Effects of dispersal on species distribution, abundance, diversity and interaction in a bacterial biofilm metacommunity.

AIMS: Dispersal effects on biofilms have not been adequately studied despite their strong potential impacts on biofilm development. We investigated the effects of dispersal on biofilm metacommunity. METHODS AND RESULTS: A bacterial consortium was allowed to form biofilms on 12 glass beads attached to disposable plates (compartmentalized or not), and biofilms were scrutinized on days 5, 10 and 15 using quantitative PCR and MiSeq sequencing. Biofilm population density was lesser by 2 orders of magnitude on day 5 when dispersal was allowed (p < 0.05). Then, the population rapidly increased by 4.4 orders with dispersal (p < 0.05) but did not change without dispersal. Community analyses revealed that dispersal increased the species diversity at all sampling times (p < 0.05). Dispersal affected the community structure and increased the homogeneity of local communities (p < 0.05). Distance-decay analysis showed that dispersal reduced the dissimilarity among local communities at all distance levels. Furthermore, dispersal reduced the variability of diversity, population and community structure. Network analysis revealed that dispersal increased the clustering coefficient, network density and connectivity. CONCLUSIONS: Dispersal increased the species diversity, population and interaction and reduced the variability of the diversity, population and structure among local communities. SIGNIFICANCE AND IMPACT OF STUDY: Our results suggest that dispersal can induce the niche complementarity and mass effects.

Spatial Variance of Species Distribution Predicts the Interspecies Interactions within a Microbial Metacommunity.

Interspecies interactions have a profound influence on spatial distribution of coexisting microbial species. We explored whether spatial variance of species distribution (SVSD) predicts the degree of interspecies interactions within a microbial metacommunity. Simulations were used to determine the relationships from random, lake, soil, and biofilm metacommunity datasets (1,000 times). All of the bacterial datasets showed a negative correlation between the habitat breadth (inverse to SVSD) and the numbers of total, positive, and negative interspecies interactions (P < 0.05); the only exception was the relationship between habitat breadth and negative interactions in the biofilm dataset. The random dataset had no significant relationships (P > 0.05). We repeated the simulations to determine the degree of correlation and reproducibility (100 times). Habitat breadth was negatively correlated with the total and positive interactions in all of the real datasets (P < 0.05), and the negative relationships persisted across repetitions. Despite variability in the slope of total interactions, the slope values of positive interactions were similar for the real datasets (- 19.9, - 19.2, and - 25.8 for lake, soil, and biofilm, respectively). In conclusion, our results demonstrate the patterns of species interaction-distribution and show that interspecies interactions are positively correlated with the SVSD.

Effects of Plants on Metacommunities and Correlation Networks of Soil Microbial Groups in an Ecologically Restored Wetland.

Plants may influence different aspects of the belowground microorganisms, including abundance, distribution, and interaction, in wetlands. Microbial communities were scrutinized in a 4-year-old restored wetland ecosystem with 5 distinct sites: a bare-soil site (10 local patches) and sites dominated by Miscanthus, Phragmites, Typha, and Zizania (20 patches per site). Ordination analysis revealed that plant-induced attributes (e.g., organic matter and total carbon and nitrogen) could explain the total environmental variance. Community comparisons showed that all groups (Bacteria, Fungi, Protista, and Metazoa) differed in community structure among the 5 sites (P < 0.05). Comparisons between the community and environmental ordination plots revealed that community structural variation among the sites correlated with the environmental change across all groups (R2 ≥ 0.61). This indicates that all groups were primarily influenced by plant detritus. In addition, correlation networks markedly varied in topology and composition among the sites across all groups. There was a strong coupling between the metacommunity and correlation network for both Bacteria and Fungi (R2 ≥ 0.58), indicating that the plants determined the spatial covariation patterns of microbial populations. Multi-group networks and group synchrony results revealed that Bacteria, Fungi, and Protista were synchronized with each other (R2 ≥ 0.52) as the key founders of the microbial systems, while Metazoa participated in the system only under Miscanthus. Our findings concluded that the plants shaped the communities by controlling the abundance and interaction of their populations.

Coordinated Metacommunity Assembly and Spatial Distribution of Multiple Microbial Kingdoms within a Lake.

