Highlights

4.2.1. Impact of C/N ratio on the microbial community structure in MBBR

Regarding the C/N ratio, several studies on lab-scale MBBR systems treating synthetic wastewater with various C/N ratios reported the dominance of Proteobacteria at phylum level [17], [26], [27], [76]. According to Jia et al. [28], Proteobacteria dominates most biological N-removal systems regardless of the C/N ratio. The authors also observed a gradual increase in abundance of Proteobacteria and Acidobacteria when C/N decreased while Bacteroidetes abundance remained unchanged at different C/N ratios. On the other hand, Pan et al. [27] reported that Proteobacteria slightly decreased while Bacteroidetes and Planctomycetes abundances were unmodified when the C/N ratio dropped from 15 to 5. However, they stated that Nitrospirae and Chloroflexi were more abundant at the C/N ratio of 5, which is the lowest C/N ratio fixed in the study. Correspondingly, Ding et al. [76] stated a similar trend where Nitrospirae was the most abundant (29.33 %) at a C/N ratio of 5.1. Hence, the C/N ratio of 5 is the optimal value that could promote the growth of nitrifying bacteria in aerobic MBBR systems. These findings are coherent because as reported in part 2, the best performance of MBBR systems in terms of TN, ammonia-N and nitrate-N removals were reported at C/N ratios of 5 and 6 [26], [27]. At the class level, Yuan et al. [17] and Pan et al. [27] identified Beta-proteobacteria as the dominant class at diverse C/N ratios. However, Jia et al. [28] recorded the dominance of Alpha-proteobacteria at highest C/N of 7.5 while Gamma-proteobacteria became the most abundant class at the lowest C/N ratio of 3.7.

In a recent work by Xiang et al. [53], they observed the decrease in Shannon, Chao1 and ACE values as the C/N ratio decreased (from 16 to 4) thus indicating that the diminution of C/N ratio affects negatively the diversity and richness of the bacterial community in the biofilm. This could be due to growth inhibition of certain bacteria, mainly heterotrophs, caused by the reduction of carbon concentration in the influent. Indeed, at C/N ratio lower than 5, accumulation of nitrate ions and suppression of ammonium removal are frequently observed due to the limited growth of heterotrophic organisms (e.g., denitrifiers and nitrifiers) [27], [28]. Nevertheless, the abundance of Proteobacteria and Bacteroidetes (>50 %) remained unchanged regardless of the C/N ratio. It was also showed that the increase in ammonia concentrations in synthetic domestic wastewater has only a slight influence on the relative abundance of phyla and almost no significant impact on the predominant phyla [17]. The authors stated that in the case of Proteobacteria, its relative abundance is less affected by the nutrient content but more easily affected by the DO concentration in the influent. On the contrary, J. Wang et al. [22] described Chloroflexi as the most dominant phylum in a lab-scale MBBR fed with synthetic wastewater and operated under different C/N ratios, pH and DO concentrations. The authors stated that Proteobacteria was the second most dominant phylum in all set of experiments except those under different HRTs in which it became the most dominant phylum replacing Chloroflexi.

4.2.2. Interdependence between HRT and the microbial community structure in MBBR

In terms of HRT, Xiong et al. [83] operated a MBBR system treating real mature landfill leachate with extremely long HRT of 10 and 20 days. They noticed that Proteobacteria (52 %) followed by Bacteroidetes (32 %) were the two most dominant phyla whereas Alpha-proteobacteria and Sphingobacteriia were the predominant classes belonging to each phylum respectively. Other works with high HRT (i.e., 32–36 h) corroborate this finding as both aforementioned phyla were identified with the highest relative abundances [29], [84]. Nevertheless, the dominance of a less common phylum, Ignavibacteriae, was observed in the biofilm of an IFAS system under SBR mode treating real pre-treated ammonia rich lagoon supernatant at a HRT of 60 h [77]. The phylum Ignavibacteriae is an organotrophic and facultative anaerobic bacterium that utilizes diverse sources of carbon substrates and electron acceptors for growth [85]. Indeed, a noteworthy proportion of Ignavibacteriae was particularly recorded in anoxic/anaerobic MBBR systems [86], [87]. In a different study by X. Liu and Li [88], a lab-scale anoxic/oxic MBBR system was operated at two HRTs (8, 20 h) and fed with synthetic wastewater to study its impact on the microbial community. The authors reported that the dominant phyla remained unchanged but their relative abundances were modified under different HRTs. It was observed that when the HRT was reduced from 20 to 8 h, the relative abundances of Proteobacteria increased significantly (36.9–74.8 %) in the anoxic reactor but it decreased (from 66.3 % to 44.8 %) in the aerobic tank. Similarly, studies conducted under short HRTs (i.e., 1–2 h) also indicated the dominance of Proteobacteria (>50 %) regardless of the type of influent or the MBBR system configurations [1], [89]. Concerning the alpha-diversity, X. Liu and Li [88] and Mazioti and Vyrides [15] deduced that a higher diversity of microbial population observed during lower HRTs could be due to higher availability of nutrients in the system. Therefore, the alpha-diversity aspects of a bacterial community in MBBR systems are more likely influenced by the C/N ratio (see previous part 4.2.1.).

