Effects of HRT and nitrite/ammonia ratio on anammox discovered in a sequencing batch biofilm reactor

Abstract

There are three key aspects of substrate effect on anaerobic ammonia oxidizing (anammox) bacteria: (1) substrate concentration-based nitrogen loading rate (NLR), (2) hydraulic retention time (HRT)-based NLR and (3) nitrite/ammonia ratio. The first part has been fully investigated in the past, while the latter two are still not properly understood. In this study, two types of substrate effects (HRT-based NLR and nitrite/ammonia ratio) were experimentally proved based on a 226 day operation of a sequencing batch biofilm reactor (SBBR) that was dominated by anammox bacteria. A modified first-order substrate removal kinetic model was developed, which efficiently fits to the experimental results. Decreasing HRTs from 72 h to 6 h were applied to the SBBR, and the HRT = 6 h was proven to be optimal for the highest nitrogen removal rate (NRR) (1.62 kg N m⁻³ d⁻¹ and the total nitrogen removal efficiency >90%). In addition, the influent nitrite/ammonia ratio of 1.2 resulted in a stable and effective operation of anammox SBBR with an improved ammonia removal efficiency (by 17%) and an enhanced NRR (from 0.93 kg N m⁻³ d⁻¹ to 1.14 kg N m⁻³ d⁻¹).

---

Introduction

Anaerobic ammonium oxidation (anammox) is an efficient and environmentally benign process for the treatment of nitrogen-rich wastewater, such as landfill leachates, rejected water, sludge digester liquids and dry-spun acrylic fiber wastewater.^1–3 Anammox bacteria are capable of utilizing nitrite (NO₂⁻) to be alternative terminal electron acceptors along with ammonia (NH₄⁺) being oxidized into nitrogen (N₂), which is principally different from conventional denitrification that employs nitrate (NO₃⁻) as electron acceptor.⁴ Compared to the conventional nitrification-denitrification technologies, the anammox process saves a huge amount of energy consumption due to the less use of aeration, carbon source and alkali, and reduces the production of excess of sludge.^5,6

Anammox bacteria are strictly anaerobic chemolithoautotrophs with extremely low growth rate, and thus they are difficult to be enriched. An effective reactor configuration can play a critical role to solve this difficulty. Previous studies based on the batch or pilot-scale reactors have proven that biofilm-based bioreactors are ecologically feasible and beneficial for slow growing anammox bacteria.^7–9 Granular biomass reactors can successfully work on anammox under a certain range of hydraulic retention times (HRTs), but the possibility of granules being washed out could be high if HRT is lower than 3 hours.¹⁰ Carrier-based biofilm reactors have higher sludge retention capacity and can perform under short HRTs without the negative influence of biomass washout.^10,11 These properties are beneficial to culture anammox biomass because the biofilms growing on a substratum can provide anammox bacteria with fine anaerobic micro-environments, where aerobic bacteria more likely grow on the outer layer as a barrier to oxygen and inhibitory substances.^12–16 Recently, sequencing batch biofilm reactors (SBBRs) that contain PVC mesh medium have been proven of high surface area, and thus regarded to be an efficient design for enriching anammox bacteria.^17–19

HRT, influent nitrite/ammonia ratio and other factors are important for the substrate effect, and thus they are critical to the anammox process.^20–24 A practical purpose while applying anammox is to pursue a shorter HRT for higher nitrogen loading rate (NLR), which is, for most of the cases, a sole way to enhance NLR. Although increasing nitrite concentrations can also bring higher NLR, for practical considerations, nitrite concentrations are always required to be within safe ranges to avoid inhibition effects.^25,26

The purpose of this study is to comprehensively evaluate substrate effect on anammox bacteria, i.e. HRT and nitrite/ammonia ratio, in a SBBR reactor. A substrate removal kinetic model was built to investigate the effect of nitrite/ammonia ratio on NRR to find the optimal HRT and nitrite/ammonia ratio, and eventually to suggest an efficient way to ensure a stable and efficient anammox process.

