Comparative Analysis of Bacterial and Archaeal Communities In

Microbes Environ. Vol. 24, No. 2, 88–96, 2009 http://wwwsoc.nii.ac.jp/jsme2/ doi:10.1264/jsme2.ME08561

Comparative Analysis of Bacterial and Archaeal Communities in Methanogenic Sludge Granules from Upflow Anaerobic Sludge Blanket Reactors Treating Various Food-Processing, High-Strength Organic Wastewaters

TAKASHI NARIHIRO1, TAKESHI TERADA1,2, KAE KIKUCHI2, AKINORI IGUCHI1, MIZUYO IKEDA2, TOSHIHIRO YAMAUCHI2, KOJI SHIRAISHI2, YOICHI KAMAGATA1,3, KAZUNORI NAKAMURA1, and YUJI SEKIGUCHI1* 1Institute for Biological Resources and Functions, National Institute of Advanced Industrial Science and Technology (AIST), 1-1-1 Higashi, Tsukuba, Ibaraki 305-8566, Japan; 2Fujikasui Engineering Co., Ltd., 1-4-3 Higashi-gotanda, Shinagawa, Tokyo 141-0022, Japan; and 3Research Institute of Genome-based Biofactory, National Institute of Advanced Industrial Science and Technology (AIST), 2-17-2-1 Tsukisamu-Higashi, Toyohiraku, Sapporo, Hokkaido 062-8517, Japan (Received December 1, 2008—Accepted February 9, 2009—Published online March 13, 2009)

A comprehensive survey of bacterial and archaeal community structures within granular sludges taken from twelve different types of full-scale, food-processing wastewater-treating, upflow anaerobic sludge blanket (UASB) reactors was performed with a 16S rRNA gene-based clone library method. In total, 1,282 bacterial 16S rRNA gene clones and 722 archaeal clones were analyzed, and their identities were determined by phylogenetic analyses. Overall, clones belonging to the bacterial phyla Proteobacteria (the class Deltaproteobacteria in particular), Firmicutes, Spirochaetes, and Bacteroidetes were observed in abundance within the bacterial clone libraries examined, indicating common bacterial denominators in such treatment systems. Within the domain Archaea, clones affiliated with the classes Methano- microbia and Methanobacteria were found to be abundant in the archaeal libraries. In relation to features of reactor performance (such as chemical oxygen demand removal, fatty acid accumulation, and sludge bulking), possible representative phylotypes likely to be associated with process failures, such as sludge bulking and the accumulation of propionate, were found in comparative analyses of the distribution of phylotypes in the sludge libraries. Key words: 16S rRNA gene clone library, granular sludge, microbial community, UASB

