Association of running manner with bacterial community dynamics in a partial short-term nitrifying bioreactor for treatment of piggery wastewater with high ammonia content

Abstract

Optimization of running parameters in a bioreactor requires detailed understanding of microbial community dynamics during the startup and running periods. Using a novel piggery wastewater treatment system termed "UASB + SHARON + ANAMMOX" constructed in our laboratory, we investigated microbial community dynamics using the Illumina MiSeq method, taking activated sludge samples at ~2-week intervals during a ~300-day period. Ammonia-oxidizing bacteria (AOB) were further investigated by quantification of AOB amoA genes and construction of gene clone libraries. Major changes in bacterial community composition and dynamics occurred when running manner was changed from continuous flow manner (CFM) to sequencing batch manner (SBM), and when effluent from an upflow anaerobic sludge blanket (UASB) reactor for practical treatment of real piggery wastewater was used as influent; differences among these three experimental groups were significant (R ² = 0.94, p < 0.01). When running manner was changed from CFM to SBM, relative abundance of the genus Nitrospira decreased sharply from 18.1 % on day 116 to 1.5 % on day 130, and to undetectable level thereafter. Relative abundance of the genus Nitrosomonas increased from ~0.67 % during the CFM period to 8.0 % by day 220, and thereafter decreased to a near-constant ~1.6 %. Environmental factors such as load ammonia, effluent ammonia, effluent nitrite, UASB effluent, pH, and DO levels collectively drove bacterial community dynamics and contributed to maintenance of effluent NH ₄ ⁺-N/NO2 ^--N ratio ~1. Theses results might provide useful clues for the control of the startup processes and maintaining high efficiency of such bioreactors.

Du et al. AMB Expr (2016) 6:76 DOI 10.1186/s13568-016-0245-5

Association ofrunning manner withbacterial community dynamics ina partial short-term nitrifying bioreactor fortreatment ofpiggery wastewater withhigh ammonia content

WeiLi Du1,2, Qiang Huang1,2, LiLi Miao1, Ying Liu1 and ZhiPei Liu1*

Introduction

Ammonia, a common aquatic pollutant, is a cause of numerous environmental problems. Wastewater from piggeries (pig farms) contains high levels of chemical oxygen demand (COD) and ammonia (Bernet etal. 2000;

Zhu etal. 2013) and is a major source of ammonia pollution (Bernet etal. 1996; Li etal. 2012). Anaerobic digestion coupled with biogas production, as utilized in upow anaerobic sludge blanket (UASB) reactors, is a widely used and eective method of COD removal (Hashimoto 1983; Llabrs-Luengo and Mata-Alvarez 1987; Lo et al. 1994). On the other hand, nitrogen removal methods generally rely on the conventional nitricationdenitrication process, which consumes huge amounts of oxygen and organic matter (Bernet etal. 1996; Boiran etal. 1996; Odegaard 1988).

*Correspondence: [email protected]

1 State Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences, No. 1 West Beichen Road, Chaoyang District, Beijing 100101, Peoples Republic of ChinaFull list of author information is available at the end of the article

2016 The Author(s). This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/

Web End =http://creativecommons.org/licenses/by/4.0/ ), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Du et al. AMB Expr (2016) 6:76

Page 2 of 16

Anaerobic ammonium oxidation (anammox/ANAMMOX) bacteria, a type of lithotrophic microorganism, were originally discovered in bioreactors of wastewater treatment plants, in which ammonium was oxidized with nitrite as electron acceptor to produce dinitrogen gas (N2)

(Mulder etal. 1995; Strous etal. 1999; van de Graaf etal. 1995). Application of anammox bacteria for treatment of wastewater containing high ammonium and low organic matter level was shown to result in operational cost savings up to 90% (Jetten etal. 2001). For this purpose, a single reactor system for high activity ammonium removal over nitrite (SHARON) reactor is necessary to transform ammonia into nitrite and ensure an effluent NH4+-N/

NO2-N ratio ~1 to meet the requirement of anammox bacteria (Kuenen 2008). Ammonia-oxidizing bacteria (AOB) and nitrite-oxidizing bacteria (NOB) are the main forces for aerobic oxidation of ammonia to nitrate in a SHARON reactor (Li etal. 2015). AOB, but not NOB, are expected to grow in a SHARON reactor, because nitrate accumulation is not acceptable in the process.

Many studies have focused on the structure of AOB communities in ammonia-contaminated sediment and in the activated sludge generated by treatment of ammonia-containing wastewater (Bai et al. 2012; Fitzgerald et al. 2015; Flood et al. 2015; Wang et al. 2012; Zhang et al. 2011, 2015). Concentrations of ammonia and dissolved oxygen (DO) are the key parameters that aect nitrogen removal processes as well as AOB community structure (Park and Noguera 2004). The predominant AOB species found in bioreactors is typically Nitrosomonas europaea (Limpiyakorn etal. 2006; Park and Noguera 2004; Wells et al. 2009), or in some cases Nitrosomonas oligotropha (Dionisi etal. 2002).

Depending on the running manner of bioreactors, differing parameters may strongly aect microbial community structure during adaptation to environmental changes (Turner et al. 1998; Wells et al. 2009). Microbial community composition in activated sludge directly determines the efficiency of wastewater treatment (Ibekwe etal. 2003; Wagner etal. 2002). A long period of time is necessary for generation of seed sludge to start a bioreactor and achieve optimal removal efficiency (Lopez etal. 2008; van der Star etal. 2007; Zheng etal. 2004). No study to date has described microbial community dynamics in activated sludge of a SHARON reactor, from the stable startup stage through wastewater treatment stage. The environmental and engineering factors that determine the dynamics of microbial community structure throughout the entire running time must be elucidated. We need to identify the relevant variables in order to design and optimize nitrication systems (Limpiyakorn etal. 2006).

We developed a novel system termed UASB+SHARON + ANAMMOX and evaluated its feasibility for treatment of piggery wastewater. In the present study, microbial community dynamics in a SHARON reactor were investigated using the Illumina MiSeq method, taking activated sludge samples at ~2-week intervals during a ~300-day period. AOB were further investigated by quantication of amoA and construction of amoA gene clone libraries. Our focus was the environmental and engineering factors that controlled the dynamics of microbial community succession. The AOB responsible for nitrite accumulation were evaluated throughout the running time. Our ndings provided new insights into microbial community dynamics and the relationships between these dynamics and bioreactor efficiency, and will be useful in optimizing running parameters for rapid startup and stable running of partial short-term nitrifying reactors.

Materials andmethods

SHARON reactor andrunning parameters

A laboratory-scale combined UASB + SHARON + ANAMMOX system (Additional le 1: Figure

S1; Table S1) was constructed to treat piggery wastewater (CODcr 6003000 mg/L, NH4+-N 5001500 mg/L)

obtained from a husbandry base affiliated with the Chinese Academy of Agriculture Sciences, Changping District, Beijing, China. In this system, the three reactors were started up separately, and subsequently connected. This system efficiently treated piggery wastewater with effluent levels that met national discharge standards. The function of SHARON was to partially transform high-concentration ammonia to nitrite and ensure effluent NH4+-N/NO2-N ratio ~1 to meet ANAMMOX requirements. The SHARON reactor was constructed of plexi-glass [poly(methyl methacrylate)] with height 600 mm, diameter 194mm, and eective volume 12.5L. Activated sludge obtained from the aeration tank of a wastewater treatment plant was used as inoculum for startup. Concentrations of volatile solids and suspended solids were 3.50 and 4.87 g/L, respectively. SHARON was run at room temperature with hydraulic retention time (HRT) 25 h. The reactor was run initially in continuous ow manner (CFM) for 120days, with DO level 0.71.5mg/L, without pH control, and run subsequently in sequencing batch manner (SBM) with DO level 7.08.0 mg/L. The running cycle of SBM was 8h, consisting of four stages: aerobic ll (130min), aerobic (318min), settle (30min), and draw (90s). On day 220, effluent from UASB treating real piggery wastewater was used as inuent. Inuent characteristics for various stages are described in Additional le1: Table S2.

Du et al. AMB Expr (2016) 6:76

Page 3 of 16

Sample collection andDNA extraction

Activated sludge samples from SHARON were collected at ~2-week intervals during startup and running period. In total, 19 samples were obtained: 7 from the CFM period and 12 from the SBM period. Three effluent samples from UASB were also obtained.

Total DNA was extracted from each sample (~0.5 g) using a PowerSoil DNA isolation kit (MO BIO Laboratories; Shenzhen, China) according to the manufacturers instructions, and stored at 80C.

