Bioresource Technology 211 (2016) 257266

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Anammox for nitrogen removal from anaerobically pre-treated

municipal wastewater: Effect of COD/N ratios on process performance
and bacterial community structure
Cntia Dutra Leal, Alyne Duarte Pereira, Fernando Terra Nunes, Lusa Ornelas Ferreira,
Aline Carolina Cirilo Coelho, Sarah Kinaip Bicalho, Erika F. Abreu Mac Conell, Thiago Bressani Ribeiro,
Carlos Augusto de Lemos Chernicharo, Juliana Calbria de Arajo
Department of Sanitary and Environmental Engineering of Federal University of Minas Gerais (UFMG), Av. Antonio Carlos, 6627, 31270-901 Belo Horizonte, MG, Brazil

h i g h l i g h t s
1
COD.
) did not affect nitrogen removal.
 Anaerobic effluent changed bacterial community structure and diversity.
 Anammox and denitrifying bacteria were both found in the reactor.
 Anammox can be applied to remove nitrogen from pre-treated municipal wastewater.

 Anammox metabolism was inhibited by the addition of 487.5 mg
L

 A COD/N ratio of 5.0 (COD of 300 mg
L

a r t i c l e

i n f o

Article history:
Received 20 January 2016
Received in revised form 18 March 2016
Accepted 19 March 2016
Available online 22 March 2016
Keywords:
Anammox
COD/N ratio
Anaerobic effluent
DGGE
Nitrogen removal

a b s t r a c t
Long-term effects of COD/N ratios on the nitrogen removal performance and bacterial community of an
anammox reactor were evaluated by adding a synthetic medium (with glucose) and real anaerobic effluent to a SBR. At a COD/N ratio of 2.8 (COD, 390 mg
L 1) ammonium removal efficiency was 66%, while
nitrite removal remained high (99%). However, at a COD/N ratio of 5.0 (COD, 300 mg
L 1), ammonium
and nitrite removal efficiencies were high (84% and 99%, respectively). High COD, nitrite, and ammonium
removal efficiencies (80%, 90% and 95%, respectively) were obtained on adding anaerobically pre-treated
municipal wastewater (with nitrite) to the reactor. DGGE revealed that the addition of anaerobic effluent
changed the bacterial community structure and selected for DNA sequences related to Brocadia sinica and
Chloroflexi. Adding glucose and anaerobic effluent increased denitrifiers concentration threefold. Thus,
the possibility of using the anammox process to remove nitrogen from anaerobically pre-treated municipal wastewater was demonstrated.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
Anaerobic ammonium oxidation (anammox) is a microbial process that can be used, in principle, for the treatment of ammoniumrich wastewaters with a low C:N ratio. Anammox bacteria oxidize

Abbreviations: Anammox, anaerobic ammonium oxidation; BOD, biochemical

oxygen demand; COD, chemical oxygen demand; DGGE, denaturing gradient gel
electrophoresis; DO, dissolved oxygen; FISH, fluorescent in situ hybridization; TNK,
total nitrogen Kjeldahl; SBR, sequencing batch reactor; TSS, total suspended solids;
UASB, upflow anaerobic sludge blanket reactor; VSS, volatile suspended solid.
Corresponding author.
E-mail address: juliana@desa.ufmg.br (J.C. de Arajo).
http://dx.doi.org/10.1016/j.biortech.2016.03.107
0960-8524/ 2016 Elsevier Ltd. All rights reserved.

ammonium (NH+4) into nitrogen gas (N2) under anoxic conditions

using nitrite (NO2 ) as the terminal electron acceptor (van de
Graaf et al., 1996; Strous et al., 1998). These are chemolitoautotrophic bacteria characterized by slow growth (Strous et al.,
1998; Kartal et al., 2011), belonging to the phylum Planctomycetes,
order Brocadiales.
The anammox process simultaneously removes two pollutants,
ammonium and nitrite, converting them into gaseous nitrogen
(Zhang et al., 2008). Inorganic carbon sources such as CO2 and
HCO3 are particularly important for the cultivation of anammox
bacteria and favor the growth and activity of these microorganisms
(van de Graaf et al., 1996). On the other hand, organic compounds

258

C.D. Leal et al. / Bioresource Technology 211 (2016) 257266

(present in almost all types of wastewaters) may give rise to

adverse effects on anammox bacteria and may impair the anammox activity, especially at high concentrations (Gven et al.,
2005; Chanchoi et al., 2008; Molinuevo et al., 2009). Inhibition of
the anammox process by organic compounds may occur due to
enzyme inactivation, and may be irreversible, resulting in cell
death (Gven et al., 2005). Another proposed mechanism to
explain anammox impairment is the competition between anammox bacteria and denitrifying heterotrophic bacteria for the electron acceptor, nitrite. Since heterotrophic bacteria are able to
grow faster than autotrophic ones, they outcompete the anammox
bacteria and hinder their chemical reactions (Gven et al., 2005;
Chanchoi et al., 2008; Lackner et al., 2008; Molinuevo et al., 2009).
There is no consensus in the literature on what is the organic
matter concentration (expressed as COD) and the COD to N ratio
that inhibits/or affects the anammox process. According to
Chanchoi et al. (2008) COD concentrations above 300 mg
L 1
(COD/N ratio of 2) may fully inhibit anammox reaction and concomitantly favor the activity of denitrifying bacteria. Ni et al.
(2012) verified that at COD/N ratios above 4 the anammox process
was impaired. Snchez-Guilln et al. (2014) investigating the
short-term effects of organic carbon addition (acetate and starch)
verified that the COD/N ratios (2 and 6) applied had no significant
influence on the anammox process.
Jenni et al. (2014) determined the influence of different COD
concentrations and hydraulic retention times on the nitrogen
removal (by the anammox process). The COD/N ratio was gradually
increased up to 1.4 g COD g N 1, while the hydraulic retention time
was reduced. After adding acetate and glucose (as carbon sources),
the efficiency reached nearly 95%. The results showed that with a
ratio of 0.8 g COD g N 1 and higher there was a reduction in the
abundance of anammox bacteria (estimated by FISH), but the number of Candidatus Brocadia fulgida cells increased. The authors also
found that alternating between acetate and glucose had no negative effect on the community profile, and that the prevalence of
Ca. Brocadia fulgida could be explained by the fact that this species
can oxidize acetate.
Despite the previous findings, these studies did not address
ammonium or nitrogen removal from real anaerobically pretreated municipal wastewater by the anammox process. In addition, no study has investigated the long-term effects of adding
COD or increasing COD/N ratios (up to 5.0) in an anammox reactor.
The anammox process has been successfully applied to remove
nitrogen at many wastewater treatment plants around the world,
especially ammonium-rich wastewaters with low C:N ratios. Nevertheless, few studies have investigated the potential application of
the anammox process to remove nitrogen from effluents with low
ammonium concentrations and high COD/N ratios, such as the
effluent from an upflow anaerobic sludge bed reactor (UASB) treating domestic wastewater. Therefore, the objectives of this study
were: to evaluate the efficiency of an anammox reactor in removing nitrogen when subjected to different COD/N ratios, and to evaluate the possibility of applying the anammox process to nitrogen
removal from anaerobically pre-treated municipal wastewater.
Moreover, the bacterial community structure in the anammox
reactor during experiments with synthetic medium (before and
after glucose addition) and with real anaerobic effluent was investigated and compared.

