide L-Lys - D-Asp; Schleifer and Kandler, 1972) and by
active population in the fermenters could be grown on 74% DNNDNA homology to B. sphaericus DSM 463.
agar plates. In the case of fermenter I the highest viable cell In the case of fermenter II fed with waste water from a
counts were, obtained on plates with Ottow-medium yeast factory, nearly the whole thermophilic micro flora,
(Ottow, 1974), supplemented with aqueous sludge extract which could be grown on plates - corresponding to 50%
(pH = 8.5), whereas in the case of fermenter II, a mixture of the physiologically active population - consisted of
of 10 fold diluted waste water of the yeast factory and 7g Gram-negative, red pigmented, non-sporeforming bacteria
KCl/l yielded the highest number of colonies on plates. identified as members of Thermus ruber (Hensel et al.,
Taxonomical characterization of the isolates. Fifty-five 1986). Mainly based on pigmentation of the colonies,
percent of the viable cell counts of fermenter I (corres- three strains could be differentiated (strain H 1, orange;
ponding to 34% of the physiologically active population strain H 2, carmine; strain H 3, cherry red) each costitut-
of that fermenter) were represented by a group of isolates ing about one third of the total.
consisting of at least 3 different but phenotypically and Bacilli were found to play only a subordinate role in this
genotypically related strains (strain S 6001, S 6025, S fermenter, representing about 0.3% of the physiologically
6026). These strains exhibited the classical features of the active bacteria. A group of seven Bacillus strains (strain H
genus Bacillus (positive Gram-staining, spore formation, 12-H18) was shown to be phenotypically and genotypi-
directly cross-linked peptidoglycan via mDpm) but dif- cally related and described as B. pallidus (Scholz et al.,
fered considerably from any Bacillus species so far 1987). A further Bacillus strain (strain H 11) also belong-
described. Therefore, these strains are described in an ing to the background flora could be identified as B.
accompanying paper as members of a new species named stearothermophilus by its 70-84% DNNDNA homology
B. thermocloacae (Demharter and Hensel, 1989). to several members of this species (DSM 22, DSM 297,
The second largest portion of the isolates from fer- DSM 456, DSM 458).
menter II with 36% (corresponding to 22% of the phy- Reexamination of the microfloral composition in the
siologically active population) was represented by strain S laboratory fermenters using the indirect immunofluoresc-
6022. This isolate exhibits unusual phenotypic and ence technique. In order to test the indirect immuno-
genotypic features not shared by any known genus. It is fluorescence technique as a tool for analyzing mixed popu-
described as representative of the only species Sphaerobac- lations grown on such inhomogeneous substrates as
ter thermophilus of a new genus (Demharter et al. 1989). sludge, we raised antisera against several members of
Strain S 6042 represented about one tenth of the viable species isolated from the laboratory fermenters and
cells. This strain belongs to the species B. sphaericus as checked their applicability as identification label in the
shown by its peptidoglycan structure typical for this original mixtures of the fermenters.

Table 3. Cross-reactions of the antisera against several isolates from the laboratory fermenters

No Strain antisera against No

1 9 10 13 14 15 16 18 19

1 B. pallidus H12 +
2 H13 +
3 H14 +
4 H15 +
5 H16 +
6 B. sphaericus DSM 461
7 DSM 493
8 DSM 1867
10 S 6042 +
11 B. stearothermophilus DSM 22
12 DSM 458
13 Hll +
14 B. thermo cloacae S 6001 +
15 S 6025 +
16 S 6026 +
17 T. ruber DSM 1279
18 HI + +
19 H2 + +
20 H3 + +
No cross-reactions were obtained with anyone of following strains:
"B. caldolyticus " DSM 405; "B. caldotenax" DSM 406; "B. caldovelox" DSM 411; B. firmus DSM 12; B. licheniformis DSM 13;
"B. thermodenitrificans" DSM 466; B. thermoglucosidasius DSM 2542; T. aquaticus DSM 625, "T. thermophilus" DSM 579.
316 R. Hensel, W. Demharter, and R. Hilpert

