Environment Protection Engineering

Vol. 40 2014 No. 1

DOI: 10.5277/epe140106

ZOU HAIMING1, 2, LU XIWU1, SAAD ABUALHAIL3, SHI JING1, GU QIAN1

ENRICHMENT OF PAO AND DPAO RESPONSIBLE

FOR PHOSPHORUS REMOVAL AT LOW TEMPERATURE

A new strategy of enrichment of polyphosphate accumulating organisms (PAO) and denitrifying

polyphosphate accumulating organisms (DPAO) at low temperature ranging from 8 °C to 11 °C was
demonstrated through two lab-scale reactors operated in sequential anaerobic-aerobic (AO) or anaer-
obic-anoxic (AA) conditions. It was found that the AO reactor is able to achieve a good phosphorus
removal performance after 40 days of operation, while a similar stable phosphorus removal can be
obtained in the AA reactor after 80 days. This result suggests that the enrichment of PAO was easier
than that of DPAO at low temperature. Through switching batch tests, when DPAO is exposed to
aerobic conditions, it can immediately exhibit a good phosphorus removal similar to that under anox-
ic conditions, while PAO can only present poor phosphorus removal when exposed to anoxic condi-
tions, suggesting that two different types of Accumulibacter were enriched both in AA and AO reac-
tors. Microbial analysis with fluorescence in situ hybridization (FISH) and DAPI (4′,6-diamidino-2
-phenylindole) staining revealed that Accumulibacter was dominant both in the two reactors, account-
ing for 61.6% and 79.3% of all bacteria in AA and AO reactors, respectively. Although the different
amount of Accumulibacter was enriched in the two reactors, the similar microbial morphologies were
observed by using scanning electron micrograph (SEM), both presenting long-rod morphology. This
kind of Accumulibacter may display affinities for sodium acetate used as the carbon sources here.
This strategy proposed in this study was shown to be effective in achieving a very high enrichment of
Accumulibacter at low temperature by linking chemical analysis with microbial observation.

1. INTRODUCTION

The discharge of nutrient materials (i.e., nitrogen and phosphorus) from

wastewater to soil and waters may adversely affect water resources in some ways,

_________________________
1
School of Energy and Environment, Southeast University, Nanjing 210096, China, corresponding
author Xiwu Lu, e-mail xiwulu@seu.edu.cn
2
Department of Resource and Environment, Anhui Science and Technology University, Fengyang
233100, China.
3
Department of Civil Engineering, College of Engineering, University of Basrah, Basra, Iraq.
68 Z. HAIMING et al.

especially potential contributions to eutrophication. Nitrogen and phosphorus polluted

surface waters often need to be pretreated prior to use in drinking water systems such
as those from Taihu Lake, Chaohu Lake and Dianchi Lake, China. As phosphorus is
the limiting nutrient in algal blooms, phosphorus removal from wastewater, hence, has
become an important need to protect public health and reduce ecological risk. In order
to meet a stringent limit for phosphorus, enhanced biological phosphorus removal
(EBPR) is generally regarded as an economical and environmentally friendly technol-
ogy for the removal of phosphorus from wastewater due to its advantages relative to
conventional chemical precipitation method such as using ferric chloride and alumini-
um oxide.
In EBPR process, a group of bacteria are generally enriched through sequential
anaerobic-aerobic conditions, known as polyphosphate accumulating organisms
(PAO) responsible for phosphorus removal. During the anaerobic period, PAO take up
carbon sources, particularly volatile fatty acids (VFA) such as acetate and propionate,
store them as poly-β-hydroxyalkanoates (PHA), supplied with energy from the hy-
drolysis of polyphosphate (poly-P) (resulting in the release of phosphorus from the
cells of PAO) and glycolysis of glycogen. In the subsequent aerobic period, PAO take
up phosphorus in excess of the anaerobic release to store them as poly-P, usually
called luxury phosphorus uptake, simultaneously accompanying the growth of bio-
mass and the regeneration of glycogen with the required energy from the oxidation of
PHA stored in cells of PAO under the anaerobic condition [1]. The details of primary
characteristics of EBPR systems have been thoroughly studied [2–4]. Although EBPR
system has many advantages, it has difficulty of application to the treatment of
wastewater with low C/N ratio under conventional anaerobic-aerobic conditions due to
the competition for carbon sources between phosphorus release and denitrification [5].
Commonly, the most appropriate solution to the problem associated with the limitation
of carbon sources is the introduction of denitrifying polyphosphate accumulating or-
ganisms (DPAO) responsible for phosphorus removal under anaerobic-anoxic condi-
tion into EBPR system [6]. Compared with conventional EBPR running under anaero-
bic-aerobic conditions, nitrate can be used as an electron acceptor instead of oxygen
for simultaneous phosphorus removal and denitrification under anaerobic-anoxic con-
ditions, which can save aeration (30%), minimize sludge production (50%) and reduce
the demand for carbon sources (50%) [2].
In recent years, studies on DPAO have been performed in lab-scale or full-scale
EBPR system [6,7], most of which have mainly focused on the electron acceptors,
chemical parameters analysis and influence factors such as pH [8], carbon sources [9]
and temperature [10,11]. In contrast, temperature, among all these factors mentioned
above, seems to be the most beyond control, especially for practical operation [11]. In
general, temperature has a strongly influence on the nutrient materials removal per-
formance of EBPR system. When temperature varying from 15 °C to 30 °C, the lower
temperature, the greater amount of microorganisms responsible for phosphorus re-
 Phosphorus removal at low temperature 69

moval seems to be enriched in EBPR system [12]. However, there is no information

available on the running performance of EBPR system at even lower temperature such
as around 10 °C, often presenting in winter in China.
Thus, a new strategy for obtaining microorganisms responsible for phosphorus
removal (i.e., PAO and DPAO) at 8–11 °C was developed in this study. This proposed
strategy was performed using two lab-scale EBPR reactors, operating under both an-
aerobic-aerobic and anaerobic-anoxic conditions for 80 days. The effectiveness of the
strategy was assessed by both monitoring carbon, nitrogen and phosphorus removal
performance using chemical analysis and investigating the change of microbial com-
munity structure characteristics in each EBPR reactor using fluorescence in situ hy-
bridization (FISH), DAPI (4′,6-diamidino-2-phenylindole) staining, and scanning elec-
tron micrograph (SEM). Comparison of PAO and DPAO was also carried out to
investigate their activities in various supplied electron acceptors.

