Enrichment of nitrifying bacteria and microbial diversity analysis by high-throughput sequencing

Abstract

To achieve nitrifying bacteria (ammonia-oxidizing bacteria (AOB) and nitrite-oxidizing bacteria (NOB)) growth, activated sludge from the A²/O process in a wastewater treatment plant (WWTP) was used as seed sludge to guide the industrial culture of nitrifying bacteria for ammonia-contaminated wastewater treatment. The bacterial community was enriched in a fully automatic miniature fermenter by an alternating mode of intermittent operation mode in the daytime and a continuous ammonia feeding mode for intermittent operation during the night. The results indicated that the nitrifying bacteria community can be enriched by this alternating operation mode, resulting in a maximum ammonia oxidation rate reaching approximately 110 mg L⁻¹ h⁻¹ and a nitrite oxidation rate of approximately 50 mg L⁻¹ h⁻¹. High-throughput sequencing demonstrated that Nitrosomonas and Nitrospira belonging to AOB and NOB, respectively, were predominant in the activated sludge after 25 days at a ratio extending from 2.94% to 60.66%.

---

1. Introduction

Traditional autotrophic nitrification processes are commonly utilized in wastewater treatment plants (WWTPs). First, ammonia-oxidizing bacteria (AOB) are used to convert ammonia into nitrite, which is called partial nitrification; then, nitrite-oxidizing bacteria (NOB) convert the nitrite to nitrate to complete the full nitrification.¹ Some previous studies have proven that the biochemical activity of AOB and NOB can be adjusted through effective regulation of dissolved oxygen (DO),^2–4 temperature,^4–6 pH,^1,5 free ammonia (FA) and free nitrous acid (FNA).^7–11 It is known that AOB is more competitive at a lower DO concentration than NOB, while NOB is suited to a relatively higher DO level.¹² The oxygen affinity coefficient of AOB is approximately 0.74 mg L⁻¹, while it is approximately 1.75 mg L⁻¹ for NOB.¹³ However, studies have suggested that a relatively high DO level improves the ammonia oxidation rate.¹⁴ Additionally, previous studies indicated that NOB was more sensitive to FA and FNA inhibition compared to AOB, and the activity of NOB became inhibited when the FA concentration was 0.1–1.0 mg L⁻¹, while AOB was not inhibited until concentrations reached 10–150 mg L⁻¹.⁷ NOB was inhibited when the FA concentration was above 6 mg L⁻¹, whereas AOB was not inhibited when the FA level reached 16 mg L⁻¹.⁸ The biosynthesis of NOB was inhibited when the FNA level was above 0.02 mg L⁻¹,⁸ while the inhibition of AOB growth started to be effective at a level of approximately 0.10 mg L⁻¹.⁹ A relatively high temperature (30–35 °C) favoured nitrite accumulation for AOB, which had a greater growth rate than NOB at high temperatures.^5,15 The ideal pH level for AOB (7.9–8.2) is higher than NOB (7.2–7.6),¹⁶ and increasing pH or temperature elevates the FA level.

Nitrifying bacteria (AOB and NOB) belong to the autotrophic bacteria, which grow slowly and are vulnerable to loss in the reactor. Therefore, cell immobilization plays an important role in resolving the loss of bacteria, enhancing resistance, maintaining stable operation and improving the reaction rate.¹ Additionally, a bacterial community has more advantages due to its biodiversity compared to a pure strain of bacteria.¹⁴ Some researchers tried to increase the concentration of bacteria by adding cell immobilization fillers or bioaugmentation in order to improve the reaction rate.^14,17–19 Additionally, nitrifying bacteria are often used in aquaculture to remove ammonia and nitrite for the growth of aquatic animals and plants.²⁰

Some researchers have been proven that the AOB and/or NOB can be enriched with a relatively high initial ammonia concentration (e.g. about 500 mg L⁻¹), a relatively long cultivation period (e.g. over 90 days) or using pure strain of bacteria for inoculation.^21,22 Few reports were available regarding rapid enrichment of both AOB and NOB community with high ammonia and nitrite oxidation rates. This study aims to explore a method for enriching the nitrifying bacteria community that have high ammonia and nitrite oxidation rates. This method can be used to guide the industrial culture of nitrifying bacteria, the community widely used for cell immobilization, bioaugmentation, and ammonia and nitrite removal in aquaculture.

2. Materials and methods

2.1. Seed sludge and synthetic feed

The seed sludge was obtained from the return sludge in an A²/O process of the Gao Bei Dian Wastewater Treatment Plant in Beijing. Details of the synthetic feed stock solution are shown in Table 1.