Freshwater planktonic communities comprise a tremendous diversity of microorganisms. This study investigated the distribution patterns of microbial kingdoms (bacteria, fungi, protists, and microbial metazoans) within a lake ecosystem. Water samples were collected from 50 sites along the shoreline in a lake during an early eutrophication period, and MiSeq sequencing was performed with different marker genes. Metacommunity analyses revealed a bimodal occupancy-frequency distribution and a Clementsian gradient persisting throughout all microbial kingdoms, suggesting similar regional processes in all kingdoms. Variation partitioning revealed that environmental characteristics, macrophyte/macroinvertebrate composition, space coordinates, and distance-based Moran's eigenvector maps (dbMEM) together could explain up to 29% of the community variances in microbial kingdoms. Kingdom synchrony results showed strong couplings between kingdoms (R2 ≥ 0.31), except between Fungi and Metazoa (R2 = 0.09). Another variation partitioning revealed that microbial kingdoms could well explain their community variances up to 73%. Interestingly, the kingdom Protista was best synchronized with the other kingdoms. A correlation network showed that positive associations between kingdoms outnumbered the negative ones and that the kingdom Protista acted as a hub among kingdoms. Module analysis showed that network modules included multi-kingdom associations that were prevalent. Our findings suggest that protists coordinate community assembly and distribution of other kingdoms, and inter-kingdom interactions are a key determinant in shaping their community structures in a freshwater lake.

Effects of Quorum Quenching on Biofilm Metacommunity in a Membrane Bioreactor.

Quorum quenching (QQ) has received attention for the control of biofilms, e.g., biofilms that cause biofouling in membrane bioreactors (MBRs). Despite the efficacy of QQ on biofouling, it is elusive how QQ influences biofilm formation on membranes. A pilot-scale QQ-MBR and non-QQ-MBR were identically operated for 4 days and 8 days to destructively sample the membranes. QQ prolonged the membrane filterability by 43% with no harmful influence on MBR performance. qPCR showed no effect of QQ on microbial density during either of these time periods. Community comparisons revealed that QQ influenced the bacterial and fungal community structures, and the fungal structure corresponded with the bacterial structure. Metacommunity and spatial analyses showed that QQ induced structural variation rather than compositional variation of bacteria and fungi. Moreover, QQ considerably enhanced the bacterial dispersal across membrane during the early development. As the dispersal enhancement by QQ counteracted the ecological drift, it eliminated the distance-decay relationship, reflecting a neutral theory archetype of metacommunity. Network analyses showed that QQ substantially reduced the amount and magnitude of interactions, e.g., competition and cooperation, for bacteria and fungi, and weakened their network structures, irrespective of time. Additionally, QQ suppressed the growth of specific microbial species (e.g., Acinetobacter), abundant and widespread at the early stage. These findings suggest that QQ influenced the community dynamics at the regional and local levels, correspondingly the ecological selection and dispersal processes, during the biofilm development.

Enhancement of electricity production in a mediatorless air-cathode microbial fuel cell using Klebsiella sp. IR21.

A novel dissimilatory iron-reducing bacteria, Klebsiella sp. IR21, was isolated from the anode biofilm of an MFC reactor. Klebsiella sp. IR21 reduced 27.8 % of ferric iron to ferrous iron demonstrating that Klebsiella sp. IR21 has electron transfer ability. Additionally, Klebsiella sp. IR21 generated electricity forming a biofilm on the anode surface. When a pure culture of Klebsiella sp. IR21 was supplied into a single chamber, air-cathode MFC fed with a mixture of glucose and acetate (500 mg L(-1) COD), 40-60 mV of voltage (17-26 mA m(-2) of current density) was produced. Klebsiella sp. IR21 was also utilized as a biocatalyst to improve the electrical performance of a conventional MFC reactor. A single chamber, air-cathode MFC was fed with reject wastewater (10,000 mg L(-1) COD) from a H2 fermentation reactor. The average voltage, current density, and power density were 142.9 ± 25.74 mV, 60.5 ± 11.61 mA m(-2), and 8.9 ± 3.65 mW m(-2), respectively, in the MFC without inoculation of Klebsiella sp. IR21. However, these electrical performances of the MFC were significantly increased to 204.7 ± 40.24 mV, 87.5 ± 17.20 mA m(-2), and 18.6 ± 7.23 mW m(-2), respectively, with inoculation of Klebsiella sp. IR21. The results indicate that Klebsiella sp. IR21 can be utilized as a biocatalyst for enhancement of electrical performance in MFC systems.

Development of droplet digital PCR assays for methanogenic taxa and examination of methanogen communities in full-scale anaerobic digesters.

