Journal of Environmental Sciences 21(2009) 1153–1161

A bioaugmentation failure caused by phage infection and weak biofilm

formation ability
FU Songzhe1,2 , FAN Hongxia1,2 , LIU Shuangjiang1 , LIU Ying3 , LIU Zhipei1,∗

1. State Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences, Beijing 100101, China.
E-mail: fusongzhe@hotmail.com
2. Graduate University of Chinese Academy of Sciences, Beijing 100039, China
3. Institute of Oceanology, Chinese Academy of Sciences, Qingdao 266701, China

Received 27 September 2008; revised 22 December 2008; accepted 07 January 2009

Abstract
Two biological aerated filters (BAF) were setup for ammonia removal treatment of the circulation water in a marine aquaculture. One
of the BAFs was bioaugmented with a heterotrophic nitrifying bacterium, Lutimonas sp. H10, where the ammonia removal was not
improved and the massive inoculation was even followed by a nitrification breakdown from day 9 to 18. The nitrification was remained
stable in control BAF operated under the same conditions. Fluorescent in situ hybridization (FISH) with rRNA-targeted probes and
cultivable method revealed that Lutimonas sp. H10 almost disappeared from the bioaugomented BAF within 3 d, and this was mainly
due to the infection of a specific phage as revealed by flask experiment, plaque assay and transmission electron observation. Analyses
of 16S rRNA gene libraries showed that bacterial groups from two reactors evolved differently and an overgrowth of protozoa was
observed in the bioaugmented BAF. Therefore, phage infection and poor biofilm forming ability of the inoculated strain are the main
reasons for bioaugmentation failure. In addition, grazing by protozoa of the bacteria might be the reason for the nitrification breakdown
in bioaugmented BAF during day 9–18.

Key words: marine culture; water circulation; ammonia removal; Lutimonas sp. H10; nitrification breakdown; fluorescent in situ
hybridization (FISH)
DOI: 10.1016/S1001-0742(08)62396-7

Introduction nia and suspended solids (SS). Moreover, BAF contains a

In aquaculture systems, ammonia is the major metabolic
waste from fish, which is excreted across the gill mem-
branes and in the urine. It is toxic to aquatic animals
even as low as 0.5 mg/L (Colt and Armstrong, 1981). The
removal of ammonia from circulation water, therefore, is
a key issue in aquaculture industry nowadays. However,
ammonia concentrations in aquaculture systems are so low
(about 0.5–1.0 mg/L) that they become the rate-limiting
factor of biological nitrification (Wheaton et al., 1994).
As a result, aquacultural nitrification biofilters operate at
a much low total ammonia nitrogen (TAN) concentra-
tion compared with those used in industrial wastewater
treatment. Under this situation, improving nitrification
efficiency of biofilters is a major challenge related to
recirculating aquaculture systems (RAS). A commonly
used biofilter in aquaculture industry to remove ammonia
is biological aerated filter (BAF). BAF has been developed
and widely applied over the world as an efficient biological
nitrification process due to its excellent performance at
much higher removal rates for organic substances, ammo-
* Corresponding author. E-mail: liuzhp@sun.im.ac.cn
granular media that provides a large specific surface area
for the development of biofilm.
One of the main drawbacks common to biological
nitrification processes is the requirement of a long start-
up period and slow recovery. These problems are mainly
due to the slow growth rate and low growth yield of
autotrophic nitrifying bacteria. In addition, high C/N ratio
also retards the accumulation of autotrophic nitrifying bac-
teria. One promising solution to short the start-up period is
inoculating heterotrophic nitrifiers. Heterotrophic nitrifiers
include taxonomically diverse groups of microorganisms,
such as Nocardia and Streptomyces (Jensen, 1952), which
are dependent on the external source of energy. Het-
erotrophic nitrification, unlike autotrophic nitrification, is
not regarded as a taxonomically significant characteristic
of heterotrophs, as it is not considered to be coupled
with energy metabolism and cell growth (Castignetti et
al., 1990). However, in contrast to autotrophic nitrifiers,
heterotrophic nitrifiers generally tend to grow more rapidly
with higher yield, require lower dissolved oxygen concen-
tration and can survive at low pH and are also favored
by high C/N ratio. Therefore, introducing and maintaining
heterotrophic nitrifiers in a complex microbial community
1154 FU Songzhe et al. Vol. 21

