Partial bioaugmentation to remove 3-chloroaniline slows bacterial species turnover rate in bioreactors

Bioaugmentation is a potentially powerful tool to direct community structure and metabolic capacities in bioreactors. Yet the outcome of bioaugmentation studies is usually unpredictable and effects on microbial community dynamics are poorly understood. We asked the question whether bioaugmentation could prevent a diversity shift induced by a model toxin, 3-chloroaniline (3-CA), regardless of whether 3-CA was degraded. Four replicate membrane bioreactors (MBRs) operating in parallel were amended with Pseudomonas putida UWC3 (pWDL7::rfp), a strain that carries the upper pathway genes necessary for partial degradation of 3-CA on its plasmid. Two MBAs served as controls and two MBRs were exposed to 3-CA for 71 days. Despite the selective pressure imposed by 3-CA, there was little or no 3-CA removal and neither the 16S rRNA gene of the augmented strain UWC3 nor the plasmid pWDL7::rfp proliferated in any of the reactors. Yet both host strain and plasmid were maintained at reduced levels (similar to 10(4) host strain cells ml(-1)) in all reactors compared to the initial inoculum (similar to 10(7) cells ml(-1); 1% of active cells). Additionally, the microbial community dynamics were evaluated for each MBR via terminal restriction fragment length polymorphism (T-RFLP) analysis (n = 15 per reactor) that targeted a portion of the 16S rRNA gene. Analysis comprised of a suite of multivariate statistics coupled with a theoretical microbial ecological approach, 'Island Biogeography', using a bacterial species time relationship (STR), within each MBR. Control MBRs had a wider range in w values than the treatment MBRs, which is attributed to the lack of a toxin selecting for biota that can withstand its toxic nature. Bioaugmentation alone strongly slowed the bacterial species turnover rate (as revealed by very low w scaling components), compared to non-bioaugmented reactors from a previous study, but did not protect the microbial community from a diversity shift caused by the toxin. Nonmetric multidimensional scaling (NMDS) analysis revealed that treatment MBRs diverged away from the control MBRs after the first 11 days, whereas control MBRs remained clustered. Individual reactors were analyzed by multi-response permutation procedures (MRPP) and a significant difference was found between each control MBR and the treatment MBRs. The study suggests that newly introduced strains can gain a foothold in established microbial communities even at low cell concentrations (about 1% of introduced concentration within the first week) regardless of selective pressure, whereas community dynamics are more affected by the presence of a selector toxin. (C) 2013 Elsevier Ltd. All rights reserved.