Response of microbial community and biological nitrogen removal to the accumulation of nonylphenol in sequencing batch reactor

Abstract

The widespread existence of nonylphenol in the environmental rendered from wastewater discharge has become a growing concern for its endocrine disrupting effects on microorganisms. In this study, the performance of nitrifying and denitrifying microbial community in a sequencing batch reactor (SBR) was investigated under different nonylphenol concentrations. The SBR was shown to be less effective in nitrogen removal at higher concentration of nonylphenol. Proteobacteria, Bacteroidetes, and Actinobacteria were characterized by 454 pyrosequencing as the dominant bacteria, nitrogen removal functional bacteria in these three phyla were inhibited by nonylphenol, and Proteobacteria and Actinobacteria were more sensitive to nonylphenol. With the accumulation of nonylphenol, the population of the most abundant denitrifying bacteria (Thauera spp.) and nitrifying bacteria (Nitrosomonas spp.) significantly reduced. Microbial diversity increased due to nonylphenol perturbation, which is indicated by the changes in microbial alpha diversity. Principal component analysis showed high similarity between microbial community in low and high concentration of nonylphenol, and the core genera involved in nitrogen removal had a low correlation with other genera shown in co-occurrence network. Moreover, linear discriminant analysis effect size analysis revealed intergroup differences in microorganisms. The mechanism of accumulated NP on the diversity and metabolism of the microbial community was examined. This paper established a theoretical foundation for the treatment of NP-containing wastewater and provided hints for further research about NP impact on biological nitrogen removal.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1007/s13762-023-04825-9.

Introduction

Endocrine disrupting compounds (EDCs) have been widely reported as exogenous chemicals that interfere with endocrine system of human and wildlife (Liu et al. 2021). Recently, it was reported that EDCs may escalate the risk of COVID-19 infection via its ability in dysregulation of immune system (Zahra et al. 2020). As one of the EDCs, nonylphenol (NP) is difficult to be degraded in wastewater treatment plants, and it will accumulation in rivers, oceans, lakes, and drinking water, where NP can persist for a long time has raised a great concern around the world in recent years (Vieira et al. 2020).

As a member of alkylphenol, NP is toxic to aquatic organisms and can accumulate in their tissue, causing DNA damage-induced mutations that ultimately cause cancer (Lalonde and Garron 2021). Studies showed that the maximum acceptable concentration of NP is 10 μg/L, and US Environmental Protection Agency also provides for the concentration of nonylphenol in freshwater to be less than 6.6 μg/L (Soares et al. 2008). The production of NP has been banned in the European Union, but this does not stop the spread of NP, as many countries still use NP as an important raw material, causing serious accumulation of NP residues. NP cannot be effectively removed by conventional wastewater treatment plants (WWTP) for its small molecule size and relatively high stability (You et al. 2019). Nevertheless, in some cities of China, industrial wastewater is often mixed with domestic sewage, leading to the accumulation of NP (Zhao et al. 2021). However, the effect of NP accumulation on the nitrogen removal performance of WWTP has been rarely reported.

The accumulation of NP in WWTP inevitably has an impact on the microbial community. Previous studies reported that high concentrations of NP significantly altered the microbial community structure and found that a variety of microorganisms were involved in NP degradation (Wang et al. 2015). It was also noted that by examining the effect of different concentrations of nonylphenol on soil microorganisms, it was found that low concentrations of NP had less impact on the microbial community structure (Mattana et al. 2019). Additionally, the impact of NP exposure on organisms and the degradation methods of NP in the environment have been extensively studied (Bhandari et al. 2021). Many studies have focused on the degradation or interaction of microorganisms on NP, but the effect of NP on the dominant functional bacterial genus in WWTP has not been investigated. Therefore, it is necessary to investigate the effect of NP accumulation on the functional microbial community associated with nitrogen removal.

This work examined the changes of microbial community structure subjected to the accumulation of NP and explored the functional mechanisms between microbial populations during the nitrification and denitrification by means of SBR. Analysis of microbial community abundance and diversity under NP accumulation was evaluated by pyrophosphate sequencing results. Visualization of differences in microbial communities of different samples was accomplished by PCA and LEfSe analysis. Network analysis revealed microbial community co-occurrence patterns under NP accumulation. This study aims to explore the accumulation of NP during nitrification and denitrification in SBR, providing some clues for the degradation of NP in water treatment and optimization of NP wastewater treatment process.