Anaerobic digestion technology has been used effectively yses have been performed for UASB granular sludges treat- to treat organic matter in waste streams. To date, various ing wastewater from a paper factory (41), a terephthalate- anaerobic processes for treating wastewater have been devel- manufacturing plant (53), a beer brewery (13), and sucrose/ oped (1, 29, 32). One of the most established technologies in propionate/acetate-based artificial wastewater (44). In addi- this field is the upflow anaerobic sludge blanket (UASB) sys- tion, the molecular characterization of UASB granules tar- tem, because of its ability to treat a broad range of organic geting specific microbial groups has also been reported (10, waste streams at high loading rates (32, 40, 45, 47). The most 20, 31). According to these studies, the 16S rRNA gene characteristic phenomenon in this process is sludge granula- clones retrieved from UASB sludges indicated the major tion, i.e., granular-shaped sludge is spontaneously formed microbial constituents to be those of the phyla Proteobacte- within the system. Granular sludge generally has superior ria, Chloroflexi, Firmicutes, Spirochaetes, and Bacteroidetes settling characteristics. Thus, the stable and efficient opera- in the domain Bacteria, and those of the classes Methano- tion of granular sludge-based systems is primarily dependent microbia, Methanobacteria, and Thermoplasmata in the on the growth and maintenance of granular sludge. Granular domain Archaea (47). In addition, such studies have also sludge is also characterized as a spherical biofilm, possessing shown that a large number of the clones assigned to candi- all the trophic groups of anaerobes necessary for the com- date phyla (known as ‘clone clusters’) were frequently found plete mineralization of organic matter. Owing to its charac- in such ecosystems (7, 8, 10, 20, 31, 48, 55). However, there teristic internal structure, granular sludge is also important are still obstacles to precisely determining and monitoring for the efficient biotransformation of organic matter into the entire microbial community of sludge because of the lim- methane (48). ited number of datasets reported so far. Understanding the ecology of anaerobes involved in gran- To create more complete microbial biodiversity maps in ular sludge is essential to the control of these bioreactors. relation to reactor performance and wastewater type, it is The microbiology of granular sludges in UASB bioreactors necessary to further increase the biodiversity data of UASB has been studied using culture-dependent and molecular- sludge granules in association with data on process failures. based approaches, particularly those targeting 16S rRNA In the present study, we attempted the exhaustive character- genes (38, 45, 47). So far, molecular-based community anal- ization of methanogenic granular sludge in full-scale UASB reactors treating various types of food-processing, high- * Corresponding author. E-mail: [email protected]; Tel: +81– strength organic wastewater. Since a more extensive applica- 29–861–7866; Fax: +81–29–861–6400. tion of UASB technology is still hampered by concerns over Microbial Community in UASB Sludges 89 operational instability such as the accumulation of volatile 16S rRNA gene clone library fatty acids (14, 24), formation of scum (23, 50), and sludge Bacterial and archaeal 16S rRNA gene clone libraries were con- bulking (3, 15, 46, 55), special attention was paid to UASB structed for all the samples listed in Table 1. DNA extraction and bioreactors associated with volatile fatty acid accumulation PCR amplification were performed as described previously with and sludge bulking in choosing representative bioprocesses. slight modifications (37, 39, 54). The partial 16S rRNA genes were amplified with the primer set EUB338mix (5'-ACWCCTACGGG- Comparative analyses of bacterial and archaeal 16S rRNA WGGCWGC-3'), which consists of an equal amount (mol) of gene clone inventories of 12 granular sludges were per- EUB338 (2), EUB338II (12), and EUB338III (12) forward primers formed, and the phylogenetic identities of the major and (possessing sequences complementary to those of the EUB338 characteristic phylotypes were determined. probe set, Escherichia coli position 338–355), and the reverse primer UNIV1492r (5'-TACGGYTACCTTGTTACGACTT-3', E. Materials and Methods coli position 1492–1513) (30) for the domain Bacteria, and ARC109f (5'-ACKGCTCAGTAACACGT-3', E. coli position 109– UASB process 125) (22) and UNIV1492r for the domain Archaea. The thermal cycle profile consisted of preheating at 95°C for 9 min and 20 UASB sludge samples were taken from 12 full-scale UASB reac- cycles of denaturation at 95°C for 30 s, annealing at 50°C for 30 s, tors that treat various types of food-processing wastewater (Table and extension at 72°C for 2 min; the final step was followed by post 1). Based on wastewater type, the reactors are categorized into extension for 10 min. PCR products were purified with a QIAquick seven groups, namely (i) isomerized sugar-processing, (ii) sugar- PCR Purification Kit (Qiagen, Valencia, CA, USA), and subcloned based food-processing, (iii) vinegar-processing, (iv) soybean-based with a TA cloning kit (Novagen, Madison, WI, USA) according to product-processing, (v) salted vegetable-processing, (vi) alcohol- the manufacturer’s instructions. Cloned 16S rRNA genes were processing, and (vii) amino-acid-processing wastewater treatments sequenced with a Quick start kit (Beckman Coulter, Fullerton, CA, (Table 1). Groups (i) and (ii) are those for the treatment of organic USA) and a CEQ 2000XL automated sequence analyzer (Beckman wastewaters discharged from production lines of isomerized sugar, Coulter). Approximately 100 bacterial and 50 archaeal gene clones and from food production lines using sugars. Group (iii) is for were randomly retrieved from each granular sludge sample. wastewater from vinegar production lines. Group (iv) is for waste- Sequence data (ca. 500 bp) were imported into the ARB program waters discharged from food production lines using soybean, such package (33) and aligned using the editing tool in the program. as those for soy source production. Group (v) is for wastewater Phylotype was defined as a group of cloned sequences with from the production of vegetables pickled in salt. Group (vi) is for >97.0% identity. For the phylotypes that are comprised of more wastewater from the production lines of alcohol (clear liquor), and than three clones, nearly full-length 16S rRNA gene sequences (ca. group (vii) is for wastewater of the production lines of artificial sea- 1,100 bp for the domain Bacteria, and ca. 1,400 bp for the domain sonings. All of these processes were located in Japan and operated under mesophilic conditions (35–40°C). The reactors N1 and N2 Archaea) were determined. Construction and sequencing of archaeal were treating identical wastewater discharged from the same sugar- 16S rRNA gene clone libraries for the reactors Sw and Dt (49 processing manufactory in parallel. Chemical oxygen demand clones for the Sw reactor and 72 clones for the Dt reactor) were (COD) was analyzed by the standard dichromate method (25). Sul- done in a previous report (39), and the same libraries and sequence fate was determined by the turbidimetric method (9). Ammonia, data were used in the present study. Phylogenetic trees based on nitrate, and nitrite were determined by the phenate method, the 16S rRNA gene sequences were constructed by the neighbor- colorimetric method, and cadmium reduction method, respectively joining method (42) with the ARB program package. To estimate (9). Organic nitrogen was determined by the Kjeldahl method (9). the confidence of the tree topologies, a bootstrap analysis (16) Phosphorus was determined by the vanadomolybdophosphoric acid for 1,000 replicates was performed with the PAUP* 4.0 program colorimetric method with sulfuric acid-nitric acid digestion pre- package (51). treatment (9). The analytical procedure for volatile fatty acid (VFA) Statistical analysis concentrations was described previously (25). Some of the perfor- The Chao1 nonparametric species richness estimator (6) and mance data (COD and VFA) for the reactors Sw and Dt were Shannon-Wiener index were calculated using the EstimateS soft- obtained from a previous report (39). Granular sludge samples were ware (version 8.0, http://viceroy.eeb.uconn.edu/estimates). The per- gently washed and immediately subjected to DNA extraction or centage of coverage was calculated using the equation [1-(n/N)], stored at −20°C. where n is the number of phylotypes represented by a single clone

Table 1. Specifications and COD data of the UASB processes

VFA conc. VFA/COD Sample Volum e COD loading rate COD removal Wastewater type in effluenta in effluent Note name (m3) (kg-COD m−3 d−1) rate (%) (mg-COD L−1) (%) Isomerized sugar-processing Sm 500 17 84 <0.01 0.002 Sw 380 10 87 290 ~100 bulking Sugar-processing Ss 230 3.2 67 29 13 N1 70 11 82 330 61 N2 70 10 5.7 2000 70 VFA accumulation Vinegar-processing Q 75 17 90 640 66 bulking Soybean-based products Hg 100 6.0 55 140 28 bulking Hn 30 24 98 18 4.3 Salted vegetables Yk 65 9.2 65 110 8.2 Ks 70 12 48 6200 ~100 VFA accumulation Alcohol-producing Hs 200 12 95 70 14 Amino-acid processing Dt 65 3.7 55 280 38 a Values for “VFA/COD in effluent” were calculated as percentages of the VFA content (as COD) in the total COD in the effluent streams. 90 NARIHIRO et al.