Droplet digital PCR (ddPCR) is a new DNA quantification platform without an external DNA calibrator. This study examined methanogen communities in four full-scale anaerobic digesters treating municipal sewage sludge, using ddPCR with taxon-specific primer/TaqMan probe sets (5 orders, 11 families, and 13 genera), many of which were developed in this study. Total methanogen abundance was positively correlated with hydraulic retention time (HRT) and temperature (p < 0.05), though the effect of HRT was stronger (r = 0.864 vs. 0.682, respectively). Moreover, total abundance was strongly correlated with biogas production rate (r = 0.896). HRT was positively correlated with seven methanogenic taxa, while temperature was positively or negatively correlated with 13 taxa (p < 0.05). For instance, the predominant genera Methanosaeta and Methanosarcina were negatively and positively associated, respectively, with temperature only (p < 0.05). Redundancy analysis and principal component analysis using the absolute-abundance dataset indicated that only temperature explained the variability in the methanogen communities at all classification levels. Therefore, HRT was the most important operational factor to influence net methanogen abundance and activity, while temperature governed the composition of the methanogen community. ddPCR enabled absolute quantification of methanogens without the external DNA standards and linked methanogen communities and operational factors, suggesting that it is a promising tool for analyzing the microbial ecology of anaerobic digestion.

The close relation between Lactococcus and Methanosaeta is a keystone for stable methane production from molasses wastewater in a UASB reactor.

The up-flow anaerobic sludge blanket (UASB) reactor is a promising method for the treatment of high-strength industrial wastewaters due to advantage of its high treatment capacity and settleable suspended biomass retention. Molasses wastewater as a sugar-rich waste is one of the most valuable raw material for bioenergy production due to its high organic strength and bioavailability. Interpretation for complex interactions of microbial community structures and operational parameters can help to establish stable biogas production. RNA-based approach for biogas production systems is recommended for analysis of functionally active community members which are significantly underestimated. In this study, methane production and active microbial community were characterized in an UASB reactor using molasses wastewater as feedstock. The UASB reactor achieved a stable process performance at an organic loading rate of 1.7~13.8-g chemical oxygen demand (COD,·L(-1) day(-1); 87-95 % COD removal efficiencies), and the maximum methane production rate was 4.01 L-CH4·at 13.8 g-COD L(-1) day(-1). Lactococcus and Methanosaeta were comprised up to 84 and 80 % of the active bacterial and archaeal communities, respectively. Network analysis of reactor performance and microbial community revealed that Lactococcus and Methanosaeta were network hub nodes and positively correlated each other. In addition, they were positively correlated with methane production and organic loading rate, and they shared the other microbial hub nodes as neighbors. The results indicate that the close association between Lactococcus and Methanosaeta is responsible for the stable production of methane in the UASB reactor using molasses wastewater.

Fate and behavior of extended-spectrum ß-lactamase-producing genes in municipal sewage treatment plants.

Extended-spectrum ß-lactamases (ESBLs) have the capability of hydrolyzing a variety of the newer ß-lactam antibiotics, including the third-generation cephalosporins and monobactams known as a rapidly evolving group of ESBLs. The purpose of this study was to investigate the occurrence and fate of ß-lactamase producing genes (CTX-M type 1, type2, CTX-M probe for all groups except CTX-M-1, and TEM, SHV, OXA) through wastewater treatment utilities. ß-lactamase producing genes in influent, digested sludge, activated sludge, and disinfected effluent were monitored. The results showed that influent contained high level of all target genes, and all CTX-M types, SHV, and OXA gene decreased significantly in biological treatment process such as activated sludge process and anaerobic digestion, however, TEM type was not effectively eliminated. Possibly, host microbes of TEM could be most resistant in target genes or to some extent gene transfer occurred in wastewater treatment processes. All target genes were significantly reduced during disinfection. Consequently, wastewater treatment process apparently reduced host microbes carrying ß-lactamase producing genes effectively, although they are selectively removed in biological processes. In addition, the significant reduction during disinfection was shown, although slightly differences of removal efficiency were observed in resistance.

Suppression of methanogenesis in hydrogen fermentation by intermittent feeding.