Materials and methods

Reactor setup and operation

The SBBR had a total exchange volume of one liter. Temperature was maintained at 35 ± 1 °C by a water jacket. A magnetic stirrer was present at the bottom of the reactor (Fig. 1). The HRT was gradually shortened from 3 days to 6 h. The synthetic wastewater (stored in a dark and cool container with pH maintained around 7.0 by adding KHCO₃) was batch fed into the reactor after periodically sparging nitrogen gas (10 minutes sparging before feeding) to minimize the growth potential of aerobic microorganisms in SBBR. The reactor was sequentially operated in cycles and each cycle contained feeding (10 min), settling (20 min), discharging (10 min) and mixing (for the remaining time). Different carriers (ring-style and sheet-style) were placed inside the SBBR with a packing rate of about 40%. The ring carriers are mainly made from high-density polyethylene (Dalian Yu Du Environmental Engineering Technology Co., Ltd, China) with a diameter of 10 mm, a specific surface area of 3 m² g⁻¹ and specific density of 965–968 kg m⁻³. In addition to the ring carriers, some sheet-style carriers (diameter of 3 cm and thickness of 1 mm) were placed at the top, middle and bottom parts of the SBBR for the close observation of the attachment. The reactor was covered by an opaque cloth to avoid the growth of algae and photosynthetic bacteria.

Fig. 1 Experimental setup of the SBBR.

Inoculating sludge and wastewater

The SBBR was inoculated by two sources of seeding sludge (in total 6.25 g VSS): (1) a bench-scale sequencing batch reactor (SBR) treating synthetic ammonia-rich wastewater under ambient temperature (5 g VSS biomass).²⁷ (2) A pilot-scale (17 m³) anammox reactor treating synthetic ammonia-rich wastewater (1.25 g VSS biomass). The SBBR used in this study was fed with synthetic medium (Table 1) with the addition of 1.25 mL L⁻¹ trace elements.^4,5,28

Table 1 The composition of the synthetic wastewater

Nutrient medium	Unit (mg L^−1)	Trace elements	Unit (mg L^−1)
NH4Cl	134–749	ZnSO4·7H2O	430
NaNO2	173–1160	CuSO4·5H2O	250
KHCO3	500	MnCl2·4H2O	990
KH2PO4	10	NiCl2·6H2O,	190
MgSO4·7H2O	60	CoCl2·6H2O	240
CaCl2·2H2O	5	H3BO4	14
FeSO4	6.25	NaSeO4·10H2O	210
EDTA	6.25	NaMoO4·10H2O	220

---

Analysis

The measurements of ammonia, nitrite and nitrate were done according to the standard methods.²⁹ Briefly, ammonia was determined with the Nessler spectrophotometric method. Nitrite was measured using the N-(1-naphthyl)-ethylenediamine spectrophotometry. Nitrate was analyzed with the nitrate electrode. DO, pH and temperatures were measured by a WTW (pH/Oxi 340i, Germany) portable multi-parameter test set. Total nitrogen was analyzed by a TN analyzer (TOC-VCPN-6000, Shimadzu, Japan).³⁰

Fluorescence in situ hybridization analysis (FISH) and scanning electron microscope (SEM)

After around 102 days, the SBBR entered the stable stage, which was characterized by the anammox populations becoming dominant and the attainment of stability in reactor performance (in nitrogen removal rate). During this period, the mature anammox in the SBBR can be characterized by FISH and SEM. Fresh biofilms were collected, fixed in paraformaldehyde and stored in 98% ethanol at −25 °C for the FISH test. The probe Amx 820, which is specific for Candidatus Brocadia anammoxidans and Candidatus Kuenenia stuttgartiensis, was purchased from TaKaRa, Dalian, China and labeled with Cy3.³¹ The hybridizations with fluorescent probes were performed according to a previous protocol.²⁷ The samples were counterstained by DAPI. A confocal laser-scanning microscope (CLSM, Carl Zeiss, Oberkochen, Germany) equipped with an Ar ion laser (488 nm) and He–Ne laser (543 nm) was used for observation.

The biofilm samples for SEM test were first washed with a phosphate buffer and fixed with 2% glutaraldehyde overnight at 4 °C, followed by a series of processes including successive dehydration, drying and gold coating, according to the previous method.²⁷ A Hitachi S-4700 (Japan) scanning electron microscope was used to capture micrographs.