4.2.3. Influence of aeration conditions on the microbial community structure of MBBR

Another deterministic factor that could affect the microbial community of biomass in MBBR systems is the aeration conditions. For example, Gilbert et al. [90] indicated the highest abundance of Planctomycetes during the entire operation of a lab-scale anoxic MBBR fed with synthetic wastewater containing only ammonium (i.e., without organic carbon). In a different work by Ma et al. [91], the influence of DO concentrations on the biofilm’s microbial composition of two different systems, MBBR and IFAS, were investigated. They observed that the changes of DO levels modified to a certain extent the dominance of phyla in both reactors. In the MBBR, Proteobacteria was the most abundant phylum at intermediate (2.13–3.02 mg/L) and highest (4.31–5.16 mg/L) DO concentrations but Actinobacteria became the most dominant phylum at the lowest DO concentrations (0.71–1.32 mg/L). On the contrary, in the IFAS biofilm, Bacteroidetes was dominant when DO level was the lowest (0.71–1.32 mg/L) and Actinobacteria was dominant at intermediate and highest DO concentrations.

In contrast, in other rare research works involving strict anoxic/anaerobic MBBR systems, Proteobacteria usually dominated the bacterial community followed by Bacteroidetes and Firmicutes as the subdominant phyla [77], [92], [93], [94]. Moreover, in a hybrid dual anaerobic-anoxic/oxic MBBR (D-A² MBBR) system treating real decentralized wastewater from the sewer network of a university, Proteobacteria (43.27–48.61 %) was the dominant phylum followed by Bacteroidetes in all the compartments during the stable operation [31]. The authors recorded the increase in abundance of Bacteroidetes and Chloroflexi in the anaerobic and anoxic reactor. Indeed, the enrichment of both phyla were typically observed in anaerobic processes as they were reported to contain phosphorus accumulation organisms (PAOs) and denitrifying PAOs (DPAOs) [31], [95]. It was also stated that Bacteroidetes was closely related to TN removal specifically for the denitrification in the anoxic reactor [93]. Alpha-proteobacteria, Beta-proteobacteria and Flavobacteriia were the commonly identified dominant classes under anoxic/anaerobic conditions [29], [93], [96]. In addition, according to X. Wang et al. [96], significant relative differences (>60 %) of ACE and Chao1 values for samples from each MBBR chambers were recorded thus distinguishing the microbial community richness of aerobic biofilms from those developed under anoxic condition. Similar findings by J. Li et al. [31] described that the alpha-diversity indexes varied significantly according to the aeration conditions applied in the reactors; the highest and lowest species richness (i.e., ACE and Chao1 indexes) were observed in the aerobic and anoxic MBBR chambers respectively.