2. MATERIALS AND METHODS

Reactor setup and operation. Two lab-scale sequencing batch reactors (SBR) with
the working volume of 3.3 dm3 (Fig. 1) were conducted for phosphorus removal, one
operated with a sequence of anaerobic-aerobic conditions (AO) for PAO enrichment
and the other in anaerobic-anoxic cycling mode (AA) for DPAO enrichment.

Fig. 1 Schematic diagram of experimental setup: 1 – influent tank, 2 – influent pump,

3 – water inlet, 4 – air pump, 5 – mixer, 6 – ORP/pH meter, 7 – DO meter, 8, 9 – sample points,
10 – water outlet, 11 – waste sludge outlet, 12 – effluent pump, 13 – effluent tank

Seed sludge was withdrawn from an aerobic basin of the Chengdong Municipal
Wastewater Treatment Plant, Nanjing, China. The cycle time was 8 h and consisted of:
a 0.5 h filling period, 2 h anaerobic period, 4 h aerobic or anoxic period, 1 h settling
70 Z. HAIMING et al.

period and 0.5 h decant period. In each cycle, 1.9 dm3 of synthetic wastewater (com-
position detailed in Table 1) was fed to the reactor during the filling phase, resulting in
a 13.9 h of hydraulic retention time (HRT) and an effluent of the same amount as in-
fluent (1.9 dm3) was discharged at the end of one cycle. In the AA reactor, sodium
nitrate solution was pumped in the first 1 min of the anoxic period to provide anoxic
conditions. Volumes of 330 cm3 and 115 cm3 mixed liquor were removed per day
from AO and AA reactors, to maintain the solids retention time (SRT) at 10 and 20
days, respectively. Air was supplied at the flow rate of 1.5 dm3/min to maintain the
dissolved oxygen (DO) concentration higher than 2 mg/dm3 during the aerobic period.
pH in the two reactors was maintained at 7.0±0.2, with the addition of 0.5 M HCl or
0.5 M NaOH when the pH was above or below this setpoint. Two reactors were oper-
ated at room temperature, ranging from 8 °C to 11 °C. The liquor in the reactors was
completely mixed with an overhead stirrer (200 rpm) during both anaerobic and aero-
bic or anoxic period.

Table 1
Composition of wastewater used for experiment
Concentration Concentration
Feeds Nutrient solution
[g/dm3] [g/dm3]
CH3COONa 1.03 FeCl3·6H2O 1.50
KH2PO4 0.18 H3BO3 0.15
(NH4)2SO4 0.47 CuSO4·5H2O 0.03
CaCl2 0.02 KI 0.18
MgSO4·7H2O 0.18 MnCl2·4H2O 0.12
Nutrient solution 0.60 cm3/dm3 Na2MoO4·2H2O 0.06
ZnSO4·7H2O 0.12
CoCl2·6H2O 0.15
EDTA 10.00
COD:P = 20:1, pH = 7.0±0.2.

Two reactors responsible for the PAO and DPAO enrichment were operated for 80
days. The entire operation strategy was divided into two phases:
• Phase 1 (0–30 days). The concentrations of COD and PO 34− -P in the feed were
800 mg/dm3 and 40 mg/dm3, respectively, with the addition of 50 mg NO 3− -N per dm3
in anoxic phase, and the ratio of COD to P was 20:1, adopting high substrate load to
promote rapid growth of microorganisms responsible for phosphorus removal; no
waste sludge was discharged from the two reactors in this phase to maintain a large
amount of biomass.
• Phase 2 (30–80 days). The concentrations of COD and PO 34− -P in the influent
were decreased to 300 mg/dm3 and 20 mg/dm3, respectively, with the decrease of ni-
 Phosphorus removal at low temperature 71

trate to 30 mg of NO 3− -N /dm3 in anoxic phase, and their ratio was correspondingly

decreased to 15:1. Waste sludge was regularly removed from the two reactors to main-
tain the concentration of MLSS in AA and AO reactor 4.5 g/dm3 and 3.7 g/dm3, re-
spectively.

Batch tests. For the batch tests, 0.5 dm3 of activated sludge was taken from both
the AA and AO reactors at the end of aerobic and anoxic process at 80 days, and was
immediately washed twice with the nutrient solution (Table 1) not containing the basic
medium. The activated sludge treated from AA reactor responsible for anoxic phos-
phorus removal was divided into two parts and filled into two 1 dm3 test devices
(transformed triangular flask), One part was operated in an anaerobic-aerobic mode
and the other under anaerobic-anoxic conditions. The AO sludge treated was per-
formed following the same rule. AO sludge was exposed to anoxic conditions to moni-
tor its capability of using nitrate as the electron acceptor, and similarly, AA sludge
was exposed to aerobic conditions to monitor its capability of using oxygen as the
electron acceptor. Before the anaerobic phase, 0.5 dm3 of synthetic wastewater was
added to achieve the initial concentration of 300 mg COD/dm3, 20 mg PO 34− -P /dm3.
After a 2h anaerobic period, oxygen and nitrogen were supplied in the aerobic and
anoxic conditions, respectively, with the addition of 30 mg NO 3− -N /dm3 in anoxic
conditions as well.