Illumina sequencing analysis of16S rRNA gene amplicons

Bacterial communities were analyzed for the 22 samples described above. The primer set used was 338F/806R, which targets the V3-V4 hypervariable region of bacterial 16S rRNA gene. MiSeq PE300 was used to obtain a 468-bp fragment. Raw data were processed with the Quantitative Insights Into Microbial Ecology (QIIME) toolkit, v. 1.8.0 (Caporaso et al. 2010). Chimeric sequences were checked and ltered with the UCHIME program (Caporaso etal. 2010). Quality reads were clustered into operational taxonomic units (OTUs) at 97 % sequence similarity using UPARSE (Edgar 2013). A representative sequence of each OTU was selected for taxonomic assignment using the Greengenes Database (Wang etal. 2007). For all OTU-based analyses, sequence number was normalized prior to statistical analysis by randomly resampling reads of each sample to the same size, based on the sample with the smallest sampling size. QIIME was also used to generate BrayCurtis distance metrics (Gauch 1973) and -diversity indices, including Chao 1 richness estimation, Shannon, ACE, Simpson, and Goods coverage. All analyzed sequences were deposited in the National Center for Biotechnology Information (NCBI; Bethesda, MD, USA) Sequence Read Archive (SRA) database under accession number SRP072716.

Realtime quantitative polymerase chain reaction (qPCR)

qPCR was performed on an ABI ViiA 7 quantitative thermocycler (Applied Biosystems, USA). Primer set amoA-1F/amoA-2R (Chen et al. 2008; Rotthauwe et al. 1997; Zhang et al. 2015) was used to amplify bacterial amoA gene. The thermal program for qPCR of bacterial amoA gene was: 3min at 94C, 40 cycles of 30s at 94C, 55s at 60C, and 45s at 72C (Chen etal. 2008). Primer set 341F/518R was used to quantify bacterial 16S rRNA gene, with thermal program: 3 min at 95 C, 40 cycles of 30s at 95C, 30s at 60C, and 40s at 72C (He et al. 2007). The 20-L reaction mixture consisted of 10 L 2 KAPA SYBR FAST qPCR Master Mix2

Universal (KAPA Biosystems; Beijing), 0.4 L Forward Primer (10 M), 0.4 L reverse primer (10 M), 0.4 L 50 ROX/Low, 2 L diluted DNA template (<20 ng),

and 6.8 L double-distilled H2O. A standard curve was constructed using recombined plasmid with bacterial 16S rRNA gene as template, and AOB amoA gene (Bai etal. 2012). All reactions, including standards and the 19 SHARON samples, were performed in triplicate.

Bacterial amoA gene cloning andsequencing

Bacterial amoA gene libraries were constructed for 10 time points (days 1, 38, 78, 98, 116, 145, 206, 220, 235, and 261) selected during the experimental period. The qPCR products (amoA gene) described above were ligated to pEASY-T1 cloning vector, and recombined products were used to transform Trans1-T1 phage resistant competent cells by heat-shock method with a pEASY-T1 Cloning Kit (TransGen Biotech; Beijing) according to the manufacturers protocol. White clones were picked randomly and re-amplied using primer set M13F/M13R to screen positive clones. Screened positive colonies were subjected to sequencing. All obtained amoA gene sequences were deposited in the National Center for Biotechnology Information (NCBI; Bethesda, MD, USA) GenBank database under accession numbers KX215988-KX216304.

All bacterial amoA gene sequences obtained were clustered into dierent aOTUs (the term signies dierentiation from those based on 16S rRNA gene sequence) with 97% similarity cuto (Gao etal. 2013; Zhang etal. 2015). AOB diversity indices were calculated by QIIME as described above.

Statistical analysis

Microbial community succession patterns in the SHARON reactor during the entire running period were determined by principle coordinate analysis (PCoA) based on BrayCurtis distance (Gauch 1973). The ANOSIM (analysis of similarity) algorithm was used to identify notable dierences among groups. Pearsons test and redundancy analysis (RDA) were used to evaluate correlations between variable environmental factors and the dynamic microbial community. Pearsons test was also used to evaluate correlations between variable environmental factors and major phyla or other taxonomic units of nitrifying bacteria, and between -diversity and running parameters. The above analyses were performed with the R software program (v. 3.2.1; http://www.r-project.org

Web End =http://www.r-project.org ). Phylogenetic trees were constructed using the MEGA 6.0 software program (Tamura etal. 2013), based on representative sequence for each OTU/aOTU, by neighbor-joining (NJ) method with bootstrap values calculated from 1000 replications.

Results

Performance ofthe SHARON reactor

The performance of the SHARON reactor was investigated during the entire experimental period, in regard to

Du et al. AMB Expr (2016) 6:76

Page 4 of 16

its expected ability to accumulate nitrite (but not nitrate) and to ensure effluent NH4+-N/NO2-N ratio ~1. During the CFM period (day 1120), with low DO level (0.71.5 mg/L) and low load ammonia content (NH4+-

N, 100200mg/L) without pH control, performance was not satisfactory; i.e., the effluent contained no nitrite, low nitrate, and high ammonia (Fig.1). This nding indicates that low DO level was not suitable for ammonia oxidation. When the running manner was changed to SBM (day 121 to end) with high DO level (7.08.0mg/L) and high load ammonia content (600900 mg NH4+-N/L,

Additional le1: Table S2) in inuent with pH ~10, performance was greatly improved. Nitrite was accumulated starting on day 130, and NH4+-N/NO2-N ratio reached ~1 on day 203, with very low nitrate content in effluent (Fig.1), reecting 50% short-term nitrication (i.e., conversion of 50% of ammonia to nitrite). When effluent from UASB-treated piggery wastewater was used as inuent on day 220, NH4+-N/NO2-N ratio was maintained at ~1 without pH adjustment, indicating the stability and efficiency of the reactor.

Microbial community dynamics inSHARON

A total of 720,982 high-quality sequences were obtained from 851,519 total sequences of raw data following sequence processing. The size of high-quality sequences for each sample ranged from 23,554 (CFM-38) to 40,512 (CFM-98). Sequences in all samples were standardized to 23,554 for further analysis. Greengenes Database (v. 13-8) core 16S reference sequences were used for evaluation of taxonomic structure of bacterial communities, resulting in classication of 44 bacterial phyla, 116 classes, 213

orders, 337 families, and 1359 OTUs in total. Detailed phylogenetic analyses at the genus level and annotated genera are shown in Additional le1: Figure S2.

Composition of phyla in the reactor varied depending on changes in environmental and engineering factors (Fig. 2a). The two major phyla, Proteobacteria and Bacteroidetes, showed opposite trends of relative abundance during the running period (R = 0.79, p < 0.001). Relative abundance of Proteobacteria was in the 6075 % range from day 1 to 145 (except for a value of 49.8% on day 98), declined to 34.2% on day 161, and showed little subsequent uctuation. In striking contrast, relative abundance of Bacteroidetes was in the 520% range from day 1 to 176, and increased to the 3762% range from day 177 to end of the experiment. Relative abundance of Planctomycetes remained ~3.3% during the CFM period and declined to undetectable level when running manner was changed to SBM. Nitrospira, the major NOB, increased gradually from 2.9% on day 1 to 18.5% on day 116, and declined to undetectable level on day 130, 10days after running manner was changed to SBM. Phylum TM7 had high relative abundance (43.3 %) on day 161, decreased to 2.4 % on day 206 (when load ammonia was adjusted from 1000 to 600 mg/L), and was never higher than 2.0% thereafter. The phyla Spirochaetae, Synergistetes, Thermotogae, WWE1, and WS6 were detected in SHARON reactor after effluent from UASB-treated piggery wastewater was used as inuent on day 220. These phyla were present in the UASB effluent, and presumably established new, stable communities in the SHARON reactor subsequently.

Du et al. AMB Expr (2016) 6:76

Page 5 of 16

Relative abundances of the major bacterial families are shown in Fig.2b. Saprospiraceae and Comamonadaceae were the predominant families in SHARON reactor. Relative abundance of Saprospiraceae was <10% during days 1 to 176, increased to 53.2 % by day 206, decreased to 13.5% by day 220, increased again to 36.6% on day 246 and 261, and decreased gradually to 12.5% by day 290. Relative abundance of Comamonadaceae decreased from 20.8% on day 21 to 9.4% on day 98, gradually increased to 24.6% by day 206, fell abruptly to 6.6% on day 220, and then increased slightly to 11.7% by day 290. Relative abundance of Rhodocyclaceae decreased gradually from

24.6 to 1.4% during the CFM period, and remained very low (~0.1%) during the SBM period. Relative abundance of Xanthomonadaceae remained in the ~1.5% range for a long time, with minor peaks of 5.1% on day 78 and 7.2% on day 145, increased gradually from 0.3% on day 206 to 13.8% on day 274, and then declined to 7.6% on day 290. Relative abundance of Pseudomonadaceae remained in the 01.0% range throughout the running period, except for peaks of 12.4% on day 1, 5.1% on day 116, and 6.4% on day 145. Relative abundance of Sphingomonadaceae was 4.0% on day 1 and 5.0% on day 21, then declined gradually to ~1 % on day 130 and to ~0.2 % thereafter.

Du et al. AMB Expr (2016) 6:76

Page 6 of 16

Chitinophagaceae had relative abundance ~2.7 % during the entire running period. Relative abundance of Cryomorphaceae was low (~0.8 %) from day 1 to 206, increased to 15.8% by day 220, then declined gradually to 0.5% by day 290. Relative abundance of Aeromonadaceae was ~5% during day 1 to 116, increased to 32.7% by day 130, and declined to nearly undetectable level thereafter. Relative abundance of Campylobacteraceae was low (~0.2 %) throughout the running period, except for a peak of 16.5% on day 116.

