Bacterial assembly in the bio-cake of membrane bioreactors: Stochastic vs. deterministic processes

Abstract

Much about assembly processes dictating bio-cake microbiota remains uncertain, leading to poor understanding of membrane biofouling in membrane bioreactors (MBRs). This work aimed to reveal the underlying mechanisms driving bio-cake community during the biofouling process under different flux conditions. On the basis of 16S rRNA sequences, the results showed that bacterial diversity decreased with increasing fouling. Additionally, low-flux bio-cake (8 LMH) communities harbored much lower diversity than high-flux (16 LMH) bio-cake microbiomes. Ecological null model analyses and phylogenetic molecular ecological networks (pMENs) revealed that environmental filtering deterministically governed low-flux bio-cake communities. In contrast, high-flux bio-cake communities were mainly shaped in a stochastic manner. This is likely due to the higher stochastic deposition of bacterial taxa from bulk sludge because of the presence of a stronger drag force. Moreover, by lowering the flux, the interactions between bacterial lineages were enhanced; this is evidenced by the greater number of links, the higher average degree, and the higher average clustering coefficients within the pMENs in low-flux bio-cakes than those in high-flux bio-cakes. Most keystone fouling-related taxa in low-flux bio-cakes were motile and involved in nitrate reduction and polysaccharide/protein metabolism. This corroborated the important role of environmental filtering in the assembly process dictating low-flux bio-cake formation. Some low-abundance taxa were observed to be key fouling-related bacteria under both flux conditions, indicating that a few populations play paramount ecological roles in triggering biofouling. In summary, our findings clearly indicate distinct bio-cake community assembly patterns under different operational conditions and highlight the importance of developing specialized strategies for fouling control in individual MBR systems.