4.3.1. Effects of carrier filling ratio on the microbial community structure in MBBR

In terms of carrier filling ratio, Feng et al. [99] evaluated the microbial community structure on polyurethane foam cubic carrier at three filling ratios (20 %, 30 %, 40 %). They reported that at phylum level, Proteobacteria was the most dominant phylum followed by Bacteroidetes and Verrucomicrobia regardless of the filling ratio. Nevertheless, the relative abundance of each phylum varies in a different manner with the filling ratio. For example, Planctomycetes decreased progressively in abundance when the carrier filling ratio increased while the abundance of Proteobacteria firstly increased when the filling ratio increased to 30 % but later decreased when it was fixed at 40 %. Hence, the authors concluded that the packing ratio only influences the abundance of phyla but not their dominance. This observation is corroborated by other studies in which Proteobacteria remained dominant despite the different filling ratios applied, but the latter had noteworthy effects on the bacterial community diversity and richness instead [100], [101]. Indeed, J.-H. Wang et al. [101] indicated that the bacterial community had the highest richness at the lowest filling ratio of 50 % while the highest diversity was attained at the highest filling ratio of 90 %. This finding is logical because at a low filling ratio, bacteria could be better enriched on the carriers because of the limited availability of surface (i.e., lesser carriers for colonization). On the other hand, at a high filling ratio, an improved diversification of bacterial community could be achieved because the attached biofilms are less exposed to shear force produced by the aeration system (i.e., longer contact time between carriers and microorganisms). The latter is possible due to the decrease in efficiency of mixing intensity resulting categorically from the increase in carriers’ quantity.

Besides, Mazioti and Vyrides [15] reported the predominance of Proteobacteria at phylum level and Alpha-proteobacteria at class level in two pilot double-staged MBBR systems, MBBR-A and MBBR-B, filled with porous PE bio-carrier at a ratio of (A) 20 % and (B) 40 % respectively. The authors noticed that instead of the carrier filling ratio, other factors could have influenced the bacterial population since several different sub-dominant phyla (e.g., Firmicutes, Spirochaete) and classes (e.g., Clostridia, Spirochaetes) were detected only in MBBRs placed on the first-stage of the systems. In a different work on a pilot-scale MBMBR operated at two carrier filling ratios (20 %, 35 %), Reboleiro-Rivas et al. [100] described that the community richness and diversity significantly increased at higher filling ratio of 35 %. The authors stated that the two predominant phyla were Proteobacteria (20.9–53.8 %) and Actinobacteria (20.6–57.6 %) in all the experiments. Moreover, the major classes detected were Alpha-proteobacteria (8.1–25.6 %) and Beta-proteobacteria (2.7–26.8 %) from the Proteobacteria phylum, and Acidimicrobiia (1.5–35.35 %) and Actinobacteria (4.5–33.04 %) belonging to the Actinobacteria phylum.

In conclusion, we deduce that the carrier filling ratio has insignificant impact on the dominance of the microbial population at phylum level since Proteobacteria was repeatedly reported to be the most abundant phylum in various reviewed studies. However, several cited articles have stated that the carrier filling ratio has an impact on the relative abundance percentage of the dominant and sub-dominant phyla, but not to the extent of modifying their dominance. Besides, in terms of diversity and richness indexes, we corroborated our deduction made in paragraph 2.4 stating that a carrier filling ratio between 30 % and 50 % has shown to be the most optimal filling ratio in achieving better hydrodynamic conditions and the best MBBR performance. Indeed, the highest microbial richness and diversity of attached biomass were reported in studies with MBBR systems operated at this range of filling ratio. Furthermore, we strongly think that the shear force exerted on the attached biomass due to the hydrodynamic turbulence has an impact on its microbial diversity and richness. It is indeed observed that a lower biofilm thickness induces a higher community diversity and richness and vice versa. This is explained by the higher diversification of the bacterial community when the biofilm is exposed to a higher shear force, thus promoting thinner biofilm (< 400 µm).

4.3.2. Interlinkage between the physicochemical properties of carriers and the microbial community structure in MBBR

As regards the material and form of bio-carriers, Chen et al. [58] observed a higher bacterial richness on plastic K1 carriers than on polyvinyl alcohol gel carriers while L. Zhu et al. [14] noticed the highest microbial richness on polyurethane (PU) sponge carriers compared to classic polypropylene (PP) hollow ball carriers and porous polymer carrier (PPC). Moreover, Zheng et al. [102] reported a more diverse and complex bacterial community in MBBR using lignite activated coke as carrier. Biofilms attached to suspended biochar also demonstrated higher microbial richness and diversity compared to carriers made out of high-density polyethylene (HDPE) [49]. Nevertheless, in terms of bacterial composition at phylum level, the typically observed dominant phylum was Proteobacteria regardless of the carriers materials or geometrical properties, and only their relatives abundances were different from one carrier model to another [47], [49], [51], [58], [102]. Indeed, in a recent work by Jang et al. [103]they also reported the dominance of Proteobacteria and its two major classes, Gamma- and Beta-proteobacteria on both types of carriers (i.e., PE and polyethylene terephthalate (PET)) used in their MBBRs despite the different carriers’ materials. However, the authors reported that a higher abundance of Proteobacteria was recorded on the PE carriers compared to the PET carriers. Hence, they concluded that the high ammonia removal efficiency (>99 %) observed in the MBBR with PE carriers was due to the aforementioned observation.