Nitrospiraceae and Nitrosomonadaceae, the two predominant families of autotrophic bacteria found in the SHARON reactor, play active roles in nitrogen cycling in natural environments. Relative abundance of Nitrospiraceae, the major NOB in SHARON, increased gradually from 2.8% on day 1 to 18.1% on day 116, decreased rapidly to 1.5% by day 130, and was nearly undetectable thereafter (Fig.2b). Relative abundance of Nitrosomonadaceae, the major AOB in SHARON, was in the 0.72.2% range from day 1 to 116, and increased slightly to the 1.04.4% range after running manner was changed to SBM, with a peak of 6.5% on day 220.

Compositions of AOB communities are of great concern in regard to SHARON functioning. In total, 15 OTUs related to AOB were dened from bacterial 16S rRNA Illumina MiSeq sequences. A phylogenetic tree was constructed based on these 15 OTUs, all of which were assigned to the genus Nitrosomonas (Fig.3a). Seven groups were generated from the phylogenetic tree: group 1 (OTU-165, OTU-170, OTU-1045) related to N. oligotropha, group 2 (OTU-273, OTU-1048) related to N. ureae, group 3 (OTU-320, OTU-521) related to Nitrosomonas spp., group 4 (OTU-278) related to Nitrosomonas spp., group 5 (OTU-202) related to N. aestuarii, group 6 (OTU-99) related to Nitrosomonas spp., and group 7 (OTU-16, OTU-51, OTU-77, OTU-123, OTU-200) related to N. europaea.

OTU-165 showed 98.4 % similarity to N. oligotropha. Relative abundance of OTU-165 increased gradually from 0.004% on day 1 to 0.65% on day 78, then declined to undetectable level on day 116 (Fig.3b). Relative abundance of OTU-170 (97 % similarity to N. oligotropha) increased from 0.1% on day 58 to 1.0% on day 98, and subsequently declined to undetectable level. OTU-1045 (94.9 % similarity to N. oligotropha) was detected only during the CFM period, with relative abundance~0.01%. OTU-273 and OTU-1048 (96.1 and 93.9 % similarity to N. ureae, respectively) were detected only during the CFM period, with very low relative abundance ~0.05%. OTU-320, OTU-521, and OTU-278 (96.1, 95.4 and 95.2 % similarity to N. aestuarii, respectively) were detected only during the CFM period, with low relative abundances 0.05, 0.04, and 0.05%). Relative abundance of

OTU-202 (97.5% similarity to N. aestuarii) was 1.9% on day 1 and declined to undetectable level thereafter. Relative abundance of OTU-99 (93.2% similarity to N. communis; possibly a new species within the genus) increased from 0.4% on day 21 to 1.9% on day 38, then decreased to 0.04% on day 78 and to undetectable level thereafter. Relative abundance of OTU-200 (99.1% similarity to N. europaea) increased gradually from undetectable level (prior to day 98) to 0.51% on day 196, and then declined to undetectable level by day 235. Relative abundance of OTU-16 (95.1 % similarity to N. europaea) increased from 0.04 % on day 98 to 3.1 % on day 130, decreased to 1.5% on day 161, increased to 3.3% on day 196, and then decreased to 0.8% on day 206 and to undetectable level thereafter. Relative abundance of OTU-77 (96.6 % similarity to N. europaea) increased from 0.05% on day 98 to 0.7% on day 196, then decreased to undetectable level thereafter. Relative abundance of OTU-51 (95.7 % similarity to N. europaea) increased sharply from undetectable level (prior to day 206) to 6.4% on day 220, then decreased to 1.0% on day 235 with little subsequent uctuation. Relative abundance of OTU-123 (96.1% similarity to N. europaea) increased gradually from 0.02 % on day 220 to 2.0% on day 290.

Local blast analysis ofnitrifying bacteria (AOB andNOB)

To elucidate relative abundances of autotrophic AOB and NOB taxa, we set up a local database for each sample based on Illumina sequencing data, and obtained a total of 19 local databases for local blast analysis. Trimmed 16S rRNA fragments for all known members of AOB and NOB were used for blast analysis of the local databases using 95 and 97% similarity cutos. Few sequences were dened at 97 % similarity cuto, but several sequences were found to be affiliated with AOB or NOB at 95% similarity cuto. For AOB, in addition to the genus Nitrosomonas as described above, the genera Nitrosospira and Nitrosovibrio were dened with extremely low relative abundance (0.042 and 0.023 %, respectively), detected only during the CFM period. The combined relative abundance of Nitrosomonas + Nitrosospira + Nitrosovibrio was nearly the same as that of Nitrosomonas by itself (from ~0.67% during the CFM period to 8.0% by day 220, and thereafter decreased to a near-constant ~1.6%, Fig.4), indicating that Nitrosomonas was the predominant AOB in SHARON. For NOB, two genera were detected: Nitrobacter and Nitrospira. Relative abundance of Nitrospira increased gradually from 1.6 % on day 1 to 16.8% on day 98, and decreased rapidly to undetectable level after running manner was changed to SBM on day 121. Relative abundance of Nitrobacter increased from 0.06 % (day 1) to 1.54 % (day 58) and decreased to undetectable level thereafter. Relative abundance of

Du et al. AMB Expr (2016) 6:76

Page 7 of 16

Nitrobacter+Nitrospira combined was only marginally (~0.1%) higher than that of Nitrospira by itself, indicating that Nitrospira was the predominant NOB during the CFM period.

Gene abundance

The functional gene amoA of AOB encodes the -subunit of ammonia monooxygenase, which catalyzes the reaction NH3 + 2[H] + O2 NH2OH + H2O (Hollocher

Du et al. AMB Expr (2016) 6:76

Page 8 of 16

etal. 1981; Li etal. 2015). The 491-bp stretch amoA gene has strong capacity for ne-scale dierentiation of closely related ammonia oxidizers and has been utilized as a functional gene marker for identication of ammonia-oxidizing microorganisms (Rotthauwe etal. 1997; Wang etal. 2012). The abundances of bacterial 16S rRNA gene and AOB amoA gene reect to some degree the relative abundances of bacteria and ammonia-oxidizing organisms (Bai etal. 2012; Gao etal. 2013; Wang etal. 2012).

We therefore quantied the abundances of bacterial 16S rRNA gene and amoA gene by qPCR (Fig.5). Copy numbers of the two genes were signicantly correlated (R2=0.58, p<0.001). Throughout the entire experimental period, bacterial 16S rRNA gene copy number ranged from 2.81011 to 2.11012 per g activated sludge, with little uctuation (Fig.5). AOB amoA gene copy number (per g activated sludge) was 8.5107 on day 1, increased gradually during the CFM period to 1.4109 on day 116, increased further to 5.8109 when running manner was changed to SBM on day 121 and then further increased to 8.7 1010 by day 176, then gradually decreased to 1.1 109 by day 206 (Fig. 5). After the inuent was replaced by UASB effluent (day 220), AOB amoA gene copy number increased to 1.71010, then decreased to ~4.5109 subsequently, with no eect on partial nitrication performance (Fig.1).

Bacterial amoA/bacterial 16S rRNA gene ratio was <0.2 % during the CFM period, increased to 2.1 % by day 145 (24days after running manner was changed to SBM), increased further to 4.1% by day 196, decreased to 0.8% by day 220, increased again to 2.1% by day 235, and nally decreased again and remained fairly stable at ~0.78% (Fig.5).

Bacterial amoA clone libraries

Ten AOB amoA gene libraries were constructed from the qPCR samples mentioned above. A total of 317 amoA gene sequences were obtained, and assigned to 21 aOTUs. Phylogenetic analysis based on 21 representative sequences indicated that these 21 aOTUs could be classied into six groups (Fig. 6a): a N. oligotropha-related group (group 1: aOTU-11, aOTU-16, aOTU-17, aOTU-18, aOTU-19, aOTU-20, aOTU-21), an unknown Nitrosomonas member-related group (group 2: aOTU-10), a second unknown Nitrosomonas member-related group (group 3: aOTU-5, aOTU-12, aOTU-13, aOTU-14), a third unknown Nitrosomonas member-related group (group 4: aOTU-15), a fourth unknown Nitrosomonas member-related group (group 5: aOTU-4, aOTU-6), and a N. europaea-related group (group 6: aOTU-1, aOTU-2, aOTU-3, aOTU-7, aOTU-8, aOTU-9). This phylogenetic tree and that of the AOB OTUs (Fig. 3a) showed very similar topological pattern.aOTU-4, aOTU-1, and aOTU-2 were respectively predominant during the CFM period, SBM period, and SBM period when UASB effluent was used as inuent (Fig. 6b). aOTU-4 was phylogenetically consistent with OTU-99 (Fig.3a) and had 77.2% similarity to N. nitrosa. Its relative abundance on day 38 was 94.4% of all amoA sequences, and it was the predominant aOTU during days 178 (Fig. 6b). aOTU-6, which clustered with aOTU-4, was at or near undetectable level during the entire running period (Fig. 6b). aOTU-1 was phylogenetically consistent with OTU-16 (Fig. 3a) and had 86.9% similarity to N. europaea. Its relative abundance increased gradually from 3.2 % at day 78 to 97.6% at 176, decreased to 52% by day 206, and further