Chemical analysis. The phosphorus removal performance of PAO and DPAO

from AO and AA reactors was monitored through chemical analytical techniques.
Samples were taken regularly from the two reactors and four batch test devices
throughout the study, sampled every two days during the PAO and DPAO enrichment
period and every 15 minutes for batch test. PO 34− -P and NO 3− -N were analyzed by the
segmented flow analysis (AutoAnalyzer3, SEAL, Germany). Chemical oxygen de-
mand (COD), mixed liquor suspended solids (MLSS) and mixed liquor volatile sus-
pended solids (MLVSS) were determined by the standard methods for the examination
of water and wastewater (APHA, 2002). Dissolved oxygen (DO), pH and oxidation-
-reduction potential (ORP) were measured by a DO meter analyzer (YSI 5100-230,
USA) and pH and ORP meter analyzer (YSI 100, USA).

Microbial analysis. Sludge samples collected from both the AA and AO reactors
were divided into two parts. One part was fixed in 4% paraformaldehyde at 4 °C for
3 h for fluorescence in situ hybridization (FISH) and the other in 2.5% glutaric
dialdehyde at 4 °C for 24 h for scanning electron micrograph (SEM). For comparison,
seed sludge afore-mentioned was also investigated by FISH and SEM. FISH was per-
formed according to the modified method of Kang et al. [13] to assess the evolution of
microbial populations in both reactors throughout the study. SEM was chosen to study
72 Z. HAIMING et al.

changes in the microbial morphologies. The 16Sr RNA oligonucleotide probes adopt-
ed for FISH are listed in Table 2. PAO651, PAO462 and PAO846 were applied to-
gether (PAOmix comprising equal amounts of those three probes) to target the
Accumulibacter, a known PAO. DAPI (4′,6-diamidino-2-phenylindole) staining was
performed for targeting the entire all bacteria [14]. FISH images were obtained with
the BioImaging Navigator fluorescence microscope (Olympus FSX100, Tokyo, Japan)
using a software of Olympus FSX-BSW.

Table 2
FISH probes used in this study
Probe mix Probe Sequence (5′–3′) rRNA Target Dye 5′
PAO651 CCCTCTGCCAAACTCCAG 651–668 Cy3
PAOmix PAO462 CCGTCATCTACWCAGGGTATTAAC 16S 462–485 Cy3
PAO846 GTTAGCTACGGCACTAAAAGG 846–866 Cy3

Quantification of the PAO community relative to the entire bacterial populations

was determined by THE FISH image analysis with the Image-Pro Plus software (ver-
sion 6.0, USA), following the method reported by Gao et al. [15]. Fluorescence in situ
hybridization was conducted following the below procedures:
• Samples fixation. Fixed sludge samples with a 4% paraformaldehyde as mentioned
above were centrifuged (5000×g RCF for 5 min) and washed twice in 0.01 M PBS, and
then resuspended in a PBS ethanol solution (1:1, v/v), being stored at –20 °C.
• Fixed samples dehydration. Before dehydration, the fixed sludge samples were
homogenized using a homogenizer fr 3 times for 30 s. Homogenized samples were
spotted onto slides treated with 3-aminopropyl-triethoxysilane (APES) acetone (1:50,
v/v) and then dried for 30 min at 40 °C. Dried slides containing sludge samples were
dehydrated for 3 min in 50%, 80% and 100% (v/v) ethanol and allowed to air dry, and
then stored in a desiccator prior to hybridization within FISH experiments.
• Hybridization. Hybridization of the treated sludge samples was performed for
2.5 h at 46 °C in a 0.001 cm3 probe (50 µg/cm3, Table 1) and in a 0.009 cm3 hybridiza-
tion buffer solution (pH 7.2) containing 0.9 M NaCl, 20 mM tris-HCl, 0.01 % sodium
dodecyl sulfate (SDS) and 35% formamide for PAOmix.
• Washing samples: when fluorescent in situ hybridization for sludge samples
completed, a stringent washing step was performed 4 times for 5 min in a washing
buffer solution (48 °C, pH 7.2) containing 20 mM tris-HCl, 5 mM EDTA, 0.01%SDS,
50 mM NaCl. And then these slides attached sludge samples were washed with double
distilled water and immediately dried at 30 °C for 2 h, and then were observed under
a fluorescence microscope.
For SEM observation, fixed AA and AO sludge samples were rinsed 3 times (for
15 min each time) with a 0.1 M phosphate buffer (PB), pH 7.0. For a better preserva-
 Phosphorus removal at low temperature 73

tion of the cell structure within microorganisms, treated sludge samples were fixed
again in 0.1% acetic acid for 2 h, and then rinsed using the same method mentioned
above. Fixed samples were dehydrated in an ethanol series, 50%, 70%, 80%, 90%,
100% ethanol (v/v) for 15 min each. Ethanol in dehydrated samples was displaced by
1:1 (v/v) of ethanol to isoamyl acetate for 30 min with slightly shaking and then by
100% isoamyl acetate for 30 min. These treated samples dried by the CO2 critical
point drying. The treated sludge samples were observed after spray-gold treatment,
using a scanning electron microscope (JSM-6360LV, Japan).