However, it is worth highlighting that the physical and chemical properties of the carriers exhibited a significant influence on the sub-dominant phyla of the microbial communities. Indeed, several studies comparing MBBR systems filled with different types of carriers recorded the sub-dominance of Actinobacteria [58], Nitrospirae [47], Planctomycetes [51], [102] and Acidobacteria [102]instead of the commonly detected Bacteroidetes. On the other hand, Chen et al. [58] identified Alpha-proteobacteria as the most abundant class on both types of carriers in their MBBR system whereas Zhu et al. [14] observed the abundance of Gamma-proteobacteria on two out of the three types of carriers utilized in their study.

On the contrary, in a work by Abu Bakar et al. [40], Chloroflexi was reported to be the dominant phylum on black polypropylene media (BPM) while Proteobacteria was the dominant phylum on PE hexafilter (HEX) carriers. The authors concluded that different types of polymer used to make the carriers and to a lesser extent their form, could have an influence on the microbial compositions because the carriers were the only highlighted variable in their study. This result is supported by a different study in which Nitrospirae was identified as the dominant phylum on sponge bio-carriers whereas Proteobacteria had the highest relative abundance on AnoxKaldnes™ K5 carriers [4]. Moreover, the authors stated that Nitrospira was the dominant class in both MBBR followed by Beta-proteobacteria and Alpha-proteobacteria. Correspondingly, T. Liu et al. [81] described that Bacteroidetes was the most abundant phylum on the chemically modified HDPE carriers of a lab-scale MBBR system. Chemical substances such as polyquaternium-10 and Fe₂O₃, which enhance the hydrophilicity and electrophilicity of the carriers have seemed to favour the proliferation of Bacteroidetes over Proteobacteria. Although there are no findings to corroborate this observation, we hypothesize that those substances could amplify the affinity between the carriers and Bacteroidetes. Furthermore, in a study conducted by Huang et al. [71] to elucidate the detachment and activity recovery of aging biofilm, the authors observed the dominance of Actinobacteria on plastic bio-carriers that were treated with enzyme and surfactants. The filamentous structure of Actinobacteria and its strong EPS secretion provided them a firm adsorption onto carriers’ surface thus resulting in a higher resistance to detachment during treatments or biofilm sloughing process [70], [71], [104]. Concerning exclusively the form of carriers, biofilms on four different configurations of cylindrical polyethylene carriers were found to be mainly dominated by Proteobacteria (52.2–64.97 %) and Bacteroidetes (7.12–10.82 %) in anoxic lab-scale MBBR systems [87]. Hence, it can be deduced that the form of carriers alone has little to no impact on the dominance of bacterial phyla, but their materials as well as the combination of both elements (i.e., material, form) were described to favour the dominance of certain phyla and classes in MBBR processes.

As regards the physicochemical properties of the carriers, we can distinguish two principal types of carriers’ materials, conventional polymers (i.e., PE, PU, PP, HDPE) and non-conventional materials such as polyvinyl alcohol gel, lignite activated coke, activated carbon and suspended biochar. Upon critical analysis of the different results presented in the cited articles, we can conclude that the majority of the carriers’ materials has inconsequential effects on the dominance of microbial populations at phylum and class level. Indeed, most conventional and non-conventional materials displayed the dominance of Proteobacteria and its corresponding classes, which is logical due to its widely known function in the depollution of wastewater. However, it is worth highlighting that abundance of this dominant Proteobacteria was observed to vary according to the types of the carriers. Based on in-depth analysis, we notice that these variations have a notable impact on the performance of the MBBR systems in terms of pollutants removal. Indeed, we can deduce that MBBR processes yield higher efficiency when a higher abundance of Proteobacteria is detected on its biomass. Furthermore, from certain cited works in this review, we also observe that some carriers can promote the growth of different phyla other than Proteobacteria. For example, it was seen that the sponge-type carriers has a high capacity to retain microorganisms due to its porous structure hence promoting the growth of slow nitrifying bacteria such as Nitrospirae. Therefore, these kind of carriers are highly recommended for MBBR systems designed for ammonium removals. Moreover, these sponge type carriers are often reported to have the highest microbial richness and diversity in the reviewed articles. Globally, the choice of carriers with varying physicochemical properties in MBBR processes has indeed a notable influence on the abundance and in some cases even on the dominance of microbial communities despite the lack of clarity on the proliferation pattern.