Material and methods

Lab-scale SBR setup and operation

The nitrification and the denitrification systems were set up with two SBRs with 3-L volume each. The SBRs were operated at the same time with identical parameters, including an operation cycle of 24 h, operating temperature of 22–25 °C, an aeration time of 14 h, a dissolved oxygen (DO) concentration around 2.5–3 mg/L, a stirring stage for 10 h under anoxic condition, and about 8–9 to promote nitrification.

The seed-activated sludge was taken from the secondary sedimentation tank of a sewage treatment plant. Upper floats and large chunks of sediment from the lower layers were removed by filtration. After aeration for several hours, the heterotrophic bacteria consume and remove the toxic substances by endogenous respiration, and the mixed liquor suspended solids (MLSS) were 3.58 g/L. Finally, the inoculated activated sludge of the equivalent volume was put into two SBRs, the sludge was acclimated, and the initial water quality index of the two SBRs was basically the same and stable. The influent is artificial domestic sewage, whose recipe is given in Table S1.

After the stabilization of two SBR reactors, NP was added to one reactor, and the other was set as a blank control. In influent, the concentration of NP was 0, 10, 20, 40, and 80 mg/L, to ensure that both reactors run under the same conditions, chemical oxygen demand (COD) (300 mg/L) and ${\text{NH}}_4^{+}$–N (250 mg/L) remained steady.

Analytical methods

A rapid digestion spectrophotometric method was adopted for determination of COD. Ammonia nitrogen (${\text{NH}}_4^{+}$–N) was measured by salicylic acid spectrophotometry, and total nitrogen was measured by alkaline potassium persulfate digestion UV spectrophotometry. Nitrate–nitrogen (${\text{NO}}_3^{-}$–N) and nitrite–nitrogen (${\text{NO}}_2^{-}$–N) were identified with ion chromatograph (Metrohm, Switzerland) using standard methods.

DNA extraction and Roche 454-pycrosequencing

Representative samples of each reactor were collected on different concentration of NP (0, 10, 20, 40, and 80 mg/L), during nitrification and denitrification periods, which were named N0, DN0, N10, DN10, N20, DN20, N40, DN40, N80, and DN80, respectively. Sampling period and description are shown in Table 1. Each sample was prepared by 30-min precipitation of 50 ml of sludge water mixture, to remove the supernatant, and the remaining sludge water mixture was mixed, and then 1 ml of the mixture was collected in a sterile centrifuge tube and centrifuged for 5 min (4000 rpm, 25 °C). After removing the supernatant, 0.25–1 g of the precipitate was collected and stored at − 80 °C for DNA extraction. The total DNA of all samples was extracted using UltraClean DNA Isolation kit (MO-BIO Laboratories, CA12800, USA), and the DNA was confirmed by electrophoresis on a 0.8% agarose gel, and operated according to the instruction manual.

Table 1.

Sample introduction

Sample	Sampling time	Description
DN0	Initiation stage	Control group, blank control
DN10	7 days	Denitrification, 10 mg/L NP
DN20	14 days	Denitrification, 20 mg/L NP
DN40	21 days	Denitrification, 40 mg/L NP
DN80	28 days	Denitrification, 80 mg/L NP
N0	Initiation stage	Control group, blank control
N10	7 days	Nitrification, 10 mg/L NP
N20	14 days	Nitrification, 20 mg/L NP
N40	21 days	Nitrification, 40 mg/L NP
N80	28 days	Nitrification, 80 mg/L NP

The V3 region of the 16S rRNA gene from each extracted DNA sample was amplified with universal primers P3 (5’-CGCCCGCCGCGCGCGGCGGGCGGGGCGGGGGCACGGGGGGCCTACGGGAGGCAGCAG-3’) and P2 (5’-ATTACCGCGGCTGCTGG-3’). The PCR reaction system is 25 µl per sample and consists of 0.25 U of platinum Pfx DNA polymerase (Invitrogen, USA), 2.5 µl of 10 × Pfx amplification buffer (Invitrogen, USA), 0.5 mM MgSO₄ (Invitrogen, USA), 0.3 mM of each dNTP, 6.25 pmol of each upstream and downstream primer (1 µl of the mixed system), and 20 ng of template DNA and was conducted in a thermocycler PCR system (PCR Sprint, Thermo electron, UK). A five-cycle reconditioning PCR method was performed in order to decrease heteroduplex formation. All the PCR products were analyzed by electrophoresis in a 1.2% agarose gel that was stained with ethidium bromide. The PCR product from each sample was purified using the Gel/PCR DNA Fragments Extraction Kit (Geneaid, UK). Thirty nanogram of each purified PCR product was mixed and purified from a 1.2% agarose gel using the Gel/PCR DNA Fragments Extraction Kit. Sequencing was performed on a GS20 platform (Roche, Shanghai, China).