could allow more rapid rate of nitrification and raised
the possibility of using single stage nitrogen removal
in wastewater treatment system regardless of their lower
specific nitrification rates (Robertson et al., 1989). In the
field of aquaculture industry, the use of nitrifying cultures
as inocula to speed up the establishment of nitrification is
not so common. Only a few practices had been reported by
using enrichment culture of autotrophic nitrifying bacteria
to enhance the nitrogen removal (Satoh et al., 2003).
Despite its long-term use in bioremediation, bioaugmen-
tation with inoculated microorganisms remains controver-
sial, especially in the full-scale wastewater treatment. Al-
though some successful full-scale applications have been
reported (Stephenson and Stephenson, 1992), still, based
on other studies conducted in full-scale and lab-scale,
the effectiveness bioaugmentation remains questionable.
The insufficiency of ecological data about the fate and
activity of the inoculated microorganisms and about the
interaction of the indigenous microbial communities make
the interpretation of the results very confused (Rittmann
and Whiteman, 1994). Numerous studies have reported the
rapid disappearance of added microorganisms and explore
the reasons for causing bioaugmentation failure (Bouchez
et al., 2000; Alexander, 1999; Mueller et al., 1992). Some
references (Goldstein et al., 1985) also suggested that
the adaptation of the inoculated microorganisms, insuffi-
ciency of substrate, competition from indigenous biomass,
and grazing by protozoa can be the possible reasons for
the failure of the bioaugmentation. Therefore, to avoid
bioaugmentation failure, the initial strain selection is quite
important for application.
In this study, a heterotrophic nitrifying bacterium, Lu-
timonas sp. H10 was inoculated to a BAF to improve
the removal efficiency of ammonia under synthetic marine
culture circulation condition. Unfortunately, the removal
of ammonia was not improved compared to control BAF.
We examined the possible reasons, including the low
nutrient, biofilm forming ability of the inoculated strain,
the competition from the indigenous bacteria and phage
infection. In order to gain new insights into the specific
criterion for initial strain screening, especially for marine
aquaculture systems, many in vitro experiments, such as
flask experiment and biofilm forming ability assay were
designed to check the suitability of Lutimonas sp. H10 for
bioaugmentation.
1 Materials and methods
1.1 Bacterial strains, oligonucleotide probes and plas-
mids
Bacterial strains, oligonucleotide probes and plasmids
used in this study are listed in Table 1. Lutimonas sp. H10,
isolated from coastal sediment near Qingdao, China, was
able to oxidize ammonia under heterotrophic conditions by
using specific medium to detect the production of NO2 −
and NO3 − (data not shown). Strain H10 was cultivated in
marine broth 2216 (MB, Difco Laboratories, USA) or on
marine agar 2216 (MA, Difco Laboratories, USA), or in
LB medium. Other strains were grown in LB medium.
1.2 Reactors and operation conditions
Two identical BAFs (1.4 m in height and 0.30 m in
diameter, with a working volume of 100 L) combining with
250-L feeding tanks were setup to form a recirculating
aquaculture system (RAS), respectively. The bioreactor
compartment was packed to a height of 1.2 m with
bamboo ball media (a ball-shape media which made by
bamboo piece) having average 12 cm diameter. Both
BAFs was operated at 18–25°C with 1.2 h hydraulic
retention time (HRT). The dissolved oxygen concentra-
tions in the reactors were always maintained above 6
mg/mL and pH was maintained between 7.2 and 7.6.
The store synthetic wastewater contains (g/L seawater):
1.377 NH4 Cl, 3.5 NaHCO3 , 1.53 K2 HPO4 , 1.59 Na2 HPO4 ,
3.6 MgSO4 ·7H2 O, 0.005 FeCl3 ·6H2 O (Zhu and Chen,
1999). Prior to use, synthetic wastewater was diluted to the
responding concentration. Every 3 days, synthetic wastew-
ater was added to both BAFs, present at a concentration
of 0.15 mg/mL NH4 + -N. A final concentration of 5.8
mg/mL C6 H12 O6 was added as carbon source (C/N, 20:1).
Both BAFs were first operated under the same strictly
autotrophic nitrifying conditions for 2 weeks. One of the
BAF was bioaugmented during start-up period with 5%

Table 1 Bacterial strains, oligonucleotide probes and plasmids used in this study