Table 1 The composition of the synthetic feed stock solution^a

Compound	Concentration (g L^−1)	The trace element stock solution	Concentration (g L^−1)
a Note: KH2PO4, K2HPO4·3H2O; MgSO4·7H2O, CaCl2; the trace element stock solutions were mixed separately.
KH2PO4	3.95	ZnSO4·7H2O	0.50
K2HPO4·3H2O	6.62	MnCl2·4H2O	0.50
MgSO4·7H2O	1.80	CoCl2·6H2O	0.40
CaCl2	0.90	CuSO4·5H2O	0.40
The trace element stock solution (L)		NiCl2·6H2O	0.20
		Na2MoO4·2H2O	0.05

---

2.2. Experimental system set-up

A fully automatic miniature fermenter (Labfors, INFORS Corp., Switzerland) with a working volume of 6 L was used in this study, as shown in Fig. S1.† Two individual high-concentration stock solutions contained in each feed bottle were supplied for the continuous ammonia feed stream (NH₄Cl; nitrogen source) and the sodium carbonate (Na₂CO₃) feed stream (200 g L⁻¹ of raw feed; pH buffer, alkalinity supplement and carbon source). Two peristaltic pumps were connected to each feed bottle to control the flow rates of the ammonia and Na₂CO₃ feed streams. The flow rate of the Na₂CO₃ feed stream was controlled automatically to maintain a pH level of 7.2–7.4. Continuous aeration was provided by an air pump and supplied through a perforated tube. The dissolved oxygen (DO) concentration was maintained using an air flowmeter. The temperature of the liquid was maintained at 23 ± 1 °C with a thermostatic heater. A stirrer (100 rpm) supplied ample mixing during the aerobic reaction phase. The sampling port on the top of the reactor allowed for sample collection, feed addition, drainage siphoning and DO measurement.

2.3. Experiment operations

The reactor was operated in an alternate mode of intermittent operation mode in the daytime and a continuous ammonia feeding mode for intermittent operation during the night. The experimental system was operated for 25 days and consisted of two cycles per day (intermittent operation mode and continuous ammonia feeding mode for intermittent operation, respectively, designed as A and B). Each cycle consisted of four phases: feed addition, aerobic reaction, sludge settling and drainage. The specific operation of the experiment is as follows:

A: The reactor was operated in intermittent mode, and reaction substrates were fed into the reactor in batches. (1) At the beginning of each cycle, the deionized water was mixed with the pre-calculated volume of synthetic feed stock solution (100 mL for each; the composition is detailed in Table 1), which was fed into the reactor, resulting in the initial concentrations: 50–300 mg L⁻¹ NH₄⁺–N (prepared and added separately for each batch to improve accuracy; with the increase of the ammonia oxidation rate during the increasing operation days, the initial NH₄⁺–N concentration was increasing when the residual NH₄⁺–N concentration was close to zero after reacting for a certain time (e.g., 2 h), or the ammonia oxidation rate was stable over consecutive days), 30 mg L⁻¹ PO₄³⁻–P, 30 mg L⁻¹ MgSO₄·7H₂O, 15 mg L⁻¹ CaCl₂, 30 mg L⁻¹ FeSO₄·7H₂O (prepared and added separately to avoid being oxidized) and 1 mL L⁻¹ trace element stock solution. The heater, pH regulator, stirrer and aeration devices were turned onto start the aerobic reaction. (2) The next batch was started by adding the pre-calculated volume of NH₄Cl solution when the ammonia and nitrite were oxidized; the number of batches was determined by the time it took each batch of ammonia and nitrite to be oxidized. (3) Liquid samples were collected from the sampling port at the beginning and after reacting for a certain time (e.g., 2 h) to calculate the ammonia oxidation rate and nitrite accumulation ratio in one batch of Cycle A. (4) When the aerobic cycle was finished, the heater, pH regulator, stirrer and aeration devices for sludge settling and drainage were stopped, and the manual drainage siphoning left the remaining liquid less than 6 L to reduce the impact on the volume for Cycle B.