Sequencing results are subject to quality control, which requires sequences that meet all four of the following restraints: The sequence has a barcode and primers are intact; the barcode of the sequence can be missing (the remaining bases must match the corresponding ratio of the barcode), while the primer of the other segment should be intact; the length of the sequence is greater than 160 bp; and there should be no controversial bases in the sequence.

Statistical analysis

The Roche 454 high-throughput sequence analysis process was mainly written in Practical Extraction and Report Language (PERL), which was combined with the corresponding bioinformatics software package, and the programming analysis tool was further embedded in the program, making it suitable for the research analysis platform. The total information analysis process is shown in Fig. S1. Heatmap were created using the pheatmap package in R. Calculated correlation coefficients using R visualize Circos using the Circos (http://mkweb.bcgsc.ca/tableviewer/visualize/) free online platform. Network analysis was assessed by Spearman’s correlation analysis in R, and its subsequent visualization was achieved by Gephi. LEfSe analysis used Galaxy (https://huttenhower.sph.harvard.edu/galaxy/) free online platform.

Results and discussion

Performance of nitrogen removal at different levels of NP accumulation

The changes in ${\text{NH}}_4^{+}$–N concentration under the influence of various concentrations of NP are shown in Fig. 1. In the control, the concentration of ${\text{NH}}_4^{+}$–N decreased to 5 mg/L after the end of domestication (7 d), and the removal efficiency reached 95%. However, in NP bioreactor, when NP (10 mg/L) is initially added (stage I), the effluent concentration of ${\text{NH}}_4^{+}$–N reduced to 10 mg/L, and the removal efficiency was less than 90%. With the increase in the concentration of NP, the removal efficiency of ${\text{NH}}_4^{+}$–N in the reactor decreased significantly, though the effluent concentration of ${\text{NH}}_4^{+}$–N at 40 mg/L NP (stage III) is lower than 20 mg/L NP (stage II). Apparently, the concentration in control group is also lower than stage II. This may be due to the long-term stable operation of the SBR, and the nitrifying bacteria can adapt well to the environment and predominated. When the concentration of NP increased to 80 mg/L (stage IV), the concentration of ${\text{NH}}_4^{+}$–N could only be reduced to 25 mg/L, and the removal rate was less than 60%. From the above four different operation periods, it can be seen that the ${\text{NH}}_4^{+}$–N concentration in the effluent of the NP group was always higher than that of the control group on the first day when different concentrations of NP were applied that the inhibition of nitrification process by NP is significant. This suggested that NP has a certain inhibitory effect on the microbial community and functional bacteria in the SBR, and it also reflected that the ammonia-oxidizing bacteria in the reactor are more sensitive to NP (Ferrer-Polonio et al. 2022).

Fig. 1.

Effect of different concentrations of NP on ${\text{NH}}_4^{+}$–N removal (stage I, II, III, and IV indicate the NP concentration of 10, 20, 40, and 80 mg/L, respectively)