(singleton) and N is the total number of clones retrieved (21). An mean diameter) formed in all the systems (data not shown). evenness index was determined using the equation H/ln R, where H A fatal accumulation of volatile fatty acids (propionate in is the Shannon-Wienar index and R is the number of phylotypes particular) was observed in the reactor N2, which treated observed (34). The rarefaction curve was determined with the Ana- wastewater discharged from a sugar-processing factory in lytic Rarefaction software (version 1.3, http://www.uga.edu/~strata/ software/). parallel with the reactor N1 (Tables 1 and 2). In addition, the reactor Ks treating wastewater discharged from a salted veg- Nucleotide sequence accession numbers etable-processing factory exhibited a low COD removal rate The 16S rRNA gene sequences presented in this study were (<50%), and a high concentration of VFAs remained in the deposited under DDBJ/EMBL/GenBank accession numbers effluent (Table 1). Therefore, reactors N2 and Ks were con- AB266889 to AB267067, AB291244 to AB291541, and AB372571 sidered to be facing serious problems associated with the and AB372572. accumulation of propionate. In association with other aspects of reactor performance, the reactors Sw, Q and Hg often Results and Discussion showed a sudden washout of fluffy sludge (bulking) in their effluent streams (data not shown), although they showed no Operational properties signs of bulking and the deterioration of effluent wastewater The operational conditions and performance of the twelve around the sampling dates. In particular, Sw, described as an full-scale UASB reactors are shown in Table 1. More UASB reactor in our previous study, frequently showed such detailed information, including sulfate, nitrogen, phosphate phenomena due to sludge bulking (55). These problems often and VFA concentrations in influent wastewaters and effluent affected performance in terms of the COD removal rate (data streams, is shown in Table 2. Except for the reactors N2 and not shown). Ks, all of the UASB processes examined in this study were operated stably with sufficient levels of COD and VFA con- Overview of 16S rRNA gene clone libraries centrations in the effluent on the sampling dates (Table 2). Bacterial and archaeal 16S rRNA gene clone libraries Reactors N2 and Ks exhibited relatively poor COD levels in were constructed for 12 UASB granular sludges (Table 3 and the effluent stream (2.8–6.0 g-COD L−1) with higher concen- Table 4). Archaeal gene clone libraries for the reactors Sw trations of VFAs in the effluent (Table 2), indicative of an and Dt were previously constructed and published elsewhere unstable system on the sampling date. Because possible (39). For the domain Bacteria, we retrieved a total of 1,282 external electron acceptors other than carbon dioxide, like clones and sequenced short (approximately 500 bp) segments sulfate and nitrate, were not sufficiently present in waste- of the 16S rRNA genes. As a result, we found 382 bacterial waters (Table 2), most of the COD removed was converted phylotypes with the criterion of >97% sequence identify. to methane in all the reactors. Granular sludge (1–2 mm, Likewise, for the domain Archaea, we found a total of 32

Table 2. The physicochemical properties of the UASB systems Isomerized Wastewater type Sugar-processing Vinegar Soybean-based products Salted vegetables plant Alcohol Amino-acid sugar-processing Sm Sw Ss N Q Hg Hn Yk Ks Hs Dt a Sample name inf. eff. inf. eff. inf. eff. inf. eff. 1beff. 2b inf. eff. inf. eff. inf. eff. inf. eff. inf. eff. inf. eff. inf. eff. pH 5.4 6.9 5.6 6.6 6.1 6.7 4.6 6.7 5.7 5.6 6.8 4.5 6.9 4.5 7.0 7.1 6.9 4.5 7.1 4.5 7.8 7.0 7.1 −1 CODCr (g L ) 3.6 0.59 2.0 0.26 0.66 0.22 3.0 0.54 2.8 9.9 0.96 1.1 0.49 20 0.42 4.0 1.4 11 6.0 10 0.50 1.6 0.73 −1 < < < < < < < < SO4 (mg L ) 72 5.0 39 21 20 5.0 2.0 2.0 2.0 5.0 5.029212713681849 5.0 51 14 14 15 T-N (mg L−1) 22 13 28 30 22 19 38 9.6 18 9.8 6.6 35 57 140 110 11 8.3 610 470 510 700 79 96 −1 NH4-N (mg L ) 1.6 7.8 9.6 12 1.3 1.3 2.4 9.3 13 4.4 6.4 9.3 32 1.5 18 1.4 3.9 52 210 230 500 43 66.1 −1 < < < < < < < < < < < < < NO2-N (mg L ) 0.01 0.01 0.01 0.01 0.03 0.01 0.01 0.01 0.01 0.04 0.01 0.01 0.03 0.01 0.02 0.01 0.01 0.3 0.01 0.01 0.01 0.2 0.05 −1 < < < < < < < < < < < < < < < < < < < < < NO3-N (mg L ) 0.01 0.01 0.01 0.01 0.2 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.09 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 Organic-N (mg L−1) 20 5.2 18 18 21 17 35 0.30 4.5 5.3 0.20 25 25 130 91 9.5 4.4 560 260 280 200 36 30 T-P (mg L−1) 14 14 27 26 1.9 1.6 8.3 1.3 1.4 14 8.6 5.8 5.8 47 37 3.3 5.5 51 48 61 74 11 12 Volatile fatty acids (mg L−1) citrate 2.4 —c 170———55——15———19———71——— —— succinate ———— 6.4———— 69—1207269— 17—75—450—14— lactate 530 <0.01360—46————530———33—430—2300 — — — — — formate 4.0—120—5.8—47———— 114.026— ——70——— —— acetate 320 — 380 — 36 9.8 620 220 480 210 440 170 50 2100 3.9 360 89 500 140 — 4.9 400 22 propionate 240 — 170 24 26 12 270 64 940 43 52 75 10 650 — 19 13 4700 3900 — — 76 — i-butyrate ——26—————— —142.8—20—————770—5.67.6 n-butyrate — — 120 110 — — 37 — 11 42 7.1 29 — 950 — — — 17 — 290 — 120 — i-valerate ———— ————— —2539—216.8———60190—118.6 n-valerate ———25————— —— 23—210————33—3253— a Wastewater samples were taken from influent (inf.) and effluent (eff.) points of each reactor. b The number indicates reactors included in the process, e.g., ‘N1’ granule sample listed in Table 1 was taken from reactor N effluent 1. c —, Not detected. Microbial Community in UASB Sludges 91