As shown in Table 3, the antisera used are specific for immunofluorescence staining of S 6042, S 6025 and 6022
the homologous antigens or at least for the species to in fermenter 1. As demonstrated in Table 4 the counts
which the homologous antigens belong. Thus the antisera obtained by immunofluorescence labelling largely corres-
against B. sphaericus S 6042 and the three B. thermo- pond to the viable cell counts of the respective isolates,
cloacae strains show strong strain-specificity whereas the even though the immunological method yields - with the
antisera developed against the T. ruber strains H 1 and H only exception of the immunolabelling of B. sphaericus -
2 as well as the antiserum against the B. pallidus strain H generally lower cell counts. This deviation between the
12 recognize all members of the respective homology counts of viable and immunolabelled cells may be
group. explained by the formation of higher cell aggregates upon
In some cases a slightly granular fluorescence appeared immunolabelling which resist disintegration by vortexing,
with unrelated strains, especially with lyzed Bacillus cells. thus yielding underestimates. An additional reason, espe-
However, this reaction pattern could usually be disting- cially for the low recovery of B. thermo cloacae, may be the
uished from the positive reactions resulting in uniformly existence of more serotypes within this species which
flourescent surfaces of the labelled bacterial cells. could not be differentiated, however, by morphological
Using the antisera listed in Table 3, we reexamined the and physiological criteria. On the other hand, the signifi-
composition of the microflora established in the labora- cantly higher counts of the immunolabelled cells of B.
tory scale fermenters. Fig. 1a-c shows examples of the shaericus S 6042 compared to its viable cell counts may
account for its incomplete recovery on plates due to inade-
quate growth conditions.
Despite these rather minor differences between the
counts of the viable and immunolabelled cells, the method
of immunolabelling proved to be suitable for identifying
certain strains or species in a complex population and
quantifying their portion of the respective population.

The microflora of several thermophilically operated

municipial sludge stabilization plants as determined by
immunofluorescence
The immunofluorescence method using the antisera
against the isolates from the laboratory fermenter fed with
sludge was applied to investigate to what extent the main
composition of the microflora established under steady
state conditions is also representative for thermophilic ally
operating sludge stabilization plants working in practice.
In all plants investigated (sewerages in Fassberg,
Gemmingen, Isenbiittel, Nettetal, Vilsbiburg) the surplus
sludge is stabilized by an aerobic treatment in two reactors
connected in series. Due to self-heating processes the reac-
Fig. 1. a-c. Cells of S 6022 (a), S 6025 (b), and S 6042 (c) in tor content reaches temperatures of 30-40 °C in the first
fermenter I after immunofluorescence labelling using the reactor and 50-65°C in the second. Within the observa-
homological antisera. tion period (August/September, 1984), both reactors
Magnification: 1200x. worked semiconinuously, with the only exception of the

Table 4. Counts of immunolabelled cells vs. viable cell counts of the laboratory scale fermenters

Isolate Immunolabeled cells/mil Viable cells/mil

Laboratory scale B. thermocloacae S 6001 1.0 X 10 9 ± 0.3 x 10 9

fermenter I S 6025 0.4 X 10 9 ± 0.2 x 10 9 2.8 X 109 ± 1.1 x 10 9
56026 0.5 X 109 ± 0.2 x 10 9
S. thermophilus S 6022 1.2 X 109 ± 0.15 x 109 1.8 X 109 ± 1.4 x 109
B. sphaericus 5 6042 1.5 X 109 ± 0.4 x 109 0.4 X 109 ± 0.2 x 109
B. pallidus H12-HI6 3.3 X 107 ± 3.0 x 107 n.d.
B. stearothermophilus Hll 4.5 X 10 7 ± 3.0 x 10 7 n.d.
Laboratory scale T. ruber HI-H3 4.5 X 108 ± 2.2 x 108 5.1 X 108 ± 1.3 x 108
fermenter II B. pallidus H12-HI6 1.8 X 106 ± 0.4 x 10 6 2.0 X 106 ± 1.6 x 10 6
B. stearothel'mophilus Hl1 < 105 < 105