2. Materials and methods

2.1. Experimental set-up
A 2.0-L glass reactor (Benchtop Fermentor & Bioreactor BioFlo/
CelliGen 115, New Brunswick Scientific Co., Enfield, CT, USA) was

used for cultivation of anammox biomass in this study. The reactor

was monitored for 558 days and was operated in sequencing batch
mode with two daily cycles, one of 7 h (short cycle) and the other
of 17 h (long cycle), resulting in a hydraulic retention time (HRT) of
24 h and total biomass retention. The SBR was fed with autotrophic
medium (van de Graaf et al., 1996; Dapena-Mora et al., 2004). The
concentrations of ammonium and nitrite in the medium ranged
from, respectively, 100 mg
L 1 to 150 mg
L 1 and 150 mg
L 1 to
200 mg
L 1 (while the nitrite to ammonium ratio applied was kept
around 1.31.4), for 323 days prior to the phase in which the reactor was exposed to different COD/N ratios.
The temperature inside the reactor was kept at 35 C and pH
was adjusted to 7.5. Influent and effluent samples were collected
to monitor the concentrations of ammonium, nitrite and the
COD. Analyses of these parameters were performed according to
the Standard Methods for the Examination of Water and Wastewater (APHA, 2012). Biomass samples were taken from the reactor at
the end of each operational phase in order to investigate the bacterial diversity inside the reactor.
2.2. Anammox biomass
The SBR was inoculated with biomass taken from two bioreactors that showed anammox activity, both seeded with sludge from
municipal wastewater treatment plants (detailed information of
these two bioreactors is presented in Pereira et al., 2014, and
Costa et al., 2014). The microbial communities of the two inocula,
characterized by 454 pyrosequencing, showed that the prevailing
phyla were Proteobacteria, Firmicutes, Chloroflexi, Verrucomicrobia,
and Planctomycetes (Pereira et al., 2014; Costa et al., 2014).
2.3. Experiment applying different COD/N ratios to the SBR
The enrichment and cultivation period (Phase I) lasted
323 days. In the following period (Phase II), different COD/N ratios
(using synthetic medium with glucose) were applied to the SBR, by
gradually increasing the COD concentration. Glucose was used as a
carbon source because it is a non-toxic organic compound and also
because anammox bacteria cannot degrade this substrate (Gven
et al., 2005; Jenni et al., 2014). Glucose was introduced into the
SBR during the long cycle (17 h) through a septum located at the
top of the reactor using 50-mL syringes. Varying amounts of a concentrated glucose solution were added to the reactor in order to
achieve the desired final concentrations of 95, 195, 390, 487.5,
and 300 mg
L 1, which corresponded to the COD/N ratios being
tested, i.e., 0.7, 1.4, 2.8, 3.5 and 5.0, respectively. In order to achieve
these ratios, the concentrations of ammonium and nitrite in the
medium were reduced to, respectively, 60 mg
L 1 and 80 mg
L 1
(while the nitrite to ammonium ratio applied was 1.33, same ratio
used in Phase I).
The results obtained in Phase I were considered reference
results for nitrogen removal by the anammox process in the SBR.
With a COD/N ratio of 3.5, a decrease in anammox activity was
observed, and, for that reason, glucose feeding was stopped to prevent irreversible inhibition of anammox bacteria activity. The reactor was held in a recovery phase for 35 days, being fed only with
synthetic medium, while the concentrations of nitrite and ammonium were reduced to 30 mg
L 1 and 30 mg
L 1 respectively, and,
after the recovery period, the test was performed with a COD/N
ratio of 5.0 (COD of 300 mg
L 1 and nitrogen concentration of
60 mg
L 1). Statistical analysis (using KruskalWallis test) was performed to assess whether different concentrations of glucose
added to the SBR altered the nitrite and ammonium removal efficiency, i.e., to verify if there would be statistical differences
between the reference values (median values of ammonium and
nitrite removal efficiencies observed in Phase I-without the

C.D. Leal et al. / Bioresource Technology 211 (2016) 257266

addition of glucose) and those values found in Phase II (in which

glucose was added to the reactor). KruskalWallis test (a = 5%)
was performed, followed by a multiple comparisons test of medians (a = 5%), using Statistica 8 software.
2.4. Experiment applying real anaerobic effluent to the SBR
In Phase III, the anammox reactor was fed with anaerobically
pre-treated municipal wastewater. A small parcel of municipal
wastewater (taken from Arrudas WWTP the main plant of Belo
Horizonte city, Minas Gerais, Brazil) was pre-treated in a demoscale UASB reactor (V = 16,800 L; average HRT = 8.6 h). The chemical composition of the anaerobically pre-treated municipal
wastewater (real anaerobic effluent) is shown in Table 1. The
anaerobic effluent feeding process comprised two steps. In the first
(Phase IIIA), 60% anaerobic effluent was added to 40% autotrophic
medium, containing 50 mg
L 1 of nitrite and 30 mg
L 1 of ammonium. In the second (Phase IIIB), only anaerobic effluent supplemented with 120 mg
L 1 of nitrite was added to the reactor,
under the same operational conditions.
Statistical analysis (using KruskalWallis test) were performed
to assess whether the addition of anaerobic effluent to the SBR
altered the nitrite and ammonium removal efficiency, i.e., to verify
if there would be statistical differences between the reference values (observed in Phase I) and those values found in Phase III. KruskalWallis test (a = 5%) was performed, followed by a multiple
comparisons test of medians (a = 5%), using Statistica 8 software.
Efficiencies in Phase III, in which anaerobic effluent (amended with
nitrite) was added to the reactor, were compared to the previous
periods (Phase II, with glucose addition; and Phase I, without glucose addition).
2.5. Characterization of bacterial community by polymerase chain
reaction-denaturing gradient gel electrophoresis (PCR-DGGE)
Biomass samples were taken from the SBR at day 349 (Phase I,
the control phase used as a reference for the anammox process),
day 448 (after the recovery step without glucose), day 482 (after
Phase II addition of glucose for different COD/N ratios), day 558
(after Phase III addition of real anaerobic effluent) and represent
each of the different phases of SBR operation. Genomic DNA was
extracted from 0.5 g of each sample using a PowerSoil DNA Isolation Kit (MO BIO Laboratories, USA). PCR was performed using
the primer set 1055F (ATGGCTGTCGTCAGCT) and 1392R
(ACGGGCGGTGTGTAC) with a GC clamp, as described previously

Table 1
Characterization of the anaerobic effluent from a UASB reactor that treats municipal
wastewater.