B: The reactor was operated in a continuous ammonia feeding mode for intermittent operation during the night to maintain the ammonia oxidation rate achieved in Cycle A in the daytime, and achieve a low nitrite accumulation ratio during the night, in order to grow both AOB and NOB. Ammonia was supplied in two ways: ammonia was added into the reactor at the beginning of each cycle, and a continuous ammonia feed stream was supplied during the aerobic reaction phase. (1) At the beginning of each cycle, the deionized water mixed with the pre-calculated volume of synthetic feed stock solution was fed into the reactor, resulting in the initial concentrations: 30 mg L⁻¹ NH₄⁺–N, 30 mg L⁻¹ PO₄³⁻–P, 30 mg L⁻¹ MgSO₄·7H₂O, 15 mg L⁻¹ CaCl₂, 30 mg L⁻¹ FeSO₄·7H₂O and 1 mg L⁻¹ trace element stock solution. The heater, pH regulator, stirrer and aeration devices were turned onto start the aerobic reaction. (2) The ammonia feed stream was supplied during the entire aerobic reaction phase. The flow rate was controlled at 20 mL h⁻¹ to reduce the impact on the volume. (3) The concentration of NH₄⁺–N feed in the bottle was determined by the actual ammonia oxidation rate, and the ammonia feed stream supply rate was increased with the increasing NH₄⁺–N concentration in the bottle for different cycles. The NH₄⁺–N concentration in the beginning of the reaction phase was maintained at 30 mg L⁻¹. The NH₄⁺–N concentration can be consistently maintained in an anticipated range during each aerobic reaction phase by using the continuous ammonia feeding mode for intermittent operation, and the FA concentration was in the range of 0.28 to 0.44 mg L⁻¹ (T = 23 °C, pH = 7.2–7.4), lower than the inhibition threshold of 10–150 mg L⁻¹ and 0.1–1.0 mg L⁻¹ of AOB and NOB, as mentioned above. The desired FA level can be controlled and stabilized by using the continuous ammonia feeding mode for intermittent operation by the regulation of the ammonia feed stream supply rate and the actual ammonia oxidation rate. (4) The residual NO₂⁻–N concentration at the end of each cycle was 100 mg L⁻¹, the calculated FNA concentration was 0.03–0.05 mg L⁻¹ (T = 23 °C, pH = 7.2–7.4), while NOB was inhibited when the FNA concentration was over 0.02 mg L⁻¹. Therefore, the desired NO₂⁻–N concentration at the end of each cycle can be controlled by using the continuous ammonia feeding mode for intermittent operation, so that the FNA level can be controlled by the regulation of the ammonia feed stream supply rate and the actual ammonia oxidation rate, thus, full nitrification can be achieved. (5) When each aerobic reaction phase was reached, the heater, pH regulator, NH₄Cl feed pump, stirrer and aeration devices for sludge settling and drainage were stopped. The manual drainage siphoning was performed. The reactor was filled with the deionized water and mixed well, and then the sludge was allowed to settle and manual drainage siphoning was taken 1–2 times, which was defined as “sludge cleaning”, to eliminate the nitrite and nitrate generated during the aerobic reaction phase. (6) Liquid samples were collected at the beginning and the end of the aerobic reaction phase to investigate the performance of ammonia oxidation and nitrite accumulation in Cycle B.

2.4. Experiment implement

The entire reactor operation was described as shown below:

• The whole experiment was divided into Phase I–VI according to the initial NH₄⁺–N concentrations (50–300 mg L⁻¹) of each batch in Cycle A on different days. The initial DO level was controlled at 1.0–1.5 mg L⁻¹ both in Cycle A and Cycle B.

• Batch test 1: take one batch of Cycle A on Day 23, for example, to investigate nitrification performance after the nitrifying bacteria community was enriched.

• Batch test 2: on Day 25, lower the initial NH₄⁺–N concentration to reduce the FA level, which will relieve the gradually increasing initial FA concentration on NOB in each batch of Cycle A.

2.5. Analytical methods

(1) The concentrations of chemical oxygen demand (COD), ammonia (NH₄⁺–N), nitrite (NO₂⁻–N), nitrate (NO₃⁻–N), mixed liquor suspended solids (MLSS) and mixed liquor volatile suspended solids (MLVSS) were all measured according to standard methods.²³ The DO levels were monitored by a DO meter (JPSJ-605F, INESA Corp., China).

(2) The FA and FNA concentrations were calculated according to eqn (1) and (2):⁷

(3) The initial and residual NH₄⁺–N concentrations with a certain reaction time were used to calculate the ammonia oxidation rate in Cycle A,²⁴ as shown in eqn (3):

where AOR_A (mg L⁻¹ h⁻¹): ammonia oxidation rate (mg L⁻¹ h⁻¹); [NH₄⁺–N]_A0: initial NH₄⁺–N concentration (mg L⁻¹); [NH₄⁺–N]_At: residual NH₄⁺–N concentration (mg L⁻¹); t: reaction time (h).