1 arithmetic mean ± standard deviation; n =; 4-5

 The Microflora Involved in Aerobic-Thermophilic Sludge Stabilization 317

Table 5. Operation parameters and counts of total, active and viable cells in the thermophilic reactors of several municipal sludge
stabilization plants

Sludge Operation Load Mean Temper- Total Viable Physiologically

stabilization mode (g arg. dry hydraulic ature cells/ml cells/ml active cells/ml
plant matterllxd) retention range I
time (d) (0C)

FaSberg semicont. 2.0 6 50-55 1.0 X 10 10 2.5 X 10 8 5x 10 9

Isenbiittel semicont. 55-60 1.0 X 10 10 7.0 X 10 7 3.6 X 108
Nettetal semicont. 14.6 2.25 60-65 1.0 X 1010 n.d. 7.0 X 10 8
Vilsbiburg semicont. 5.14 5 50-55 2.4 X
10
10 4.0 x 10 8 2.0 X 109
Gemmingen discont. 3.2 7 40-45 1.7 X 10 10 1.3 X 10 9 4.0 X 10 8
I
within 7 days before the samples were taken (8.19. 1984)

Table 6. Counts of the immunolabelled cells (cells/ml) in the thermophilic reactors of several municipial sludge stabilization plants

Strain FaSberg Isenbiittel Nettetal Vilsbiburg Gemmingen

B. thermocloacae S 6001 3.4 x 108 0.8 X 10 7 1.2 X 10 7 0 1.7 X 10 7

S 6025 0 1.5 x 10 7 4.5 X 10 7 0 1.4 X 107
S 6026 2.1 x 10 8 5.5 X 10 7 6.5 X 10 7 6.6 X 10 8 1.4 X 10 7
S. thermophilus S 6022 0 5.0 x 10 7 1.6 X 10 8 0 1.9 X 107
B. sphaericus S 6042 1.0 x 10 7 4.0 X 107 1.0 X 10 8 0 0.7 X 10 7
B. pallidus H12-H16 0 1.5 x 10 7 3.5 X 10 7 1.6 X 10 8 0
B. stearothermophilus Hll 0 0 7.0 X 10 7 1.4 X 10 7 0
Sum of the immunolabelled cells 2.2 x 10 8 1.8 X 10 8 4.9 X 10 8 9.6 X 10 8 0.9x 10 8
Portion of the immunollabeled
cells within the physiologically 4.4 51 70 43 6.9
active population (%)

plant in Gemmingen. In the latter case the second reactor nounced in the microbial population of the plant of Vils-
was filled up semicontinuously over a one-we~k period biburg, in which even B. pallidus - together with B. ther-
and completely emptied at the end of the week. Depending mocloacae - is predominant. The micro flora of the model
on the operation mode this reactor reached thermophilic plant does not seem representative for the main flora of
temperatures only in the second half of the week. the plants of Fassberg and Gemmingen.
All microbial analyses were confined to the second,
thermophilically operated reactors of the respective plants.
The main technological operation parameters and the cell
counts of the respective reactors are summerized in Table Discussion
5. As shown, on the average, only about 20% of the phy-
siologically active population could be grown without For analyzing complex microbial populations we prop-
further optimization of the medium composition on plates ose a strategy comprising both, the classical methods, like
with Ottow-medium. taxonomical characterization requiring isolation and culti-
The counts of the immunolabelled cells varied between vation of the respective strains, and the indirect immuno-
0.1-8.6 X 10 8 cells/ml corresponding to 43-70% of the fluorescence technique, an identification method allowing
physiologically active population in the case of the reac- a qualitative and quantitative characterization of the flora
tors in Isenbiittel, Nettetal and Vilsbiburg, but only to in situ.
4.4-6.9% of that in the plant of Fassberg and the discon- This in situ identification seems to be suitable especially
tinuously operated reactor of Gemmingen (Table 6). in such cases where a rapid analysis is required (e. g. when
The microflora in the thermophilically operated reac- the microfloral analysis should be used as a rapid diagnos-
tors in Nettetal and Isenbiittel showed the highest similar- tic for th~ operation state of a biotechnological process or
ity to that of the laboratory fermenter I, as characterized when the micro flora to be investigated undergoes rapid
by the dominance of B. thermocloacae, S. thermophilus changes which should be monitored) - or where the mic-
and B. sphaericus. In these plants, however, the portion of robial community consists of members which can only be
B. pallidus and B. stearothermophilus - compared to the isolated with difficulty (Macario and Conway de Macario,
model plant - is significantly higher. This tendenGY is pro- 1985).
318 R. Hensel, W. Demharter, and R. Hilpert