First-order substrate removal model

Following the online data recording, we compared and screened the fitting results of various models, and then established a first-order substrate removal model to simulate the SBBR performance, which was simple and capable of properly matching the observations. Within the first-order substrate removal model, the change rate of substrate concentration in a complete mixed system can be expressed as follows:^32,33

Some assumptions of the SBBR system are as follows: (1) it keeps a pseudo-steady-state condition, (2) the influent filling is instantaneous, and (3) there is no diffusion limitation within the biofilms.³⁴ Because the change rate (−ds/dt) is negligible, the equation can be transitioned as follows:

Further described as follows:

where Q and V are the inflow rate (L h⁻¹) and the reactor volume (L), Si and Se are influent and effluent substrate (ammonia and nitrite) concentrations (mg L⁻¹), k is the first-order substrate removal rate constant (1/h), HRT is the hydraulic retention time (h).

The HRT can be considered to be the reaction time (t) for each batch. To solve the equation closer to the actual situation of the reactor, the first-order substrate removal constant b was used to modify the equation and it was changed to

where b is the first-order substrate removal constant.

Results and discussion

Observation of anammox bacteria

A mature anammox community was observed after about 100 days and during this period the reactor performance was stable as well. The clear and large area of red fluorescence, corresponding to the anammox bacteria, was observed by CLSM (Fig. 2A–C), which indicated that the large amounts of anammox bacteria exist in the biofilms. The SEM proves that the heterogeneous surface of the carriers (Fig. 2D) helped to harbor biofilms, and the round shape anammox bacterial cells (Fig. 2E) can be clearly seen in the biofilms. All these evidences indicated a suitable time to perform the online monitoring to build the first-order substrate removal model, and eventually to evaluate and predict the SBBR performance.

Fig. 2 Molecular and microscopic evidences of anammox bacterial cells in SBBR. (A) Fluorescence in situ hybridization (FISH) micrograph of Cy3-labeled Amx820 (targeting two anammox bacterial species Candidatus Brocadia anammoxidans and Candidatus Kuenenia stuttgartiensis). (B) FISH micrograph of DAPI stained sample (targeting total bacteria). (C) FISH micrograph of Cy3-labeled Amx820 (targeting anammox bacteria) and counter-stained with DAPI. The dominance of anammox bacterial community can be seen in this figure based on the percentage of red fluoresence versus blue. (D) Scanning electron microscopy image (SEM) of the surface of a virgin carrier; (E) SEM of a mature biofilm growing on the surface of a carrier. Heterogeneous surface of the carriers (Fig. 2D) help to harbour biofilms (Fig. 2E) and the round shape anammox bacterial cells (pointed by red arrows) can be clearly seen in the biofilms.

Kinetics of ammonia and nitrite removal

Selecting a suitable HRT is very important for the successful culturing of anammox bacteria. The reactor was tested under different substrate concentrations to obtain the optimal HRT and data set for modelling. HRT was decreased stepwise from 3 days to 6 hours. During this process, the reactor went through three stages: period of instability (stage I), transition period (stage II), and robust + stable period (stage III).²⁷ The reactor was in a very stable period during HRT of 12 hours, and thus the substrate concentration was monitored online using a real-time recording mode (Fig. 3A and B). The low concentrations (80 mg L⁻¹) of nitrite and ammonia were initially fed to the reactor. The concentration of ammonia and nitrite decreased to 23 mg L⁻¹ and 0 mg L⁻¹ respectively after 6 h; enough amount of nitrite was not supplied to anammox for the next 6 hour, thus the initial medium concentration was increased to 140 mg L⁻¹ (ammonia and nitrite each). The concentration of ammonia decreased to 31.8 mg L⁻¹ and that of nitrite to 8.2 mg L⁻¹ for 6 h.

Fig. 3 The variation of different substrate concentrations at HRT 12 h (A: 80 mg L⁻¹; B: 140 mg L⁻¹).