Compound

Unit

Median value

pH
Conductivity
DO
Temperature
Total BOD
Total COD
Soluble COD
TSS
VSS
Total nitrogen
TNK
NH+4
NO2
NO3
PO34
SO24

7.2 (0.1)a
744 (64)
0.3 (0.1)
25.7 (2.3)
72 (22)
150 (92)
80 (35)
71 (43)
51 (30)
46 (8)
45 (8)
27 (4)
0.0 (0.0)
0.1 (0.1)
6.9 (4.0)
14.2 (4.5)

ls
cm
mg
L
C
mg
L
mg
L
mg
L
mg
L
mg
L
mg
L
mg
L
mg
L
mg
L
mg
L
mg
L
mg
L

1
1

1
1
1
1
1
1
1
1
1
1
1
1

The values in brackets are standard deviation.

259

(Ferris et al., 1996). PCR products were subjected to DGGE. DGGE

was performed with a Bio-Rad DCode Universal Mutation Detection System (Hercules, CA, USA) using an 8% polyacrylamide gel
with 5075% denaturing gradient for 16.5 h at 75 volts. Gels were
stained with SybrGold (LifeTechnologies) and analyzed with the
BioNumerics 7.1 software (Applied Maths, Austin, TX, USA). Band
profiles were compared using Dice similarity coefficient and the
dendrogram was generated with UPGMA method, with 1% position
tolerance.
DNA gel bands were excised and re-amplified. The PCR products
were sent to Macrogen, Inc. (Seoul, Korea) for purification and unidirectional sequencing in a 3730XL Sequencer. These sequences
were compared with sequences in the Ribosomal Database Project
by RDP Classifier (https://rdp.cme.msu.edu/classifier/classifier.jsp),
reaching a reliability level of 80%, and with sequences from the
National Center for Biotechnology Information database using the
Basic Local Alignment Search Tool (Altschul et al., 1990).
2.6. Quantitative PCR
The abundance of anammox bacteria, denitrifiers, and total bacteria were investigated by real time quantitative PCR (qPCR) using
SybrGreen assays on biomass samples taken from the SBR at the
end of Phases I, II and III (at day 349, 482 and 558, respectively).
qPCR assays were conducted on a real-time PCR thermal cycler
(Applied Biosystems 7500 instrument). The total bacterial abundance was quantified using eubacterial 16S rRNA-targeted primers
1055F (ATGGCTGTCGTCAGCT) and 1392R (ACGGGCGGTGTGTAC)
(Ferris et al., 1996), and anammox bacteria were quantified using
primers Pla46F (GGATTAGGCATGCAAGTC) and Amx667R (ACCAGAAGTTCCACTCTC). The abundance of denitrifiers was quantified
using primers targeting the nitrous oxide reductase gene nosZF
(CGYTGTTCMTCGACAGCCAG) and nosZ1622R (CGSACCTTSTTGCCSTYGCG) (Enwall et al., 2005). Each 20 lL reaction mixture contained 10 ng of template DNA, 375 nM forward and reverse
primers (each), and 10 lL of MaximaTM SybrGreen/ROX qPCR Master Mix 2 (Thermo Scientific, USA). Standard curves for qPCR were
generated by 10-fold serial dilutions of plasmid DNA containing
specific target gene inserts.
3. Results and discussion
3.1. Nitrogen removal during operational Phases I, II and III
The performance of the SBR over 558 days is shown in Fig. 1. To
evaluate the effect of different COD/N ratios and the addition of
real anaerobic effluent on the anammox process, the profiles of
nitrogenous compound concentrations were divided into three
phases: the control phase (Phase I), the operational phase, with different COD/N ratios due to glucose addition (Phase 2), and the real
anaerobic effluent addition phase (Phase III).
3.1.1. Anammox cultivation phase (Phase I)
The first 323 days of operation corresponded to the anammox
enrichment and cultivation period after inoculation of biomass
(Phase I). During this phase, the reactor showed an average pH
value of 7.3 0.8, a steady temperature of 35 C, and DO between
0 and 0.5 mg
L 1. The ammonium and nitrite concentrations in
the influent ranged from, respectively, 100 mg
L 1 to 150 mg
L 1
and 150 mg
L 1 to 200 mg
L 1 due to variations in the preparation
of culture medium (but the nitrite to ammonium ratio applied was
kept around 1.31.4). Over the monitoring period in Phase I, the
average removal efficiencies for ammonium and nitrite were 92%
and 97%, respectively (Fig. 1b and c), and the average nitrate production value was 39.6 mg
L 1 (Table 2).

260

C.D. Leal et al. / Bioresource Technology 211 (2016) 257266

(a)

(b)

(c)

Fig. 1. (a) Performance of the SBR during operation with autotrophic medium (Phase I, anammox cultivation), increasing influent COD/N ratios (Phase II, glucose addition),
and the addition of real anaerobic effluent (Phase III). Box plot of ammonium (b) and nitrite (c) removal efficiencies of the SBR during operation with autotrophic medium
(Phase I), increasing COD/N ratios (Phase II), and the addition of real anaerobic effluent (Phase III).