As shown by the low portion of the viable cell counts the operation mode of the plant in Gemmingen, which
related to the counts of the physiologically active cells in prevents a stable thermophilic flora to be established, must
the different thermophilically operated municipial plants, be the cause of this discrepancy, other causes are relevant
the aerobic-thermophilic flora resists normal isolation for the atypical floral composition in the plant of Fassberg.
procedures and requires time-consuming optimization for Possibly, this deviation is related to an unusual substrate
its cultivation. Thus the immunological in situ identifica- composition, of which the strikingly high organic matter
tion is unequivocally advantageous for characterizing the content due to the large faecal portion of this sludge may
core population of that habitat as present in the different be indicative.
plants. Despite these exceptions and despite the rather limited
Since this method requires a collection of specific antis- data collection, the flora of the laboratory fermenter fed
era against the representatives of the population in ques- with sludge seems to be representative for the habitat of
tion, the respective strains must first be isolated and recog- the aerobe-thermophilic sludge stabilization.
nized as representative, which on the other hand requires a According to the presented data, B. thermocloacae rep-
sufficiently stable population and thus constant growth resents the most wide-spread and the most dominant
conditions. Therefore, to analyse the typical microflora species of this thermophilic microflora. S. thermophilus,
involved in the aerobic-thermophilic sludge stabilization B. sphaericus, B. pallidus, and B. staerothermophilus also
the respective studies were conducted in laboratory scale involved in the typical flora playa rather subordinate role.
fermenters as model plants imitating the technological The indirect immunofluorescence has proven to be a
process in its main characteristics but working under suitable in situ differentiation method for microorganisms
steady-state conditions. applicable even in such heterogenic substrates like sludge.
The isolation of 60% of the physiologically active popu- Specific antisera against additional strains of this habitat,
lation showed that the main flora of the laboratory fer- however, would be desirable not only for a more detailed
menter fed with surplus sludge could be characterized. knowledge of this microflora but also to provide a biologi-
Mainly, but not exclusively, this dominant population is cally based diagnostic method for optimally operating the
represented by bacilli: a majority (corresponding to 34% aerobic-thermophilic sludge stabilization process.
of the physiologically active population) could be assigned
to the newly described species Bacillus thermocloacae Acknowledgements. We thank Prof. O. Kandler for stimulat-
(Demharter and Hensel, 1989), a minority (corresponding ing discussions and Miss Silvia Laumann, H. Renz and K. Felkl
to 5% of the physi<1iogically active flora) were identified for excellent technical assistance. The help of Dr. Hornberger in
as members of B. sphaericus. Twenty-two percent of the immunizing procedures is greatly acknowledged. Thanks are also
physiologically active flora consist of S. thermophilus, a due to Dr. D. Grogan for critical reading of the manuscript. The
work was supported by the government of the Federal Republic
representative of a new, deep branch within the actinomy- of Germany (ministry of research and technology) under contract
cetes subdivision (Demharter et aI., 1989). PTB 8163.
The influence of the quality of the substrate on the mic-
rofloral composition of the aerobic-thermophilic biotope
becomes apparent by comparing the flora adapted to ther- References
mophilically 'treated sludge to that grown on the heavily
loaded waste water of a yeast factory: under comparable Claus, D., Lack, P., Neu, B.: Catalogue of strains. Deutsche
cultivation conditions, on this substrate a population has Sammlung von Mikroorganismen, Gottingen, Federal Repub-
been established which mainly consists of different strains lic of Germany.
of T. ruber whereas bacilli play only a subordinate role Demharter, W., Hensel, R.: Bacillus thermo cloacae sp. nov., A
(less than 1 % of the physiologically active flora). New Thermophilic Species from Sewage Sludge. System. App!.
When the floral composition of the laboratory fer- Microbio!. 11, in press (1989)
menter fed with surplus sludge was compared to that of Demharter, W., Hensel, R., Smida, J., Stackebrandt, E.:
several municipial plants working under similar tempera- Sphaerobacter thermophilus gen. nov., sp. nov., A Deeply
ture and substrate conditions, largely corresponding mic- Rooting Member of the Actinomycetes Subdivision Isolated
from Thermophilically Treated Sewage Sludge. System. App!.
robial populations could be found in 3 out of 5 plants Microbio!. 11, in press (1989)
(sewerages of Isenbiittel, Nettetal, Vilsbiburg). Hensel, R., Demharter, W., Kandler, 0., Kroppenstedt, R. M.,
Strikingly, the main flora of these municipial plants dis- Stackebrandt, E.: Chemotaxonomic and molecular-genetic
playa greater variety of species than that of the laboratory studies on the genus Thermus: Evidence for a phylogenetic
fermenter, indicating that the less constant operation con- relationship of Thermus aquaticus and Thermus ruber to the
ditions in practice prevent the establishing of a highly genus Deinococcus. Int. ]. System. Bact. 36, 444-453 (1986)
specialized dominant flora as favoured in the model plant Macario, A. J. L., Conway de Macario, E.: Antibodies for
working under nearly ideal steady-state conditions. Obvi- methanogenic biotechnology. Trends in Biotechno!' 3,
204-208 (1985)
ously, the load per volume of the plants - at least in the
.Qttow, J. c. G.: Detection of hippurate hydrolase among Bacil-
range of 2-13 g organic dry matter/lxd - does not severely lus species by thin layer chromatography and other methods.].
affect the floral composition. App. Bact. 37, 15-30 (1974)
On the other hand, the flora of the laboratory fermenter Ottow, J. c. G., Forster, W., Riiprich, W., Breitenbiicher, K.:
is not representative for the municipial plants of Mikroorganismen der exotherm-aeroben Schlamm-
Gemmingen and Fassberg. Whereas the discontinuity in stabilisierung. Landwirtsch. Forsch. 37, 268-276 (1984)
 The Microflora Involved in Aerobic-Thermophilic Sludge Stabilization 319