(4) Based on the continuous ammonia feeding mode for intermittent operation of Cycle B, the calculation formulas were established by Renda Yao as follows:

where AOR_B (mg L⁻¹ h⁻¹): ammonia oxidation rate (mg L⁻¹ h⁻¹); [NH₄⁺–N]_B0: initial NH₄⁺–N concentration (mg L⁻¹); [NH₄⁺–N]_add: ammonia feed stream supply rate added into the reaction liquid (mg L⁻¹ h⁻¹); [NH₄⁺–N]_Bend: residual NH₄⁺–N concentration (mg L⁻¹); t₁: the supply time of ammonia feed stream (h); t₂: reaction time (h), t₁ = t₂ in this study.where m: the quality of NH₄Cl in the bottle for each cycle (g); [NH₄⁺–N]_add: ammonia feed stream supply rate added into the reaction liquid (mg L⁻¹ h⁻¹); t₁: the supply time of ammonia feed stream (h); V: the volume of reaction liquid (L); S: the concentration of NH₄⁺–N in 1 L of reaction liquid provided by NH₄Cl, 100 mg L⁻¹ as a basic calculation value, ; the value is 0.382 in this study.

(5) For the residual nitrite and nitrate from the previous cycle after the “sludge cleaning”, the output of NO₂⁻–N and NO₃⁻–N in the batch of Cycle A and Cycle B were determined to evaluate the nitrite accumulation ratio,²⁵ as shown in eqn (6):

where NAR: nitrite accumulation ratio; [NO₂⁻–N]₀: initial NO₂⁻–N concentration (mg L⁻¹); [NO₂⁻–N]_end: residual NO₂⁻–N concentration (mg L⁻¹); [NO₃⁻–N]₀: initial NO₃⁻–N concentration (mg L⁻¹); [NO₃⁻–N]_end: residual NO₃⁻–N concentration (mg L⁻¹); note: the value is zero if the output is a negative value.

2.6. High-throughput 16S rRNA gene sequencing

Sludge samples were collected for DNA extraction, and DNA integrity was checked by agarose gel electrophoresis. The extracted DNA was used for PCR amplification of the 16S rRNA gene (V3–V4 region). PCR products were verified by electrophoresis, and then the selected bands were excised for DNA purification. Then, sequencing libraries were generated and assessed. The library was sequenced on an Illumina HiSeq2500 platform for high-throughput sequencing (SinoGenoMax Corp., China), and the raw reads were generated, and further filtration to form the database of high quality reads. These reads were clustered into operational taxonomic units (OTUs) with a setting 0.03 distance limit (equivalent to 97% similarity). The OTUs were used to depict the similarities and differences between different communities.

3. Results and discussion

3.1. Process performance of Cycle A

The reactor was operated in intermittent mode to investigate the ammonia oxidation rate of the seed sludge. The initial NH₄⁺–N concentration was 48.32 mg L⁻¹. After reacting for 2 h, the residual concentration was 14.16 mg L⁻¹; thus, the calculated ammonia oxidation rate was 17.08 mg L⁻¹ h⁻¹. Due to the intermittent operation mode, the output of nitrite and nitrate accumulated in the reactor to form the background value. For the variations of background values at the beginning of each batch, the initial and residual NH₄⁺–N concentrations and the output values of nitrite and nitrate in the first 2 h in the selected batch in Cycle A are used to reflect the nitrification performance of the activated sludge, as shown in Fig. 1(a). The calculated ammonia oxidation rate and nitrite accumulation ratio are shown in Fig. 1(b). The ammonia oxidation rate reached 107.26 mg L⁻¹ h⁻¹ on Day 24, which is over 6 times the seed sludge rate of 17.08 mg L⁻¹ h⁻¹.

Fig. 1 The performance of the activated sludge during the operation period of one batch of Cycle A for the enriched nitrifying bacteria community: (a) the concentrations of NH₄⁺–N, NO₂⁻–N, NO₃⁻–N at the beginning of and 2 h into the aerobic reaction phase during one batch of Cycle A; (b) the calculated ammonia oxidation rate and nitrite accumulation ratio.

With the increasing ammonia oxidation rate in the intermittent operation mode, the initial NH₄⁺–N concentration of each batch in Cycle A was gradually enhanced. The FA concentration and different initial NH₄⁺–N concentrations are shown in Fig. 1. The FA level increased from below 1 mg L⁻¹ to approximately 5 mg L⁻¹ (T = 22–24 °C, pH = 7.2–7.4). The initial FA concentration of each batch enhanced gradually to reach the level above the inhibition threshold of 0.1–1.0 mg L⁻¹ of NOB. As shown in Fig. 1(a), the initial NH₄⁺–N concentrations of each batch in Cycle A were 50 mg L⁻¹ and 100 mg L⁻¹ in Phase I and II, and the initial FA concentrations were 0.59 mg L⁻¹ and 1.18 mg L⁻¹ (T = 23 °C, pH = 7.3). Therefore, the NOB was not inhibited by FA, almost all of the ammonia converted to nitrate. However, the initial NH₄⁺–N concentrations of each batch in Cycle A increased from 150 mg L⁻¹ to 300 mg L⁻¹ from Phase III to Phase VI, and the corresponding FA level increased from 1.77 mg L⁻¹ to 3.54 mg L⁻¹. Thus, the NOB was inhibited by the increasing FA level, resulting in the increasing nitrite accumulation ratio, as shown in Fig. 1(b).