Table 3. Bacterial 16S rRNA gene library of various types of UASB granular sludges Isomerized sugar- Vinegar- Soybean-based Wastewater type Sugar-processing Salted vegetables Clear-liquor Amino-acid processing processing products processing processing Sample name Sm Sw Ss N1 N2 Q Hg Hn Yk Ks Hs Dt Clones analyzed 105 106 105 84 108 104 117 110 101 122 103 117 Phylotypesa 57 41 53 43 41 42 61 43 42 56 46 65 Chao1 estimatorb 98 104 121 84 73 87 133 61 105 145 107 124 Coverage (%)c 67 76 67 68 78 76 68 81 75 67 74 67 Evenness indexd 0.92 0.85 0.92 0.92 0.81 0.88 0.92 0.83 0.90 0.85 0.91 0.95 Bacteria e Proteobacteria Alphaproteobacteria ● 1.9f — ● 1.9 ● 1.2 ● 2.8 ———● 4.0 ——● 1.7 Betaproteobacteria ——● 3.8 ● 7.1 ———○ 0.9 ● 1.0 ● 3.3 —— Gammaproteobacteria ● 5.7 ● 2.8 ● 1.0 ——————● 8.2 —— Deltaproteobacteria ● 6.7 ●● 17.0 ●● 10.5 ●● 17.9 ●● 11.1 ●● 22.1 ●● 23.9 ●● 16.4 ●● 16.8 ● 2.5 ●● 20.4 ●● 20.5 Epsilonproteobacteria ● 2.9 — ● 4.8 ● 1.2 ○ 0.9 ————●● 16.4 —— Chloroflexi ● 6.7 ● 8.5 ●● 17.1 ●●● 32.1 ●● 24.1 ●● 10.6 ●● 16.2 ● 5.5 ● 7.9 — ● 4.9 ● 8.5 Firmicutes ●●● 34.3 ●● 10.4 ● 5.7 ● 8.3 ● 6.5 ● 3.8 ●● 17.9 ● 8.2 ●● 11.9 ●●● 55.7 ● 9.7 ●●● 33.3 Spirochaetes ● 2.9 ● 5.7 ● 8.6 ● 6.0 ● 6.5 ● 1.9 ● 2.6 ● 7.3 ● 2.0 ○ 0.8 ● 1.9 ●● 10.3 Bacteroidetes ● 4.8 ● 6.6 ●● 13.3 ●● 10.7 ● 4.6 ● 8.7 ●● 10.3 ●● 10.0 ●● 15.8 ● 4.9 ●●● 32.0 ● 6.8 Actinobacteria ● 3.8 ● 5.7 ● 9.5 ●● 13.1 ●●● 33.3 ● 1.9 ● 2.6 ● 4.5 ●● 19.8 ○ 0.8 ● 4.9 — Defferibacteres ● 1.0 ● 1.9 ——● 1.9 ● 1.0 ● 7.7 ● 1.8 ●● 10.9 ● 4.9 ●● 13.6 ●● 11.1 Planctomycetes ● 1.0 ● 2.8 —————● 2.7 — ○ 0.8 — ● 2.6 Chlorobi ● 2.9 ● 1.9 ———●● 14.4 — ○ 0.9 ——● 2.9 — Nitrospira ●● 13.3 ●● 23.6 ● 6.7 ● 2.4 ● 2.8 ●● 10.6 ● 2.6 ●●● 29.1 —— — — Acidobacteria ● 2.9 ○ 0.9 ● 6.7 ——● 1.0 — ○ 0.9 ● 2.0 ——○ 0.9 Verrucomicrobia-Chlamydiae ——● 3.8 ——● 1.0 ● 2.6 ○ 0.9 ● 1.0 ——○ 0.9 Thermotoga ———————● 1.8 ——● 1.0 ○ 0.9 Gemmatimonadetes —————● 1.0 ———— — — Caldothix ——● 1.0 ——————— — — Caldicellulosiruptor ——● 1.0 ——————— — — KSB3 (clone cluster)g ● 1.0 ● 6.6 ———● 5.8 ● 6.0 ——— — — OP10 (clone cluster) — — ● 1.0 — ● 1.9 ● 1.0 ——● 1.0 —— — OP8 (clone cluster) — — — — — ● 1.0 ———— — — TM6 (clone cluster) ● 3.8 ———————● 3.0 ——○ 0.9 OP9 (clone cluster) — — ● 1.0 ——● 4.8 — ○ 0.9 ——● 3.9 — Termite group I (clone cluster) — — — — — — — — — ● 1.6 ● 1.0 — OS-K (clone cluster) — — — — — — ● 2.6 ——— — — TM7 (clone cluster) — — — — ○ 0.9 ———● 1.0 —— — NKB19 (clone cluster) ● 1.0 ————————— —○ 0.9 WS6 (clone cluster) ● 1.0 ————————— —○ 0.9 OD2 (clone cluster) ● 1.9 — ● 2.9 —————● 1.0 —— — Others ● 1.0 ● 5.7 ——● 2.8 ● 9.6 ● 5.1 ● 8.2 ● 1.0 — ● 3.9 — a A phylotype was defined as >97.0% sequence identity. b Chao1 nonparametric species richness estimator was calculated according to Chao (1987) (6). c Coverage index was calculated according to Good (1953) (21). d Evenness index was calculated according to Magurran (2004) (34). e Bacterial phyla and classes are named according to Taxonomic Outline of the Bacteria and Archaea (http://www.taxonomicoutline.org/). f Frequency of clones assigned with a phylogenetic group in percentage of the total number of sequence analyzed: —, 0%; ○, 0.1–0.9%; ●, 1–9%; ●●, 10–29%; ●●●, >29%. g Candidate phyla (order) are named according to a review reported by Hugenholtz (2002) (27). phylotypes from 722 cloned 16S rRNA gene sequences, lation of phylotypes within the domain Bacteria was insuffi- including 121 previously published gene clones (39). Nearly cient to achieve the goal of saturation (Fig. 1A). In addition, full sequences (ca. 1,100 bp) of the phylotypes that appeared relatively low coverage values (67–81%, Table 3) also sup- frequently in the libraries (148 phylotypes, phylotypes of ported the underestimation of species richness estimated on tripletons and over) were sequenced and subjected to more the basis of only the libraries examined. According to the detailed phylogenetic analyses. The rank abundances of the Chao1 nonparametric estimator, it can be estimated that 1.4– bacterial and archaeal phylotypes are shown in Figs. S1 and 2.6 fold the number of the phylotypes detected here were S2, respectively. present in the granules (Table 3). On the other hand, the rar- The results of rarefaction curves indicate that the accumu- efaction curves for members of the domain Archaea seemed 92 NARIHIRO et al.