3. RESULTS

The EBPR performance of AO and AA reactors running under anaerobic-aerobic

and anaerobic-anoxic conditions, respectively, was investigated throughout the en-
richment to PAO and DPAO at low temperature. During this process, the change of
microbial community structure in each EBPR reactor was observed. Moreover, com-
parison of PAO and DPAO was also carried out to investigate their activities through
switching tests. For comparison, seed sludge was simultaneously conducted to investi-
gate its capacity of phosphorus removal and microbial characteristics.
3.1. PERFORMANCE OF AA AND AO REACTOR

The phosphorus removal performance throughout the process of microorganisms

acclimatization in the both EBPR reactors at low temperature (8–11 °C) over the
80 day time period is shown in Fig. 2. During the AO reactor operation under anaero-
bic-aerobic conditions, a stable phosphorus removal performance was presented after
40 days, as given by the variation of phosphorus concentration in the effluent, while
the AA reactor reached the similar stable state phase after approximately 80 days running
under anaerobic-anoxic conditions. At the same operation parameters, each reactor re-
sponded differently throughout the PAO and DPAO enrichment experiments, which
DPAO acclimatization to attain stable state required one times more than PAO, indicating
the higher activities of PAO at low temperature than DPAO. At the stable state phase, both
the AA and AO reactors exhibited a good phosphorus removing performance and the ef-
fluent phosphorus concentrations were both lower than 0.5 mg P/dm3. Consistently, the
stable concentrations of MLSS, MLVSS, phosphorus release and uptake and the con-
stant ratios of MLVSS to MLSS and phosphorus release to phosphorus uptake were
also observed both in the AA and AO reactors.
The amount of phosphorus stored in the microorganisms such as PAO and DPAO
can be generally implied according to the ratio of MLVSS to MLSS, and the lower
ratio, the greater amount of phosphorus may be stored in microbes responsible for
phosphorus removal. At the end of acclimatization of PAO and DPAO in the two reac-
tors, the average MLVSS and MLSS concentrations were 2.6 g/dm3 and 3.7 g/dm3,
3.5 g/dm3 and 4.5g/dm3, respectively, exhibiting that their ratios were 0.70 and 0.78,
74 Z. HAIMING et al.

respectively. These results implied that a higher amount of phosphorus was stored in
the AO sludge than in AA sludge per gram of biomass.
50
P release

Amount of P release or uptake

45 3
20 P uptake
40 mg P. /dm of influent
PO4 -P concentration of effluent / (mg¡¤dm )

0~30 days
-3

/ (mg P g MLSS)
40

-1
35 · 10

30

25 0
AO AA
Types of sludge
20
3
20 mg P. /dm of influent
15 30~80 days
3-

10
Steady state, AA
5 Steady state, AO

0
0 10 20 30 40 50 60 70 80
Time/day

Fig. 2. Phosphorus profiles during the enrichment of PAO and DPAO

A typical key phosphorus biochemical transformation responsible for EBPR was ob-
served both in the AA and AO reactors through a cycle batch test performed at the end of
acclimatization study, as shown in Figs. 2–5, strongly suggesting that PAO and DPAO
were predominant in their respective reactors proposed in this study. However, a signifi-
cant difference in the amount of phosphorus release and uptake per MLSS between AO
sludge and AA sludge was monitored (Fig. 2), probably due to the fact that
Accumulibacter exhibits a different metabolic process based on the different running
modes, namely anaerobic-aerobic and anaerobic-anoxic. For AO sludge, the anaerobic
phosphorus release rate and aerobic phosphorus uptake rate were 19.46 mg P/(g MLSS)
and 24.74 mg P/(g MLSS), respectively, both higher than the phosphorus release rate
and anoxic phosphorus uptake rate of AA sludge, which were 13.56 mg P/(g MLSS)
and 17.33 mg P/(g MLSS), respectively. These results demonstrated the PAO and
DPAO phenotypes responsible for phosphorus removal from wastewater. Although
a significant difference existing in the amount of phosphorus release/uptake between
PAO and DPAO was observed, the ratio of the phosphorus release to the phosphorus
uptake (0.786) in AO sludge was quite consistent with that in AA sludge (0.782), fur-
ther suggesting that both PAO and DPAO were dominant in their respective reactor at
the end of enrichment period. This explanation was also demonstrated by the linear
 Phosphorus removal at low temperature 75

relationship between the amount of COD consumption and that of phosphorus release
under anaerobic conditions, as given in Fig. 3.

COD uptake VS P release (AA)

COD uptake VS P release (AO)
60
Phosphorus release/(mg·dm )
–3

50
2
y = 0.3256 x + 0.0068, R = 0.9411
40
2
y = 0.2975 x + 1.1067, R = 0.9712

30

20

10

50 100 150 –3 200

COD consumption/(mg·dm )

Fig. 3. Dependences of phosphorus release on COD uptake

for AA and AO sludge in anaerobic conditions

3.2. ANAEROBIC-AEROBIC BATCH TEST WITH DPAO SLUDGE

The phosphorus release and uptake capacities of DPAO during two different cycles
(anaerobic-aerobic and anaerobic-anoxic) at the end of acclimatization phase were investi-
gated through two batch tests proposed here. Typical profiles (time dependence of carbon,
nitrogen and phosphorus) monitored in these batch tests are shown in Fig. 4. Under the
anaerobic conditions, sodium acetate was mostly taken up, which was accompanied by
phosphorus release, additionally showing a good correlation between sodium acetate up-
take and phosphorus release here (Fig. 3). The phosphorus anaerobic release rate of DPAO
obtained here was 13.56 mg P/g MLSS lower than that of PAO (19.46 mg P/g MLSS),
likely due to the less amount of Accumulibacter enriched in AA reactor compared to
AO reactor (Fig. 6). After a two hour anaerobic phase, DPAO sludge exhibited a good
phosphorus uptake performance both under anoxic and aerobic conditions. The phospho-
rus uptake rates obtained in these batch tests were 17.33 mg P/g MLSS in anoxic mode
and 17.76 mg P/g MLSS under aerobic conditions, indicating that DPAO was able to im-
mediately use oxygen as the electron acceptor when exposed to aerobic conditions, as
evidenced by the rapidly phosphorus uptake rate (Fig. 4).
76 Z. HAIMING et al.