Characteristics or usage Source or reference

Bacteria
Lutimonas sp. H10 Inoculated strain This study
Escherichia coli DH10B Positive control of biofilm forming ability assay Grant et al., 1990
Pseudoalteromonas piscicida For biofilm forming ability assay This study
Vibrio natriegens For biofilm forming ability assay This study
Enterobacter sp. For biofilm forming ability assay This study
Cyclobacterium marinum For biofilm forming ability assay This study
Pseudoalteromonas sp. For biofilm forming ability assay This study
Bacillus aquimaris For biofilm forming ability assay This study
Oligonucleotide probes
S-G-Lni-1237-a-A-18 Formamide content is 40% and target site is 1237–1254 This study
Plasmids
pBS-T Cloning of PCR products Tiangen
Lutimonas sp. H10 is deposited in China General Microbiological Culture Collection Center (CGMCC) as CGMCC 1. 3529, the GenBank/EMBL/DDBJ
accession number for 16S rRNA gene sequence of strain Lutimonas sp. H10 is AY500144.
No. 8 A bioaugmentation failure caused by phage infection and weak biofilm formation ability
(V/V) of the culture of Lutimonas sp. H10, another BAF
was used as a control. Prior to inoculating, Lutimonas sp.
H10 was grown in LB broth for 72 h at room temperature
(15–20°C) to a concentration of 8.9×107 CFU/mL.
1.3 Determination of biomass in BAF
The numbers of chemo-autotrophic ammonia-oxidizing
bacteria (AOB) and nitrite-oxidizing bacteria (NOB) were
determined by the most probable number (MPN) method
(Levin, 1992). The number of Lutimonas sp. H10 and the
total number of heterotrophic bacteria were determined
by colony forming units (CFU) method (Levin, 1992) on
marine agar 2216 (MA, Difco Laboratories, USA), in the
circulation water and in the biofilm formed at day 1, 3, 5,
7, 10, 15, 20, 30 during the experimental period.
1.4 Bacterial isolation and identification
For isolation of the heterotrophic bacteria, serially
diluted samples were spread onto low-organic seawater
medium (LOSWM) contained (per liter of seawater) 1.0
g peptone and 0.20 g yeast extract with 1.5% (W/V)
agar and incubated at 30°C for 3 d. Single colonies on
LOSWM plates were picked up on LOSWM slant and
incubated at 30°C for 2 d, and the purity of the isolates
was examined by microscopy. Then the 16S rRNA gene
of the isolates was amplified using two bacterial universal
primers, 27F and 1492R (Lane, 1991), and sequenced. The
similarity of the 16S rRNA gene sequence was performed
by using BLAST program (Altschul et al., 1990) on NCBI
(http://www.ncbi.nlm.nih.gov).
1.5 Fluorescence in situ hybridization (FISH)
A specific probe for Lutimonas sp. H10, S-G-Lni-
1237-a-A-18 (5 -TTAAGGATTCGCTCCCCC-3 ), was
designed by using ARB program (Strunk and Ludwig,
1997) and its specificity was also checked with the PROBE
MATCH tool of the Ribosomal Database Project (Wheel-
er et al., 1996). Time-dependent changes of Lutimonas
sp. H10 in the biofilm were monitored by FISH using
pure culture as a positive control. Sample fixation and
cryosectioning were made following a protocol (Lydmark
et al., 2006). Epifluorescence microscope (Olympus BH-
2, Japan) was used to examine the biofilm specimens.
Processed images were printed by using the software
package Adobe Photoshop 7.0. Meanwhile, the cultivable
bacteria in the circulation water and biofilm were also
counted by CFU method on LOSWM plates.
1.6 Biofilm forming ability assay
Biofilm forming ability assay were carried out according
to the modified microtiter plate test (Stepanović et al.,
2000; Challan Belval et al., 2006; Wakimoto et al., 2004).
In this experiment, Lutimonas sp. H10 and six indigenous
microorganisms of the circulation water (Table 1) were
used. These indigenous bacteria were the most common
and predominant ones in the circulation water. All ex-
periments were performed in triplicates. Biofilm forming
ability of a strain was presented as OD595 . Bacteria were
classified as introduced by Stepanović et al. (2000).
1155
Influence of yeast extract (YE, Difco Laboratories,
USA), glucose (Amresco, USA), sucrose (Amresco, USA),
EPS, and polysaccharide on the kinetics of bacterial groups
from non-bioaugmented reactor (natural bacterial groups)
and Lutimonas sp. H10 biofilm formation was also as-
sessed on 48-well polystyrene microplates. Free EPS of
biofilm were extracted according to the method of Michel
et al. (2007). Extraction of polysaccharide was performed
under stringent condition as described by Cordeiro et al.
(2003).
1.7 Biofilm sampling and DNA extraction
Biofilm was sampled after Lutimonas sp. H10 was
inoculated for 1, 20 and 30 d. The sampled depths were
approximately 0.2 m below the surface of the media. The
media were gently rinsed in sterile seawater, then biofilm
(in duplicates) was removed from the support media with
sterile blade and collected into sterile bottles with 50 mL
phosphate-buffer (130 mmol/L NaCl, 7 mmol/L Na2 HPO4 ,
3 mmol/L NaH2 PO4 , pH 7.4). Then the genomic DNA
extraction was performed as described by Lyautey et al.
(2005).
1.8 Construction and analysis of 16S rRNA gene li-
braries
The DNAs extracted from biofilm samples were sub-
jected to PCR amplification of 16S rRNA genes with
the universal primers, 27F and 1492R (Lane, 1991).
The amplified products were recovered from gel by
using gel extraction kit (Tiangen, China), and were
cloned into the pBS-T vector (Tiangen, China) and
transformed to E. coli TOP 10 in accordance with the
manufacturer’s instructions. Randomly selected clones
were sequenced by Hetaibio Company (Beijing, Chi-
na). Alignments of 16S rRNA gene sequences were
carried out with Cluster X software and sequence simi-
larities were obtained by performing BLASTN searches
(http://www.ncbi.nlm.nih.gov/BLAST/).
1.9 Flask experiments
To evaluate whether the rapid disappearance of Luti-
monas sp. H10 could be a result of competition from
indigenous biomass, 5 mL inoculum was added into the
100 mL sterile wastewater or non-sterile wastewater, re-
spectively. All combinations were performed in duplicates.
The co-culture plates were incubated at 28°C and samples
withdrawn daily for the determination of Lutimonas sp.
H10 densities by using LB agar plates and LOSWM plates.
1.10 Isolation and characterization of Lutimonas sp.
H10 bacteriophages
To isolate Lutimonas sp. H10-specific phage, inoculated
cultures from non-sterile wastewater were observed for
clearing, which signaled phage lysis. Cleared cultures were
treated with chloroform to a 2.5% final volume, shaken
vigorously to mix, and held overnight at 4°C. Supernatants
containing phages were decanted and were filtered by
using a 0.22-μm Millipore filter (Bedford, USA). The
