Evaluation of granulation processes in upflow anaerobic sludge blanket reactors using oligonucleotide probe hybridizations

The potentials for Methanosaeta concilii and propionate-degrading syntrophic consortia to serve as nuclei for granulation were evaluated by monitoring granulation processes starting from non-granular sludge in two laboratory-scale upflow anaerobic sludge blanket reactors (Reactor G and Reactor GP). New probes for Methanosaeta spp. and Syntrophobacter spp. were developed. The influent of Reactor G contained glucose as the only energy source, while Reactor GP was fed glucose and propionate. The two reactors were started following the general upflow anaerobic sludge blanket reactor startup guidelines and the effluent acetate concentrations were controlled below 200 mg/l. Quantitative membrane hybridizations and fluorescent in situ hybridizations were used to monitor changes in microbial population levels and cell aggregate structures during the granulation processes. Methanosaeta concilii demonstrated good settling ability and a high abundance in the microbial communities. Increases in their levels were associated with significant increases in cell aggregate size during the early stage of granulation. Methanosaeta concilii cells were found to serve as backbones in the small cell aggregates with other cells of archaea and bacteria attached to them, and they remained dominant in larger cell aggregates and the mature granules. These observations support the hypothesis that Methanosaeta cells can serve as nuclei for granulation. Syntrophobacter and Desulfobulbus, on the other hand, exhibited poor settling ability and could be easily washed out from the system. They were not suitable as nuclei but may enhance granulation processes by immobilizing on the nuclei. The ribosomal RNA levels of Syntrophobacter and Desulfobulbus were only slightly higher in Reactor GP than in Reactor G even though the biomass from Reactor GP exhibited a much higher propionate-degrading metabolic activity than the biomass from Reactor G. This may indicate the existence of unknown syntrophic propionate-oxidizing bacteria that are not detectable by the probes used here. Alternatively, it is possible that the syntrophic propionate-oxidizing bacteria in Reactor GP are much more active than those in Reactor G even though their rRNA levels were slightly different. Nevertheless, it was demonstrated that the microbial community with high propionate-degrading ability was more stable and recovered faster from long-term overloading than the one with low propionate-degrading ability.