Ruprich, W., Srauch, D.: Technologische und hygienische Hygienisierung von Klarschlamm. 7. Mitteilung: Weitere
Aspekte der aerob-thermophilen Schlammstabilisierung. Kor- Untersuchungen an einem Verfahren zur aerob-thermophilen
resp. Abwasser 31, 946-952 (1984) Schlammstabilisierung. gwf-Wasser/Abwasser 121, 552-555
Scholz, T., Demharter, W., Hensel, R., Kandler, 0.: Bacillus (1980)
pallidus sp. nov. A new thermophilic species from sewage. Strauch, D., Wassen, H.: Hygienische Untersuchungen bei der
System. Appl. Microbiol. 9, .91-96 (1987) Umwalzbeliiftung von kommunalem Klarschlamm. g\vf-Was-
Sonnleitner, B., Fiechter, A.: Bacterial diversity in thermophilic ser/Abwasser 117, 129-135 (1976)
aerobic sewage sludge. II. Types of organisms and their Wolf, P., Nordmann, W.: Eine Feldmethode rur die Messung des
capacities. Eur. J. Appl. Microbiol. Biotechnol. 18, 174-180 CSB von Abwasser. Korresp. Abwasser 77, 277-279 (1977)
(1983) Zimmermann, R., Iturriaga, R., Becker-Birde, J.: Simultaneous
Steldern, D. von, Ottow, J. c. G., Loll, U.: Thermophile aerobe determination of the total number of aquatic bacteria and the
Sporenbildener bei der biologischen Reinigung hochkonzen- number therof involved in respiration. Appl. Environ. Mic-
trierter Abwasser. Z. aUg. Microbiol. 14,229-236 (1974) robiol. 36, 926-935 (1978)
Strauch, D.: Mikrobiologische Untersuchungen zur

Dr. Reinhard Hensel, Max-Planck-Institut fur Biochemie, D-8033 Martinsried