The
stoichiometric
coefficient
(consumption
of
NNO2 /consumption of N-NH+4) determined during Phase I was 1.46
(as shown in Table 2), which is close to that (1.32) reported for
an anammox reaction by Strous et al. (1998). Quan et al. (2008)
reported a coefficient of 1.46 in an UASB reactor operated in anammox conditions. The average value found for the coefficient of NNO3 production/N-NH+4 consumption was 0.34, slightly higher than

the one reported by Strous et al. (1998), of 0.26. Date et al. (2009)
reported a coefficient of 0.33 (production of N-NO3 /consumption
of N-NH+4) in an upflow reactor with biomass from a wastewater
treatment tank from a pig production facility. A similar result
(0.36) was reported by Pereira et al. (2014).
The results obtained in Phase I demonstrated that the anammox
process was established in the SBR, thus confirming that the two

261

C.D. Leal et al. / Bioresource Technology 211 (2016) 257266

Table 2
Nitrogen mass balance evaluation of participation of different processes.
COD/N
ratio

NH+4-N removal
(mg N/L)a

Anammox

Denitrification

Production

Removal

Final

116.0

33.6

169.4

0.0

39.6

0.0

II

0.7:1
1.4:1
2.8:1
3.5:1
R**
5.0:1

49.1
46.4
43.2
11.4
35.0
31.5

5.5
9.0
5.3
0.9
16.9
8.0

71.7
67.8
63.1
16.6
51.2
37.7

10.9
1.0
16.9
61.6
6.1
1.4

16.7
15.8
14.7
3.9
11.9
9.0

5.3
6.8
11.0
1.1
5.3
3.3f

Phase

SD

NO2-N removal (mg N/L)

COD influent
(mg/L)

COD removal
(mg/L)

Stoichiometrye

39.6

0.0

0.0

1:1.46:0.34

11.5
9.0
3.7
2.8
6.6
12.3

97.5
195.0
390.0
487.5
0.0
300.0

67.8
162.2
261.0
89.2
0.0
174.4

1:1.30:0.31

NO3 -N (mg N/L)

b

III A

3.0:1

31.7

15.8

34.8

0.3

6.3

0.9

5.4

238.2

204.4

1:1.10:0.20

III B

3.5:1

48.5

10.1

82.4

32.1

9.7

0.9

8.8

172.2

133.1

1:1.70:0.20

SD: Standard deviation.

**
Recovery period (without glucose).
a
Mean values.
b
Values denote nitrate produced via anammox process.
c
Values mean nitrate removed by denitrification.
d
Values are the nitrate concentrations determined in the effluent.
e
The stoichiometry of removed NH+4-N: NO2-N: produced NO3 -N obtained during Phase I (control phase) was used to calculate the nitrite removal via anammox and
denitrification processes during Phase II (COD/N ratios of 0.7, 1.4, 2.8, 3.5; and R period). For other periods the stoichiometry used is shown in the table.
f
Negative value indicates that nitrate balance did not close, as more nitrate was determined in the effluent.

inocula used and the operational conditions applied were suitable

for the enrichment and cultivation of anammox biomass. This
phase was considered the control phase, and the nitrogen
removal values were used as references for comparison with
results obtained from subsequent phases.
3.1.2. Performance of the SBR during operation with different COD/N
ratios (Phase II)
For 131 days, from day 351 to 482 (Phase II), the reactor was
operated with increasing influent COD/N ratios due to glucose
addition. To perform this experiment, the concentrations of ammonium and nitrite in the autotrophic medium were reduced to
60 mg
L 1 and 80 mg
L 1, respectively, (while the nitrite to ammonium ratio was kept around 1.31.4) for COD/N ratios of 0.7, 1.4,
2.8, and 3.5. Fig. 1 shows the concentrations of nitrogenous compounds in the influent and effluent of the SBR in Phase II. Increasing COD/N ratios were found to negatively affect ammonium
removal efficiency (Fig. 1b) and to partially inhibit the anammox
process (as confirmed by the mass balance results shown in
Table 2). This was most notable for the COD/N ratio of 3.5 (COD
of 490 mg
L 1), when the average ammonium removal efficiency
values dropped sharply to 21.7%. For this reason, the test was
stopped and the reactor was fed only with autotrophic medium
(without glucose) for 33 days, from the 415th to the 448th day,
to recover nitrogen removal efficiency (this period was called the
recovery period, or Phase R). The concentrations of nitrite and
ammonium in the autotrophic medium were reduced to 30 mg
L 1
and 30 mg
L 1, respectively, to carry out the test with COD/N ratios
up to 5.0 after the recovery period (since COD was 300 mg
L 1 and
nitrogen was 60 mg
L 1). Thus, in this period the nitrite to ammonium ratio applied was 1.0. During Phase II, the average pH value
was 6.83 0.25, which was lower than the pH values observed in
Phase I (7.4 0.8); however, the pH values remained within the
ideal range for anammox bacteria, i.e., between 6.7 and 8.3
(Strous et al., 1999). The nitrate concentrations in the SBR effluent
were also monitored during Phase II, and the average value was
15.8 mg
L 1.
Regarding nitrite removal efficiency in Phase II, the results
showed that the average values remained close to 99%, indicating
that nitrite removal was not affected by increasing COD/N ratios
applied to the reactor (as shown in Fig. 1c). The nonparametric
KruskalWallis statistical test did not detect differences between

these values and those obtained during Phase I, followed by multiple comparisons of medians (a = 5%; p = 0.1531). However, for
ammonium removal efficiency, statistical differences between
the average values obtained from Phase II and Phase I were
detected by KruskalWallis test (a = 5%; p = 0.01356). In addition,
for multiple comparisons of median values, the KruskalWallis test
(a = 5%) was performed and statistical differences were observed
between periods with COD/N ratios of 0.7 and 1.4 (p = 0.0412),
0.7 and 2.8 (p = 0.0453), 0.7 and 3.5 (p = 0.0167), 0.7 and the recovery phase (p = 0.0102) and 0.7 and 5.0 (p = 0.0201). Thus, the addition of glucose (at concentrations of 195 mg
L 1, 390 mg
L 1,
487.5 mg
L 1 and 300 mg
L 1, corresponding to COD/N ratios of
1.4, 2.8, 3.5 and 5.0, respectively) affected ammonium removal efficiency and consequently the anammox process. Moreover, these
results indicate that the COD concentration of 487.5 mg
L 1 was
the one that severely affected and inhibited the anammox process
since ammonium removal was significantly reduced (as shown in
Fig. 1b) and the color of the biomass changed from brownishyellow to a darker color (as shown in Supplementary Fig. S1).
The nitrogen mass balance was calculated for all operational
phases to evaluate the participation of different processes (Table 2).
The results indicated that the addition of glucose at different concentrations (during Phase II) gradually reduced the participation of
anammox process in the nitrite removal (from 87% at COD/N ratio
of 0.723% at COD/N of 3.5) while favored the denitrification process, being the highest consumption of nitrite by denitrification
was observed at the COD/N ratio of 3.5 (Table 2). Nitrate removal
during Phase II was also observed, being the highest nitrate consumption via denitrification was observed at COD/N ratio of 2.8
(Table 2).
The COD was monitored during Phase II to determine whether
there was glucose consumption (COD removal) by other bacterial
groups or whether COD was accumulating inside the reactor. The
average COD removal efficiencies in each of the steps (shown in
Fig. 2) were 71% (COD of 97.50 mg
L 1, COD/N = 0.7), 82% (COD of
195 mg
L 1, COD/N = 1.4), 62% (COD of 390.0 mg
L 1, COD/
N = 2.8), 16% (COD of 487.50 mg
L 1, COD/N = 3.5), and 55% (COD
of 300.00 mg
L 1, COD/N = 5.0), indicating that COD was being consumed inside the reactor probably by denitrifying bacteria (confirming the mass balance results).
Phase II may be viewed as the most critical phase of the whole
study, as it was the one in which nitrogen removal performance by