Du et al. AMB Expr (2016) 6:76

Page 9 of 16

decreased to undetectable level thereafter (Fig. 6b). aOTU-2 was phylogenetically consistent with OTU-51 (Fig. 3a) and had 86.7 % similarity to N. europaea. Its relative abundance was at or near undetectable level during days 1206, increased to nearly 100 % during days 220235, then decreased to ~80% by day 261 (Fig. 6b). Four aOTUs (aOTU-5, aOTU-12, aOTU-13, aOTU-14) were grouped together and detected only on day 1. They were N. ureae-related (83.686.6 % similarity), and had relative abundance 48.39, 0.06, 0.03, and 0.03 %, respectively (Fig. 6b). They may be phylo-genetically consistent with OTU-273 and OTU-1048 (Fig. 3a). Seven aOTUs (aOTU-11, aOTU-16, aOTU-17, aOTU-18, aOTU-19, aOTU-20, aOTU-21) were detected only on day 78, with low relative abundance (0.030.16%). They were clustered with N. oligotropha (91.894.2 % similarity) (Fig. 6b), and may be phylo-genetically consistent with OTU-165, OTU-170, and OTU-1045 (Fig. 3a). aOTU-10 (85.1 % similarity to N. oligotropha) was detected only on day 38, with relative abundance 0.03 % (Fig. 6b). aOTU-15 (79.6 % similarity to N. aestuarii) was detected only on day 116, with relative abundance 0.06 % (Fig. 6b). The four remaining aOTUs (aOTU-3, aOTU-7, aOTU-8, aOTU-9) were N. europaea-related (78.694.7 % similarity). aOTU-8 was detected only on day 261, with relative abundance 0.21%. aOTU-9 was detected only on day 206, with relative abundance 0.04%. aOTU-7 was detected on days 78, 116, and 206, with relative abundance 3.32, 33.33, and 12 %, respectively. Relative abundance of aOTU-3 increased from 8.33% on day 116 (the day it was rst

detected) to 11.11% on day 145, decreased to 2.44% by day 176, increased to 32% by day 206, and declined to undetectable level thereafter (Fig.6b).

Bacterial diversity

Goods coverage estimates of 22 samples were all >99% (Table1), indicating that nearly all bacterial species in the reactor were well represented. -diversity indices were used to assess diversity dynamics of microbial communities in the bioreactor during the entire experimental period (Table1). The number of OTUs (97% similarity cuto) ranged from 185 to 634. Shannon index decreased gradually from 6.45 on day 1 to 2.95 on day 206, and increased further after UASB effluent was used as inuent on day 220, up to 6.06 on day 290 (Table1). All running parameters showed strongly negative correlations with -diversity calculated based on 16S rRNA genes by Pearsons test (Additional le1: Table S3). For -diversity indices calculated based on AOB amoA genes, Goods coverage estimates were all >93.6% (except for 87.10% on day 78), indicating that nearly all AOB in SHARON reactor were included (Table2). According to Pearsons test, all running parameters were negatively correlated with -diversity indices (Additional le1: Table S4). Running parameters had a positive, but not signicant, eect on relative abundance of AOB (Table3). Relative abundance of NOB had signicant negative correlations with numerous running parameters: load ammonia (R = 0.65, p<0.01), effluent ammonia (R=0.51, p<0.05), effluent nitrite (R=0.60, p<0.01), DO (R=0.80, p<0.001), and pH (R=0.57, p<0.05) (Table3).

Du et al. AMB Expr (2016) 6:76

Page 10 of 16

Du et al. AMB Expr (2016) 6:76

Page 11 of 16

Table 1 OTU richness anddiversity indices ofmicrobial communities inthe SHARON reactor

Samplea OTU numberb Goods coverage (%) ACE Chao I Shannon Simpson

CFM1 617 99.50 719.3 752.8 6.45 0.96 CFM21 536 99.48 637.5 646.3 5.60 0.93 CFM38 634 99.34 781.1 779.7 6.33 0.96 CFM58 599 99.62 651.3 661.2 6.48 0.97 CFM78 563 99.56 640 651.1 6.49 0.97 CFM98 569 99.55 646.8 648.5 6.23 0.95 CFM116 456 99.50 567.6 581.5 5.39 0.93 SBM130 461 99.46 587 578.7 5.03 0.90 SBM145 374 99.61 460.1 465 5.72 0.96 SBM161 299 99.57 424.4 435.5 3.86 0.79 SBM176 213 99.69 314.6 292.6 3.71 0.83 SBM196 185 99.77 247.8 247.2 4.17 0.88 SBM206 155 99.80 210 215.1 2.95 0.69 SBM220 452 99.62 516.7 554.7 5.58 0.94 SBM235 455 99.61 522.3 552.5 5.69 0.93 SBM246 437 99.62 502.8 494.6 4.96 0.87 SBM261 410 99.62 481 501 4.91 0.86 SBM274 465 99.62 527.3 550.2 5.58 0.93 SBM290 456 99.72 491.4 503.3 6.06 0.96 UASB288 455 99.82 491.7 508.3 5.38 0.93 UASB304 457 99.77 509.2 511.2 5.30 0.92 UASB334 488 99.79 529.4 542.6 5.91 0.93

a Numbers following CFM, SBM, and UASB in this column indicate sampling date

b 97% similarity cuto

Table 2 aOTU richness anddiversity indices ofAOB communities based onamoA genes inthis SHARON

Samplea Clone number aOTU numberb Goods coverage (%) Chao I Shannon

CFM1 38 5 93.6 5.5 1.61 CFM38 36 2 94.44 4.0 0.36 CFM78 32 10 87.10 12 2.86 CFM116 37 6 94.44 6.5 1.86 SBM145 27 2 100 2 0.50 SBM176 41 2 97.56 2 0.17 SBM206 26 4 96 4 1.57 SBM220 29 1 100 1 0 SBM235 22 1 100 1 0 SBM261 29 2 100 2 0.74

a Numbers following CFM and SBM in this column indicate sampling date

b 97% similarity cuto

Application of PCoA resulted in clustering of 19 SHARON samples into three groups (days 1116, days 130196, days 220290) related to the time points when running manner was changed from CFM to SBM, and to SBM with UASB effluent used as inuent (Fig.7a). Samples from days 220 to 290 clustered with UASB effluent samples when UASB effluent was used as inuent

for SHARON on day 220 (Fig. 7a). ANOSIM (analysis of similarity) revealed signicant dierences between the three groups (R2 = 0.94, p < 0.01). RDA (redundancy analysis) revealed the dependence of community dynamics on environmental factors (high load ammonia, effluent ammonia, effluent nitrite, pH, DO) during the entire running period (Additional le 1: Figure S3).

Du et al. AMB Expr (2016) 6:76

Page 12 of 16

Table 3 Correlations (R values) between AOB/NOB andrunning parameters fromPearsons test

*p<0.05, **p<0.01, *** p < 0.001

DO pH

AOB 0.30 0.17 0.43 0.36 0.33 NOB 0.65** 0.51* 0.60** 0.80*** 0.57*

Bacteria NH4+N inuent

NH4+N effluent

NO2N effluent

Pearsons test showed that running parameters had signicant eects on relative abundance of the major phyla (Table 4). PCoA based on 317 AOB amoA sequences showed nearly the same dynamic pattern as PCoA based on 16S rRNA sequences (Fig.7b). Samples obtained during the SBM period were separated from those obtained during the CFM period, except for the day 116 sample (Fig. 7b). Samples obtained from day 220 to 261 (when UASB effluent was used as inuent) were separated from those obtained with articial wastewater used as inuent (Fig.7b).

Discussion

Most studies to date on bacterial communities and AOB in nitrifying bioreactors have focused on activated sludge samples obtained from a single time point (Gao et al. 2013; Limpiyakorn et al. 2006; Wells et al. 2009; Zhang etal. 2011), which does not reect bacterial community succession during the startup period or entire running period, and does not allow determination of the environmental factors that drive community dynamics. A novel treatment system termed UASB+SHARON+ANAM

MOX was constructed in our laboratory for treatment

of piggery wastewater with high contents of COD and ammonia. The SHARON reactor was expected to achieve 50% short-term nitrication and ensure effluent NH4+-

N/NO2-N ratio ~1. In the present study, we investigated dynamics of bacterial community diversity, structure, and composition during the entire running period, including startup, domestication, and replacement of articial wastewater with real wastewater. We analyzed the predominant functional bacterial taxa during dierent running periods, and the eects of various environmental and engineering factors on reactor performance.