100
300
Anaerobic Aerobic or anoxic

250 80
COD (AA)
COD (AO)
3-
200 PO4 -P..(AA)
3-
60
COD/(mg·dm )-

N, P/(mg·dm )
PO4 -P. (AO)
–3

–3
-
NO3-N (AA)
150

40

100

20
50

0 0
0 30 60 90 120 150 180 210 240 270 300 330 360
Time/min
Fig. 4. Carbon, nitrogen and phosphorus profiles of DPAO sludge
exposed to anaerobic-aerobic or anoxic conditions

Concurrently, residual sodium acetate from anaerobic phase was completely con-
sumed by the denitrifying bacteria or by the heterotrophic bacteria. Obviously, nitrate
added in the anoxic phase was removed from wastewater by the denitrifying phospho-
rus removing bacteria, namely Accumulibacter, with the function of simultaneous
denitrificaiton and phosphorus removal.

3.3. ANAEROBIC-ANOXIC BATCH TEST WITH PAO SLUDGE

Two batch tests similar to those conducted in Sect. 3.2 were performed to compare the
phosphorus uptake capacity of PAO from AO reactor in anaerobic-anoxic and anaerobic-
aerobic modes at the end of acclimatization phase. This result obtained in these batch tests
is shown in Fig. 5. During a 2 h anaerobic phase, a good performance of both phosphorus
release and sodium acetate uptake was present for PAO sludge from the AO reactor, as
also illustrated in Fig. 3, where the phosphorus release rate was 19.46 mg P/g MLSS and
the sodium acetate uptake rate was 61.49 mg COD/g MLSS, both higher than that of
DPAO (13.56 mg P/g MLSS, 47.56 mg COD/g MLSS, as shown in Fig. 4). In contrast
with DPAO (see Fig. 4), however, a significant difference of phosphorus uptake perfor-
mance of PAO between aerobic and anoxic was clearly present in Fig. 5. Here, the aerobic
phosphorus uptake rate was 24.74 mg P/g MLSS, while that was 4.86 mg P/g MLSS
in anoxic conditions, suggesting that the phosphorus uptake ability of PAO sludge was
 Phosphorus removal at low temperature 77

inhibited when exposed to anoxic conditions, as also evidenced by the less nitrate
reduction in this batch test.
100
300
Anaerobic Aerobic or anoxic

250 80

COD (AA)
200 COD (AO)
60

N, P/(mg·dm )
COD/(mg·dm )

3-

–3
PO4 -P..(AA)
–3

3-
PO4 -P. (AO)
150
-
NO3-N (AA)
40

100

20
50

0 0
0 30 60 90 120 150 180 210 240 270 300 330 360
Time/min
Fig. 5. Carbon, nitrogen and phosphorus profiles of PAO sludge
exposed to anaerobic-aerobic or anoxic conditions

3.4. MICROBIAL CHARACTERISTICS ANALYSIS

Throughout the entire start-up period of AA and AO reactors, different phosphorus

removal performance observed in the two reactors could be due to the variations in the
microbial population. For this, the techniques of both fluorescence in situ hybridization
(FISH) and scanning electron micrograph (SEM) were adopted to obtain a better under-
standing of the microbial community shift both in AA and AO reactors during the accli-
matization phase. The results from AA and AO reactors are shown in Fig. 6.
FISH analysis showed the Accumulibacter population clearly increasingd with
time, during the start-up period, from 9.3 % of all bacteria in seed sludge, to 61.6 % in
AA sludge and to 79.3 % in AO sludge, indicating that Accumulibacter was substan-
tially dominant at the end of acclimatization phase, which was consistent with phos-
phorus removal performance in respective reactor as described above. As shown in
Fig. 6, the dominance was much higher in the AO reactor than in the AA reactor.
From this FISH quantification, phosphorus removal microorganisms can be consider-
ably enriched in EBPR systems at low temperature when a strategy suitable for the
growth of Accumulibacter is provided, as the new strategy proposed in this study.
78 Z. HAIMING et al.
a

a
a) b) c)

d
d) e) f)

Fig. 6. FISH and SEM images of activateed sludge in seed slu

udge and at the endd of acclimatizationn period:
a) seed
s sludge, PAO (9.3%) + DAPI, 500×, b) AA sludge, DPAO
D (61.6%) + DAPI, 50×, c) AO O sludge,
PAO (79.3%) + DAPI,
D 50×, d) seedd sludge, 5000×, e) AA sludge, 5000×, f) AO sludge, 50000×

From SEM imaages (see Fig. 6),6 it is clear thaat significant diifferences in miicrobial
morrphologies weree observed betw ween the seed sludge
s from A2O and the AA A or AO
sluddge from EBPR R systems studieed here. Long-rrod morphologyy microbes were abun-
danttly enriched both in the AA annd AO reactorss, while seed sluudge has a highher pro-
porttion of cocci oro short-rod moorphology micrroorganisms. S Similar microbees were
enriched in the tw wo reactors durring the acclim matization proceess suggested that t the
longg-rod morpholoogy of Accumuulibacter respon nsible for phosphorus removval may
prefferably take up sodium acetatee, as supplied in n the influent inn this study, reggardless
of thhe types of electron acceptors.