filtered supernatants were used for plaque assay to confirm
1156 FU Songzhe et al.
the existence of Lutimonas sp. H10-specific phages. The
experiment of purification of phage, plaque assay, spot
test and single-step growth were conducted as described
by Michael et al. (2007) using double-agar-layer MA.
Host specificity was checked with all 40 isolated marine
bacteria. Purified phage stocks were stored at 5°C and were
examined by transmission electron microscopy (JEM-
1400, JEOL, Japan).
1.11 Chemical analyses
Analysis of ammonia, nitrite and nitrate was performed
according to the standard methods (APHA, 1998).
2 Results
2.1 Performance of the BAFs
Both the BAFs were firstly operated under the same
strictly autotrophic nitrifying conditions for 2 weeks. After
stabilization of the reactors’ performances, one BAF was
inoculated with Lutimonas sp. H10 and operated for 35
d. The results (Fig. 1) showed that no improvement was
observed on ammonia removal for bioaugmented BAF in
comparison with control BAF. The massive inoculation
was followed by a nitrification breakdown from day 9 to
day 18, in contrast, the nitrification remained stable in the
control BAF operated under the same conditions. Due to
the nitrification breakdown, a backwash was conducted
in both BAFs at day 18. From day 21, NH4 + -N removal
rate of both BAFs were improved gradually. At the end
of experiment, 88.5% and 86.9% of NH4 + -N removal rate
were reached for control BAF and bioaugmented BAF,
respectively.
2.2 Biomass in the BAFs
The number of total heterotrophic bacteria, AOB and
NOB were monitored and are presented in Fig. 2. Over the
start-up period, the number of total heterotrophic bacteria
remained stable. However, it was noticed that the number
of attached AOB and NOB decreased significantly at day 8
in the bioaugmented BAF, whereas they were still stable in
the control BAF. In addition, no Lutimonas sp. H10 could
be detected after day 3.
Fig. 1 Ammonia removal rate of the two biological aerated filters
(BAFs).
Vol. 21
Fig. 2 Variations in the number of bacteria in the two BAFs. AOB:
ammonia-oxidizing bacteria; NOB: nitrite-oxidizing bacteria.
2.3 Change of inoculated strain number
As revealed by FISH (Fig. 3), before inoculation, no
hybridization signal could be detected on samples taken
from both BAFs, showing that no Lutimonas sp. H10
exists in the BAFs. Positive control exhibited a strong
signal, indicating that Lutimonas sp. H10 can be detected
(Fig. 3a). Just after 2 h and 1 d of inoculation in the
bioaugmented BAF, the number of Lutimonas sp. H10 in
biofilm decreased significantly (Figs. 3b and 3c). Two days
later, the inoculated Lutimonas sp. H10 was still present,
but in such low numbers they could no longer be quantified
accurately (Fig. 3d). Five days later, only a weak signal can
be detected in the biofilm samples (Fig. 3e), indicating a
rapid disappearance of Lutimonas sp. H10.
The result also showed that the number of Lutimonas sp.
H10 in the liquid phase and biofilm both decreased signifi-
cantly within 5 days as revealed by cultivable method (Fig.
4). The total number of heterotrophic bacteria decreased
in liquid phase and increased in the biofilm. The result
was consistent with the observation of FISH. Furthermore,
the cultivation of strain H10 in wastewater revealed that
the number of Lutimonas sp. H10 in sterile wastewater
gradually increase and stabilized at day 4 (Fig. 5). In
contrast, the number of Lutimonas sp. H10 in non-sterile
wastewater decreased very sharply within 5 days, indicat-
ing the existence of inhibition from indigenous biomass.
2.4 Isolation and characterization of Lutimonas sp.
H10-specific bacteriophages CM-1
Phages were discovered in non-sterile wastewater. Inoc-
ulated cultures were observed for clearing, which signaled
phage lysis. Thereafter, following purification of phage,
plaque assay, spot test and single-step growth were con-
ducted to characterize the phage CM-1. Plaque forming
units (pfu) for CM-1 were 7.0×109 and spot test reaction
is positive. Host specificity experiment showed that phage
CM-1 could infect strain H10 (Fig. 6a), but not other
isolated strains, indicating that CM-1 may be a Lutimonas
sp. H10-specific phage. Its morphology was exhibited in
Fig. 6b. Cell lysis of Lutimonas sp. H10 was exhibited in
No. 8 A bioaugmentation failure caused by phage infection and weak biofilm formation ability
Fig. 3 Simultaneous in situ hybridization of vertical sections of the bioaugmented biofilm at 0 h (a), 2 h (b), day 1 (c), day 2 (d), and day 5 (e) with
FITC-labeled probe S-G-Lni-1237-a-A-18.
Fig. 4 Time-dependent changes of Lutimonas sp. H10 and total het-
erotrophic bacteria in biofilm and liquid phase.
Fig. 6c.
2.5 Biofilm formation ability of single and dual species
The comparisons between single and dual species
biofilms showed the existence of interspecies microbial
interactions. Biofilm interspecies relationships were based
on the comparison between dual-species biofilm character-
istics (ranking) and those from each single species biofilm.
The existence of synergistic or antagonistic interactions
in dual species biofilm formation was considered whether
the biofilm formation category of each single bacterium
was lesser or greater than that found for dual species
biofilms (Table 2). Evident antagonistic interactions were
found for groups of Lutimonas sp. H10 + Enterobacter
sp., Vibrio natriegens + Enterobacter sp., V. natriegens
+ Pseudoalteromonas sp., Pseudoalteromonas piscicida
+ V. natriegens, Enterobacter sp. + Pseudoalteromonas
1157
Fig. 5 Time-dependent changes of Lutimonas sp. H10 in sterile wastew-
ater and non-sterile wastewater. Duplicate test samples were plated for
each time interval and means of data were analyzed and plotted using
Excel (Microsoft Corp., USA).
sp.; mutualism/synergy in biofilm formation was found
with the association of P. piscicida + Enterobacter sp.,
P. piscicida + Cyclobacterium marinum, Lutimonas sp.
H10 + P. piscicida, Lutimonas sp. H10 + C. marinum,
Lutimonas sp. H10 + Bacillus aquimaris, V. natriegens +
C. marinum, P. piscicida + B. aquimaris, B. aquimaris
+ Pseudoalteromonas sp., V. natriegens + C. marinum,
V. natriegens + B. aquimaris. Neutral interaction was
apparently existent for groups of Lutimonas sp. H10 +
V. natriegens, Lutimonas sp. H10 + Pseudoalteromonas
sp., C. marinum + B. aquimaris, C. marinum + Pseudoal-
teromonas sp., P. piscicida + Pseudoalteromonas sp., V.
natriegens + Pseudoalteromonas sp., Enterobacter sp. +
C. marinum, Enterobacter sp. + B. aquimaris. The results
showed that the biofilm formation of inoculated strain was
inhibited by some indigenous bacteria.
1158 FU Songzhe et al. Vol. 21