Table 4. Archaeal 16S rRNA gene library of various types of UASB granular sludges Isomerized Vinegar- Soybean-based Wastewater type Sugar-processing Salted vegetables Clear-liquor Amino-acid sugar-processing processing products processing processing Sample name Sm Swh SsN1N2QHgHnYkKsHs Dth Clones analyzed 56 49 50 52 57 75 67 49 60 68 64 75 Phylotypesa 8 6 7 9107119 9127 7 Chao1 estimatorb 9 6 7.5 13.5 14 10 12 15.3 11 30 7.5 7 Coverage (%)c 96 98 98 94 93 96 97 90 97 91 98 97 Evenness indexd 0.83 0.88 0.85 0.83 0.71 0.72 0.80 0.50 0.71 0.65 0.71 0.61 Archaea e Euryarchaeota Methanomicrobia ●●● 85.7f ●●● 69.4 ●●● 66.0 ●●● 50.0 ●● 19.3 ●●● 41.3 ●●● 79.1 ●●● 87.8 ●●● 83.3 ●● 20.6 ●●● 92.2 ●●● 85.3 Methanobacteria ●● 14.3 ●●● 30.6 ●●● 34.0 ●●● 50.0 ●●● 80.7 ●●● 57.3 ●● 19.4 ●● 12.2 ●● 15.0 ●●● 79.4 ● 7.8 ●● 14.7 WSA2 (clone cluster)g — ————● 1.3 ● 1.5 — ● 1.7 —— — a A phylotype was defined as >97.0% sequence identity. b Chao1 nonparametric species richness estimator was calculated according to Chao (1987) (6). c Coverage index was calculated according to Good (1953) (21). d Evenness index was calculated according to Magurran (2004) (34). e Archaeal phyla and classes are named according to Taxonomic Outline of the Bacteria and Arhaea. f Frequency of clones assigned with a phylogenetic group in percentage of the total number of sequence analyzed: —, 0%; ○, 0.1–0.9%; ●, 1–9%; ●●, 10–29%; ●●●, >29%. g Candidate phyla (order) are named according to a review reported by Hugenholtz (2002) (27). h Data of archaeal 16S rRNA gene clone libraries for the reactors Sw and Dt were obtained from a previous report (39).

to reach a plateau, and the coverages were calculated to be 90–98%, suggesting that the phylotypes retrieved here were sufficient to estimate the biodiversity of members of the Archaea in the bioreactors. Bacterial population in granular sludge The comparative analysis of the 16S rRNA clones showed common microbial denominators in the anaerobic granular sludges in mesophilic UASB processes treating food-processing organic wastewaters. Within the domain Bacteria, clones belonging to the phyla Proteobacteria (especially the class Deltaproteobacteria), Firmicutes, Spirochaetes, and Bacteroidetes were commonly observed in all the sludges (Table 3). This is well consistent with information available on the microbial community structure of anaerobic granular sludge (47). Within the class Deltaproteobacteria, the major phylotypes were identified as being of the order Desulfuromonadales (3 phylotypes) and the families Syntrophobacteraceae (6 phylotypes) and Syntrophaceae (10 phylotypes) (Fig. S1). The latter two families are well- known as proton-reducing, syntrophic substrate-degrading bacteria (e.g., Syntrophobacter, syntrophic propionate-degrader; Syntrophus, syntrophic benzoate-degrader), which play critical roles in anaerobic (methanogenic) bioconversion in association with methanogenic archaea (26, 43). Interestingly, the most predominant phylotype in the overall bacterial clone libraries was a Nitrospirae phylotype (bacterial phylotype No. 114, accession AB266934) (Fig. 2), which was affiliated with the phylum Nitrospirae, but had no closely related cultured bacteria (the closest cultured neigh- Fig. 1. Rarefaction curves of (A) bacterial and (B) archaeal 16S rRNA gene clones of 12 UASB granular sludges. Closed triangle, Sm; bor of the phylotype was Thermodesulfovibrio islandicus, dark gray line, Sw; black line, Ss; closed square, N1; plus, N2; open X96726; sequence similarity, 87%). The phylotype, together triangle, Q; open circle, Hg; light gray line, Hn; cross, Yk; open with similar phylotypes, formed a coherent clone cluster, square, Ks; open diamond, Hs; closed diamond, Dt. named the anaerobic sludge group (Fig. 2). The group consisted of environmental rRNA gene clones with no cultured representatives. The most characteristic trait of this group is Microbial Community in UASB Sludges 93

Fig. 2. Phylogenetic position of the most abundantly observed bacterial phylotype (No. 114, indicated in boldface, 96 of the 1,282 bacterial 16S rRNA gene clones) assigned to the phylum Nitrospirae. The sequence of the phylotype and related reference sequences were aligned using the ARB program. Phylogenetic trees were constructed by the neighbor-joining method. The 16S rRNA gene sequences of Chloroflexus aurantiacus J-10-flT (D38365), Oscillochloris trichoides DG-6T (AF093427) and Roseiflexus castenholzii HLO8T (AB041226) were used to root the tree. Branching points supported by probability values above 95% in bootstrap analyses (based on 1000 replicates, estimated using the neighbor-joining method) are indicated by solid circles, whereas nodes supported by bootstrap values >85% are indicated by open circles. The accession number of each reference sequence is shown after the name of the strain or clone. The scale bar represents the number of changes of nucleotide per sequence position. that their environmental clones were retrieved from anaero- Thermovirga), their ecophysiological traits may not be limi- bic wastewater treatment systems (13, 28). This may indicate ted to the degradation of proteinous substrates. In fact, a the importance of these phylotypes in such biological treat- recent study using FISH-microautoradiography (MAR) with ment processes; however, the ecophysiology of these organ- an anaerobic sludge suggested more physiological versatility isms remains largely unknown. of these uncultured organisms (4). The second most abundant phylotype was one related with A phylotype (bacterial phylotype No. 182, AB266987) the suborder Propionibacterineae (bacterial phylotype No. classified in the phylum Chlorobi made up 14.4% of the total 146, AB266973, which was closely related to Brooklawnia number of clones for the reactor Q, which had been treating cerclae, DQ196625; similarity, 97%). This phylotype was wastewater from a vinegar production process (Table 3). abundant in the bacterial library for the reactor N2, in which Cultivated members of the phylum Chlorobi are known as propionate is accumulated at high concentrations (Table 2): anaerobic phototrophic bacteria (green sulfur bacteria) (18); this suggests a certain association of this organism with the all of these cultivated bacteria are classified within the order propionate accumulation problem (described below). Chlorobiales. However, the phylotype found in this study is Besides these taxa, high percentages (>10%) of the phyla distantly related to this order (similarity with the 16S rRNA Deferribacteres (formerly known as the ‘Synergistes’) and gene of Chlorobium chlorochromatii (AJ578461), 83%) but Chlorobi were observed in some bacterial libraries. For forms a coherent sister clade with it with only environmental example, phylotypes of the phylum Deferribacteres rRNA gene clones (35, 44). This clade was suggested to con- accounted for 10.9%, 13.6% and 11.1% of the total number tain a variety of environmental clones from activated sludge, of bacterial clones in the reactors Yk, Hs, and Dt, respec- soil, sediment, and anaerobic sludge (48). Considering their tively (Table 3). On the basis of the physiological character- abundance in the bacterial library, they may play a signifi- istics of previously known Deferribacteres species, it was cant role in wastewater treatment processes. However, their found that known members of the phylum have the ability to actual ecophysiology remains unknown. degrade amino acids under anoxic conditions (17, 52); this 16S rRNA gene clones associated with uncultured clone suggests that other uncultivated members may also be clusters at the phylum level, such as those of the putative involved in the degradation of proteinous substrates in phyla KSB3, OP10, and TM6, were also detected in 11 of the the treatment processes. However, since 13 of the 14 Deferri- 12 UASB sludges examined (Table 3). For example, the bacteres phylotypes detected here have low sequence simi- rRNA gene clones assigned in the candidate phylum KSB3 larities (<90%) with those of previously known species (such were found at frequencies of 6.6%, 5.8%, and 6.0% in the as those of the genera Anaerobaculum, Synergistes, and bacterial libraries for the reactors Sw, Q and Hg, respectively 94 NARIHIRO et al.