4. DISCUSSIO
ON

4.1. DEVELOPMEN
NT OF THE OPER
RATIONAL STRAT
TEGY
TO PROMOTE THE GROWTH OF
O PAO AND DPA
AO

The strategy off enrichment PA AO and DPAO under anaerobic-aerobic and anaero-
bic-anoxic mode reespectively at loower temperatu
ure was developped based on thee previ-
ous reports that Acccumulibacter, a known PAO,, contains two different types:: one is
 Phosphorus removal at low temperature 79

capable of not only aerobic phosphorus uptake by using oxygen as the electron accep-
tor, but also anoxic phosphorus uptake by using nitrate, namely DPAO, and the other
only using oxygen instead of nitrate as the electron acceptor for phosphorus removal
[16, 17], and that temperature seems to be one of the most important influence factors
on wastewater systems containing EBPR in practical operation, particularly at low
temperature [11].
It can be seen from Fig. 1 that both AO and AA reactors operated in anaerobic-
-aerobic and anaerobic-anoxic conditions confirmed the phenotypes of PAO and
DPAO responsible for phosphorus removal and reached a similar stable state, as evi-
denced by the effluent phosphorus concentrations, MLSS, MLVSS, the ratio of
MLVSS/MLSS, phosphorus release rate and uptake rate and their ratio. The AO and
AA reactors reached the stable state after 40 and 80 days, respectively, suggesting the
higher activities of PAO at low temperature than DPAO, probably due to the fact that
the energy generated from the oxidative phosphorylation with NO3− is about 40% low-
er than that with O2 [5]. The ratio of MLVSS to MLSS from 0.86 (3.6/4.2) of the start-
up phase (namely, seed sludge collected from an aerobic basin within an A2O process)
decreased to 0.70 (2.6/3.7) of stable-state phase in the AO reactor and to 0.78 (3.5/4.5)
in the AA reactor which are in agreement with the reports [18] indicating that the
higher amount of phosphorus was stored in PAO or DPAO than in seed sludge, sug-
gesting that this strategy studied here dramatically promoted the growth of PAO and
DPAO in their respective reactor.
The specific phosphorus release and uptake rates for PAO were estimated to be 19.46
and 24.74 mg P/g MLSS both higher than that for DPAO, 13.56 and 17.33 mg P/g MLSS,
respectively. This was likely due to the following two reasons: the energy produced by
PAO with oxygen was higher than that of DPAO with nitrate (as discussed above) and
the size of DPAO was higher than that of PAO, causing limited transfer of carbon,
nitrogen and phosphorus to the active biomass [19]. Indeed, SEM conducted in this
study showed that DPAO grow with the aggregation of biomass into similar granules,
while PAO grow with flocs. Overall, these results obtained here demonstrated that the
operational strategy at low temperature proposed in this study is rather effective in
acclimatization of PAO and DPAO in EBPR systems, therefore providing a practical
strategy for stable state operation of this process at low temperature such as in winter.

4.2. CHARACTERISTICS OF PAO AND DPAO DURING SWITCHING TESTS

Based on characteristics of Accumulibacter, it was expected that sludge enriched

with phosphorus removing bacteria would show various phosphorus removal abilities
according to the change in electron acceptors. Figures 4 and 5 show the correlation
between the types of electron acceptor and the phosphorus removal performance. For
DPAO, no considerable difference in the phosphorus uptake rate with nitrate and oxy-
gen as electron acceptors was observed by switching the mode from normal anoxic to
80 Z. HAIMING et al.

aerobic, namely that when DPAO is exposed to aerobic conditions it can take up
phosphorus immediately, which agrees well with the report [20]. However, the phos-
phorus uptake performance of PAO was obviously inhibited when it was exposed to
anoxic rather than aerobic conditions. These results obtained through the switching
batch tests suggested that DPAO can readily produce the quantity of enzymes for aer-
obic metabolisms similar to anoxic metabolisms, while PAO lacks the enzymes re-
quired for anoxic metabolisms using nitrate instead of oxygen as an electron acceptor
[21]. The phosphorus uptake rate of PAO was very low in anoxic conditions as com-
pared with the aerobic conditions (Fig. 5). Interestingly, some studies [19, 22] have
demonstrated that when the anoxic phase was extended to 30 h rather than to 4 h, the
phosphorus uptake rate of PAO can be obviously improved, suggesting that a several
hour lag phase may exist in phosphorus uptake when PAO is exposed to anoxic condi-
tions. From these studies, we hypothesize that PAO could gradually develop the re-
quired amount of enzymes for anoxic metabolisms during the lag time, which may be
in agreement with the above explanation (PAO lacks the enzymes required for anoxic
metabolisms using nitrate).
Through four batch tests, comparison of the phosphorus removal performance of
PAO between aerobic and anoxic conditions, and similar comparison to DPAO were
conducted, demonstrating that Accumulibacter has, at least, two different types, which
supports the reports [16, 17] described above. Nevertheless, based on Carvalho et al.
findings [1], a new type of Accumulibacter was found not capable of using nitrate or
oxygen as electron acceptors but able to use nitrite for phosphorus removal under an-
oxic conditions, which may further promote the development of EBPR techniques in
practice, especially in the wastewater treatment containing phosphorus. Similarly, the
nitrite accumulation and then elimination under anoxic conditions was also observed
both in the acclimatization phase and batch tests studied here, probably supporting the
existence of three different types of Accumulibacter in EBPR systems according to the
provided various electron acceptors.