Bacterial community dynamics

In this SHARON reactor, bacterial diversity showed a general decreasing trend during days 1206, and increased slightly following day 220, when UASB effluent was used as inuent (Table1). Pearsons test showed that our running parameters were negatively correlated with -diversity of the microbial community (Additional le1: Table S3). The decrease in bacterial diversity may have resulted from domestication of articial wastewater with high ammonia content, similarly to the observations of Whittenbury etal. (1970). The slight increase in diversity after UASB effluent was used as inuent may have resulted from the introduction of groups such as Spiro-chaetae, Synergistetes, Thermotogae, WWE1, and WS6 present in UASB effluent (Fig.2a). Entire microbial community dynamics and AOB community dynamics displayed very similar succession trends, according to PCoA (Fig. 7). Entire microbial community dynamics underwent two major changes associated with change of running manner and use of UASB effluent as inuent, and

Du et al. AMB Expr (2016) 6:76

Page 13 of 16

Table 4 Correlations (R values) betweenmajor phyla andrunning parameters fromPearsons test

*p<0.05, **p<0.01, ***p<0.001

phylum inuent NH4+N effluent NH4+N effluent NO2N DO pH

Bacteroidetes 0.55* 0.23 0.86*** 0.65** 0.64** Proteobacteria 0.60** 0.32 0.86*** 0.70*** 0.78***

Nitrospirae 0.69** 0.53** 0.62** 0.83*** 0.60**

TM7 0.4 0.53* 0.06 0.29 0.60** Chloroexi 0.50* 0.54* 0.19 0.33 0.48*

Chlorobi 0.09 0.14 0.27 0.17 0.38

Firmicutes 0.45 0.46* 0.15 0.56* 0.09

Actinobacteria 0.23 0.05 0.70*** 0.49* 0.32

Synergistetes 0.17 0.04 0.46* 0.39 0.18

Gemmatimonadetes 0.35 0.56** 0.18 0.05 0.22

Planctomycetes 0.79*** 0.63** 0.70*** 0.86*** 0.68***

bacterial communities were clustered into three groups (Fig. 7a) having signicantly dierent compositions (R2=0.94; p<0.01). This observation was not surprising, because change of running manner has been shown to greatly alter community composition for adaptation to environmental changes (Turner et al. 1998; Wells et al. 2009). In PCoA based on AOB amoA genes, the day 116 sample was separated from the CFM samples and clustered with SBM samples (Fig.7b). This observation suggests that evolutionary trends of the AOB community diered from those of the entire microbial community, consistently with the ndings of Zhang etal. (2011).

Bacteroidetes and Proteobacteria were the primary bacterial phyla found in this SHARON, as they generally are in activated sludge (Juretschko etal. 1998, 2002; Kong etal. 2002; Xia etal. 2008). The relative abundances of these two phyla were inversely correlated (Table 4), presumably because of changes in running parameters; high DO levels and pH values favor growth of Bacteroidetes (Gao etal. 2011; Hu etal. 2015). The low relative abundance of Planctomycetes during the CFM period and its further decline to undetectable level after change of running manner to SBM (Fig.2a) may have been due to the increase of DO level from 0.71.5 to 7.08.0mg/L) (R=0.86, p<0.001), because these bacteria are anaerobic ammonia oxidizers (Innerebner et al. 2007; Strous etal. 1999). The gradual increase of relative abundance of Nitrospirae, a NOB (Daims etal. 2015), during the CFM period, with little nitrate accumulation in effluent (Figs.1, 2a), and its decrease following change to SBM, may have been related to inhibitory eects of high pH (~10; R=0.57, p<0.01) and high ammonium concentration (600900 mg/L; R = 0.65, p < 0.01) in inuent. High pH and load ammonia were previously reported to be the major factors that inhibit NOB growth (Balmelle et al. 1992; Bernet etal. 2001; Garrido etal. 1997). The phylum

TM7 showed high relative abundance only during days 161196 (Fig.2a), perhaps because of high values of both load and effluent ammonia, which are positively correlated with TM7 abundance (Table 4). The mechanisms underlying such short-term high abundance, and its eects in this SHARON, await further investigation.

Dynamics ofammonia oxidizers, nitrite oxidizers, andnitrite accumulation

Ammonia can be oxidized to nitrite by both AOB and ammonia-oxidizing archaea (AOA). AOB populations (Fig.3a) and AOB amoA genes (Figs.5, 6) were detected in the present study, but AOA populations and AOA amoA genes were not. AOB is much more abundant than AOA in activated sludge (Mussmann et al. 2011; Wang etal. 2012; Zhang etal. 2009). AOA are typically found in habitats with low ammonia concentration, low DO level, and/or acidic conditions (low pH) (Gao etal. 2013; Zhang et al. 2012). They were barely detectable in this SHARON, which had high DO level and ammonia content. Pearsons test showed that the high DO and ammonia also greatly aected -diversity of AOB (Additional le1: Table S4). These ndings are consistent with those of Lydmark etal. (2007) and Wang etal. (2012).

The real relative abundance of AOB in this SHARON may be much higher than our estimate based on 16S rRNA gene sequencing through Illumina MiSeq method (Figs.3 and 4). The average number of 16S rRNA operons in heterotrophic bacteria has been estimated as 5.5 in nutritional environments (Klappenbach et al. 2000, 2001; Nadkarni etal. 2002), and there is only one copy of 16S rRNA in AOB (Aakra etal. 1999). Surprisingly, the AOB amoA gene/bacterial 16S rRNA gene ratio calculated based on qPCR was less than the relative abundance of AOB estimated based on 16S rRNA genes. There are 23 amoA copies in every AOB cell for -subdivision of

Du et al. AMB Expr (2016) 6:76

Page 14 of 16

Proteobacteria (Norton et al. 2002; Okano et al. 2004). The dierence may be due to biases arising from the quantitative process, since the AOB amoA primers used were designed based on N. europaea (He etal. 2007; Rotthauwe etal. 1997).

The topologies of phylogenetic trees based on OTUs and aOTUs were similar, and indicate that Nitrosomonas was the predominant genus in this SHARON. Groups generated from the two trees were generally congruent but not identical (Figs.3a, 6b), because of the inconsistency in analysis of AOB through 16S rRNA and amoA genes (Purkhold et al. 2003). Many AOB OTUs and aOTUs showed <97% sequence similarity to known AOB species, suggesting that many novel AOB taxa were present in this SHARON.

Average relative abundance of AOB was ~1 % during CFM with no nitrite accumulation during days 1120, with low DO (0.71.5mg/L), low inuent ammonia concentration (NH4+-N, 100200mg/L), and no pH control (Fig. 1), demonstrating that these running parameters were not favorable for AOB. The predominant AOB during this CFM period were unidentied ones represented by OTU-99/aOTU-4 and the N. oligotropha-related group (OTU-165/170/1045; aOTU-11/16-21), which were able to adapt to the acidic environment (Gieseke etal. 2006). The presence of Nitrospira may have led to the failure of nitrite accumulation despite the presence of N. oligotropha-related AOB. When running manner was changed to SBM, high inuent ammonia concentration (NH4+-N, 600900 mg/L) with high pH (~10)

and high DO level (7.08.0mg/L) increased the relative abundance of AOB, and the N. oligotropha-related group was replaced by the N. europaea-related group, with nitrite accumulation in the effluent (Figs.1, 3, 5). These changes may have been due to altered parameters resulting from the change of running manner. AOB activity is inhibited under acidic conditions because the bioavailability of ammonia is reduced by ionization (Downing and Nere 1964; Gerardi 2003; Suzuki et al. 1974), and high pH increases the concentration of molecular ammonia. The Ks value of NH3 oxidation is much lower for N. oligotropha (2.44.2M) than for N. europaea (3056M) (Koops et al. 1991; Koops and Pommerening-Rser 2005). Nitrosomonas europaea has higher affinity for oxygen than does N. oligotropha (Park and Noguera 2007).

Ammonia-oxidizing activity was enhanced when UASB effluent was used as inuent, despite the reduced relative abundance of AOB (Fig. 4). One possible explanation is that alkaline organic materials in the UASB effluent maintained an alkaline pH during the entire SBR batch. AOB activity may also have been promoted by organic compounds (e.g., formate, acetate, pyruvate, glucose, peptone) present in the UASB effluent (Krummel and Harms 1982).

The families Rhodocyclaceae, Comamonadaceae, Xanthomonadaceae, Pseudomonadaceae, and Sphingomonadaceae were detected at moderate levels in this SHARON (Fig. 2b). Certain members of these families have been identied as heterotrophic AOB (Bal Krishna etal. 2013; Huang etal. 2015; Khardenavis etal. 2007; Kim etal. 2008), and may have contributed to nitrite accumulation in this SHARON. This possibility requires further investigation.

In conclusion, the novel UASB + SHARON + ANAMMOX system described here utilized

SBM for partial nitrication, following CFM, to transform 50% of ammonia to nitrite. The pattern of microbial community dynamics was nearly the same as that of AOB community dynamics. Nitrosomonas europaea-related bacteria were the autotrophic AOB primarily responsible for partial nitrication during SBM. Ammonia-oxidizing activity was enhanced by the high DO level, high ammonia (NH4+-N) concentration, high pH, and presence of organic materials in UASB effluent used as inuent.