As shown in Fig. 2a, with the stable operation of bioreactor, the concentration of ${\text{NO}}_3^{-}$–N and ${\text{NO}}_2^{-}$–N in the effluent decreased in the control, and the concentration of ${\text{NO}}_3^{-}$–N was higher than ${\text{NO}}_2^{-}$–N. When the concentration of dissolved oxygen is the same, nitrifying bacteria have a better access to oxygen (Zhang et al. 2020), so ${\text{NO}}_3^{-}$–N is produced at a higher rate than ${\text{NO}}_2^{-}$–N. During denitrification, most of the ${\text{NO}}_2^{-}$–N reductase activity is greater than that of nitrate reductase (Balotf et al. 2016). Therefore, ${\text{NO}}_2^{-}$–N is reduced to N₂ before ${\text{NO}}_3^{-}$–N, as a result of which the concentration of ${\text{NO}}_3^{-}$–N is greater than ${\text{NO}}_2^{-}$–N. In NP bioreactor, shown in Fig. 2b, the accumulation of ${\text{NO}}_3^{-}$–N was evident. Nevertheless, with the increase in the concentration of NP, the concentration of ${\text{NO}}_2^{-}$–N in effluent increased. However, when the NP concentration was 20 mg/L and 40 mg/L, the ${\text{NO}}_2^{-}$–N concentration first increased to 98 and 99 mg/L and then decreased to 33 and 55 mg/L. When the NP concentration was 40 mg/L, the ${\text{NO}}_2^{-}$–N effluent concentration even reached 100 mg/L. Accumulation of ${\text{NO}}_2^{-}$–N is evident in NP group, and as the concentration of NP increased, the accumulation of ${\text{NO}}_2^{-}$–N became more and more significant. In all four periods, on the first day of NP dosing, the concentrations of ${\text{NO}}_3^{-}$–N in the NP group were all lower than those in the control group, and the concentrations of ${\text{NO}}_2^{-}$–N were all higher than control group. Thus, NP has a significant inhibitory effect on denitrification, and the higher concentration of ${\text{NO}}_2^{-}$–N is caused by the above-mentioned inhibition of nitrification by NP. These results indicated that the increase in NP concentration may lead to a decrease carbon source, which is glucose in our work, and the lack of carbon source significantly inhibits the denitrification process. Moreover, NP is difficult to degrade under anaerobic conditions (Graca et al. 2016), thus the inhibitory effect of NP on the denitrification process is significantly higher than that of the nitrification.

Fig. 2.

a Changes of ${\text{NO}}_2^{-}$–N and ${\text{NO}}_3^{-}$–N in the control group. b Changes of ${\text{NO}}_2^{-}$–N and ${\text{NO}}_3^{-}$–N in NP group (stage I, II, III, and IV indicate the NP concentration of 10, 20, 40, and 80 mg/L, respectively)

The dynamics of microbial community analysis at different taxonomic levels

As shown in Fig. 3, Proteobacteria, Bacteroidetes, and Actinobacteria were the top three phyla. As prevalent phylum of bacteria in wastewater treatment, Proteobacteria plays an important role in biological treatment, especially the class of Betaproteobacteria, which is considered to play a dominant role in the activated sludge process (Rehman et al. 2020). Bacteroidetes, with a high metabolic potential, are widely inhabit in a variety of water environments. Although Actinobacteria are considered to be soil bacteria, they may be more abundant in freshwater and possess metabolic versatility (van Bergeijk et al. 2020). All three phyla mentioned above are in high abundance in the aqueous environment. Proteobacteria was the most abundant phylum in almost all samples except DN10, whereas Bacteroidetes was the dominant phylum in DN10. Both Proteobacteria and Bacteroidetes have a strong connection to nitrogen cycle activities (Shu et al. 2015). Actinobacteria was clearly dominated in DN20, indicating that Actinobacteria thrive in oligotrophic conditions within the tolerable concentration of NP. It was suggested that low concentration of NP (10 mg/L) may inhibit the activity of denitrifying bacteria in Proteobacteria and Actinobacteria; however, maybe due to higher prevalence under anaerobic environment, low concentration of NP had minor inhibition (Hu et al. 2012); nevertheless, it is not particularly resistant to high NP levels.

Fig. 3.

Microbial community at phylum-level correlated to different samples were visualized via Circos (N0, N10, N20, N40, and N80 indicate that the concentration of NP was injected as 0, 10, 20, 40, and 80 in nitrification; DN0, DN10, DN20, DN40, and DN80 indicate that the concentration of NP was injected as 0, 10, 20, 40, and 80 in denitrification, respectively)