Fig. 6 Infection of Lutimonas sp. H10 by the isolated phage (a), transmission electron micrograph of the isolated phage (b), and cell lysis of Lutimonas
sp. H10 by the phage (c).

Table 2 Biofilm-forming abilities of single and dual strains in synthetic 0.25 ± 0.01 to standardize the number of bacteria (107–108
wastewatera CFU/mL). Then, a sterile 48-well flat tissue culture plate
Bacterial group Sampling time (h) (Costar, USA) were filled under aseptic conditions with
24 48 72 180 μL of cell suspension and 20 μL solution of related
Escherichia coli DH10B ++++b +++++ +++++
carbon sources (20 mg/mL), and incubated at 30°C without
Lutimonas sp. H10 (H10) ++ +++ +++ shaking. Natural bacterial groups 200 μL and Lutimonas
Pseudoalteromonas piscicida (Y) +++ +++++ +++ sp. H10 cell suspension without adding related carbon
Vibrio natriegens (WG) +++++ +++++ +++ sources were added as a control. After 24, 48, 72, 96 h, the
Enterobacter sp. (CW) ++++ +++++ ++++
Cyclobacterium marinum (P) ++ +++ ++ medium was removed from each well and conducted with
Pseudoalteromonas sp. (FS) +++ +++++ +++++ the same procedure as mentioned in Section 1.6. All ex-
Bacillus aquimaris (R) +++ +++ + periments were performed in triplicates with three repeats.
H10 + Y ++++ +++++ +++ Kinetic studies show that the addition of glucose, sucrose,
H10 + WG ++++ +++++ +++
H10 + CW +++ ++++ ++ EPS and polysaccharide enhances natural bacterial groups
H10 + P +++ ++++ +++ adhesion more rapidly than Lutimonas sp. H10 (Fig. 7b and
H10 + FS ++ ++++ ++++ 7c). On the opposite, weak adhesion of Lutimonas sp. H10
H10 + R ++ +++ ++++
was observed in the control and in the presence of glucose,
Y + WG ++ ++++ ++++
Y + CW ++++ ++++ +++++ sucrose, EPS and polysaccharide. This could be explained
Y+P ++++ +++ +++++ by the fact that introduced species may lack of mechanism
Y + FS +++ +++++ ++++ for EPS or polysaccharides secretion.
Y+R ++++ +++++ +++
WG + CW ++ +++ ++++
Only in the presence of yeast extract, natural bacterial
WG + P +++ +++ ++++ groups and Lutimonas sp. H10 showed a similar trend
WG + FS +++ ++++ +++++ (Fig. 7a). This suggests that yeast extract and EPS or
WG + R ++ ++++ +++++ polysaccharide could play a different role: yeast extract
CW + P ++ +++ +++
CW + FS ++ ++++ ++++
could rather act as a nutrient source, whereas EPS or
CW + R +++ +++ +++ polysaccharide could rather be involved in cell adhesion
P + FS +++ ++++ ++++ (Wingender et al., 1999). Sugars such as glucose and
P+R ++ ++ +++ sucrose also seem to be involved mainly in the cell at-
R + FS +++ ++++ +++++
tachment of natural bacterial groups rather than Lutimonas
a According to a modified classification proposed by Stepanović et al.
sp. H10. Indeed, analysis of EPS extracted from a biofilm
(2006). b +: OD595 = 1∼2 ODc , weak biofilm producer; ++: OD595 = 2∼4
ODc : moderate biofilm producer; +++: OD595 = 4∼6 ODc : strong biofilm
of non-bioaugmented reactor showed that sugars were the
producer; ++++: OD595 = 6∼8 ODc , very strong biofilm producer; main compound of EPS. The addition of yeast extract led
+++++: OD595 > 8 ODc , super strong biofilm producer. ODc was defined to a rapid attachment of both bacterial consortium, but a
as the mean OD595 of three standard deviation values of the negative decrease attachment appeared after 48 h. As yeast extract
control.
is a rapid carbon source, it can be assumed that the bacteria
2.6 Adhesion of bacteria on a solid surface attached to polystyrene in the presence of yeast extract
and after 48 h, the consumption of yeast extract led to
We also studied the influence of biochemical factors on the decrease of the attachment. The results reveal that the
the adhesion of natural bacterial groups and Lutimonas sp. biochemical factors involved in the formation of a biofilm
H10 to a support substrate. Polystyrene 48-well microtiter are mainly polysaccharides or EPS.
plates were employed to study the factors that positively
influence the attachment of the bacteria to a growth sup- 2.7 Difference of bacterial communities between the
port. Briefly, the biofilm from non-bioaugmented reactor biofilms from control BAF and bioaugmented BAF
was sampled, then inoculated into 1/10 LB liquid medium. Bacterial communities of both BAFs were analyzed by
Biofilm samples and Lutimonas sp. H10 then incubated construction of 16S rRNA gene libraries and randomly
overnight until the absorbance at 600 nm of inoculum was sequence of the clones. The results (Table 3) showed that
No. 8 A bioaugmentation failure caused by phage infection and weak biofilm formation ability 1159