(Table 3). Interestingly, sludge bulking phenomena have different from that for the reactor N1 (evenness index=0.92) been frequently observed in those three processes (Table 1). (Fig. S1); i.e., although the most predominant phylotype for Recently, we have reported that KSB3-type filamentous bac- both the reactors was assigned to the same phylotype in the teria triggered sludge bulking in a mesophilic UASB system suborder Propionibacterineae (bacterial phylotype No. 146, treating wastewater discharged from a sugar-processing fac- AB266973), the bacterial library of the reactor N2 indicates a tory (55). 3.8-fold higher proportion of this phylotype in the library. This suggests the outgrowth of these organisms (or decreases Archaeal population in granular sludge in the numbers of organisms other than those in Propioni- All of the archaeal clones examined were classified in the bacterineae) in the system. Members of the family, such as phylum Euryarchaeota, particularly in the classes Methano- those of the genera Propionibacterium and Brooklawnia, microbia and Methanobacteria (Table 4). Within the class produce propionate as an end product of the fermentation Methanomicrobia, 6 of all the archaeal phylotypes detected process from various types of carbohydrate (5, 11). Consider- (32 phylotypes) were affiliated with the family Methano- ing these community changes in association with the physio- saetaceae, which accounted for 48.6% of the total number of logical properties of closely related organisms, the phylotype archaeal clones (722 clones). This family is known to consist No. 146 (possibly an outgrowth of this bacterium) may be of aceticlastic methanogens like Methanosaeta species (19). tightly associated with the accumulation of propionate. Within the class Methanobacteria, 10 of the 32 phylotypes In addition, we found that the bacterial community of the were assigned with the family Methanobacteriaceae, which propionate-accumulated reactor Ks was mainly composed of includes hydrogenotrophic methanogenic archaea like those members of the phylum Firmicutes and the class Epsilon- of the genus Methanobacterium (19). The most predomi- proteobacteria (Table 3). The major phylotypes in this nant phylotype in the overall archaeal clone library was a reactor were the Campylobacteraceae-like phylotype No. 59 Methanosaeta-like phylotype (archaeal phylotype No. 1, (AB266960) and unidentified Firmicutes phylotypes Nos. AB266919), which was closely related to the 16S rRNA 318 and 322 (Fig. S1, AB267033 and AB267032, respec- gene of Methanosaeta concilii (similarity with the sequence tively). These phylotypes were scarcely detected within the X16932, 99%). The second most abundant phylotype was a other sludge samples, implying the rarity of this bacterial Methanobacteriaceae-like phylotype (archaeal phylotype community. The Campylobacteraceae-like phylotype No. 59 No. 23, AB266900), which was closely related to the 16S is closely related with the 16S rRNA gene of Arcobacter rRNA gene of Methanobacterium formicicum (similarity butzleri (CP000361) (97% similarity). This organism is known with the sequence AF028689, 97%). These results support to grow on lactate under anaerobic conditions but cannot uti- our current knowledge that methanogenic archaea predomi- lize carbohydrates under these conditions (36, 49), suggest- nate in UASB sludge granules, playing roles in the last step ing that the phylotype No. 59 plays a role in the degradation in the anaerobic bioconversion of organic matter to methane of lactate. Actually, all the lactate (2,300 mg L−1) in the influ- (47). One phylotype (archaeal phylotype No. 22, AB266909) ent stream was removed in the reactor Ks (Table 2). We also was assigned to the candidate subphylum WSA2 (also called detected no Syntrophobacteraceae-like phylotypes (bacterial ‘Arc I’), which is known as a clone cluster in the phylum phylotype Nos. 23–28), which in the other reactors made up Euryarchaeota as functionally uncertain archaea (7, 48). No 3.4–15.5% of the bacterial clones (Fig. S1), indicating the phylotypes assigned to the class Thermoplasmata were found lack of general syntrophic propionate-oxidizers assigned in this study, although the primer set used for PCR covers to this group. Although the actual mechanism by which most of the members of the class, including clone lineages propionate accumulates remains unknown, the lack of such often retrieved from wastewater treatment systems. syntrophic degraders is likely to be one cause of the accumulation. Population associated with volatile fatty acid accumulation In addition to the changes in the bacterial communities, Although over-speculating the ecophysiological functions the proportion of methanogenic archaea seemed to be of phylotypes based solely on molecular phylogenetic analy- affected by the efficiency with which propionate was ses should be avoided, 16S rRNA gene-based biodiversity removed, i.e., the Methanobacteriaceae phylotypes (archaeal maps of sludge granules in association with reactor perfor- phylotype Nos. 23, 29, and 30) predominated in the reactors mance indicate possible links between microbial community N2 and Ks, which were less capable of removing propionate constituents and process failures. As mentioned above, the in effluent. On the other hand, the proportion of Methano- reactors N2 and Ks had poor quality effluent streams, con- saetaceae phylotypes (archaeal phylotype No. 1) was always taining high concentrations of volatile fatty acids, particu- high in the reactors well capable of removing propionate. larly propionate (940–3,900 mg L−1). The reactors N1 and Although the mechanism underlying the relation between the N2 were treating identical wastewater under the same opera- removal of propionate and archaeal community composition tional conditions. However, for unknown reasons, the quality remains unclear, this type of relation may be used for better of effluent from the reactor N2 suddenly worsened, with a monitoring of the process through the quantitative detection significant increase in the concentration of propionate in the of these microbial populations. effluent stream and concomitant decreases in the COD removal rate, methane formation, and the effluent pH (Table Conclusions 2). On the basis of the evenness index and rank abundance analysis, it was found that the species (phylotype) evenness In summary, we constructed biodiversity maps of anaero- for the reactor N2 (evenness index=0.81) was significantly bic granular sludges obtained from 12 mesophilic UASB Microbial Community in UASB Sludges 95 reactors treating food-processing wastewaters by using 16S and A. Sghir. 2005. Novel predominant archaeal and bacterial groups rRNA-based molecular inventories. The main conclusions revealed by molecular analysis of an anaerobic sludge digester. Envi- ron. Microbiol. 7:1104–1115. drawn from our results are as follows: 8. Chouari, R., D. Le Paslier, C. Dauga, P. Daegelen, J. Weissenbach, 1. Within the domain Bacteria, members of the phyla and A. Sghir. 2005. Novel major bacterial candidate division within Proteobacteria (especially the class Deltaproteobacteria), a municipal anaerobic sludge digester. Appl. Environ. Microbiol. Firmicutes, Spirochaetes, and Bacteroidetes were commonly 71:2145–2153. 