Heatmap of phylum-, order-, genus-level taxa in bioreactor is shown in Fig. 5. At the phylum level, 12,441 sequences fragments measured in ten sets of samples (N0, DN0, N10, DN10, N20, DN20, N40, DN40, N80, and DN80) cover a total of seven known bacterial phyla. The dominant phyla in bioreactor are Proteobacteria (58%), Bacteroidetes (21.3%), and Actinobacteria (15.8%), consistent with the results in Fig. 4. Bacteria with the highest abundance in the bioreactor are Sphingobacteriales (18.9%), Actinomycetales (14.4%), Rhodocyclales (13.3%), Xanthomonadales (7.6%), Burkholderiales (7.7%), and Rhizobiales (6.3%). From the genus level, Thauera spp. (11.6%), Pseudoxanthomonas spp. (4.2%), Acidovorax spp. (3.4%), Brevundimonas spp. (2.7%), Hydrogenophaga spp. (2%), and Nitrosomonas spp. (1.7%) belonging to Proteobacteria, and Ferruginibacter spp. (3%) belonging to Bacteroidetes, were found to be seven main genera. Notably, almost all of these genera with a high abundance belong to the phylum of Proteobacteria, because most of these bacteria are involved in the metabolism of nitrification and denitrification, belonging to Alpha-, Beta-, Gamma-, and Deltaproteobacteria, which are major orders belonging to the phylum of Proteobacteria.

Fig. 5.

Bacterial community diversity shifts. a Alpha diversity analysis of bacterial communities. b PCA analysis of bacterial communities

Fig. 4.

Heatmap of microbial community at phylum, order, genus level

In the presence of low concentration of NP, the abundance of Proteobacteria and Actinobacteria both had a notable increase in the system of nitrification, while the abundance of Bacteroidetes had no significant change. These results are consistent with the abundance of Bacteroidetes mentioned above. Interestingly, the abundance of these phyla reduced in denitrification. Taken together, low concentration of NP may have a greater adverse effect on anaerobic or parthenogenic anaerobic bacteria; thus, aerobic bacteria have a favorable advantage in competition with anaerobic bacteria under low concentration of NP. However, from the change of ${\text{NH}}_4^{+}$–N, it is suggested that some aerobic bacteria involved in nitrification are not the dominant functional bacteria. With the stepwise increase in NP concentration, the abundance of Proteobacteria, Bacteroidetes, and Actinobacteria showed increase to decrease both in nitrification and denitrification in general. Under the stress of high concentration of NP, the abundance of Bacteroidetes and Actinobacteria is lower than the initial, in the system of nitrification and denitrification. This is because NP interfered regular metabolism of Actinobacteria (Palmer and Horn 2012). Furthermore, high concentration of NP inhibits hydrolytic activity of Ferruginibacter spp. from the phylum Bacteroidetes (Kang et al. 2015), leads to difficulty in degradation of some organic matter in the water, and affects the subsequent aerobic nitrification. Some bacteria of Bacteroidetes have been identified to have denitrification ability (Su et al. 2019), whereas the abundance of Proteobacteria was increased at high concentration of NP in denitrification and decreased in nitrification compared with the performance at low concentration of NP. It may be that under the toxic effect of high concentration of NP, some aerobic nitrifying bacteria are in a disadvantageous position in the competition against denitrifying bacteria, and gradually undergo community succession; the shift from aerobic to anaerobic metabolism is caused by changes in the expression levels of a few genes (Giannopoulos et al. 2017). At high concentration of NP, most of them are non-denitrifying bacteria with increased genus abundance.

As the concentration of NP in the reactor gradually increased, the abundance of Pseudoxanthomonas spp., Acidovorax spp., and Brevundimonas spp. went higher. This may be due to their greater environmental adaptability (Lukhele et al. 2021). Pseudoxanthomonas spp. are the dominant denitrifying bacteria in SBR and resistant to toxic substances through EPS secretion function (Zhang et al. 2019), and Acidovorax spp. were shown to have an outstanding ability to degrade aromatic hydrocarbons (Benedek et al. 2018). Brevundimonas spp. were shown to have the potential to degrade estrogen (Muller et al. 2010). Therefore, NP as an estrogen also inhibited the activity of Brevundimonas spp.. Above all, the abundance of Proteobacteria increased. Thauera spp. is the most abundant genus in the bioreactor in the control group for nitrification and denitrification (N0, DN0). The abundance of Thauera spp. has decreased significantly as NP concentrations increased, whether through nitrification or denitrification. It is concluded that Thauera spp. became vulnerable under the effect of NP in SBR. In autotrophic, heterotrophic, and mixotrophic environments, Thauera spp. were plentiful and widespread, which led to the gradual deterioration of the removal of ${\text{NH}}_4^{+}$–N in the SBR and the accumulation of a large amount of ${\text{NO}}_2^{-}$–N. Hydrogenophaga spp., belonging to Betaproteobacteria, are involved in denitrification (Xu et al. 2020). The abundance of Hydrogenophaga spp. increased with the increasing concentration of NP in the system of denitrification. As an autotrophic bacterium, Hydrogenophaga spp. may have no adverse effect on the reduction of carbon sources caused by high concentration of NP. It was observed that the abundance of Nitrosomonas spp. decreased. Nitrosomonas spp. is an obligate chemolithoautotrophy belonging to Betaproteobacteria, and it is an ammonia-oxidizing bacteria, which is essential for the nitrification process. These results suggested that some nitrifying and denitrifying bacteria were inhibited by NP, especially at higher concentrations.