Fig. 7 Effect of substrates on the kinetics of natural bacterial groups (NBG) and Lutimonas sp. H10 attachment to polystyrene as a growth support. (a)
control and YE, (b) glucose and sucrose, (c) EPS and polysaccharide.

bacterial groups from two BAFs evolved very differently.
At the beginning of the experiment, the bacteria in both
BAFs were similar, CFB group were the dominant taxa and
present in large numbers, and then the α-Proteobacteria
and Planctomycetes; γ-Proteobacteria was also represent-
ed (Table 3). In the non-bioaugmented BAF, CFB group
decreased and γ-Proteobacteria increased in contrast from
day 1 to day 30. From day 20 to day 30, the majority
of dominant γ-Proteobacteria in the non-bioaugmented
BAF could be assigned to genus Acinetobacter; and α-
Proteobacteria remained a relatively stable proportion
during the whole experiment. However, in the bioaugment-
ed BAF, Planctomycetes decreased and α-Proteobacteria
increased from day 1 to day 30 and they presented in large
numbers at the end. It should, however, be noted that the
sequence of Thraustochytrids mitochondrial DNA consti-
tute almost 1/5 proportion of 16S rRNA gene libraries in
the bioaugmented BAF of day 20.
3 Discussion
Bioaugmentation of a BAF with induced a transient
nitrogen loss only during the inoculation. Comparing with
non-bioaugmented BAF, no significant improvement in
nitrogen removal was obtained in the long term run. In
our experiment, the strain was inoculated in a recirculating
system, suggesting that it would be not washout.
Our results suggest that colonization on substrate of in-
oculated strain is relatively weak, especially when EPS and
polysaccharide were added as carbon sources. On contrary,
natural bacteria showed strong colonization abilities under
all conditions, indicating that they will preferential attach
the packing media and out-compete the added strain. In
low-nutrient environments, such as seawater, it may be as-
sumed that oligocarbophilic-obligate bacteria are the first
colonizers of the system. Therefore, the weak attachment
Table 3 Bacterial communities in the two BAFs (% of the total clones sequenced)
of added strain to media may count for an important reason
for the bioaugmentation failure. Furthermore, there was
an antagonism interaction between the inoculated species
and indigenous ones, this also inhibited the growth of the
inoculated strain.
In marine ecosystems, phages are highly abundant,
ranging from 104 to 108 per mL (Bergh et al., 1989). It
has been estimated that 10%–20% of planktonic marine
bacteria are lysed by phages per day (Suttle, 1994). Phages
are a main mortality factor for marine bacterioplankton
(Fuhrman and Noble, 1995). Some references suggested
that these viruses may play a key role in nutrient cycling
in marine environments, such as increase the pool of
dissolved organic matter and limit the flow of nutrients
to higher trophic levels (Fuhrman, 1999; Suttle, 2005).
It has been estimated that up to a quarter of the photo-
synthetically fixed carbon in ocean is recycled back to
dissolved organic material by viral lysis (Wilhelm and
Suttle, 1999). In addition, by selective infection and lysis
of predominant bacterial populations, lytic phages control
bacterial community composition and sustain coexistence
of bacterial species (Thingstad, 2000). Therefore, phages
affect the abundance, diversity, and seasonal succession
of their hosts and regulate bacterial community through
host-specific infection and lysis (Suttle, 2005; Scanlan et
al., 2005; Wommack and Colwell, 2000). Lutimonas sp.
H10, isolated from coastal sediment, may also have its
own specific phage. In fact, it disappeared rapidly (within
3 days) in non-sterile wastewater due to the infection of a
Lutimonas sp. H10-specific phage CM-1.
In addition, the analysis of 16S rRNA gene libraries
confirmed a large existence of mitochondrial DNA of
some protozoa in the bioaugmented BAF. Protozoa are
feed on bacteria, and their overgrowth causes the decrease
of bacteria including AOB and NOB. As a result, the
nitrification breakdown of the bioaugmented BAF was
from day 9 to day 18.