Additional le

Abbreviations

UASB: upow anaerobic sludge blanket; SHARON: single reactor system for high activity ammonium removal over nitrite; anammox/ANAMMOX: ANaero bic AMMonium OXidation; AOB: ammonium oxidation bacteria; AOA: ammo nium oxidation archaea; NOB: nitrite oxidation bacteria; CFM: continuous ow manner; SBM: sequencing batch manner; HRT: hydraulic retention time; DO: dissolved oxygen; COD: chemical oxygen demand; QIIME: quantitative insights into microbial ecology; qPCR: quantitative polymerase chain reaction; OTU: operational taxonomic units; PCoA: principal coordinates analysis; RDA: redun dancy analysis; ANOSIM: analysis of similarity; NJ: neighborjoining.

Authors contributions

WLD performed the experiments assisted by QH. WLD contributed to analyze the data as well as results and wrote the manuscript. ZPL, LLM and YL conceived this study and constructed the bioreactors. ZPL supervised all the experiments and revised the manuscript. All authors read and approved the nal manuscript.

Author details

1 State Key Laboratory of Microbial Resources, Institute of Microbiology, Chinese Academy of Sciences, No. 1 West Beichen Road, Chaoyang District, Beijing 100101, Peoples Republic of China. 2 University of Chinese Academy of Sciences, Beijing 100049, Peoples Republic of China.

Acknowledgements

The authors are grateful to Dr. S. Anderson for English editing of the manuscript.

Competing interests

The authors declare that they have no competing interests.

Funding

This work was supported by the National Science and Technology Support Program of China (Grant No. 2013BAD21B0204) and National Natural Science Foundation of China (Grant No. 31500103).

Received: 24 August 2016 Accepted: 7 September 2016

Du et al. AMB Expr (2016) 6:76

References

Aakra , Utaker JB, Nes IF. RELP of rRNA genes and sequencing of the 16S23S rDNA intergenic spacer region of ammoniaoxidizing bacteria: a phyloge netic approach. Int J Syst Bacteriol. 1999;49:12330.

Bai Y, Sun Q, Wen D, Tang X. Abundance of ammoniaoxidizing bacteria and archaea in industrial and domestic wastewa ter treatment system. FEMS Microbiol Ecol. 2012;80:32330. doi:http://dx.doi.org/10.1111/j.1574-6941.2012.01296.x

Web End =10.1111/j.15746941.2012.01296.x .

Bal Krishna KC, Sathasivan A, Ginige MP. Microbial community changes with decaying chloramine residuals in a labscale system. Wat Res. 2013;13:466679. doi:http://dx.doi.org/10.1016/j.watres.2013.04.035

Web End =10.1016/j.watres.2013.04.035 .

Balmelle B, Nguyen K, Capdeville B, Cornier J, Deguin A. Study of factors con trolling nitrite buildup in biological processes for water nitrication. Wat Sci Technol. 1992;26:101725.

Bernet N, Dangcong P, Delgens JP, Moletta R. Nitrication at low oxygen concentration in biomlm reactor. J Environ Eng. 2001;127:26671.

Bernet N, Delgenes N, Akunna JC, Delgenes JP, Moletta R. Combined anaerobicaerobic SBR for the treatment of piggery wastewater. Wat Res. 2000;34:6119. doi:http://dx.doi.org/10.1016/S0043-1354(99)00170-0

Web End =10.1016/S00431354(99)001700 .

Bernet N, Delgenes N, Moletta R. Denitrication by anaerobic sludge in piggery wastewater. Environ Technol. 1996;17:293300. doi:http://dx.doi.org/10.1080/09593331708616387

Web End =10.1080/09593331708616387 .

Boiran B, Couton Y, Germon JC. Nitrication and denitrication of liq uid lagoon piggery waste in a biolm inltrationpercolation aer ated system (BIPAS) reactor. Bioresour Technol. 1996;55:6377. doi:http://dx.doi.org/10.1016/0960-8524(95)00142-5

Web End =10.1016/09608524(95)001425 .

Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK,

Fierer N, Pena AG, Goodrich JK, Gordon JI, Huttley GA, Kelley ST, KnightsD, Koenig JE, Ley RE, Lozupone CA, McDonald D, Muegge BD, Pirrung M, Reeder J, Sevinsky JR, Tumbaugh PJ, Walters WA, Widmann J, Yatsunenko T, Zaneveld J, Knight R. QIIME allows analysis of highthroughput community sequencing data. Nat Methods. 2010;7:3356. doi:http://dx.doi.org/10.1038/nmeth.f.303

Web End =10.1038/nmeth.f.303 . Chen XP, Zhu YG, Xia Y, Shen JP, He JZ. Ammoniaoxidizing archaea: impor tant players in paddy rhizosphere soil? Environ Microbiol. 2008;10:197887. doi:http://dx.doi.org/10.1111/j.1462-2920.2008.01613.x

Web End =10.1111/j.14622920.2008.01613.x .

Daims H, Lebedeva EV, Pjevac P, Han P, Herbold C, Albertsen M, Jehmlich N, Palatinszky M, Vierheilig J, Bulaev A, Kirkegaard RH, von Bergen M, Rattei T, Bendinger B, Nielsen PH, Wagner M. Complete nitrication by Nitrospira bacteria. Nature. 2015;528:5049. doi:http://dx.doi.org/10.1038/nature16461

Web End =10.1038/nature16461 .

Dionisi HM, Layton AC, Harms G, Gregory IR, Robinson KG, Sayler GS. Quan tication of Nitrosomonas oligotrophalike ammoniaoxidizing bacteria and Nitrospira spp. from fullscale wastewater treatment plants by competitive PCR. Appl Environ Microbiol. 2002;68:24553. doi:http://dx.doi.org/10.1128/aem.68.1.245-253.2002

Web End =10.1128/ http://dx.doi.org/10.1128/aem.68.1.245-253.2002

Web End =aem.68.1.245253.2002 .

Downing LS, Nere R. Nitrication in the activated sludge process. J Inst Sew age Purication (Great Britain). 1964;63:13058.

Edgar RC. UPARSE: highly accurate OTU sequences from microbial amplicon reads. Nat Methods. 2013;10:9968. doi:http://dx.doi.org/10.1038/nmeth.2604

Web End =10.1038/nmeth.2604 .

Fitzgerald CM, Camejo P, Oshlag JZ, Noguera DR. Ammoniaoxidizing micro bial communities in reactors with efficient nitrication at lowdissolved oxygen. Wat Res. 2015;70:3851. doi:http://dx.doi.org/10.1016/j.watres.2014.11.041

Web End =10.1016/j.watres.2014.11.041 .

Flood M, Frabutt D, Floyd D, Powers A, Ezegwe U, Devol A, TiquiaArashiro SM. Ammoniaoxidizing bacteria and archaea in sediments of the Gulf of Mexico. Environ Technol. 2015;36:12435. doi:http://dx.doi.org/10.1080/09593330.2014.942385

Web End =10.1080/09593330.2014.9 http://dx.doi.org/10.1080/09593330.2014.942385

Web End =42385 .

Gao DW, Fu Y, Tao Y, Li XX, Xing M, Gao XH, Ren NQ. Linking microbial community structure to membrane biofouling associated with varying dissolved oxygen concentrations. Bioresour Technol. 2011;102:562633. doi:http://dx.doi.org/10.1016/j.biortech.2011.02.039

Web End =10.1016/j.biortech.2011.02.039 .

Gao JF, Luo X, Wu GX, Li T, Peng YZ. Quantitative analyses of the composition and abundance of ammoniaoxidizing archaea and ammoniaoxidizing bacteria in eight fullscale biological wastewater treatment plants. Biore sour Technol. 2013;138:28596. doi:http://dx.doi.org/10.1016/j.biortech.2013.03.176

Web End =10.1016/j.biortech.2013.03.176 . Garrido JM, van Benthum WAJ, van Loosdrecht MCM, Heijnen JJ. Inuence of dissolved oxygen concentration on nitrite accumulation in a biolm airlift suspension reactor. Biotechnol Bioeng. 1997;53:16878. doi:http://dx.doi.org/10.1002/(sici)1097-0290(19970120)53:2%3c168:aid-bit6%3e3.0.co;2-m

Web End =10.1002/ http://dx.doi.org/10.1002/(sici)1097-0290(19970120)53:2%3c168:aid-bit6%3e3.0.co;2-m

Web End =(sici)10970290(19970120)53:2<168:aidbit6>3.0.co;2m .

Gauch HG. A quantitative evaluation of the BrayCurtis ordination. Ecology. 1973;54:82936. doi:http://dx.doi.org/10.2307/1935677

Web End =10.2307/1935677 .

Gerardi MH. Nitrication and denitrication in the activated sludge process. Environmental Protection Magazine Series. New York: Wiley; 2003.

Page 15 of 16

Gieseke A, Tarre S, Green M, de Beer D. Nitrication in a biolm at low pH values: role of in situ microenvironments and acid tolerance. Appl Environ Microbiol. 2006;72:428392. doi:http://dx.doi.org/10.1128/AEM.00241-06

Web End =10.1128/AEM.0024106 .

Hashimoto AG. Thermophilic and mesophilic anaerobic fermentation of swine manure. Agric Wastes. 1983;6:17591. doi:http://dx.doi.org/10.1016/0141-4607(83)90085-9

Web End =10.1016/01414607(83)900859 .