Changes in microbial diversity during NP accumulation

Shannon index shown in Fig. 5a, which is the most widely used metric of Alpha diversity, was considered to be positively related to the microbial diversity. It can be observed that the impact of NP on bacterial diversity is greater. However, bacterial diversity rose dramatically following NP dosage; both low and high concentrations of NP can promote the increase in key transport substances for coenzyme synthesis and expression of key genes (Duan et al. 2021). Nevertheless, low NP concentrations (N10, DN20) resulted in high bacterial richness. Thus, when NP concentration increased, both of their microbial diversity dropped compared to low concentration. When the concentration of NP is at the highest level (N80, DN80), bacterial diversity showed the lowest rate. These revealed that the addition of a small amount of NP caused a change in the composition of carbon source, favoring the growth of new microorganisms over the toxic effect on existing microorganisms. And toxicity to microorganisms gradually increased with the accumulation of NP, research have confirmed the difficulty of bacterial survival in the presence of high NP concentrations, and affected by the high concentration of NP, no microbial communities were found to be resilient (Mattana et al. 2019). Even at high NP concentration, bacterial diversity was greater than the control group, suggesting that some bacteria may have developed resistance to NP.

According to principal component analysis (PCA), samples could be clustered into three main groups (Fig. 5b), i.e., (1) low concentration of NP (N10, N20, DN20), (2) high concentration of NP (N80, DN80), and (3) N0 and DN10. In the control group, microbial communities in denitrification are less correlated with other samples, but in nitrification, microbial communities are similar to those in DN10. Indicated that the addition of NP may have caused microbial community succession, the effect of low concentration of NP on nitrification was not as strong as denitrification, and the system of denitrification was more sensitive to NP. The greater resemblance between microbial communities was seen under the low concentration of NP. However, as the concentration of NP increased, the distance between samples grew longer, and the degree of similarity between microbial communities decreased. When concentration of NP is N80 or DN80, the relevance of microbial community is higher. It is demonstrated that the increasing concentration of NP will have a clear effect on nitrifying and denitrifying bacteria. In the presence of high NP concentration, microbial communities were discovered to have a high degree of resemblance. As previous research showing that cell growth is entirely inhibited by high doses of NP (Yang et al. 2021), it is proposed that only a microbial fraction with resistance to the toxicity of NP was not inhibited during nitrification and denitrification, thus revealing a similar microbial community.

Co-occurrence network analysis of microbial communities

In order to identify bacterial interactions under the effect of NP, we have established the co-occurrence network of bacteria. The co-occurrence network of bacterial community is shown in Fig. 6, the networks consisted of 69 nodes (OTUs) and 296 edges, and eight modules (I–VIII) make up the network partitioning. The abundance of each module bacteria showed clusters of bacterial prosperities. Module I is mainly the phylum of Actinobacteria, and module II and module III mostly belonged to Proteobacteria. There is a strong correlation between module I and module III, demonstrating that the bacteria belonged of the Actinobacteria phylum, and some bacteria from the Proteobacteria phylum have a potential connection in functional metabolism. Actinobacteria and Proteobacteria contain the majority of the denitrifying bacteria, and the O₂ tension affects the bacterial community (Chu et al. 2020). Moreover, the most abundant genus Thauera sp. from module V is related to Lysobacter sp. and Rhizobium sp., illustrating that these two bacteria may have the same functional as Thauera sp.. Previous research have shown that Lysobacter sp. and Rhizobium sp. were nitrogen-fixing bacteria which are related to ammonia transfer (Iwata et al. 2010). Pseudoxanthomonas sp. was correlated with Brevundimonas sp., and their abundance varied consistently as shown in Fig. 6, because they all have the function of degrading certain environmental pollutants (Song et al. 2019). It is indicated that both of these genera have a high abundance in the phylum of Proteobacteria, and they may play the same role in metabolism. Above all, modules with abundant bacteria are clearly discrete and were poorly correlated with any other bacteria. Therefore, these crucial functional genera are self-contained.