This study investigated whether intermittent feeding by using a concentrated carbon source is an appropriate method for selective enrichment of hydrogenesis by means of methanogen suppression. In a conventional reactor fed continuously for 10 d, methanogens increased from 2.8 × 10(7) to 1.1 × 10(9) gene copy number (GCN)/mg-cell dry weight, and methane concentration in the resulting biogas was 5.8%. However, when a carbon source was intermittently supplied for 10 d to the reactor, the number of methanogens was reduced 98.9% from 2.77 × 10(7) to 1.2 × 10(3) GCN/mg-cell dry weight, and methane was not detected during this period of intermittent feeding. Intermittent feeding shifted the dominants in the reactor from Clostridiaceae (70.5%) and Lactobacillaceae (11.0%) to Acetobacteraceae (62.0%) and Clostridiaceae (38.0%). In the reactor operated in continuous feeding mode after intermittent feeding, methane concentration was below 0.3% and the portion of methanogens in the bacterial community was maintained below 0.2%. These results suggest that the intermittent feeding of a carbon source during hydrogen production processes is a suitable method to suppress the activity of methanogens.

Effects of granular activated carbon on methane removal performance and methanotrophic community of a lab-scale bioreactor.

Two identical lab-scale bioreactor systems were operated to examine the effects of granular activated carbon (GAC) on methane removal performance and methanotrophic community. Both bioreactor systems removed methane completely at a CH4 loading rate of 71.2 g-CH4·d(-1) for 17 days. However, the methane removal efficiency declined to 88% in the bioreactor without GAC, while the bioreactor amended with GAC showed greater methane removal efficiency of 97% at a CH4 loading rate of 107.5 g-CH4·d(-1). Although quantitative real-time PCR showed that methanotrophic populations were similar levels of 5-10 × 10(8) pmoA gene copy number·VSS(-1) in both systems, GAC addition changed the methanotrophic community composition of the bioreactor systems. Microarray assay revealed that GAC enhanced the type I methanotrophic genera including Methylobacter, Methylomicrobium, and Methylomonas of the system, which suggests that GAC probably provided a favorable environment for type I methanotrophs. These results indicated that GAC is a promising support material in bioreactor systems for CH4 mitigation.

Long-term survival of methanogens of an anaerobic digestion sludge under starvation and temperature variation.

To investigate starvation effect on methanogen community, two identical membrane reactors were continuously operated for 84 consecutive days, with a temperature change from 50 degrees C to 20 degrees C. Continuous feeding washed out 97% biomass from reactors during the experimental period. Quantitative PCR, using mcrA genes, indicated that the methanogen abundance decreased from 7.0 x 10(7) to 1.2 x 10(7) mcrA copies ml(-1) (volume basis) at 50 degrees C, and then increased to 4.4 x 10(7) mcrA copies ml(7) at 20 degrees C (p<0.05). Correspondence analysis indicated that methanogen communities were distinctly grouped by each temperature. Canonical correspondence analysis indicated that temperature showed a significant correlation with the methanogen community composition. These results suggest that methanogens can survive for a long time (at least more than 84 days) under starvation conditions, and that temperature could be a primary factor determining the density and community of methanogens.

Comparison of droplet digital PCR and quantitative real-time PCR for examining population dynamics of bacteria in soil.

The newly developed droplet digital PCR (DD-PCR) has shown promise as a DNA quantification technology in medical diagnostic fields. This study evaluated the applicability of DD-PCR as a quantitative tool for soil DNA using quantitative real-time PCR (qRT-PCR) as a reference technology. Cupriavidus sp. MBT14 and Sphingopyxis sp. MD2 were used, and a primer/TaqMan probe set was designed for each (CupMBT and SphMD2, respectively). Standard curve analyses on tenfold dilution series showed that both qRT-PCR and DD-PCR exhibited excellent linearity (R (2) = 1.00) and PCR efficiency (≥92 %) across their detectable ranges. However, DD-PCR showed a tenfold greater sensitivity than qRT-PCR. MBT14 and MD2 were added to non-sterile soil at 0 ~ 5 × 10(8) and 0 ~ 5 × 10(7) cells per gram of soil, respectively (n = 5). This bacterial load test indicated that DD-PCR was more sensitive and discriminating than qRT-PCR. For instance, DD-PCR showed a gradual DNA increase from 14 to 141,160 MBT14 rDNA copies µL DNA extract(-1) as the bacterial load increased, while qRT-PCR could quantify the DNA (6,432 copies µL DNA(-1)) at ≥5 × 10(5) MBT14 per gram of soil. When temporal DNA changes were monitored for 3 weeks in the amended soils, the two technologies exhibited nearly identical changes over time. Linearity tests (y = a · x) revealed excellent quantitative agreement between the two technologies (a = 0.98, R (2) = 0.97 in the CupMBT set and a = 0.90, R (2) = 0.94 in the SphMD2 set). These results suggest that DD-PCR is a promising tool to examine temporal dynamics of microorganisms in complex environments.