The reactor performed in an effective and stable mode after about 100 days of start-up. During this period (HRT 12 h), the initial substrate concentration was 70 mg L⁻¹. To clearly express the relationship of removed substrate, the dynamic equation of substrate removal was derived as a linear equation. The constituted model efficiently fitted to the experimental values under both initial ammonia concentrations of 80 mg L⁻¹ and 140 mg L⁻¹ (the experimental values and calculated values are listed in Fig. 3A and B), with the r² values being 0.962 and 0.965, respectively (Fig. 4A and B). The model also enabled a fine predictability of nitrite concentration with r² of 0.934 and 0.955, respectively (Fig. 4C and D). The above information demonstrates that the established first-order substrate removal model was suitable to characterize the kinetics for ammonia and nitrite depletion in the anammox SBBR, which can also be applicable to other types of reactors according to the previous studies.^32,33

Fig. 4 Kinetic characteristic and correlation coefficient. (A) kinetic model of ammonia removal; (B) correlation coefficient between calculated values and experimental values under different ammonia concentrations (80 mg L⁻¹ and 140 mg L⁻¹); (C) kinetic model of nitrite removal; (D) correlation coefficient between calculated values and experimental values under different nitrite concentrations (80 mg L⁻¹ and 140 mg L⁻¹).

Results showed that the nitrite was 0 mg L⁻¹ and 8.2 mg L⁻¹ after 6 h for the groups with initial nitrite concentrations of 80 mg L⁻¹ and 140 mg L⁻¹, respectively (Fig. 3). The calculated values of substrate also showed similar results (Fig. 3). The remaining ammonia and nitrite were not sufficient to support the growth of anammox bacteria in the following six hours (considering the HRT of 12 h). Consequently, it is necessary to shorten the HRT to 6 h to save half of the time and get a higher nitrogen load. This indicates that the SBBR reactor had an excellent nitrogen removal capacity as well. In addition, it is generally accepted by others that the mode with a lower concentration of the substrate is superior to the one with the higher concentration under the same HRT conditions.²³ Based on the substrate concentration model, a suitable HRT for the reactor and the limitation of substrate to the anammox were also identified to be the crucial factors in recent studies. In practice, this model was significant to predict the treatment plant performance and optimize the plant design.^33,35

Effect of HRT

Anammox was enriched under different HRTs. The initial HRT was 72 h, which was then shortened step by step from 72 h to 48 h, 24 h, 12 h and 6 h. The initial substrate concentration was 70 mg L⁻¹ and the removal efficiency of ammonia and nitrite were closely observed to promptly increase or decrease the concentration of the substrate. 100 days after the start-up, the reactor stayed in an effective and stable period (HRT was 12 h), and the HRT was mainly studied during this period. When the HRT decreased from 12 h to 6 h, both the ammonia removal efficiency and nitrogen loading rate were improved. The SBBR reactor performed for about 30 days under HRT of 12 h, during this period the ammonia and nitrite concentration was increased in a stepwise manner (70 mg L⁻¹, 84 mg L⁻¹, 112 mg L⁻¹ and 140 mg L⁻¹). The ammonia removal efficiency was 77%.When the HRT was set at 6 h; the substrate concentration was elevated from 140 mg L⁻¹ to 196 mg L⁻¹. Accordingly, the ammonia removal efficiency reached to 92% (Fig. 5) and the TN removal efficiency increased from 78.6% to 87.1%. The SBBR exhibited a performance as stable as before without any negative impact. The nitrogen loading rate was increased by a factor of four from 0.28 kg N m⁻³ d⁻¹ to 1.18 kg N m⁻³ d⁻¹ at HRT of 6 h (Fig. 5). Shortening the HRT was an indirect but effective way to improve the anammox efficiency to meet a high nitrogen loading rate, whereas the increased substrate concentration may stimulate anammox bacteria growth, yielding sufficient biomass to support the increasing loading rate.²⁰ It is important to note that the medium concentration of nitrite should be carefully controlled because high nitrite (e.g. >210 mg L⁻¹ (15 mM)) may result in the inhibition of anammox cells.^36–38

Fig. 5 Nitrogen transformation at different HRTs (left axis indicates nitrogen removal efficiency (%) and the right one indicates the TN removal rate (kg N m⁻³ d⁻¹)).