4.3. IDENTIFICATION OF ACCUMULIBACTER IN AA AND AO REACTORS

A clear shift in bacterial populations towards Accumulibacter responsible for

phosphorus removal was observed both in AA and AO reactors during the 80 day
operation. Similar findings have been reported in other EPBR systems [13, 23]. Anal-
ysis by FISH showed that Accumulibacter enriched in AA and AA reactors accounted
for 61.6%, 79.3% of all bacteria, respectively, lower than that of the report (greater
than 90%) by Lu et al. [18] but significantly higher than those by Kang et al. (41.49%)
[10] and Crocetti et al. (41%) [24]. It is likely due to the fact that different adopted
operational conditions, such as temperature, pH, carbon sources, etc., result in the
difference in enrichment capacities of phosphorus removal microbes. The relative
abundance of Accumulibacter in AA and AO reactors linked well with their respective
 Phosphorus removal at low temperature 81

EBPR performance observed in this study (as described above).These results demon-
strated that the operational strategy proposed in this study may be highly effective in
enrichment of Accumulibacter, and therefore may provide a new practical method for
obtaining a good phosphorus removal performance at low temperature.
Analysis of SEM showed that identical microbial morphologies (long-rod mi-
crobes) were present in the two reactors at the end of acclimatization period, suggest-
ing that this kind of Accumulibacter may display good affinity to sodium acetate as the
carbon source. Indeed, two different types (rods or cocci) of Accumulibacter were
found by Martin et al. [21] in two EBPR systems, where one was feed with sodium
acetate and the other with propionate. Similarly, He et al. [25] also found the distribu-
tion of different types of Accumulibacter in one lab-scale reactor and six full-scale
reactors both presenting good phosphorus performance. These studies may support the
hypothesis that different carbon sources feed to the phosphorus removal microorgan-
isms could promote the growth of different types of Accumulibacter, probably regard-
less of electron acceptors which is also partially supported by the results obtained in
this study. Overall, the combination of chemical analysis with microbial analysis sug-
gested that Accumulibacter, both PAO and DPAO, with a long-rod morphology was
more preferably enriched with sodium acetate as compared with other carbon sources
such as sodium propionate.

5. CONCLUSIONS

A strategy for obtaining microorganisms responsible for phosphorus removal (i.e.,

Accumulibacter including PAO and DPAO) at low temperature ranging from 8 °C to
11 °C has been demonstrated in this study. The approach is as follows:
• Switching the ratio of COD to P of two EBPR systems operated in anaerobic-
aerobic (AO) and anaerobic-anoxic (AA) modes respectively, from 20:1 to 15:1; with
the decrease of concentration COD, PO34− -P and NO3− -N in the influent from
800 mg/dm3, 40 mg/dm3 and 50 mg/dm3 to 300 mg/dm3, 20 mg/dm3 and 30 mg/dm3,
respectively.
• Adopting different solids retention time (SRT) for enrichment of different types
of Accumulibacter; no excess of activated sludge was wasted at the beginning of start-
up, and then 10 days SRT for PAO and 20 days SRT for DPAO.
• Maintaining a high and different MLSS concentrations in AA and AO reactors,
based on the metabolic characteristics of Accumulibacter supplied the different elec-
tron acceptors, 4.5 g/dm3 MLSS for the AA reactor and 3.7 g/dm3 MLSS for the AO
reactor.
This proposed strategy here was shown to be effective in achieving a very high en-
richment of Accumulibacter at low temperature by linking chemical analysis with
microbial observation. From chemical analysis, similar stable state was achieved both
82 Z. HAIMING et al.

in AO and AA reactors when different enrichment time provided. Chemical analysis

of batch tests indicated the existence of two types of Accumulibacter, one using oxy-
gen as the electron acceptor and the other – nitrates. Through microbial observation,
a high abundance of Accumulibacter was found in both AA and AO reactors. The
identical microbial morphology observed both in the two reactors suggested the long-
rod phosphorus removal microorganisms may display affinities to sodium acetate used
as the carbon sources in this study. Although the strategy may not be the unique meth-
od for the enrichment of phosphorus removal microorganisms at low temperature, it is
recommendable for the future studies in practical application.

ACKNOWLEDGEMENT

We wish to thank M.M.SU for assistance with activated sludge treatment and Dr. Wang for enlight-
ening discussion. This research is supported by grant 2012ZX07101-005 from National Key Technology
in Water Pollution Control and Treatment in 12th Five-year Plan of China and grant 51078074 from
National Natural Science Foundation of China.