Fig. 6.

Networks of co-occurring bacterial community

Intergroup variability of microbial communities between high concentration and other concentration levels of NP

LEfSe analysis emphasizes statistical significance and biological relevance; biomarkers that are statistically different between subgroups can be detected. LDA scores when performing important biomarker analysis are at least 2.0. The results of the LEfSe analysis include two parts: a Cladogram, to show the taxonomic hierarchy distribution of marker species that are significantly enriched in each group of community samples, the circles radiating from the inside to the outside represent the taxonomic rank, and the diameter of the small circles represents the relative abundance and a Histogram of the distribution of LDA values for significantly different species. The length of the bar graph represents LDA score, as well as indicates the extent of the effect of significantly different species between groups. Twenty-six bacterial clades were identified in LEfSe analysis as shown in Fig. 7b, namely two phyla (Actinobacteria and Proteobacteria), two classes (Actinobacteria and Gammaproteobacteria), three orders(Burkholderiales, Xanthomonadales, and Actinomycetales), six families (Comamonadaceae, Xanthomonadaceae, Caulobacteraceae, Propionibacteriaceae, Chitinophagaceae, and Micrococcaceae), and 13 genera (Pseudoxanthomonas, Sphingomonas, Acidovorax, Shinella, Simplicispira, Caulobacter, Ensifer, Thermomonas, Bosea, Sphingopyxis, Tessaracoccus, Ferruginibacter, and Arthrobacter). The results showed significant differences between the levels in the classification plots of the five levels (Fig. 7a). The highest abundance of biomarkers belonged to Actinobacteria in A and Proteobacteria in B. Although it showed the largest abundance with high concentrations of NP, Proteobacteria was not the dominating bacterium. It means that whereas the nitrifying and denitrifying bacteria in Proteobacteria were decreased in reaction to NP, nitrifying and denitrifying bacteria in Actinobacteria were dominating.

Fig. 7.

a Description of microbial intergroup variability under the effect of different concentrations of NP that have marked associations with Group1 [control group and the group of low concentration of NP (0, 10, 20, and 40 mg/L)] and Group2 [the group of high concentration of NP (80 mg/L)]. b 26 identified significantly divergent taxonomy with linear discriminant analysis LDA scores larger than 2.0

Conclusion

This study is the first report about the mechanisms of NP impact and accumulation on the microbial community structure in SBR. The effectiveness of SBR for nitrogen removal showed that both nitrification and denitrification are inhibited with the increase in concentration of NP. Four hundred and fifty-four pyrosequencing analysis was microbial community structure, diversity, interspecific similarity, and species relatedness. Under the impact of low concentration of NP, microbial diversity increased in SBR with high interspecific similarity. The bacteria in Proteobacteria and Actinobacteria were more sensitive to NP, but the abundance of bacteria in Bacteroidetes did not change significantly. With the accumulation of high concentration of NP, microbial diversity decreased compared to low concentration of NP, but interspecific similarity was higher. With the accumulation of NP, the microbial diversity increased. The bacteria of nitrifying and denitrifying in Proteobacteria, Actinobacteria, and Bacteroidetes were all inhibited. Although the abundance of Proteobacteria increased during denitrification and then decreased during nitrification, the abundance of the major denitrifying genus Thauera spp. was distinct declined, and nitrifying and denitrifying bacteria in Actinobacteria become the dominant species. The major nitrifying genus Nitrosomonas spp. was also unable to survive under high NP accumulation. Meanwhile, these nitrifying and denitrifying functional genera have low correlation with other genera. Above all, the impact of NP on SBR increased the diversity of microbial community, and the accumulation of NP not only inhibited nitrification and denitrification, but also the succession of microbial communities occurred. This study provides theoretical foundation focusing on the diversity, abundance, and community structure of nitrifying and denitrifying bacteria in response to NP accumulation.

Supplementary Information

Below is the link to the electronic supplementary material.

Authors’ contribution

KC and YH conceived of the presented idea. Material preparation, data collection, and analysis were performed by YC. The first draft of the manuscript was written by XY. WX and PL commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. All data generated or analyzed during this study are included in this published article (and its supplementary information files).

Declarations

Conflict of interest

The authors declare no conflicts of interest.