1 2 3 4 5 6 7 8 9

Day 1 Control BAF 48.21 26.79 ND ND 8.93 7.14 ND ND 7.13

Bioaugomented BAF 34.55 20.00 5.45 7.27 16.36 ND ND 1.82 14.55
Day 20 Control BAF 18.18 24.24 10.60 34.85 4.55 ND ND ND 7.58
Bioaugomented BAF 8.82 33.82 5.88 16.18 4.41 4.41 16.18 4.41 5.89
Day 30 Control BAF 9.38 20.31 10.94 48.44 3.13 ND ND ND 7.81
Bioaugomented BAF 17.91 29.85 11.94 4.48 4.48 4.48 ND 8.96 17.90
1: CFB group; 2: α-Proteobacteria; 3: β-Proteobacteria; 4: γ-Proteobacteria; 5: Planctomycetes; 6: high G+C Gram-positive bacteria; 7: Thraus-
tochytriidae sp. mitochondrial DNA; 8: Verrucomicrobia; 9: others. ND: not detected.
1160 FU Songzhe et al.
We can thus conclude that phage infection and the
difficulty to attach to the media are the main mechanisms
for the failure of bioaugmentation of the studied BAF. Sea-
water environment is alien to the inoculated strains.
From this bioaugmentation failure, we can learn that
ecological considerations influence the outcome of bioaug-
mentation. The ecological background constitutes a major
barrier in the successful bioaugmentation performance.
The rapid decline in population size of active exogenously
inoculated microbial cells is also described in the past few
years (van Veen et al., 1997).
Therefore, to increase the probability of the initial
establishment and long-term efficacy of an inoculum in
wastewater, future strategies for bioaugmentation must
provide both predictable ecological selectivity (such as
adding a metabolic that is uniquely utilized by added
microorganisms and not being utilized by the indigenous
microbial or encapsulation of added microorganisms) and
protection from adverse circumstances (such as the resis-
tance to phage and production of antagonism compounds).
These ecological considerations should be taken seriously
and a thorough selection and investigation procedures are
required before applying these effective inoculants. The
selection of strains should be taken on the basic knowl-
edge of the kind of microbial communities. Therefore,
we are suggesting that the first phase of the selection
procedure should check the basic constrains for bioaug-
mentation, such as insufficiency of substrate, competition
from indigenous biomass, phage infection and grazing
by protozoa. If there is no obvious constrains for the
proliferation of added strains, then the second phase of
the selection procedure should be conducted through in
situ experiment. Inevitably, with most biological systems,
there will always be some elements of unpredictability
and uncertainty. However, this investigation procedure will
definitely reduce the failure risk for implementation of
bioaugmentation. From the research results obtained in this
study, we can learn that it is critically important to check
if there is phage infection and the competition from the
indigenous bacteria for added strain, especially for a ma-
rine aquaculture system. Therefore, we proposed that for a
marine aquaculture system, specifically, no phage-infected
or phage-resisted and no obvious competition from the
indigenous bacteria is very important criterion for strain
selection and should be seriously checked beforehand.
For our experiment, the selection of no phage-infected or
phage-resisted inoculants may have much higher success
probability.
Future studies will further characterize Lutimonas sp.
H10-specific phage CM-1. Basic phage-host interactions
and dynamics in their natural environment will also be
examined.
Acknowledgments
This work was supported by the Hi-Tech Re-
search and Development Program (863) of China (No.
2006AA100305). We would especially like to thank Mr.
Fangyong Shi and Mr. Benben Song for sampling assis-
tance and pilot plant maintenance.
Vol. 21
References
Alexander M, 1999. Biodegradation and Bioremediation (2nd
ed.) San Diego, CA: Academic Press. 453.
Altschul S F, Gish W, Miller W, Myers E W, Lipman D J, 1990.
Basic local alignment search tool. Journal of Molecular
Biology, 215: 403–410.
APHA, AWWA, WPCF, 1998. Standards Methods for the Exam-
ination of Water and Wastewater (9th ed.). Washington DC,
USA.
Bergh O, Borsheim K Y, Bratbak G, Heldal M, 1989. High abun-
dance of viruses found in aquatic environments. Nature,
340: 467–468.
Bouchez T, Patureau D, Dabert P, Juretschko S, Dore J, Delgenes
P et al., 2000. Ecological study of a bioaugmentation
failure. Environmental Microbiology, 2: 179–190.
Castignetti D, Palutsis D, Turley J, 1990. An examination
of proton translocation and energy conservation during
heterotrophic nitrification. FEMS Microbiology Letters,
54(1-3): 175–181.
Challan Belval S, Gal L, Margiewes S, Garmyn D, Piveteau P,
Guzzo J, 2006. Assessment of the roles of LuxS, S-ribosyl
homocysteine, and autoinducer 2 in cell attachment during
biofilm formation by Listeria monocytogenes EGD-e. Ap-
plied and Environmental Microbiology, 72: 2644–2650.
Colt J E, Armstrong D A, 1981. Nitrogen toxicity to crustaceans,
fish, and mollusks. In: Proceedings of the Bioengineering
Symposium for Fish Culture, Fish Culture Section. Ameri-
can Fisheries Society, Bethesda MD. 34–47.