9. Clesceri, L.S., A.E. Greenberg, and A.D. Eaton. 1998. Standard observed in the sludges examined. In addition, functionally Methods for the Examination of Water and Wastewater, American unknown phylotypes assigned with the uncultured lineages Public Health Association, Washington, DC., USA. of phyla Nitrospirae, Deferribacteres, and Chlorobi were 10. Collins, G., L. O’Connor, T. Mahony, A. Gieseke, D. de Beer, and V. frequently observed in the libraries for some sludge samples. O’Flaherty. 2005. Distribution, localization, and phylogeny of abundant populations of Crenarchaeota in anaerobic granular sludge. Interestingly, the most predominant phylotype in all the bac- Appl. Environ. Microbiol. 71:7523–7527. terial clone libraries was classified in the clone clade in the 11. Cummins, C.S., and J.L. Johnson. 1986. Genus I. Propionibacterium phylum Nitrospirae, which is composed of only environmen- Orla-Jensen 1909, 337AL, p. 1346–1353. In P.H.A. Sneath, N.S. Mair, tal rRNA gene clones with no cultured organisms. M.E. Sharpe, and J.G. Holt (ed.), Bergey’s Manual of Systematic Bacteriology, vol. 2. Williams & Wilkins, Baltimore, USA. 2. Within the domain Archaea, phylotypes of the classes 12. Daims, H., A. Bruhl, R. Amann, K.H. Schleifer, and M. Wagner. Methanomicrobia and Methanobacteria of the phylum 1999. The domain-specific probe EUB338 is insufficient for the Euryarchaeota were found in abundance. Archaeal popula- detection of all bacteria: Development and evaluation of a more tions were relatively simple in terms of genetic diversity comprehensive probe set. Syst. Appl. Microbiol. 22:434–444. based on rRNA genes, and most of the phylotypes detected 13. Diaz, E.E., A.J.M. Stams, R. Amils, and J.L. Sanz. 2006. Phenotypic properties and microbial diversity of methanogenic granules from a were closely related with known methanogenic archaea. full-scale upflow anaerobic sludge bed reactor treating brewery 3. The outgrowth of members of the Propionibacterineae wastewater. Appl. Environ. Microbiol. 72:4942–4949. was suggested to be associated with the accumulation of 14. Duran, M., and R.E. Speece. 1998. Staging of anaerobic processes for propionate in UASB reactors. reduction of chronically high concentrations of propionic acid. Water Environ. Res. 70:241–248. 4. Phylotypes assigned to the candidate bacterial phylum 15. Endo, G., and Y. Tohya. 1988. Ecological study on anaerobic sludge KSB3 are thought to be associated with sludge bulking bulking caused by filamentous bacterial-growth in an anaerobic con- phenomena in different UASB systems. tact process. Water Sci. Technol. 20:205–211. The information obtained here will be used as a basis for 16. Felsenstein, J. 1985. Confidence-limits on phylogenies—An approach using the bootstrap. Evolution 39:783–791. advancing more appropriate monitoring of the performance 17. Garrity, G.M., and J.G. Holt. 2001. Phylum BIX. Deferribacteres of reactor through quantitative or qualitative microbial detec- phy. nov., p. 465–474. In D.R. Boone, R.W. Castenholz, and G.M. tion during processing. The elucidation of the ecophysiologi- Garrity (ed.), Bergey’s Manual of Systematic Bacteriology, 2nd ed., cal functions of unidentified and uncultured microorganisms vol. 1. Springer-Verlag, New York, USA. 18. Garrity, G.M., and J.G. Holt. 2001. Phylum BXI. Chlorobi phy. nov., detected in this study may be needed to provide more valu- p. 601–623. In D.R. Boone, R.W. Castenholz, and G.M. Garrity able insights into the basis of the anaerobic (methanogenic) (ed.), Bergey’s Manual of Systematic Bacteriology, 2nd ed., vol. 1. bioconversion and granulation of sludge in UASB systems. Springer-Verlag, New York, USA. 19. Garrity, G.M., and J.G. Holt. 2001. Phylum AII. Euryarchaeota phy. nov., p. 211–355. In D.R. Boone, R.W. Castenholz, and G.M. Acknowledgements Garrity (ed.), Bergey’s Manual of Systematic Bacteriology, 2nd ed., vol. 1. Springer-Verlag, New York, USA. This study was supported by the Project “Development of Tech- 20. Godon, J.J., J. Morinière, M. Moletta, M. Gaillac, V. Bru, and J.P. nologies for Analyzing and Controlling the Mechanism of Biode- Delgènes. 2005. Rarity associated with specific ecological niches in grading and Processing”, of the New Energy and Industrial Tech- the bacterial world: The ‘Synergistes’ example. Environ. Microbiol. nology Development Organization (NEDO). 7:213–224. 21. Good, I.J. 1953. The population frequencies of species and the esti- References mation of population parameters. Biometrika 40:237–264. 22. Grosskopf, R., P.H. Janssen, and W. Liesack. 1998. Diversity and 1. Aiyuk, S., I. Forrez, D.K. Lieven, A. van Haandel, and W. Verstraete. structure of the methanogenic community in anoxic rice paddy soil 2006. Anaerobic and complementary treatment of domestic sewage in microcosms as examined by cultivation and direct 16S rRNA gene regions with hot climates—A review. Bioresour. Technol. 97:2225– sequence retrieval. Appl. Environ. Microbiol. 64:960–969. 2241. 23. Halalsheh, M., J. Koppes, J. den Elzen, G. Zeeman, M. Fayyad, and 2. Amann, R.I., B.J. Binder, R.J. Olson, S.W. Chisholm, R. Devereux, G. Lettinga. 2005. Effect of SRT and temperature on biological con- and D.A. Stahl. 1990. Combination of 16S ribosomal-RNA-targeted versions and the related scum-forming potential. Water Res. 39:2475– oligonucleotide probes with flow-cytometry for analyzing mixed 2482. microbial-populations. Appl. Environ. Microbiol. 56:1919–1925. 24. Han, S.K., S.H. Kim, and H.S. Shin. 2005. UASB treatment of waste- 3. Angenent, L.T., and S.W. Sung. 2001. Development of anaerobic water with VFA and alcohol generated during hydrogen fermentation migrating blanket reactor (AMBR), a novel anaerobic treatment of food waste. Process Biochem. 40:2897–2905. system. Water Res. 35:1739–1747. 25. Harada, H., S. Uemura, A.C. Chen, and J. Jayadevan. 1996. Anaero- 4. Ariesyady, H.D., T. Ito, and S. Okabe. 2007. Functional bacterial and bic treatment of a recalcitrant distillery wastewater by a thermophilic archaeal community structures of major trophic groups in a full-scale UASB reactor. Bioresour. Technol. 55:215–221. anaerobic sludge digester. Water Res. 41:1554–1568. 26. Hattori, S. 2008. Syntrophic acetate-oxidizing microbes in methano- 5. Bae, H.S., W.M. Moe, J. Yan, I. Tiago, M.S. da Costa, and F.A. genic environments. Microbes Environ. 23:118–127. Rainey. 2006. Brooklawnia cerclae gen. nov., sp. nov., a propionate- 27. Hugenholtz, P. 2002. Exploring prokaryotic diversity in the genomic forming bacterium isolated from chlorosolvent-contaminated ground- era. Genome Biol. 3:reviews0003.1–0003.8. water. Int. J. Syst. Evol. Microbiol. 56:1977–1983. 28. Kaksonen, A.H., J.J. Plumb, P.D. Franzmann, and J.A. Puhakka. 6. Chao, A. 1987. Estimating the population size for capture-recapture 2004. Simple organic electron donors support diverse sulfate-reduc- data with unequal catchability. Biometrics 43:783–791. ing communities in fluidized-bed reactors treating acidic metal- and 7. Chouari, R., D. Le Paslier, P. Daegelen, P. Ginestet, J. Weissenbach, sulfate-containing wastewater. FEMS Microbiol. Ecol. 47:279–289. 96 NARIHIRO et al.