3.2. Process performance of Cycle B

A continuous ammonia feeding mode for intermittent operation was used for Cycle B. Due to the variations of background values at the beginning of each cycle, the initial and residual NH₄⁺–N concentrations, the output values of nitrite and nitrate in the 12 h of Cycle B are used to reflect the nitrification performance, as shown in Fig. 2(a). The ammonia feed stream supply rate, calculated ammonia oxidation rate and nitrite accumulation ratio are shown in Fig. 2(b). The relationship of the ammonia oxidation rate in Cycle A and B is clearly reflected in Fig. 2(b).

Fig. 2 The performance of the activated sludge during the operation period of Cycle B for the enriched nitrifying bacteria community: (a) the concentrations of NH₄⁺–N, NO₂⁻–N, NO₃⁻–N at the beginning and the end of the aerobic reaction phase (12 h); (b) the ammonia feed stream supply rate, calculated ammonia oxidation rate and nitrite accumulation ratio.

As shown in Fig. 1(b), in Phase I and II (Day 1–8), the nitrite accumulation ratio in Cycle A was always below 10%, and the ammonia feed stream supply rate in Cycle B can be determined according to the ammonia oxidation rate of Cycle A. However, from Phase III to Phase VI (Day 9–24), the nitrite accumulation ratio increased in Cycle A from approximately 20% to approximately 70% due to the increasing FA level inhibition of NOB. Therefore, the high nitrite accumulation ratio can be obtained during the 12 h operation period of Cycle B if the ammonia feed stream supply rate is determined according to the ammonia oxidation rate in that day of Cycle A, which has already been proven by the authors' other studies.

From Phase III to Phase VI (Day 9–24), the ammonia feed stream supply rate of Cycle B was adjusted according to both the ammonia oxidation rate and nitrite accumulation ratio of Cycle A. With the increase in ammonia oxidation rate, the outputs of nitrite and nitrate were also increasing. The output of nitrite was almost less than 100 mg L⁻¹ on these days, and the nitrite accumulation ratio was always almost below 25% in order to reach full nitrification to lower nitrite accumulation. Therefore, the residual NH₄⁺–N concentration was always less than 10 mg L⁻¹ at the end of Cycle B, and the ammonia oxidation rate was close to the ammonia feed stream supply rate in Cycle B. The ammonia oxidation rate increased from 18.57 mg L⁻¹ h⁻¹ to 61.76 mg L⁻¹ h⁻¹ during the enrichment for nitrifying bacteria for 24 days. Based on the analysis, the ammonia feed stream supply rate in Cycle B can be determined according to [NH₄⁺–N]_add = (0.9–1.3) × AOR_A × (1 − NAR_A).

3.3. Reactor performance after enrichment of the nitrifying bacteria community

As shown in Fig. 3(a), the initial NH₄⁺–N and FA concentrations are 278.39 mg L⁻¹ and 3.29 mg L⁻¹ (T = 23 °C, pH = 7.3), respectively. After reacting for 180 min, the ammonia is fully oxidized and the nitrite accumulation ratio is 58.60%. The average ammonia oxidation rate in the first 2 h reached 109.41 mg L⁻¹ h⁻¹. After reacting for 180 min, only NOB played a role in the system because there was no ammonia left in the system, only the nitrite resulting from ammonia oxidation. The generated nitrite in the previous 180 min could be almost fully converted to nitrate in the subsequent 3 h. The maximum nitrite oxidation rate reached 47.91 mg L⁻¹ h⁻¹. Additionally, the DO level increased sharply due to the low residual NH₄⁺–N concentration, therefore, the dissolved oxygen (DO) concentration should be properly decreased to less than 3 mg L⁻¹, otherwise, a sharp increase in DO concentration to the level suitable for bacterial endogenous respiration would occur.

Fig. 3 Ammonia oxidation and nitrite accumulation performance after the enrichment of the nitrifying bacteria community: (a) the initial NH₄⁺–N concentration of 300 mg L⁻¹; (b) the initial NH₄⁺–N concentration of 80 mg L⁻¹.

The activity of NOB was investigated after the initial FA level was decreased by lowering the initial NH₄⁺–N concentration on Day 25, as shown in Fig. 3(b). The initial NH₄⁺–N and FA concentration were 74.42 mg L⁻¹ and 0.88 mg L⁻¹, respectively. After reacting for 50 min, the ammonia is fully oxidized and the nitrite accumulation ratio is 32.12%. Compared with the results in Fig. 3(a), the nitrite accumulation ratio is reduced from 58.60% to 32.12%; the peak value of the nitrite output accounted for the initial NH₄⁺–N concentration is reduced from 65.15% to 51.03%. This indicates that the activity of NOB recover rapidly after the release of FA inhibition on NOB.