Cordeiro L M C, Stocker-Wörgötter E, Gorin P A J, Iacomini
M, 2003. Comparative studies of the polysaccharides from
species of the genus Ramalinae lichenized fungi of three
distinct habitats. Phytochemistry, 60: 967–975.
Fuhrman J A, Noble R T, 1995. Viruses and protists cause
similar bacterial mortality in coastal seawater. Limnology
Oceanography, 40: 1236–1242.
Fuhrman J A, 1999. Marine viruses and their biogeochemical and
ecological effects. Nature, 399: 541–548.
Goldstein R M, Mallory L M, Alexander M, 1985. Reasons for
possible failure of inoculation to enhance biodegradation.
Applied and Environmental Microbiology, 50: 977–983.
Grant S G, Jessee J, Bloom F R, Hanahan D, 1990. Differential
plasmid rescue from transgenic mouse DNAs into Es-
cherichia coli methylation-restriction mutants. Proceedings
of the National Academy of Sciences, 87: 4645–4649.
Jensen H L, 1952. The coryneform bacteria. Annual Review of
Microbiology, 6: 77–90.
Lane D J, 1991. 16S/23S rRNA sequencing. In: Nucleic
Acid Techniques in Bacterial Systematics (Stackebrandt E,
Goodfellow M, eds.). Chichester, United Kingdom: John
Wiley. 115–175.
Levin M A 1992. Microbial Ecology: Principle, Methods and
Applications. New York, USA: McGraw-Hill Inc.
Lyautey E, Ten-Hage L, Lacoste B, Rols J L, Garabetian F,
2005. Analysis of bacterial diversity in river biofilms using
16S rDNA PCR-DGGE: methodological settings and fin-
gerprints interpretation. Water Research, 39: 380–388.
Lydmark P, Lind M, Sörensson F, Hermansson M, 2006. Vertical
distribution of nitrifying populations in bacterial biofilms
from a full-scale nitrifying trickling filter. Environmental
Microbiology, 8: 2036–2049.
Michel C, Jean M, Coulon S, Dictor M C, Delorme F, Morin
D, Garrido F, 2007. Biofilms of As(III)-oxidising bacteria:
formation and activity studies for bioremediation process
No. 8 A bioaugmentation failure caused by phage infection and weak biofilm formation ability
development. Applied Microbiology and Biotechnology, 77:
457–467.
Mueller J G, Resnick S M, Shelton M E, Pritchard P H, 1992.
Effect of inoculation on the biodegradation of weathered
Prudhoe Bay crude oil. Journal of Industrial Microbiology,
10: 95.
Rittmann B E, Whiteman R, 1994. Bioaugmentation: a coming
of age. Biotechnology, 1: 12–16.
Robertson L A, Cornelisse R, De Vos P, Hadioetomo R, Kuenen
J G, 1989. Aerobic denitrification in various heterotrophic
nitrifiers. Antonie van Leeuwenhoek, 56: 289–299.
Satoh H, Okabe S, Yamaguchi Y, Watanabe Y, 2003. Evaluation
of the impact of bioaugmentation and biostimulation by in
situ hybridization and microelectrode. Water Research, 37:
2206–2216.
Scanlan D J, Post A F, Joint I, Mann N H, 2005. Genetic diversity
of marine Synechococcus and co-occurring cyanophage
communities: evidence for viral control of phytoplankton.
Environmental Microbiology, 7: 499–508.
Stepanović S, Vuković D, Davić I, Savić B, Ŝvabić-Vlahović M,
2000. A modified microtiter-plate test for quantification of
staphylococcal biofilm formation. Journal of Microbiologi-
cal Methods, 40: 175–179.
Stephenson D, Stephenson T, 1992. Bioaugmentation for en-
hancing biological wastewater treatment. Biotechnology
Advances, 10: 549–559.
Strunk O, Ludwig W, 1997. ARB software package.
http://www.biol.chemie.tu-muenchen.de/pub/ARB/.
Suttle C A, 1994. The significance of viruses to mortality in
aquatic microbial communities. Microbial Ecology, 28:
237–243.
Suttle C A, 2005. Viruses in the sea. Nature, 437: 356–361.
1161
Thingstad T F, 2000. Elements of a theory for the mechanisms
controlling abundance, diversity, and biogeochemical role
of lytic bacterial viruses in aquatic systems. Limnology
Oceanography, 45: 1320–1328.
van Veen J A, van Overbeek L S, van Elsas J D, 1997. Fate and
activity of microorganisms introduced into soil. Microbiol-
ogy and Molecular Biology Reviews, 61: 121–135.
Wakimoto N, Nishi J, Sheikh J, Nataro J, Sarantuya J, Iwashita
M et al., 2004. Quantitative biofilm assay using a mi-
crotiter plate to screen for enteroaggregative Escherichia
coli. American Journal of Tropical Medicine and Hygiene,
71: 687–690.
Wheaton F W, Hochheimer J N, Kaiser G E, Krones M J,
Libey G S, Easter C C, 1994. Nitrification filter principles.
In: Aquaculture Water Reuse Systems: Engineering Design
and Management (Timmons M B, Losordo T M, eds.).
Netherlands: Elsevier. 101–126.
Wheeler Alm E, Oerther D B, Larsen N, Stahl D A, Raskin
L, 1996. The oligonucleotide probe database. Applied and
Environmental Microbiology, 62: 3557–3559.
Wilhelm S W, Suttle C A, 1999. Viruses and nutrient cycles in
the sea. BioScience, 49: 781–788.
Wingender J, Neu T, Flemming H, 1999. What are bacterial extra-
cellular polymeric substances? In: Microbial Extracellular
Polymeric Substances (Wingender J, Neu T, Flemming H,
eds.). New York: Springer. 1–19.
Wommack K E, Colwell R R, 2000. Virioplankton: viruses in
aquatic ecosystems. Microbiology and Molecular Biology
Reviews, 64: 69–114.
Zhu S M, Chen S L, 1999. An experimental study on nitrification
biofilm performances using a series reactor system. Aqua-
cultural Engineering, 20: 245–259.