Functional rigidity of a methane biofilter during the temporal microbial succession.

Temporal microbial succession was investigated in relation to the performance of a methane biofilter. A laboratory-scale biofilter packed with perlite was operated for 108 days, without a deliberate biomass control. The system performance was stable over the period with a mean elimination capacity of 1,563 g m(-3) day(-1), despite a temporal deterioration (45-56 days). Ribosomal-tag pyrosequencing showed that bacterial communities at days 14-28 were distinct from those of days 68-108. The accumulation of nonviable substances strongly coincided with the community change (R (2) > 0.97). Rhodobacter, Hydrogenophaga, and Methylomonas were dominated in the earlier period, while Methylocaldum and Methylococcus were abundant in the later period. The methanotrophic proportion gradually increased to 41 %, and type I methanotrophs became predominant over time. However, community structure and methanotrophic population density stably retained over time, allowing the system to keep the similar performance. Therefore, the perlite biofilter system was functionally rigid against the temporal microbial succession.

Effects of water temperature and backwashing on bacterial population and community in a biological activated carbon process at a water treatment plant.

Bacterial community dynamics was examined in an actual biological activated carbon (BAC) process for four consecutive seasons, using quantitative polymerase chain reaction and pyrosequencing. The BAC stably removed organic carbons for the period, although the water temperature substantially varied over the study period. Neither the population density nor community organization was correlated with time and temperature. However, the similarity degree between communities significantly reduced with time and temperature differences. Community analyses indicated that the community evolved over time, resulting in four distinct groups, and that the abundances of particular bacteria were significantly correlated with time and temperature, as well as their interaction. Additionally, backwashing did not affect the BAC bacterial population, community organization (diversity, evenness, and richness), or composition, although backwashing dislodged a large number of bacteria from the BAC (≈10(15) · m(-3)). These results suggest that water temperature is an important factor driving community dynamics and that backwashing is a harmless management option for biomass control.

Tobermolite effects on methane removal activity and microbial community of a lab-scale soil biocover.

Three identical lab-scale biocovers were packed with an engineered soil (BC 1), tobermolite only (BC 2), and a mixture of the soil and tobermolite (BC 3), and were operated at an inlet load of 338-400 g-CH4 m(-2) d(-1) and a space velocity of 0.12 h(-1). The methane removal capacity was 293 ± 47 g-CH4 m(-2) d(-1) in steady state in the BC 3, which was significantly higher than those in the BC 1 and BC 2 (106 ± 24 and 114 ± 48 g-CH4 m(-2) d(-1), respectively). Quantitative PCR indicated that bacterial and methanotrophic densities (6.62-6.78 × 10(7) 16S rDNA gene copy number g-dry sample(-1) and 1.37-2.23 × 10(7) pmoA gene copy number g-dry sample(-1) in the BC 1 and BC 3, respectively) were significantly higher than those in the BC 2. Ribosomal tag pyrosequencing showed that methanotrophs comprised approximately 60 % of the bacterial community in the BC 2 and BC 3, while they only comprised 43 % in the BC 1. The engineered soil favored the growth of total bacteria including methanotrophs, while the presence of tobermolite enhanced the relative abundance of methanotrophs, resulting in an improved habitat for methanotrophs as well as greater methane mitigation performance in the mixture. Moreover, a batch experiment indicated that the soil and tobermolite mixture could display a stable methane oxidation level over wide temperature (20-40 °C, at least 38 µmol g-dry sample(-1) h(-1)) and pH (5-8, at least 61 µmol g-dry sample(-1) h(-1)) ranges. In conclusion, the soil and tobermolite mixture is promising for methane mitigation.

Two-stage biogas production by co-digesting molasses wastewater and sewage sludge.