Additionally, the accumulated nitrite and nitrate was eliminated by “sludge cleaning” which was susceptible to the washout of some of the bacteria of the system, including a small proportion of nitrifying bacteria and to imprecisely control the age of the sludge. The MLSS decreased from 4573 mg L⁻¹ to 4064 mg L⁻¹ during the operation period, while the ammonia oxidation rate and nitrite oxidation rate were significantly increased, indicating that AOB and NOB were effectively enriched and with highly active ability after sludge cultivation. The ratio of MLVSS/MLSS was lower, between 0.73 and 0.62, and the level of SVI was reduced from 63 mL g⁻¹ to 16 mL g⁻¹, which indicates that the inorganic substance in the system increased, and the sludge properties changed through the transformation of floc sludge to granular sludge, as shown in Fig. S2.† This can be explained by the crystal nucleus hypothesis principle, where inorganic salt precipitation acts as the crystal nucleus in the formation of granular sludge during the cultivation process.²⁶

3.4. High-throughput sequencing for microbial diversity analysis

High-throughput sequencing was applied to analyze the microbial diversity of the sludge samples to investigate the effect of the enrichment of the nitrifying bacteria community. Good's coverage of three samples were greater than 99%, indicating that the result represented the microbial components of the sludge samples. Richness estimators Chao1 and ACE were used to estimate the total number of OTUs. A higher value shows greater richness. Diversity indexes Shannon and Simpson were used to reflect species diversity. A higher value indicates greater diversity. As shown in Table 2, Chao1, ACE, Shannon and Simpson of the sample of seed sludge are 1919, 1989, 8.40 and 0.99, respectively. These values decreased to 1289, 1431, 4.35 and 0.87, respectively, during the period for nitrifying bacteria cultivation on Day 17, and they further decreased to 1111, 1252, 3.70 and 0.81, respectively, after the nitrifying bacteria community was enriched on Day 25. In this study, an alternate mode of intermittent operation mode in the daytime and a continuous ammonia feeding mode for intermittent operation during the night was applied to enrich the nitrifying bacteria community. The results indicate that the ammonia oxidation rate and nitrite oxidation rate were greatly increased, while the microbial diversity was decreased in the activated sludge system. This is in agreement with previous investigations that the performance of the system improved with reduced microbial diversity and the richness of some specific functional species enhanced.^27,28

Table 2 Richness and diversity assessment of the nitrifying bacteria community in the seed sludge and cultivated sludge^a

Sample	Good's coverage	Chao1	ACE	Shannon	Simpson
a Notes: yao2: the seed sludge. yao4: the sludge during the period of enrichment of the nitrifying bacteria community (Day 17). yao5: the enriched nitrifying bacteria community sludge (Day 25).
yao2	0.9949	1919	1989	8.40	0.99
yao4	0.9910	1289	1431	4.35	0.87
yao5	0.9928	1111	1252	3.70	0.81

---

The bacteria community structure at the phylum level is exhibited in Fig. 4(a), showing the variations in microbial diversity during the operation periods. There is high microbial diversity in the seed sludge, while the diversity is lower after cultivation of the nitrifying bacteria community. The predominant phyla such as Proteobacteria, Actinobacteria, Bacteroidetes, Firmicutes, Chloroflexi and Nitrospirae, were detected in the seed sludge. Their relative abundances were 41.70%, 16.63%, 14.90%, 5.11%, 4.54% and 2.71%. This is in agreement with previous investigations which showed that the Proteobacterial microorganisms are a large part of activated sludge in municipal wastewater treatment systems, related to biological nitrification and organic matter degradation.^29,30 Actinobacteria and Bacteroidetes are often reported to be related to biological phosphorus removal and polysaccharide degradation, respectively.^31,32 Previous studies have showed that Firmicutes and Chloroflexi are often detected in anaerobic digestion sludge.^33,34 Molecular biology analysis demonstrates that most AOB existing in the soil and freshwater belong to β-Proteobacteria, including Nitrosomonas and Nitrosospira.³⁵ Nitrobacter (α-Proteobacteria) and Nitrospira (distinct phylum) affiliated with NOB are often found in wastewater treatment plants.³⁶ As shown in Fig. 4(a), during the operation period of 25 days, the ratio of Proteobacteria increased from 41.70% to 48.73%, while Nitrospirae was found to be more abundant in the sample, increasing from 2.71% to 25.80%. Bacteroidetes accounted for 20.60%, Actinobacteria was reduced to 4.36%, and other phyla of microorganisms almost gradually washed out of the reactor, resulting from the enrichment of the AOB and NOB community.