REFERENCES

[1] CARVALHO G., LEMOS P.C., OEHMEN A., REIS M.A.M., Denitrifying phosphorus removal: linking the
process performance with the microbial community structure, Water Res., 2007, 41 (19), 4383.
[2] JIANG X.X., YANG J.X., MA F., YANG F.F., WEI J., YING J., Denitrifying phosphorous removal in an-
aerobic/anoxic SBR system with different startup operation mode, Journal of Harbin Institute of
Technology, 2010, 17 (6), 824.
[3] CUI D., LI A., ZHANG S., PANG C., YANG J., GUO J., MA F., WANG J., REN N., Microbial community
analysis of three municipal wastewater treatment plants in winter and spring using culture-
dependent and culture-independent methods, World Journal of Microbiology and Biotechnology,
2012, 28, 2341.
[4] SHI X.Y., SHENG G.P., LI X.Y., YU H.Q., Operation of a sequencing batch reactor for cultivating
autotrophic nitrifying granules, Bioresource Technol., 2010, 101 (9), 2960.
[5] AHN J., DAIDOU T., TSUNEDA S., HIRATA A., Characterization of denitrifying phosphate-
-accumulating organisms cultivated under different electron acceptor conditions using polymerase
chain reaction-denaturing gradient gel electrophoresis assay, Water Res., 2002, 36 (2), 403.
[6] KONG Y., NIELSEN J.L., NIELSEN P.H., Identity and ecophysiology of uncultured actinobacterial
polyphosphate-accumulating organisms in full-scale enhanced biological phosphorus removal
plants, Appl. Environ. Microb., 2005, 71 (7), 4076.
[7] TIAN W.D., LI W.G., AN K.J., KANG X.R., MA C., RAN Z.L., Denitrifying dephosphatation perfor-
mance link to microbial community structure, Journal of Water Sustainability, 2011, 1 (3), 269.
[8] SEMERCI N., BAKICI N., KOCAMEMI B.A., Effects of nitrite, oxygen and initial pH on biological phos-
phorus removal in a post-denitrification system, J. Biotechnol., 2010, 150, 292.
[9] GUERRERO J., GUISASOLA A., BAEZA J.A., The nature of the carbon source rules the competition be-
tween PAO and denitrifiers in systems for simultaneous biological nitrogen and phosphorus removal,
Water Res., 2011, 45 (16), 4793.
[10] KANG H., LI N., REN N.Q., Competition between phosphate-accumulating organisms and glycogen-
accumulating organisms and the phosphate removal efficiency in EBPR reactor at low temperature,
Journal of Harbin Institute of Technology, 2010, 42 (6), 881.
 Phosphorus removal at low temperature 83

[11] PANSWAD T., DOUNGCHAI A., ANOTAI J., Temperature effect on microbial community of enhanced
biological phosphorus removal system, Water Res., 2003, 37 (2), 409.
[12] LOPEZ-VAZQUEZ C.M., OEHMEN A., HOOIJMANS C.M., BRDJANOVIC D., GIJZEN H.J., YUAN Z., VAN
L.M., Modeling the PAO–GAO competition: Effects of carbon source, pH and temperature, Water
Res., 2009, 43 (2), 450.
[13] KANG H., WANG X., LI N., REN N., Optimization and application of fluorescent in situ hybridization
(FISH) process in EBPR fed with municipal wastewater, Journal of Harbin Institute of Technology,
2011, 18 (3), 55.
[14] MIELCZAREK A.T., NGUYEN H.T.T., NIELSEN J.L., NIELSEN P.H., Population dynamics of bacteria
involved in enhanced biological phosphorus removal in Danish wastewater treatment plants, Water
Res., 2012, 47 (4), 1529.
[15] GAO J.F., CHEN R.N., SUO K., Formation and reaction mechanism of simultaneous nitrogen and
phosphorus removal by aerobic granular sludge, Environmental Science, 2010, 31 (4), 1021.
[16] ZENG W., LI L., YANG Y., WANG X., PENG Y., Denitrifying phosphorus removal and impact of nitrite
accumulation on phosphorus removal in a continuous anaerobic–anoxic–aerobic (A2O) process treating
domestic wastewater, Enzyme Microb. Tech., 2011, 48 (2), 134.
[17] ACEVEDO B., OEHMEN A., CARVALHO G., SECO A., BORRAS L., BARAT R., Metabolic shift of poly-
phosphate-accumulating organisms with different levels of polyphosphate storage, Water Res., 2012,
46 (6), 1889.
[18] LU H.B., OEHMEN A., VIRDIS B., KELLER J., YUAN Z.G., Obtaining highly enriched cultures of
Candidatus Accumulibacter phosphates through alternating carbon sources, Water Res., 2006, 40
(20), 3838.
[19] ZENG R.J., SAUNDERS A.M., YUAN Z., BLACKALL L.L., KELLER J., Identification and comparison of aero-
bic and denitrifying polyphosphate-accumulating organisms, Biotechnol. Bioeng., 2003, 83 (2), 140.
[20] GEBREMARIAM S.Y., BEUTEL M.W., CHRISTIAN D., HESS T.F., Research Advances and Challenges in
the Microbiology of Enhanced Biological Phosphorus Removal. A Critical Review, Water Environ.
Res., 2011, 83 (3), 195.
[21] MARTIN H.G., IVANOVA N., KUNIN V., WAMECKE F., BARRY K.W., MCHARDY A.C., YEATES C.,
HE S., SALAMOV A.A., SZETO E., Metagenomic analysis of two enhanced biological phosphorus re-
moval (EBPR) sludge communities, Nat. Biotechnol., 2006, 24 (10), 1263.
[22] ZAFIRIADIS I., NTOUGIAS S., NIKOLAIDIS C., KAPAGIANNIDIS A.G., AIVASIDIS A., Denitrifying poly-
phosphate accumulating organisms population and nitrite reductase gene diversity shift in
a DEPHANOX-type activated sludge system fed with municipal wastewater, J. Biosci. Bioeng.,
2011, 111 (2), 185.
[23] GE Y.H., ZHAO L., ZHANG R.C., LIU Y.J., Optimization and Application of Fluorescence in Situ
Hybridization Assay for Detecting Polyphosphate-Accumulating Microorganisms, Advanced Mate-
rials Research, 2011, 183, 1369.
[24] CROCETTI G.R., HUGENHOLTZ P., BOND P.L., SCHULER A., KELLER J., JENKINS D., BLACKALL L.L.,
Identification of polyphosphate-accumulating organisms and design of 16S rRNA-directed probes
for their detection and quantitation, Appl. Environ. Microb., 2000, 66 (3), 1175.
[25] HE S., GU A.Z., MCMAHON K.D., Fine-scale differences between Accumulibacter-like bacteria in
enhanced biological phosphorus removal activated sludge, Water Sci. Technol., 2006, 54 (1), 111.