4.4.1. Correlation between the type of influent and the microbial community structure in MBBR

A study conducted by An et al. [18] on lab-scale MBBR systems treating real stormwater mixed with different proportions of river water according to the defined strategy showed that Proteobacteria (>70 %) was the dominant phylum and its relative abundance greatly increased at the end of experimental cycles. They stated that the abundance of subdominant phyla such as Bacteroidetes and Acidobacteria decreased after long-term operation due to the frequent change in environments. Furthermore, Proteobacteria (35.1–46.4 %) was also identified as dominant phylum in MBBR processes fed with two different real effluents, raw and ozone-treated oil sands process-affected water (OSPW) collected from a water pumping station [105]. The authors’ major observations were the similar relative abundance of Nitrospirae in both MBBRs (24.6 %, 23.7 %) and the high abundance of Acidobacteria (21.9 %) in the MBBR treating raw OSPW. However, Zhuang et al. [106] reported the dominance of Firmicutes (15.2 %) followed by Bacteroidetes (7.92 %) and Proteobacteria (2.5 %) on attached biomass community of a MBBR co-treating real pre-treated produced water (PW) from a typical shale gas well and synthetic domestic wastewater. On the other hand, a recent work by Chen et al. [107] demonstrated similar dominant microbes among all the samples despite certain variations in terms of relative abundances. They noticed that Proteobacteria was the dominant phylum in the MBBR treating synthetic swine wastewater but Actinobacteria had the highest abundance during the treatment of real swine wastewater. Correspondingly, the dominant classes in synthetic and real swine wastewater system were Gamma-proteobacteria and Actinobacteria respectively.

Moreover, in MBBR processes treating seawater or wastewater with high salinities, commonly identified predominant phyla were Proteobacteria followed by Bacteroidetes and Nitrospirae [108], [109], [110]. On the contrary, the study by Ganesan et al. [111] reported that Firmicutes became the most abundant phylum in a MBBR when the salinity of the synthetic wastewater attained 21.35 g/L NaCl but Proteobacteria regained its dominance at the salinity of 31.53 g/L NaCl. This shows that Proteobacteria is the prevalent species in high salinity water. Besides, several works were carried out in MBBR processes treating synthetic wastewater containing diverse sources of carbon to investigate their influence on the performance of the systems especially on the biological phosphorus removal (BPR) and their microbial community structure. Indeed, the latter could be affected because the change in carbon sources might alter the metabolism pathways of certain microorganisms, thus influencing the taxonomic structure of the bacterial community. It was observed that MBBRs fed with acetate, ethanol or starch were typically dominated by Proteobacteria at varying relative abundance (39–88.92 %) [95], [104]. A different study by M. Zhang et al. [112] corroborated this observation as they recorded the dominance of Proteobacteria (77.6–81.79 %) in MBBR fed with propionate as the carbon source. As regard the phyla playing important role in BPR, Bacteroidetes and Chloroflexi were reported to dominate enhanced BPR systems and include DPAOs [95], [104]. In addition, at the class level, the microbial communities of MBBRs were dominated by Gamma-proteobacteria (20–42 %) and Alpha-proteobacteria (15–32 %) regardless of the carbon sources [95]. According to Peng et al. [113], Proteobacteria was the dominant phylum while Bacteroidetes was the second most abundant phylum in lab-scale MBBR systems treating two different synthetic wastewater (i.e., low C/N ratio, antibiotic).