262

C.D. Leal et al. / Bioresource Technology 211 (2016) 257266

Fig. 2. Box plot of COD removal efficiencies at different COD/N ratios during Phase II (glucose addition) and Phase III (addition of anaerobic effluent amended with nitrite).

the anammox process was significantly reduced in comparison

with Phase I (Fig. 1b and Table 2). After testing the COD/N ratio
of 3.5, the reactor underwent a recovery period with no organic
compound (glucose) added. During recovery phase and the subsequent step (COD/N ratio of 5) as well, the anammox process was
reestablished in the reactor (as can be seen by the increase in
nitrite removal by this process, showed in Table 2).
3.1.3. Performance of the SBR during operation with real anaerobic
effluent (Phase III)
For 57 days, from days 483 to 540 (Phase III), the reactor was
operated with real anaerobic effluent. The effluent was taken from
a UASB reactor that treats municipal wastewater and showed median COD values of 150 mg
L 1 and ammonium concentration of
27 mg
L 1 (see Table 1), corresponding to a COD/N ratio of 5.5.
Over the Phase III monitoring period, the average pH inside the
reactor was 6.9 0.4, while the temperature was kept at 35 C and
the dissolved oxygen ranged from 0.0 to 0.4. The results of the SBR
performance during Phase III are presented in Fig. 1. Phases IIIA
and IIIB had different results for ammonium and nitrite removal
efficiencies (Fig. 1b and c, respectively). In step A, in which the
reactor was operated with real anaerobic effluent diluted with
autotrophic medium (containing 50 mg
L 1 of NO2 -N and
30 mg
L 1 of NH+4-N), instability of the anammox process was
observed, as nitrite was completely consumed (the average
removal efficiency was 99.8%, Fig. 1c), while the average ammonium removal efficiency was 51% (Fig. 1b). These results might
indicate that there was not sufficient nitrite available for the anammox reaction. In fact, in this step (IIIA) the nitrite to ammonium
ratio applied was below 1.0 and therefore was not completely
favorable to the anammox. Moreover, some denitrification
occurred in this phase (as shown in Table 2) which contributed
to leave no terminal electron acceptor for anammox. For this reason, in step B, the reactor was operated with real anaerobic effluent
(undiluted) supplemented with 120 mg
L 1 of nitrite, so that the
anammox bacteria would have sufficient electron acceptors available to oxidize the ammonium. Tang et al. (2010) studied the inhibition of the anammox process when treating effluent with high
COD concentrations (from 50 to 700 mg
L 1), and observed that
anammox reactions were inhibited by denitrifying bacteria
because they competed for electron acceptors. The authors added
40 mg
L 1 of nitrite to the reactor to favor the anammox process.
In Phase IIIB, in which anaerobic effluent supplemented with
nitrite was added to the reactor, the average ammonium removal

efficiencies were higher (85%) than in Phase IIIA (51%) (Fig. 1b).
Statistical differences were observed between these values when
they were compared by MannWhitney test (a = 5%;
p = 0.00016). By the end of Phase IIIB, an increase in ammonium
removal efficiencies to higher than 95% was observed.
Previous studies have shown that anammox bacteria enriched
in reactors fed with organic compounds such as propionate and
acetate are capable of competing with denitrifying heterotrophic
bacteria for the electron acceptor (nitrite) because of their affinity
for the substrate (Kartal et al., 2012). Anammox bacteria have very
low growth rate and cellular yield (coefficient Y = 0.066 0.01)
when compared to denitrifying heterotrophic bacteria
(Y = 0.27 0.3), according to Strous et al., 1998. Therefore, denitrifying bacteria may grow at a higher rate when organic compounds
are combined with ammonium and nitrite in the wastewater being
treated.
Nitrate production (via anammox process) was monitored during Phase III, and the average values obtained were 6.3 and
9.7 mg
L 1 in steps A and B, respectively (Table 2). These values
were lower than those obtained in Phases I and II. This might be
explained by an increase in denitrifying activity following the addition of anaerobic effluent to the SBR. In fact, the mass balance
results (Table 2) confirmed this observation and showed that 28%
of the nitrite added to the reactor was consumed by denitrification
process (in Phase IIIB).
A COD consumption assessment was performed during Phase
III, and the results varied depending on the composition of the
anaerobic effluent used to feed the SBR. The average COD value
determined for the anaerobic effluent was 208.44 54 mg
L 1.
During Phase III, COD removal efficiency was nearly 85% in step
A and 74% in step B (Fig. 2). Therefore, along with the anammox
process, heterotrophic denitrification was also taking place inside
the reactor (as shown in Table 2), as a large part of COD (from
the real anaerobic effluent) was consumed, and that COD removal
explained the high nitrite consumption, and, consequently, the
higher observed stoichiometric coefficient (N-NO2 /N-NH+4) of 1.7
(Table 2). In Phase IIIB, nitrite consumption via denitrification
was 28% while via anammox was 72%, indicating that denitrification occurred in the reactor but anammox process prevailed.
The results from Phase III suggest that the anammox process
can be applied to the post-treatment of anaerobic effluents with
high COD and low ammonium concentrations in order to remove
nitrogen. An important aspect of the present study was the application of the anammox process to the treatment of real anaerobic

263

C.D. Leal et al. / Bioresource Technology 211 (2016) 257266

effluent with COD/N ratios between 4 and 6, while most other

studies have applied COD/N ratios of less than 3, except for Ni
et al. (2012), who also applied COD/N ratios of up to 4 (400 mg
L 1
COD). However, in their study, the authors observed inhibition of
the anammox process with COD/N ratios of 3 and above. As far
as we know, this is the first study that operated an SBR (with
anammox enriched biomass) using real anaerobic effluent (unfiltered and undiluted), rather than synthetic effluent or diluted
effluent, to a reactor with pure anammox sludge (as was done in
most of the previous studies, such as Snchez-Guilln et al.,
2014; Jenni et al., 2014). This is another highlight of the present
study, since anammox bacteria were able to coexist with denitrifying bacteria as reported in the literature (Molinuevo et al., 2009;
Oshiki et al., 2011; Kartal et al., 2012). Furthermore, the biomass
developed in the reactor was capable of simultaneously removing
nitrogen and COD.