He S, Gall DL, McMahon KD. Candidatus Accumulibacter population structure in enhanced biological phosphorus removal sludges as revealed by polyphosphate kinase genes. Appl Environ Microbiol. 2007;73:586574. doi:http://dx.doi.org/10.1128/AEM.01207-07

Web End =10.1128/AEM.0120707 .

Hollocher TC, Tate ME, Nicholas DJ. Oxidation of ammonia by Nitrosomonas europaea. Denite 18Otracer evidence that hydroxylamine formation involves a monooxygenase. J Biol Chem. 1981;256:108346.

Hu D, Zhou Z, Shen X, Wei H, Jiang LM, Lv Y. Eects of alkalinity on membrane bioreactors for reject water treatment: performance improvement, foul ing mitigation and microbial structures. Bioresour Technol. 2015;197:21726. doi:http://dx.doi.org/10.1016/j.biortech.2015.08.082

Web End =10.1016/j.biortech.2015.08.082 .

Huang TL, Zhou SL, Zhang HH, Bai SY, He XX, Yang X. Nitrogen removal characteristics of a newly isolated indigenous aerobic denitrier from oligotrophic drinking water reservoir, Zoogloea sp. N299. Int J Mol Sci. 2015;16:1003860. doi:http://dx.doi.org/10.3390/ijms160510038

Web End =10.3390/ijms160510038 .

Ibekwe AM, Grieve CM, Lyon SR. Characterization of microbial communities and composition in constructed dairy wetland wastewater effluent. Appl Environ Microbiol. 2003;69:50609. doi:http://dx.doi.org/10.1128/aem.69.9.5060-5069.2003

Web End =10.1128/aem.69.9.50605069.2003 . Innerebner G, Insam H, FrankeWhittle IH, Wett B. Identication of anammox bacteria in a fullscale deammonication plant making use of anaerobic ammonia oxidation. Syst Appl Microbiol. 2007;30:40812. doi:http://dx.doi.org/10.1016/j.syapm.2007.02.001

Web End =10.1016/j. http://dx.doi.org/10.1016/j.syapm.2007.02.001

Web End =syapm.2007.02.001 .

Jetten MSM, Wagner M, Fuerst J, Van Loosdrecht MCM, Kuenen G, Strous M.

Microbiology and application of the anaerobic ammonium oxidation (anammox) process. Curr Opin Biotechnol. 2001;12:2838. doi:http://dx.doi.org/10.1016/S0958-1669(00)00211-1

Web End =10.1016/ http://dx.doi.org/10.1016/S0958-1669(00)00211-1

Web End =S09581669(00)002111 .

Juretschko S, Loy A, Lehner A, Wagner M. The microbial community composi tion of a nitrifyingdenitrifying activated sludge from an industrial sew age treatment plant analyzed by the fullcycle rRNA approach. Syst Appl Microbiol. 2002;25:8499. doi:http://dx.doi.org/10.1078/0723-2020-00093

Web End =10.1078/0723202000093 .

Juretschko S, Timmerman G, Schmid M, Schleifer KH, PommereningRoser A,

Koops HP, Wagner M. Combined molecular and conventional analyses of nitrifying bacterium diversity in activated sludge: Nitrosococcus mobilis and Nitrospiralike bacteria as dominant populations. Appl Environ Micro biol. 1998;64:304251.

Khardenavis AA, Kapley A, Purohit HJ. Simultaneous nitrication and deni trication by diverse Diaphorobacter sp. Appl Microbiol Biotechnol. 2007;77:4039. doi:http://dx.doi.org/10.1007/s00253-007-1176-5

Web End =10.1007/s0025300711765 .

Kim M, Jeong SY, Yoon SJ, Cho SJ, Kim YH, Kim MJ, Ryu EY, Lee SJ. Aerobic deni trication of Pseudomonas putida AD21 at dierent C/N ratios. J Biosci Bioeng. 2008;106:498502. doi:http://dx.doi.org/10.1263/jbb.106.498

Web End =10.1263/jbb.106.498 .

Klappenbach JA, Dunbar JM, Schmidt TM. rRNA operon copy number reects ecological strategies of bacteria. Appl Environ Microbiol. 2000;66:132833. doi:http://dx.doi.org/10.1128/AEM.66.4.1328-1333.2000

Web End =10.1128/AEM.66.4.13281333.2000 .

Klappenbach JA, Saxman PR, Cole JR, Schmidt TM. rrndb: the ribosomal RNA operon copies number database. Nucleic Acids Res. 2001;29:1814. doi:http://dx.doi.org/10.1093/nar/29.1.181

Web End =10.1093/nar/29.1.181 .

Kong YH, Ong SL, Ng WJ, Liu WT. Diversity and distribution of a deeply branched novel proteobacterial group found in anaerobicaerobic activated sludge processes. Environ Microbiol. 2002;70:538390. doi:http://dx.doi.org/10.1046/j.1462-2920.2002.00357.x

Web End =10.1046/j.14622920.2002.00357.x .

Koops HP, Bttcher B, Mller UC, PommereningRoeser A, Gtehr G. Classica tion of eight new species of ammoniaoxidizing bacteria: Nitrosomonas communis sp. nov., Nitrosomonas ureae sp. nov., Nitrosomonas aestuarii sp. nov., Nitrosomonas marina sp. nov., Nitrosomonas nitrosa sp. nov., Nitrosomonas eutropha sp. nov., Nitrosomonas oligotropha sp. nov. and Nitrosomonas halophila sp. nov. J Gen Microbiol. 1991;137(16891):699. doi:http://dx.doi.org/10.1099/00221287-137-7-1689

Web End =10.1099/0022128713771689 .

Koops HP, PommereningRser A. Nitrosomonas winogradsky 1892, 127 AL

(Nom. Cons. Opin. 23 Jud. Comm. 1958, 169). Bergeys manual of system atic bacteriology; 2005: 864867.

Krummel A, Harms H. Eect of organic matter on growth and cell yield of ammoniaoxidizing bacteria. Arch Microbiol. 1982;133:504. doi:http://dx.doi.org/10.1007/BF00943769

Web End =10.1007/ http://dx.doi.org/10.1007/BF00943769

Web End =BF00943769 .

Kuenen JG. Anammox bacteria: from discovery to application. Nat Rev Micro biol. 2008;6:3206. doi:http://dx.doi.org/10.1038/nrmicro1857

Web End =10.1038/nrmicro1857 .

Du et al. AMB Expr (2016) 6:76

Page 16 of 16

Li HS, Zhou SQ, Ma WH, Huang GT, Xu B. Fast startup of ANAMMOX reactor: operational strategy and some characteristics as indicators of reactor per formance. Desalination. 2012;286:43641. doi:http://dx.doi.org/10.1016/j.desal.2011.11.038

Web End =10.1016/j.desal.2011.11.038 .

Li J, Nedwell DB, Beddow J, Dumbrell AJ, McKew BA, Thorpe EL, Whitby C. amoA gene abundances and nitrication potential rates suggest that benthic ammoniaoxidizing bacteria and not archaea dominate N cycling in the Colne Estuary, United Kingdom. Appl Environ Microbiol. 2015;81:15965. doi:http://dx.doi.org/10.1128/AEM.02654-14

Web End =10.1128/AEM.0265414 .

Limpiyakorn T, Kurisu F, Yagi O. Quantication of ammoniaoxidizing bacteria populations in fullscale sewage activated sludge systems and assess ment of system variables aecting their performance. Wat Sci Technol. 2006;54:919. doi:http://dx.doi.org/10.2166/wst.2006.376

Web End =10.2166/wst.2006.376 .

LlabrsLuengo P, MataAlvarez J. Kinetic study of the anaerobic diges tion of strawpig manure mixtures. Biomass. 1987;14:12942. doi:http://dx.doi.org/10.1016/0144-4565(87)90015-1

Web End =10.1016/01444565(87)900151 .

Lo K, Liao P, Gao Y. Anaerobic treatment of swine wastewater using hybrid UASB reactors. Bioresour Technol. 1994;47:1537. doi:http://dx.doi.org/10.1016/0960-8524(94)90114-7

Web End =10.1016/09608524(94)901147 .

Lopez H, Puig S, Ganigue R, Ruscalleda M, Balaguer MD, Colprim J. Startup and enrichment of a granular anammox SBR to treat high nitrogen load wastewaters. J Chem Technol Biot. 2008;83:23341. doi:http://dx.doi.org/10.1002/jctb.1796

Web End =10.1002/jctb.1796 .

Lydmark P, Almstrand R, Samuelsson K, Mattsson A, Sorensson F, Lindgren PE, Hermansson M. Eects of environmental conditions on the nitrifying population dynamics in a pilot wastewater treatment plant. Environ Microbiol. 2007;9:222033. doi:http://dx.doi.org/10.1111/j.1462-2920.2007.01336.x

Web End =10.1111/j.14622920.2007.01336.x .

Mulder A, Graaf A, Robertson L, Kuenen J. Anaerobic ammonium oxidation discovered in a denitrifying uidized bed reactor. FEMS Microbiol Ecol. 1995;16:17784. doi:http://dx.doi.org/10.1111/j.1547-6941.1995.tb00281.x

Web End =10.1111/j.15476941.1995.tb00281.x .