The stoichiometry ratios of nitrite/ammonia and nitrate/ammonia are the key factors for evaluating the health of an anammox process.³⁹ The corresponding stoichiometric values 1.32 (nitrite/ammonia) and 0.26 (nitrate/ammonia) have been widely proven and accepted to be the indicators of a typical anammox process.⁶ In this study, when HRT was decreased from 12 h to 6 h, nitrite/ammonia and nitrate/ammonia ratio reached 1.26 and 0.26 (Fig. 6), respectively, which are close to the theoretical values. However, when the HRT was longer than 12 h (period of instability), high accumulated nitrate was observed, which may be the result of a strong or a weak denitrification activity.²⁷ Then, AOB and NOB were most probably inactive due to the strict control of medium DO and the washout of some loosely attached AOB/NOB from the out layer of biofilm.⁴⁰ The real-time experimental results also showed that the linear fitting nitrite/ammonia ratio was 1.25 with an R² value of 0.996. According to the previous studies, the ratios observed in an upflow biofilter were 1.0 ± 0.171 and 0.2 ± 0.105. The value found in an anammox upflow column reactor was 1.03–1.17.^39,41 The stoichiometric data strongly indicated a typical anammox process in the SBBR, which was in accordance to the previous molecular biological results that showed anammox bacteria to be dominant with a relative abundance of about 32%.²⁷

Fig. 6 Stoichiometric ratio of nitrite/ammonia and nitrate/ammonia at different HRTs.

A distinct advantage of biofilm-based reactors (such as the SBBR in this study) is its ability to maintain a fine-tuned and self-adapted micro-environment, which can benefit both the fast growing microorganisms (such as aerobic ones) and slow growers (such as anaerobic ones). In this study, the carrier substratum provided fine conditions for anammox bacteria to grow and the out-layer biomass played an important role to be a barrier to oxygen and inhibitory substances. The SBBR reactor used in this study was exposed to open air during the entire operation (Fig. 1). However, the continuous penetration of oxygen did not significantly affect the anammox process, neither was there any inhibition to anammox bacteria observed. On the contrary, anammox bacteria became dominant after three months. Compared to other reactor configurations, such as suspended sludge or granular sludge, SBBR is cost-saving in terms of building and power-saving during practical use as well.

Effect of influent nitrite/ammonia ratio

The effect of influent nitrite/ammonia ratio was investigated under controlled substrate concentrations. Considering the fact that a high concentration of nitrite (e.g. >15 mM) may inhibit anammox bacteria and lead to incomplete conversion,^42,43 the SBBR reactor was first fed with nitrite/ammonia ratio of 1 : 1 (10 mM/10 mM). The HRT was fixed at 6 h. The real-time online results showed that the amount of nitrite was insufficient to support the growth of anammox after 6 h. Then, the ratio was increased to 1.1 : 1 (11 mM/10 mM), and the average ammonia removal efficiency increased by 4%. A further increase in nitrite/ammonia ratio to 1.2 : 1 (12 mM/10 mM) led to an increase in ammonia removal efficiency by 17% (Table 2). It is notable that the NRR was improved from 0.93 to 1.14 kg N m⁻³ d⁻¹ during this period under the fixed concentration of ammonia, but nitrite increasing concentration increased, however, the nitrite in the effluent was continuously lower than 1 mM. The reactor performance was not inhibited by the high nitrite concentration, and this is probably attributed to the advantageous biofilm architectures of SBBR carriers.⁴⁴

Table 2 -Nitrogen removal efficiencies at different influent ratios of nitrite/ammonia

Nitrite/ammonia ratio	1	1.1	1.2
Ammonia removal (%)	78.9 ± 3.3	88.4 ± 1.8	97.0 ± 2.9
Nitrite removal (%)	99.8 ± 0.4	100.0 ± 0.0	98.6 ± 1.9
TN removal (%)	80.2 ± 1.7	86.6 ± 4.1	89.8 ± 1.9
NRR (kg N m^−3 d^−1)	0.93 ± 0.04	1.01 ± 0.08	1.14 ± 0.04

---

It is worthwhile to mention that an even higher mole ratio of nitrite/ammonia (e.g. >1.5 : 1) may negatively influence the anammox process because it will lead to a higher residual of nitrite, which may promote the growth of NOB and denitrifying bacteria, which can strongly compete with anammox bacteria.⁴⁵ Previous researchers have found that when the influent ratio of nitrite/ammonia increased from 1.5 : 1 to 1.8 : 1, the anammox process was severely affected, and most of the studies have concluded that an optimal ratio level should be around 1.2 : 1.³⁹

Conclusions

The study demonstrates the co-existence of aerobic bacteria and anammox bacteria in the SBBR, and it displayed that the anammox bacteria became dominant after three months. The performance of the reactor was also significantly satisfactory. Compared to other reactor configurations, such as suspended sludge or granular sludge, SBBR is cost-effective for building and power-saving applications.