Table 3
Identification of DNA band sequences obtained from DGGE.

3.2. Bacterial community structure in the SBR before and after the
addition of glucose and real anaerobic effluent
DGGE profiles and band patterns from biomass developed in the
SBR and sampled at day 349 (after Phase I), day 482 (after Phase II,
with glucose addition), day 448 (recovery phase-R, without glucose
addition), and day 558 (Phase III, after the addition of anaerobic
effluent) were evaluated (Fig. 3). The bacterial community structure changed after the addition of glucose (Phase II) and anaerobic
effluent (Phase III) to the reactor (Fig. 3a).
Representative bands were excised, and the DNA was
sequenced to identify microorganisms present in each phase of
reactors operation (Table 3). Most of the DGGE bands yielded
sequences related to those of bacteria within the phyla Proteobacteria, Chloroflexi, and Planctomycetes.
DGGE results showed that bands A and H, with sequences closely related to the anammox bacteria Candidatus Brocadia caroliniensis and Candidatus Brocadia sinica, respectively, were found in
every phase of the study (Fig. 3a and Table 3). These species are
commonly found in wastewater treatment systems (Hu et al.,
2010). Ca. Brocadia sinica has been reported in the biomass of
anammox reactors from China, Japan, and Germany, indicating
the worldwide distribution of these bacteria (Oshiki et al., 2011).
Physiologic characteristics of the species Ca. Brocadia sinica
were compared to the characteristics of Ca. Brocadia anammoxidans
and Ca. Kuenenia stuttgartiensis by Oshiki et al. (2011). They found
that Ca. Brocadia sinica had a higher growth rate with less affinity

Band

RDP classifier

Blast

Similaritya
(%)

Acc. No.

KF810110.1

97

KF810114.1

C
D

Myxococcales
Chloroflexi

Ca. Brocadia
caroliniensis
Denitratisoma
oestradiolicum
Myxococcales
Anaerolineaceae

98

B, J

Ca.
Brocadiaceae
Rhodocyclaceae

92
97

FJ552616.1

Rhodospirillales

Acetobacteraceae

91

EU193085.1

Chloroflexi

Caldilinea sp.

84

HQ043268.1

Acetobacteraceae

Acetobacteraceae

93

EU921207.1

Ca. Brocadia sp.

Ca. Brocadia sinica

99

I
K

Chlorobi
Actinomycetales

Chlorobi
Propioniferax sp.

88
90

AB565477.1
KJ941776.1

Anaerolineaceae

Chloroflexi

97

JF703582.1

Caldilinea sp

Chloroflexi

94

AY921707.1

Planctomycetes

Planctomycetales

80

FJ517124.1

Thermoflexus sp.

Chloroflexi

95

AY921707.1

Chloroflexi

Chloroflexi

96

AY921913.1

HE648186.1

KP419698.1

Percentages indicate the similarity between the DGGE band sequences and the
closest matched sequences in GenBank. Words in bold indicate: DNA bands A and H
related to anammox bacteria (Brocadia); bands B, J and K are related to bacteria able
to degrade glucose and might be involved in COD removal.

for the substrate compared to the other two species studied. In

their study, many experimental conditions were tested, such as
temperature, dissolved oxygen, pH, salinity, and inhibition by
organic compounds (acetate, propionate, glucose, and methanol).
Oshiki et al. (2011) concluded that in reactors under adverse
conditions and high ammonium and nitrite loads, Ca. Brocadia
sinica prevailed over other anammox bacteria. This was the case
in the present study since band H (closely related to Ca. Brocadia
sinica) were favored in the absence of glucose and prevailed in
the reactor after the addition of the anaerobic effluent.
Bacteria, within the phylum Chloroflexi, were observed in the
present study (DGGE bands D, F, L, M, O, and P, Table 3). This phylum includes bacteria with diversified metabolism (Hug et al.,
2013) and are frequently found in anammox reactors (Cho et al.,
2010; Costa et al., 2014, and Pereira et al., 2014). They can degrade
starch, sugars and peptides (Hug et al., 2013), and thus they might
be involved in COD removal observed in the present study.

Fig. 3. Bacterial community analysis by denaturating gradient gel electrophoresis (DGGE). (a) DGGE profile of the bacterial community in biomass sampled at day 349 (I)
(after Phase I, anammox cultivation), day 482 (II) (after Phase II, glucose addition), day 448 (R) (after 35 days without glucose-recovery phase), and day 558 (III) (Phase III,
after the addition of anaerobic effluent). (b) Dendrogram based on the DGGE profiles.

264

C.D. Leal et al. / Bioresource Technology 211 (2016) 257266

(a)

(b)

Fig. 4. (a) Nitrogen removal rate and abundance of anammox, denitrifying (inferred from the nosZ gene), and total bacterial populations in the biomass of the SBR in Phases I
(anammox cultivation), II (after glucose addition), and III (after anaerobic effluent addition). (b) Relative abundance (in percentage) of anammox bacteria and denitrifiers in
relation to total bacterial populations.

Within the phylum Proteobacteria, DGGE bands B and J (found in

every phase of this study), yielded sequences closely related to that
of Denitratisoma oestradiolicum, which is a denitrifying bacteria isolated from a municipal wastewater treatment plant in Germany
(Fahrbach et al., 2006). This bacterium can degrade 17 betaestradiol as a carbon source, oxidizing it into CO2 and H2O, while
NO3 is reduced to N2O and N2 (Fahrbach et al., 2006). Band K in
the DGGE profile showed sequence related to bacteria of the genus
Propioniferax (Actinomycetales). They are facultative anaerobic
microorganisms that can ferment glucose and produce propionic
acid and reduce nitrate as well (Yokota et al., 1994). Therefore,
bands B, J, and K might be involved in the degradation of glucose
added to the reactor and consequently in COD removal.
The changes in bacterial community structure, during different
phases of reactor operation, were investigated by using 16SrDNADGGE analyses. The similarities between the DGGE profiles for
each sample were calculated and visualized as a dendrogram
(Fig. 3b). Two distinct clusters (with 49.9% similarity) were identified (Fig. 3b). The first cluster consists of samples I (biomass taken
from the SBR after Phase I, day 349) and II (sampled at day 482
after the increasing COD/N ratios) which had high similarity with
each other (above 85%). This result apparently indicates that the
addition of glucose did not change the bacterial community. However, when the DGGE profile of sample R (biomass sampled at day