Mussmann M, Brito I, Pitcher A, Damste JSS, Hatzenpichler R, Richter A, Nielsen JL, Nielsen PH, Muller A, Daims H, Wagner M, Head IM. Thaumarchaeotes abundant in renery nitrifying sludges express amoA but are not obligate autotrophic ammonia oxidizers. Proc Natl Acad Sci USA. 2011;108:167716. doi:http://dx.doi.org/10.1073/pnas.1106427108

Web End =10.1073/pnas.1106427108 .

Nadkarni MA, Martin FE, Jacques NA, Hunter N. Determination of bacterial load by realtime PCR using a broadrange (universal) probe and primers set. MicrobiolSgm. 2002;148:25766. doi:http://dx.doi.org/10.1099/00221287-148-1-257

Web End =10.1099/002212871481257 . Norton JM, Alzerreca JJ, Suwa Y, Klotz MG. Diversity of ammonia monooxyge nase operon in autotrophic ammonia oxidizing bacteria. Arch Microbiol. 2002;177:13949. doi:http://dx.doi.org/10.1007/s00203-001-0369-z

Web End =10.1007/s002030010369z .

Odegaard H. Treatment of anaerobically pretreated effluents. In: Hall ER, Hob son PN, editors. Processing of 5th international symposium on anaerobic digestion. Bologna: Pergamon Press; 1988. p. 22538.

Okano Y, Hristova KR, Leutenegger CM, Jackson LE, Denison RF, Gebreyesus B, Lebauer D, Scow KM. Application of realtime PCR to studyeects of ammonium on population size of ammoniaoxidizingbacteria in soil. Appl Environ Microbiol. 2004;70:100816. doi:http://dx.doi.org/10.1128/aem.70.2.1008-1016.2004

Web End =10.1128/ http://dx.doi.org/10.1128/aem.70.2.1008-1016.2004

Web End =aem.70.2.10081016.2004 .

Park HD, Noguera DR. Evaluating the eect of dissolved oxygen on ammo niaoxidizing bacterial communities in activated sludge. Wat Res. 2004;38:327586. doi:http://dx.doi.org/10.1016/j.watres.2004.04.047

Web End =10.1016/j.watres.2004.04.047 .

Park HD, Noguera DR. Characterization of two ammoniaoxidizing bacteria iso lated from reactors operated with low dissolved oxygen concentrations. J Appl Microbiol. 2007;102:140117. doi:http://dx.doi.org/10.1111/j.1365-2672.2006.03176.x

Web End =10.1111/j.13652672.2006.03176.x . Purkhold U, Wagner M, Timmermann G, PommereningRoser A, Koops HP.

16S rRNA and amoAbased phylogeny of 12 novel betaproteobacterial ammoniaoxidizing isolates: extension of the dataset and proposal of a new lineage within the nitrosomonads. Int J Syst Evol Microbiol. 2003;53:148594. doi:http://dx.doi.org/10.1099/ijs.0.02638-0

Web End =10.1099/ijs.0.026380 .

Rotthauwe JH, Witzel KP, Liesack W. The ammonia monooxygenase structural gene amoA as a functional marker: molecular nescale analysis of natural ammoniaoxidizing populations. Appl Environ Microbiol. 1997;63:470412.

Strous M, Fuerst JA, Kramer EHM, Logemann S, Muyzer G, Van de Pas

Schoonen KT, Webb R, Kuenen JG, Jetten MS. Missing lithotroph identi ed as new planctomycete. Nature. 1999;400:4469. doi:http://dx.doi.org/10.1038/22749

Web End =10.1038/22749 . Suzuki I, Dular U, Kwok S. Ammonia or ammonium ion as substrate for oxidation by Nitrosomonas europaea cells and extracts. J Bacteriol. 1974;120:5568.

Tamura K, Stecher G, Peterson D, Filipski A, Kumar S. MEGA6: molecular evolutionary genetics analysis version 6.0. Mol Biol Evol. 2013;30:27259. doi:http://dx.doi.org/10.1093/molbev/mst197

Web End =10.1093/molbev/mst197 .

Turner MG, Baker WL, Peterson CJ, Peet RK. Factors inuencing succession les sons from large, infrequent natural disturbance. Ecosystems. 1998;1:51123. doi:http://dx.doi.org/10.1007/s100219900047

Web End =10.1007/s100219900047 .

Van de Graaf AA, Mulder A, de Bruijn P, Jetten MS, Robertson LA, Kuenen JG.

Anaerobic oxidation of ammonium is a biologically mediated process. Appl Environ Microbiol. 1995;61:124651.

Van der Star WR, Abma WR, Blommers D, Mulder JW, Tokutomi T, Strous M,

Picioreanu C, van Loosdrecht MC. Startup of reactors for anoxic ammo nium oxidation: experiences from the rst fullscale anammox reactor in Rotterdam. Wat Res. 2007;41:414963. doi:http://dx.doi.org/10.1016/j.watres.2007.03.044

Web End =10.1016/j.watres.2007.03.044 . Wagner M, Loy A, Nogueira R, Purkhold U, Lee N, Daims H. Microbial commu nity composition and function in wastewater treatment plants. Antonie Van Leeuwenhoek. 2002;81:66580. doi:http://dx.doi.org/10.1023/A:1020586312170

Web End =10.1023/A:1020586312170 .

Wang Q, Garrity GM, Tiedje JM, Cole JR. Naive Bayesian classier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol. 2007;73:52617. doi:http://dx.doi.org/10.1128/AEM.00062-07

Web End =10.1128/AEM.0006207 .

Wang X, Wen X, Xia Y, Hu M, Zhao F, Ding K. Ammonia oxidizing bacteria com munity dynamics in a pilotscale wastewater treatment plant. PLoS ONE. 2012;7:e36272. doi:http://dx.doi.org/10.1371/journal.pone.0036272

Web End =10.1371/journal.pone.0036272 .

Whittenbury R, Phillips KC, Wilkinson JF. Enrichment, isolation and some prop erties of methaneutilizing bacteria. J General Microbiol. 1970;61:20518. Wells GF, Park HD, Yeung CH, Eggleston B, Francis CA, Criddle CS. Ammo niaoxidizing communities in a highly aerated fullscale activatedsludge bioreactor: betaproteobacterial dynamics and low relative abundance of Crenarchaea. Environ Microbiol. 2009;11:231028. doi:http://dx.doi.org/10.1111/j.1462-2920.2009.01958.x

Web End =10.1111/j.14622920.2009.01958.x .

Xia Y, Kong Y, Thomsen TR, Nielsen PH. Identication and ecophysiological characterization of epiphytic proteinhydrolyzing saprospiraceae (Can didatus Epiobacter spp.) in activated sludge. Appl Environ Microbiol. 2008;74:222938. doi:http://dx.doi.org/10.1128/AEM.02502-07

Web End =10.1128/AEM.0250207 .

Zhang LM, Hu HW, Shen JP, He JZ. Ammoniaoxidizing archaea have more important role than ammoniaoxidizing bacteria in ammonia oxidation of strongly acidic soils. ISME J. 2012;6:103245. doi:http://dx.doi.org/10.1038/ismej.2011.168

Web End =10.1038/ismej.2011.168 . Zhang T, Jin T, Yan Q, Shao M, Wells G, Criddle C, Fang HHP. Occurrenceof ammoniaoxidizing Archaea in activated sludges of a laboratoryscale reactor and two wastewater treatment plants. J Appl Microbiol. 2009;107:9707. doi:http://dx.doi.org/10.1111/j.1365-2672.2009.04283.x

Web End =10.1111/j.13652672.2009.04283.x .

Zhang T, Ye L, Tong AH, Shao MF, Lok S. Ammoniaoxidizing archaea and ammoniaoxidizing bacteria in six fullscale wastewater treatment bioreactors. Appl Microbiol Biotechnol. 2011;91:121525. doi:http://dx.doi.org/10.1007/s00253-011-3408-y

Web End =10.1007/ http://dx.doi.org/10.1007/s00253-011-3408-y

Web End =s002530113408y .

Zhang Y, Chen L, Sun R, Dai T, Tian J, Wen D. Ammoniaoxidizing bacteriaand archaea in wastewater treatment plant sludge and nearby coastal sediment in an industrial area in China. Appl Microbiol Biotechnol. 2015;99:4495507. doi:http://dx.doi.org/10.1007/s00253-014-6352-9

Web End =10.1007/s0025301463529 .

Zheng P, Lin FM, Hu BL, Chen JS. Startup of anaerobic ammonia oxidation bio reactor with nitrifying activated sludge. J Environ SciChina. 2004;16:136. Zhu L, Wang Z, Shu Q, Takala J, Hiltunen E, Feng P, Yuan Z. Nutrient removal and biodiesel production by integration of freshwater algae cultivation with piggery wastewater treatment. Wat Res. 2013;47:4294302. doi:http://dx.doi.org/10.1016/j.watres.2013.05.004

Web End =10.1016/j.watres.2013.05.004 .