The HRT and nitrite/ammonia ratio effects on the anammox process were also studied. The results show that an optimal HRT for anammox SBBR is 6 h, under which the highest NRR (1.62 kg N m⁻³ d⁻¹) can be reached. The stoichiometric ratio of nitrite/ammonia was also proven to be critical to anammox and the optimal ratio was found to be 1.2. The kinetic parameters of a first-order substrate (ammonia and nitrite) removal model suitable for SBBR was established and each model efficiently fits to the experimental results (r² = 0.962 and 0.965 for ammonia, r² = 0.934 and 0.955 for nitrite). The study demonstrates that the substrate effect, in terms of HRT and stoichiometric ratio of nitrite/ammonia, is of great importance for a stable and efficient anammox process.

Acknowledgements

This research was supported by National Natural Science Foundation of China (no. 21177033).

References

1. D. Scaglione, M. Ruscalleda, E. Ficara, M. D. Balaguer and J. Colprim, Chem. Eng. J., 2012, 209, 62–68.
2. I. Zekker, E. Rikmann, T. Tenno, V. Lemmiksoo, A. Menert, L. Loorits, P. Vabamae, M. Tomingas and T. Tenno, Biodegradation, 2012, 23, 547–560.
3. P. An, X. Xu, F. Yang, L. Liu and S. Liu, Chem. Eng. J., 2013, 222, 32–40.
4. A. A. van de Graaf, P. de Bruijn, L. A. Robertson, M. S. M. Jetten and J. Gijs Kuenen, Microbiology, 1996, 142, 2187–2196.
5. M. Strous, E. Van Gerven, P. Zheng, J. G. Kuenen and M. S. Jetten, Water Res., 1997, 31, 1955–1962.
6. M. S. M. Jetten, M. Strous, K. T. van de Pas-Schoonen, J. Schalk, U. G. J. M. van Dongen, A. A. van de Graaf, S. Logemann, G. Muyzer, M. C. M. van Loosdrecht and J. G. Kuenen, FEMS Microbiol. Rev., 1998, 22, 421–437.
7. L. Chu and J. Wang, Chem. Eng. J., 2011, 170, 220–225.
8. H. De Clippeleir, X. Yan, W. Verstraete and S. E. Vlaeminck, Appl. Microbiol. Biotechnol., 2011, 90, 1537–1545.
9. K. Egli, F. Bosshard, C. Werlen, P. Lais, H. Siegrist, A. J. B. Zehnder and J. R. van der Meer, Microb. Ecol., 2003, 45, 419–432.
10. U. Imajo, T. Tokutomi and K. Furukawa, Water Sci. Technol., 2004, 49, 155–163.
11. H. Okamoto, K. Kawamura, T. Nishiyama, T. Fujii and K. Furukawa, Biodegradation, 2013, 24, 99–110.
12. T. Kindaichi, I. Tsushima, Y. Ogasawara, M. Shimokawa, N. Ozaki, H. Satoh and S. Okabe, Appl. Environ. Microbiol., 2007, 73, 4931–4939.
13. Z. Gong, S. T. Liu, F. L. Yang, H. Bao and K. Furukawa, Bioresour. Technol., 2008, 99, 2749–2756.
14. Y. Xiao, G. M. Zeng, Z. H. Yang, Y. S. Liu, Y. H. Ma, L. Yang, R. J. Wang and Z. Y. Xu, J. Appl. Microbiol., 2009, 106, 496–505.
15. S. W. H. Van Hulle, H. J. P. Vandeweyer, B. D. Meesschaert, P. A. Vanrolleghem, P. Dejans and A. Dumoulin, Chem. Eng. J., 2010, 162, 1–20.
16. S. E. Vlaeminck, A. Terada, B. F. Smets, H. De Clippeleir, T. Schaubroeck, S. Bolca, L. Demeestere, J. Mast, N. Boon, M. Carballa and W. Verstraete, Appl. Environ. Microbiol., 2010, 76, 900–909.
17. Z. J. Zhang, S. H. Chen, P. Wu, L. F. Lin and H. Y. Luo, Bioresour. Technol., 2010, 101, 6309–6314.