448, after 35 days without glucose) was compared with sample II

(biomass sampled at day 482 after glucose addition), changes in
the bacterial community structure were observed indicating that
the addition of glucose was the driving force for bacterial selection.
The presence of glucose selected against bands H, M, N, O, and P
which were related to anammox bacteria and Chloroflexi (see
Fig. 3a and Table 3). These DNA bands were prominent in the subsequent phases, indicating that they were favored in the absence of
glucose and prevailed after the addition of the anaerobic effluent
(Fig. 3a). Members of Chloroflexi can utilize microbial products
derived from cell decay (Okabe et al., 2005). This seemed to be
the case in the SBR after the stress caused by the addition of high
concentrations of COD.
The second cluster consists of samples R (biomass sampled at
day 448, without glucose) and III (sampled at day 558 after the
addition of anaerobic effluent) which were more similar to each
other (81.1% similarity) than to the other profiles (samples I and
II) (Fig. 3b). This might be explained by the fact that the anaerobic
effluent had low concentration of biodegradable organic compounds (median value for soluble COD was 80 mg
L 1) and thus
this condition was similar to the Phase R (recovery period without
glucose). Moreover, the addition of anaerobic effluent changed the
bacterial community profile and increased the diversity inside the
reactor, since the Shannon diversity indices for samples I, II and III

C.D. Leal et al. / Bioresource Technology 211 (2016) 257266

were 2.49, 2.50 and 2.82, respectively. The addition of anaerobically pre-treated municipal wastewater to the SBR did not select
against anammox bacteria, on the contrary, it allowed for the coexistence of anammox, denitrifying bacteria and Chloroflexi.
The presence of denitrifying bacteria in the SBR was also confirmed by qPCR (Fig. 4a and b). Higher concentrations of denitrifying bacteria (as measured by nosZ gene abundance) were observed
in the SBR biomass with increasing COD/N ratios (Phase II, with
glucose addition) and after the addition of real anaerobic effluent
(Phase III). These results indicate that the addition of glucose and
the anaerobic effluent favored the enrichment of denitrifiers and
increased their concentration threefold (from 2.57  109 to
7.57  109 copies/g of sludge) (Fig. 4a). However, the concentration
of anammox bacteria (as measured by 16S rRNA gene abundance)
did not change with an increase in COD or application of anaerobic
effluent, indicating that long-term operation of the SBR enriched
and sustained the anammox population (ranging from 2.08  109
to 1.90  109 copies/g of sludge). Ni et al. (2010), investigating an
anammox UASB reactor, observed an anammox concentration of
4.6  108 copies/g of sludge when the reactor reach a nitrogen
removal efficiency of 94%.
The relative abundances of anammox bacteria and denitrifiers
in relation to total bacteria are presented in Fig. 4b. The results also
showed that the addition of glucose and the real anaerobic effluent
(Phases II and III, respectively) changed the bacterial community
by increasing the abundance of denitrifiers more than fourfold in
relation to the anammox population.
4. Conclusions
High concentrations of COD (above 487 mg
L 1) inhibited the
anammox process. However, a COD/N ratio of 5.0 (with COD concentrations up to 300 mg
L 1) did not inhibit the process and
allowed for the coexistence of anammox and denitrifying bacteria.
High COD, nitrite, and ammonium removal efficiencies (80%, 90%,
and 95%, respectively) were obtained with the addition of real
anaerobic effluent to the SBR. This changed the bacterial community structure and selected for DNA sequences related to Brocadia
sinica and Chloroflexi. Thus, the feasibility of applying the anammox
process to nitrogen removal from anaerobically pre-treated municipal wastewater was demonstrated.
Conflict of interest
All authors declare that they have no conflict of interest.
Acknowledgements
We are thankful to the following brazilian agencies: Coordenao de Aperfeioamento de Pessoal de Nvel Superior (CAPES),
to Fundao de Amparo a Pesquisa do Estado de Minas Gerais
(FAPEMIG) and to Conselho Nacional de Desenvolvimento Cientfico e Tecnolgico do Brasil (CNPq).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.biortech.2016.03.
107.
References
Altschul, S.F., Gish, W., Miller, W., Myers, E.W., Lipman, D.J., 1990. Basic local
alignment search tool. J. Mol. Biol. 2015, 403410.
APHA, 2012. Standard Methods for the Examination of Water and Wastewater,
22nd ed. American Public Health Association, Washington, DC, USA.

