高効率な嫌気・好気逐次型膜分離活性汚泥法（MBR）システムによるエネルギー含有廃水の処理 - パート2：微生物群集の動態

A two-stage anaerobic-aerobic membrane bioreactor (MBR) effectively treated synthetic and real munition constituent wastewaters consisting of various combinations of traditional (2,4,6-trinitrotoluene (TNT), 1,3,5-trinitro-1,3,5-triazine (RDX), octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX)) and insensitive (nitroguanidine (NQ), 2,4-dinitroanizole (DNAN), 3-nitro-1,2,4-triazole-5-one (NTO)) explosives, as well as the oxidants perchlorate (ClO₄^-) and nitrate. Microbial community dynamics in the anaerobic and aerobic MBRs were analyzed. In the anaerobic MBR, Pleomorphomonas spp. and Saccharimonadales dominated and established within the community network. Increased ClO₄^- promoted Pleomorphomonas, Propionibacteriaceae, and Christensenellaceae; these populations also thrived during IMX-101 (NQ, NTO, DNAN) treatment. Switching the main carbon feed to the ANO MBR from fructose to invert sugar increased NQ biotransformation with a corresponding increase of unknown Propionibacteriaceae. Coincident with the feed of actual IMX-104 (RDX, NTO, DNAN) wastewater, a unique group of Rhodococcus erythropolis proliferated, and high levels of xenA, xenB, and nitronate monooxygenase orthologs were detected, which could be associated with degradation of insensitive high explosives and RDX. The aerobic MBR showed high percentages of Proteobacteria as the main phylum. Beijerinckiaceae most frequently dominated and formed a singular submodule, suggesting a unique role in the degradative process. Adding an NQ-oxidizing inoculum increased Reyranella spp., Leucobacter spp., and unknown Beijerinckiaceae which enhanced NQ degradation. Overall, minor microbial community shifts were driven by variations of munition wastewater compositions but the abilities of the MBRs to degrade munitions constituents was maintained even after individual or multiple constituents were absent from the feed for weeks and then re-added. This demonstrates the high resilience of the dual MBR system and may provide insights for performance optimization.