Fig. 4 Distributions of bacterial community in the seed sludge and cultivated sludge at different levels: (a) at the phylum level; (b) at the genus level.

The results can be further proven by the analyses based on the genus level, depicted in Fig. 4(b). From Day 1 to Day 25, Nitrosomonas (β-Proteobacteria), belonging to AOB, increased from 0.24% to 34.87%, and Nitrospira, affiliated with NOB, increased from 2.70% to 25.79%. Both were found to be the predominant microbial groups in the cultivated sludge and accounted for 60.66% of the total microorganisms. This result is consistent with the excellent biological nitrification performance of the high ammonia oxidation rate (approximately 110 mg L⁻¹ h⁻¹) and nitrite oxidation rate (approximately 50 mg L⁻¹ h⁻¹). Additionally, the samples collected on Day 17 and Day 25 were compared, and the results show that the AOB were further enriched. The ratio of Nitrosomonas increased from 21.75% to 34.87%, while the proportion of Nitrospira remained almost at the same level, between 25.95% and 25.79%. This can be explained by the following two aspects: (1) the initial NH₄⁺–N concentration in each batch in Cycle A increased gradually, 250 mg L⁻¹ and 300 mg L⁻¹ on Day 17 and 24, respectively, resulting in an increase in FA inhibition on NOB (Fig. 1). (2) The output of NO₂⁻–N concentration in Cycle B was almost in the range of 100–200 mg L⁻¹ during Day 17–24, which was higher than the value of almost less than 30 mg L⁻¹ during Day 1–16, leading to FNA inhibition on NOB (Fig. 2).

In addition, the top 50 most abundant genera in each sample were selected, and the richness heat map of the top 50 most abundant genera is shown in Fig. 5. Each column represents one sample, each line represents one genus, and the colour reflects the richness of each genus. We observed that Nitrosomonas and Nitrospira were both highly enriched, as shown by the variations of colour of these three samples. Moreover, the more similar the microbial community structure is, the closer the distance between samples in the clustering tree (shown above the colour pieces), indicating the close relationship of the microbial species in different samples. Thus, the results confirmed that the nitrifying bacteria community, including AOB and NOB, were both gradually and highly enriched during the enrichment period and predominated in the cultivated sludge.

Fig. 5 Richness heat map of the top 50 most abundant genera.

4. Conclusions

The nitrifying bacteria community were successfully enriched by an alternate mode in which intermittent operation in the day time increased the initial concentration of each batch to improve the ammonia oxidation rate, and a continuous ammonia feeding mode for intermittent operation during the night maintained the ammonia oxidation rate and resulted in a low nitrite accumulation ratio. The results in the growth of both AOB and NOB, a maximum ammonia oxidation rate reached approximately 110 mg L⁻¹ h⁻¹ and a maximum nitrite oxidation rate of approximately 50 mg L⁻¹ h⁻¹. After the ammonia oxidation rate increased significantly, FA or FNA inhibition on NOB can be alleviated effectively by lowering the initial ammonia concentration by intermittent operation to reduce the nitrite peak and nitrite accumulation rate during the reaction process. High-throughput sequencing was applied to investigate the effect of the enrichment of AOB and NOB on the total microorganisms, and a specific comparison at the phylum and genus level indicated that the diversity of the microorganisms declined during the operational period. Nitrosomonas belonging to AOB and Nitrospira affiliated with NOB became the most dominant communities, starting at a ratio of 2.94% and increasing to 60.66%.

Acknowledgements

This work was supported by the Program “The industrialization of bioactive fillers with excellent nitrifying performance used in wastewater treatment” (PXM2016 014204 500040) from Beijing Municipal Commission of Education for developing creativity and innovation in the universities owned by the municipal government of Beijing.