265

Chanchoi, N., Nitisoravut, S., Schmidt, J.E., 2008. Inactivation of anammox

communities under concurrent operation of anaerobic ammonium oxidation
(anammox) and denitrification. Bioresour. Technol. 99, 33313336.
Cho, S., Takahashi, Y., Fujii, N., Yamada, Y., Satoh, H., Okabe, S., 2010. Nitrogen
removal performance and microbial community analysis of an anaerobic upflow granular bed anammox reactor. Chemosphere 78, 11291135.
Costa, M.C.M.S., Carvalho, L., Leal, C.D., Dias, M.F., Martins, K.L., Garcia, G.B.,
Mancuelo, I.D., Hiplito, T., Mac Conell, E.A., Okada, D., Etchebehere, C.,
Chernicharo, C.A., Araujo, J.C., 2014. Impact of inocula and operating
conditions on the microbial community structure of two anammox reactors.
Environ. Technol. 35, 112.
Dapena-Mora, A., Van Hulle, S.W.H., Campos, J.L., Mendez, R., Van Rolleghem, P.A.,
Jetten, M., 2004. Enrichment of anammox biomass from municipal activated
sludge: experimental and modelling results. J. Chem. Technol. Biotechnol. 79,
14211428.
Date, Y., Isaka, K., Ikuta, H., Sumino, T., Kaneko, N., Yoshie, S., Tsuneda, S., Inamori, Y.,
2009. Microbial diversity of anammox bacteria enriched from different types of
seed sludge in an anaerobic continuous-feeding cultivation reactor. J. Biosci.
Bioeng. 103, 281286.
Enwall, K., Philippot, L., Hallin, S., Salin, S., 2005. Activity and composition of the
denitrifying bacterial community respond differently to long-term fertilization.
Appl. Environ. Microbiol. 71, 83358343.
Fahrbach, M., Kuever, J., Meinke, R., Kmpfer, P., Hollender, J., 2006. Denitratisoma
oestradiolicum gen. Nov., sp. Nov., a 17beta-oestradiol-degrading, denitrifying
betaproteobacterium. Int. J. Syst. Evol. Microbiol. 56, 15471552.
Ferris, M.J., Muyzer, G., Ward, D.M., 1996. Denaturing gradient gel electrophoresis
profiles of 16 rRNA defined populations inhabiting a hot spring microbial mat
community. Appl. Environ. Microbiol. 62, 340346.
Gven, D., Dapena, A., Kartal, B., Schimid, M.C., Maas, B., Van De Pas-Schoonem, K.,
Sozen, S., Mendez, R., Op Den Camp, H.J.M., Jetten, M., Strous, M., Schimidt, I.,
2005. Propionate oxidation by and methanol inhibition of anaerobic
ammonium oxidizing bacteria. Appl. Environ. Microbiol. 71, 10661071.
Hu, B.L., Zheng, P., Tang, C.J., Chen, J.W., Van Der Biezen, E., Zhang, L., NI, B.J., Jetten,
M.S.M., Yan, J., Yu, H.Q., Kartal, B., 2010. Identification and quantification of
anammox bacteria in eight nitrogen removal reactors. Water Res. 44, 5014
5020.
Hug, L.a., Castelle, C.J., Wrighton, K.C., Thomas, B.C., Sharon, I., Frischkorn, K.R.,
Williams, K.H., Tringe, S.G., Banfield, J.F., 2013. Community genomic analyses
constrain the distribution of metabolic traits across the Chloroflexi phylum and
indicate roles in sediment carbon cycling. Microbiome 1, 22. http://dx.doi.org/
10.1186/2049-2618-1-22.
Jenni, S.S., Vlaeminck, E., Morgenroth, E., Ubert, K.M., 2014. Successful application of
nitritation/anammox to wastewater with elevated organic carbon to
ammonium ratios. Water Res. 49, 316326.
Kartal, B., Maalcke, W.J., Almeida, N.M., Cirpus, I., Gloeric, J., Geerts, W., Camp, J.M.,
Harhangi, H.R., Janssen-Megens, E.M., Francoijs, K., Stunnenberg, H.G., Keltjens,
J.T., Jetten, M.S.M., Strous, M., 2011. Molecular mechanisms of anaerobic
ammonium oxidation. Nature 479, 127130.
Kartal, B., Van Niftrik, L., Keltjens, J.T., Op Den Camp, H.J.M., Jetten, M.S.M., 2012.
Anammox-growth physiology, cell biology and metabolism. Microb. Physiol. 60,
211262.
Lackner, S., Terada, A., Smets, B.F., 2008. Heterotrophic activity compromises
autotrophic nitrogen removal in membrane-aerated biofilms: results of a
modeling study. Water Res. 42, 11011112.
Molinuevo, B., Garca, M.C., Karakashev, D., Angelidaki, I., 2009. Anammox from
ammonia removal from pig manure effluents: effect of matter organic content
on process performance. Bioresour. Technol. 100, 21712175.
Ni, B., Hu, B., Fang, F., Xie, W., Kartal, B., Liu, X., Sheng, G., Jetten, M., Zheng, P., Yu, H.,
2010. Microbial and physicochemical characteristics of compact anaerobic
ammonium-oxidizing granules in a upflow anaerobic sludge blanket reactor.
Appl. Environ. Microbiol. 76, 26522656.
Ni, S.Q., Ni, J.Y., Hu, D.L., Sung, S., 2012. Effect of organic compound on the
performance of granular anammox process. Bioresour. Technol. 110, 701705.
Okabe, S., Kindaichi, T., Ito, T., 2005. Fate of 14C-labeled microbial products derived
from nitrifying bacteria in autotrophic nitrifying biofilms. Appl. Environ.
Microbiol. 71, 39873994.
Oshiki, M., Shimokawa, M., Fujii, N., Satoh, H., Okabe, S., 2011. Physiological
characteristics of the anaerobic ammonium-oxidizing bacteria Candidatus
Brocadia sinica. Microbiology 157, 17061713.
Pereira, A.D., Leal, C.D., Dias, M.F., Etchebehere, C., Chernicharo, C.A.L., Arajo, J.C.,
2014. Effect of phenol on the nitrogen removal performance and microbial
community structure and composition of an anammox reactor. Bioresour.
Technol. 166, 103111.
Quan, Z., Rhee, S., Zuo, J., Yang, Y., Bae, J., Park, J.R., Lee, S., Park, Y., 2008. Diversity of
ammonium-oxidizing bacteria in a granular sludge anaerobic ammoniumoxidizing (anammox) reactor. Environ. Microbiol. 10, 31303139.
Snchez-Guilln, J.A., Yimman, Y., Lopez Vazquez, C.M., Brdjanovic, D., Van Lier, J.B.,
2014. Effects of organic carbon source, chemical oxygen demand/N ratio and
temperature on autotrophic nitrogen removal. Water Sci. Technol. 69, 2079
2084.
Strous, M., Heijnen, J.J., Kuenen, J.G., Jetten, M.S.M., 1998. The sequencing batch
reactor as a powerful tool for the study of slowly growing anaerobic
ammonium-oxidizing microorganisms. Appl. Microbiol. Biotechnol. 50, 589
596.
Strous, M., Kuenen, J.G., Jetten, M.S.M., 1999. Key physiology of anaerobic
ammonium oxidation. Appl. Microbiol. Biotechnol. 65, 32483250.

266

C.D. Leal et al. / Bioresource Technology 211 (2016) 257266

Tang, C.J., Zheng, P., Wang, C.H., Mahmood, Q., 2010. Suppression of anaerobic
ammonium oxidizers under high organic content in high-rate Anammox UASB
reactor. Bioresour. Technol. 101, 17621768.
Van de Graaf, A.A., Bruijn, P., Robertson, L.A., Jetten, M.S.M., Kuenen, J.G., 1996.
Autotrophic growth anaerobic ammonium-oxidation micro-organisms in a
fluidized bed reactor. Microbiology 142, 21872196.

Yokota, A., Tamura, T., Takeuchi, M., Weiss, N., Stackebrandt, E., 1994. Transfer of
Propionibacterium innocuum Pitcher and Collins 1991 to Propioniferax gen. nov.
as Propioniferax innocua comb. nov. Int. J. Syst. Bacteriol. 44, 579582.
Zhang, L., Zheng, P., Tang, C., Jin, R., 2008. Anaerobic ammonium oxidation for
treatment of ammonium-rich wastewaters. J. Zhejiang Univ. Sci. B 9, 16426.