3.3. Changes in microbial community structure
Bacterial and archaeal community structures were investigated by DGGE and phylogenetic analysis. We
especially focused on the changes in bacterial community structure during the initial period of each cycle
where the most dynamic growth of bacterial population occurred along with the most accumulation of
acidogenic products. Acidogenic period was defined in each cycle as the initial incubation period until the
total accumulation of acidogenic products reaches its maximum: 1.3 d in C1, 0.5 d in C2 and 0.5 d in C3 (Fig.
1). Eight bands of interest were excised from the bacterial DGGE gel and 7 of them were closely related
(>97% sequence similarity) to Clostridum species producing organic acids from carbohydrate, amino acids,
and/or H2/CO2. Among a total of 6 bands retrieved from the archaeal DGGE gel, three were affiliated with a
hydrogenotrophic genus Methanospirillum belonging to of the order Methanomicrobiales. The remaining
three were affiliated with Methanosaeta, Methanosarcina, and Methanobacterium genera, respectively. These
results suggested Methanospirillum populations as the dominant methanogens in the reactor. NMS plot
showed significant changes in bacterial community structure over repeated cycles (Fig. 2). It was evident that
the temporal variations within a cycle decreased significantly (i.e., closely clustered points) over repeated
cycles. Corresponding to this, the similarity between bacterial community profiles (Sorensen distance measure)
increased over cycles from 93% in C3, indicating that the acidogen community structure was
stabilized with minimal changes after three cycles. Such shifts may be ascribed to the acclimation of bacterial
community to the reactor conditions. The changes in bacterial community structure, together with the
increased bacterial population size with cycle number, may have contributed to the enhanced acidogenic
reaction rate over cycles (Fig. 1). On the other hand, methanogen community structure little changed during
the experiment (data not shown). All 6 DGGE profiles (i.e., 2 analysis points per cycle) shared a high
similarity of >90%, indicating that the methanogen community structure formed in C1 remained stable. These
suggest that the dynamic changes in bacterial community structure over cycles had no significant influence on
the formation of methanogen community structure. This may reflect much less diverse and dynamic nature of
methanogens as compared to bacteria in anaerobic environments [11], which may be related to the very
narrow substrate range of methanogens [12]. Our results suggest that the increase in methanogenesis rate with
cycle number was more likely due to the increase in methanogen abundance rather than the alteration of
community structure.
Acknowledgement
This work was supported by the 2012 Research Fund of Ulsan National Institute of Science and
Technology (UNIST) through a Future Challenge Project.