18. C. Wantawin, J. Juateea, P. L. Noophan and J. Munakata-Marr, Water Sci. Technol., 2008, 58, 1889–1894.
19. E. H. Koupaie, M. A. Moghaddam and S. Hashemi, Water Sci. Technol., 2013, 67, 1816–1821.
20. K. Isaka, T. Sumino and S. Tsuneda, J. Biosci. Bioeng., 2007, 103, 486–490.
21. S. Bagchi, R. Biswas and T. Nandy, J. Ind. Microbiol. Biotechnol., 2010, 37, 871–876.
22. C.-J. Tang, P. Zheng, L.-Y. Chai and X.-B. Min, Chem. Eng. J., 2013, 230, 149–157.
23. I. Tsushima, Y. Ogasawara, T. Kindaichi, H. Satoh and S. Okabe, Water Res., 2007, 41, 1623–1634.
24. D. X. Liao, X. M. Li, Q. Yang, Z. H. Zhao and G. M. Zeng, World J. Microbiol. Biotechnol., 2007, 23, 1015–1020.
25. I. Fernández, J. Dosta, C. Fajardo, J. L. Campos, A. Mosquera-Corral and R. Méndez, J. Environ. Manage., 2012, 95, S170–S174.
26. R.-C. Jin, G.-F. Yang, J.-J. Yu and P. Zheng, Chem. Eng. J., 2012, 197, 67–79.
27. Y. C. Yu, D. W. Gao and Y. Tao, Appl. Microbiol. Biotechnol., 2013, 97, 6057–6064.
28. C. Trigo, J. L. Campos, J. M. Garrido and R. Méndez, J. Biotechnol., 2006, 126, 475–487.
29. L. S. Clescerl, A. E. Greenberg and A. D. Eaton, Standard Methods for the Examination of Water and Wastewater, American Public Health Association, Washington D.C., 21st edn, 2005.
30. Y. Tal, J. E. M. Watts and H. J. Schreier, Appl. Environ. Microbiol., 2006, 72, 2896–2904.
31. N. Chamchoi and S. Nitisoravut, Chemosphere, 2007, 66, 2225–2232.
32. R. C. Jin and P. Zheng, J. Hazard. Mater., 2009, 170, 652–656.
33. S.-Q. Ni, P.-H. Lee and S. Sung, Bioresour. Technol., 2010, 101, 5767–5773.
34. D. Gao, Y. Peng and W.-M. Wu, Environ. Sci. Technol., 2010, 44, 5015–5021.
35. A. K. Vangsgaard, M. Mauricio-Iglesias, K. V. Gernaey, B. F. Smets and G. Sin, Bioresour. Technol., 2012, 123, 230–241.
36. C.-J. Tang, P. Zheng, C.-H. Wang, Q. Mahmood, J.-Q. Zhang, X.-G. Chen, L. Zhang and J.-W. Chen, Water Res., 2011, 45, 135–144.
37. J. C. Yang, L. Zhang, Y. Fukuzaki, D. Hira and K. Furukawa, Bioresour. Technol., 2010, 101, 9471–9478.
38. T. L. G. Hendrickx, Y. Wang, C. Kampman, G. Zeeman, H. Temmink and C. J. N. Buisman, Water Res., 2012, 46, 2187–2193.
39. R.-C. Jin, B.-S. Xing, J.-J. Yu, T.-Y. Qin and S.-X. Chen, Ecol. Eng., 2013, 53, 130–137.
40. A. Gali, J. Dosta, M. C. M. van Loosdrecht and J. Mata-Alvarez, Process Biochem., 2007, 42, 715–720.
41. S. T. Liu, F. L. Yang, F. A. Meng, H. H. Chen and Z. Gong, J. Biotechnol., 2008, 138, 96–102.
42. A. Dapena-Mora, I. Fernandez, J. L. Campos, A. Mosquera-Corral, R. Mendez and M. S. M. Jetten, Enzyme Microb. Technol., 2007, 40, 859–865.
43. C. J. Tang, P. Zheng, Q. Mahmood and J. W. Chen, J. Ind. Microbiol. Biotechnol., 2009, 36, 1093–1100.
44. T.-T. Chen, P. Zheng, L.-D. Shen, C.-J. Tang and S. Ding, Biotechnol. Bioprocess Eng., 2012, 17, 1093–1102.
45. Y. Tao, D. W. Gao, Y. Fu, W. M. Wu and N. Q. Ren, Bioresour. Technol., 2012, 104, 73–80.

---