We evaluated the feasibility of co-digesting molasses wastewater and sewage sludge in a two-stage hydrogen- and methane-producing system. The highest energy was recovered at the 21-h hydraulic retention time (HRT) of the first hydrogenic reactor and at 56-h HRT of the secondary methanogenic reactor. Hence, the two-stage system recovered 1,822 kJ from 1 L of the mixed wastes (19.7: hydrogenic reactor plus, 1,802 kJ L(-1): methanogenic reactor). Despite the overloaded VFA-run with a short HRT of 56 h, the GAC-CH4 reactor increased methane production rate and yields due to enhanced pH buffer capacity. An RNA-based community analysis showed that the Ethanoligenens and Methanosaeta dominated the hydrogen and methane bioreactor, respectively. The two-stage system of co-digesting molasses and sewage sludge is particularly cost-effective due to non-pretreatment of sewage sludge.

Effects of ultrasonic pretreatment on quantity and composition of bacterial DNA recovered from granular activated carbon used for drinking water treatment.

Effects of ultrasonic pretreatment on bacterial DNA recovery from granular activated carbon (GAC) were investigated. GAC (Calgon F400), biologically activated, was sampled from an actual drinking water plant. Different ultrasonic energy densities (0-400 J·cm(-3)) were applied with agitation (250 rpm for 30 min), and recovered bacterial DNA was quantified using quantitative PCR. Energy density was linearly correlated with the concentration of carbon fines produced from GAC during ultrasonication. Ultrasonication alone had no effect on DNA recovery at ≤60 J·cm(-3), but a strongly adverse effect at >67 J·cm(-3) due to the produced carbon fines. Agitation along with ultrasonication strongly enhanced the bacterial DNA recovery when ≤40 J·cm(-3) was applied, although it did not affect the production of carbon fines. Ribosomal tag pyrosequencing was used to compare recovered bacterial communities (0, 20 and 30 J·cm(-3) with or without agitation). Ultrasonication allowed for obtaining a more diverse and richer bacterial community from GAC, compared with the control. Agitation did not show a positive effect on community organization (richness and diversity). Consistently, canonical correspondence analysis indicated that the energy density was associated with the relative abundances of particular bacterial members (P < 0.05), while agitation did not. Correspondence analysis revealed that the recovered bacterial communities were grouped according to the applied energy densities. In conclusion, ultrasonication and agitation influence the recovered DNA in quality and quantity, respectively, and carbon fines as a by-product by ultrasonication interfere with the DNA recovery.

Effects of nonmethane volatile organic compounds on microbial community of methanotrophic biofilter.

Effects of nonmethane volatile organic compounds (NMVOCs) on methanotrophic biofilter were investigated. Laboratory-scale biofilters packed with pumice and granular-activated carbon (10:1, w/w) were operated with CH4 and NMVOCs including dimethyl sulfide (DMS) and benzene/toluene (B/T). DMS alone exhibited a positive effect on the methanotrophic performance; however, the coexistence of B/T removed this effect. B/T alone exerted no effect on the performance. Pyrosequencing and quantitative PCR revealed that the NMVOCs strongly influenced the bacterial and methanotrophic communities but not the population density of methanotrophs. DMS alone diversified and changed both bacterial and methantrophic communities, but its effects were nullified by the presence of B/T. Canonical correspondence analysis revealed significant correlations between the NMVOCs and community composition and significant interaction between DMS and B/T. DMS did not affect the distribution of types I/II methanotrophs (60/40), while B/T increased the abundance of type I to 82 %. DMS and B/T favored the growth of the methanotrophic bacteria Methylosarcina and Methylomonas, respectively. These results suggest that NMVOCs can be a significant abiotic factor influencing methane biofiltration.

Comparison of RNA- and DNA-based bacterial communities in a lab-scale methane-degrading biocover.

Methanotrophs must become established and active in a landfill biocover for successful methane oxidation. A lab-scale biocover with a soil mixture was operated for removal of methane and nonmethane volatile organic compounds, such as dimethyl sulfide (DMS), benzene (B), and toluene (T). The methane elimination capacity was 211±40 g m(-2) d(-1) at inlet loads of 330-516 g m(-2) d(-1). DMS, B, and T were completely removed at the bottom layer (40-50 cm) with inlet loads of 221.6±92.2, 99.6±19.5, and 23.4±4.9 mg m(-2) d(-1), respectively. The bacterial community was examined based on DNA and RNA using ribosomal tag pyrosequencing. Interestingly, methanotrophs comprised 80% of the active community (RNA) while 29% of the counterpart (DNA). Types I and II methanotrophs equally contributed to methane oxidation, and Methylobacter, Methylocaldum, and Methylocystis were dominant in both communities. The DNA vs. RNA comparison suggests that DNA-based analysis alone can lead to a significant underestimation of active members.