Bacteria capable of complete ammonia oxidation (comammox) are widespread and contribute to nitrification in wastewater treatment facilities. However, their roles in partial nitrification–anaerobic ammonium oxidation (anammox) systems remain unclear. In this study, a bench-scale bioreactor with continuous stirring was operated for more than 1000 days with limited oxygen supply to achieve efficient nitrogen removal (70.1 ± 2.7%) at a low ammonium loading of 35.2 mg-N/L/day. High-throughput amplicon sequencing analysis of the comammox ammonia monooxygenase subunit A (amoA) gene revealed seven sequence types from two clusters in clade A of comammox Nitrospira. Quantitative polymerase chain reaction analyses suggested that the comammox species dominated the ammonia-oxidizing community, with an abundance as high as 89.2 ± 7.9% in total prokaryotic amoA copies. Multiple linear regression further revealed the substantial contribution of the comammox Nitrospira to ammonia oxidation in the bioreactor. The investigation with bioreactor and batch experiments consistently showed that activities of comammox Nitrospira were inhibited by free ammonia far more severely than other ammonia-oxidizing microbes. Overall, this study provided new insight into the ecology of comammox Nitrospira under hypoxic conditions and suggested comammox-associated partial nitrification–anammox as a potential method for treating low-strength ammonium-containing wastewater.

Performance of a bioreactor is affected by complex microbial consortia that regulate system functional processes. Studies so far, however, have mainly emphasized the selective pressures imposed by operational conditions (i.e., deterministic external physicochemical variables) on the microbial community as well as system performance, but have overlooked direct effects of the microbial community on system functioning. Here, using a bioreactor with ammonium as the sole substrate under controlled operational settings as a model system, we investigated succession of the bacterial community after a disturbance and its impact on nitrification and anammox (anaerobic ammonium oxidation) processes with fine-resolution time series data. System performance was quantified as the ratio of the fed ammonium converted to anammox-derived nitrogen gas (N2) versus nitrification-derived nitrate (npNO3−). After the disturbance, the N2/npNO3− ratio first decreased, then recovered, and finally stabilized until the end. Importantly, the dynamics of N2/npNO3− could not be fully explained by physicochemical variables of the system. In comparison, the proportion of variation that could be explained substantially increased (tripled) when the changes in bacterial composition were taken into account. Specifically, distinct bacterial taxa tended to dominate at different successional stages, and their relative abundances could explain up to 46% of the variation in nitrogen removal efficiency. These findings add baseline knowledge of microbial succession and emphasize the importance of monitoring the dynamics of microbial consortia for understanding the variability of system performance. IMPORTANCE Dynamics of microbial communities are believed to be associated with system functional processes in bioreactors. However, few studies have provided quantitative evidence. The difficulty of evaluating direct microbe-system relationships arises from the fact that system performance is affected by convolved effects of microbiota and bioreactor operational parameters (i.e., deterministic external physicochemical forcing). Here, using fine-resolution time series data (daily sampling for 2 months) under controlled operational settings, we performed an in-depth analysis of system performance as a function of the microbial community in the context of bioreactor physicochemical conditions. We obtained statistically evaluated results supporting the idea that monitoring microbial community dynamics could improve the ability to predict system functioning, beyond what could be explained by operational physicochemical variables. Moreover, our results suggested that considering the succession of multiple bacterial taxa would account for more system variation than focusing on any particular taxon, highlighting the need to integrate microbial community ecology for understanding system functioning.