4.4.2. Influence of MPs on the microbial community structure in MBBR

In a recent work by Atasoy et al. [16], a real effluent of a secondary clarifier was spiked with a mixture of 16 MPs at a concentration of 2 µg/L, and then used as feed for three different bioreactors including a MBBR. The authors reported the abundance of Proteobacteria (24 ± 11 %) followed by Cyanobacteria (23 ± 17 %) and Actinobacteria (12 ± 10 %) on the attached biomass of the MBBR. This finding is supported by various studies in which the dominance of Proteobacteria was notably recorded in MBBR systems fed with wastewater containing MPs [11], [32], [52], [114]. Besides, Song et al. [32] noticed that the presence of pharmaceuticals stimulated the proliferation of Actinobacteria as the second dominant phylum thus indicating that they could withstand the biological toxicity of such products. Correspondingly, Liang et al. [52] reported the enrichment of Actinobacteria and Bacteroidetes with the addition of diverse antibiotics in the MBBR system. In contrast, X. Zhang et al. [114] described the decrease in abundance of Actinobacteria, the third most abundant phylum in their system, when the concentration of MPs increased but opposite trend was observed with Bacteroidetes, particularly in the MBBR spiked with carbamazepine. This indicates that the phylum Bacteroidetes could be particularly resistant to this pharmaceutical molecule while the growth of other phyla were notably inhibited in the presence of carbamazepine. Other reported tendencies at phylum level include the proportional increase in relative abundance of Firmicutes with MPs concentration [32], [115] and the decline of Acidobacteria abundance at higher concentration of MPs [52], [114]. Concerning the microbial community at class level, Gamma-proteobacteria was identified as the most dominant class in MBBR systems treating MPs [52], [74], [115]. Nevertheless, the dominance of certain classes such as Alpha-proteobacteria [11], [116] and Actinobacteria [52] were also observed in other MBBR systems developed to eliminate MPs. In terms of the alpha-diversity of microbial populations, X. Zhang et al. [114] stated that the OTUs increased when MPs concentration increased from 1 mg/L to 2 mg/L but they reduced when the concentration of MPs reached 5 mg/L due to the microorganisms tolerance limit towards MPs. Furthermore, it is noticed that the Shannon index was the lowest (i.e., lowest biodiversity) in MBBRs containing pharmaceuticals at the highest concentration while the highest Chao1 index (i.e., highest microbial richness) was recorded in MBBRs with the intermediate concentration of MPs [32].

To summarize, Proteobacteria is the most commonly detected dominant phylum on the attached biofilm of MBBR processes regardless of the nature of influent and the type of external carbon source added into the influent. This is highly likely due to its excellent adaptability to constant changes in environment and its high resistance to influent with high salinity or oligotrophic characteristic. Nevertheless, it is worth highlighting that the composition of a real swine wastewater has favoured the proliferation of Actinobacteria over Proteobacteria while a real pre-treated PW from a shale gas well has promoted the abundant growth of Firmicutes instead of Proteobacteria. As regards the real swine wastewater, we believe that the presence of pharmaceutical products such as antibiotics, which are generally used in livestock industry, could have enhanced the dominance of Actinobacteria. Furthermore, this hypothesis is corroborated in other reviewed articles because Actinobacteria was abundantly detected on fixed biofilms and its abundance has been positively altered in MBBR systems treating wastewater with MPs. We also observe a similar trend for Bacteroidetes in such systems thus indicating that some MPs could improve its growth while inhibiting other phyla. However, more studies need to be executed to confirm the effect of MPs on these aforementioned phyla. It would also be interesting to evaluate the impact of one specific MP, instead of a cocktail of MPs, on the microbial communities in order to distinguish and validate the effect (i.e., inhibition or enhancement) of that particular MP on the bacterial growth. At class level, we can deduce that Gamma-proteobacteria and Actinobacteria dominated the majority of the cited works. Once again, we could correlate this findings to the fact that both classes are either resistant to changes in composition of influents or favour the presence of atypical pollutants such as pharmaceutical compounds for its proliferation. Nonetheless, at this state of the research, we are still unable to firmly conclude on the participation of these bacteria in the degradation of such MPs. As regards the alpha-diversity aspects, it can be inferred that a very high concentration of MPs (> 5 mg/L) negatively affects the microbial diversity and richness. Therefore, it is important to monitor the concentration of MPs in a bioreactor because it could disrupt with the development of functional microorganisms.