References

1. G. Ciudad, R. González, C. Bornhardt and C. Antileo, Water Res., 2007, 41, 4621–4629.
2. Y. Ma, Y. Z. Peng, S. Y. Wang, Z. G. Yuan and X. L. Wang, Water Res., 2009, 43, 563–572.
3. J. Wu, G. Yan, G. J. Zhou and T. Xu, RSC Adv., 2015, 5, 100299–100308.
4. Y. Xue, F. L. Yang, S. T. Liu and Z. M. Fu, Bioresour. Technol., 2009, 100, 1055–1060.
5. C. Hellinga, A. A. J. C. Schellen, J. W. Mulder, M. C. M. van Loosdrecht and J. J. Heijnen, Water Sci. Technol., 1998, 37, 135–142.
6. J. A. Torà, J. Lafuente, J. Carrera and J. A. Baeza, Environ. Technol., 2012, 33, 1361–1366.
7. A. C. Anthonisen, R. C. Loehr, T. B. S. Prakasam and E. G. Srinath, J. - Water Pollut. Control Fed., 1976, 48, 835–852.
8. V. M. Vadivelu, J. Keller and Z. G. Yuan, Water Sci. Technol., 2007, 56, 89–97.
9. V. M. Vadivelu, J. Keller and Z. G. Yuan, Biotechnol. Bioeng., 2006, 95, 830–839.
10. D. J. Kim, D. L. Lee and J. Keller, Bioresour. Technol., 2006, 97, 459–468.
11. J. A. Torà, J. Lafuente, J. A. Baeza and J. Carrera, Bioresour. Technol., 2010, 101, 6051–6058.
12. C. Rongsayamanont, T. Limpiyakorn and E. Khan, Bioresour. Technol., 2014, 164, 254–263.
13. A. Guisasola, I. Jubany, J. A. Baeza, J. Carrera and J. Lafuente, J. Chem. Technol. Biotechnol., 2005, 80, 388–396.
14. J. Yan and Y. Y. Hu, Bioresour. Technol., 2009, 100, 2341–2347.
15. S. B. Gu, S. Y. Wang, Q. Yang, P. Yang and Y. Z. Peng, Bioresour. Technol., 2012, 112, 34–41.
16. J. Alleman, Water Sci. Technol., 1985, 17, 409–419.
17. M. A. Head and J. A. Oleszkiewicz, Water Res., 2004, 38, 523–530.
18. S. Z. Ahammad, R. J. Davenport, L. F. Read, J. Gomes, T. R. Sreekrishnan and J. Dolfing, RSC Adv., 2013, 3, 774–781.
19. H. Khan and W. Bae, Water Sci. Technol., 2014, 70, 517–523.
20. H. Shan and J. P. Obbard, Appl. Microbiol. Biotechnol., 2001, 57, 791–798.
21. A. Bollmann and H. J. Laanbroek, FEMS Microbiol. Ecol., 2001, 37, 211–221.
22. R. D. Chen and T. M. Lapara, Process Biochem., 2008, 43, 33–41.
23. State Environmental Protection Administration (SEPA), Monitoring and Analysis Methods for Water and Wastewater, China Environmental Science Press, Beijing, China, 4th edn, 2002, in Chinese.
24. X. H. Liu, Y. Peng, T. Ma, C. H. Liu and Y. Z. Peng, Environ. Sci., 2008, 29, 660–664 , in Chinese.
25. Z. Y. Zhang, X. Chen and L. C. An, Chem. Bioeng., 2007, 24, 63–67 , in Chinese.
26. B. Long, C. Z. Yang, W. H. Pu, J. K. Yang, G. S. Jiang, J. F. Dan, C. Y. Li and F. B. Liu, Bioresour. Technol., 2014, 166, 57–63.
27. J. X. Ma, Z. W. Wang, Y. Yang, X. J. Mei and Z. C. Wu, Water Res., 2013, 47, 859–869.
28. L. W. Shen, Y. Yao and F. G. Meng, J. Membr. Sci., 2014, 462, 139–146.
29. J. Snaidr, R. Amann, I. Huber, W. Ludwig and K. H. Schleifer, Appl. Environ. Microbiol., 1997, 63, 2884–2896.
30. Z. G. Yuan and L. L. Blackall, Water Res., 2002, 36, 482–490.
31. G. R. Crocetti, P. Hugenholtz, P. L. Bond, A. Schuler, J. Keller, D. Jenkins and L. L. Blackall, Appl. Environ. Microbiol., 2000, 66, 1175–1182.
32. C. Kragelund, C. Levantesi, A. Borger, K. Thelen, D. Eikelboom, V. Tandoi, Y. Kong, J. van der Waarde, J. Krooneman, S. Rossetti, T. R. Thomsen and P. H. Nielsen, FEMS Microbiol. Ecol., 2007, 59, 671–682.
33. G. Zhao, F. Ma, L. Wei and H. Chua, Bioresour. Technol., 2012, 113, 83–88.
34. D. Riviere, V. Desvignes, E. Pelletier, S. Chaussonnerie, S. Guermazi, J. Weissenbach, T. Li, P. Camacho and A. Sghir, ISME J., 2009, 3, 700–714.
35. S. Siripong and B. E. Rittmann, Water Res., 2007, 41, 1110–1120.
36. H. Daims, J. L. Nielsen, P. H. Nielsen, K. H. Schleifer and M. Wagner, Appl. Environ. Microbiol., 2001, 67, 5273–5284.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra24213h

---