29. Kleerebezem, R., and H. Macarie. 2003. Treating industrial waste- 44. Sekiguchi, Y., Y. Kamagata, K. Syutsubo, A. Ohashi, H. Harada, water: Anaerobic digestion comes of age. Chem. Eng. 110:56–64. and K. Nakamura. 1998. Phylogenetic diversity of mesophilic and 30. Lane, D.J. 1991. 16S/23S rRNA sequencing, p. 115–175. In E. thermophilic granular sludges determined by 16S rRNA gene Stackebrandt, and M. Goodfellow (ed.), Nucleic Acid Techniques in analysis. Microbiology 144:2655–2665. Bacterial Systematics, John Wiley & Sons, Chichester, UK. 45. Sekiguchi, Y., Y. Kamagata, and H. Harada. 2001. Recent advances 31. Leclerc, M., J.P. Delgènes, and J.J. Godon. 2004. Diversity of the in methane fermentation technology. Curr. Opin. Biotechnol. 12:277– archaeal community in 44 anaerobic digesters as determined by single 282. strand conformation polymorphism analysis and 16S rDNA sequenc- 46. Sekiguchi, Y., H. Takahashi, Y. Kamagata, A. Ohashi, and H. ing. Environ. Microbiol. 6:809–819. Harada. 2001. In situ detection, isolation, and physiological properties 32. Lettinga, G. 1995. Anaerobic-digestion and waste-water treatment of a thin filamentous microorganism abundant in methanogenic systems. Antonie Van Leeuwenhoek 67:3–28. granular sludges: A novel isolate affiliated with a clone cluster, the 33. Ludwig, W., O. Strunk, R. Westram, and et al. 2004. ARB: A soft- green non-sulfur bacteria, subdivision I. Appl. Environ. Microbiol. ware environment for sequence data. Nucleic Acids Res. 32:1363– 67:5740–5749. 1371. 47. Sekiguchi, Y., and Y. Kamagata. 2004. Microbial community struc- 34. Magurran, A.E. 2004. Measuring Biological Diversity, Blackwell ture and functions in methane fermentation technology for wastewater Publishing, Oxford, UK. treatment, p. 361–384. In M.M. Nakano, and P. Zuber (ed.), Strict and 35. Mchugh, S., M. Carton, G. Collins, and V. O’Flaherty. 2004. Reactor Facultative Anaerobes: Medical and Environmental Aspects, Horizon performance and microbial community dynamics during anaerobic Bioscience, Wymondham, UK. biological treatment of wastewaters at 16–37 degrees C. FEMS 48. Sekiguchi, Y. 2006. Yet-to-be cultured microorganisms relevant to Microbiol. Ecol. 48:369–378. methane fermentation processes. Microbes Environ. 21:1–15. 36. Miller, W.G., C.T. Parker, M. Rubenfield, and et al. 2007. The com- 49. Smibert, R.M. 1984. Genus Campylobacter Sebald and Veron 1963, plete genome sequence and analysis of the epsilonproteobacterium 907AL, p. 111–118. In N.R. Krieg, and J.G. Holt (ed.), Bergey’s Arcobacter butzleri. PLoS ONE 2:e1358. Manual of Systematic Bacteriology, vol. 1. Williams & Wilkins, 37. Miyata, R., N. Noda, H. Tamaki, K. Kinjyo, H. Aoyagi, H. Uchiyama, Baltimore, USA. and H. Tanaka. 2007. Phylogenetic relationship of symbiotic archaea 50. Souza, C.L., S.Q. Silva, S.F. Aquino, and C.A.L. Chernicharo. 2006. in the gut of the higher termite Nasutitermes takasagoensis fed with Production and characterization of scum and its role in odour control various carbon sources. Microbes Environ. 22:157–164. in UASB reactors treating domestic wastewater. Water Sci. Technol. 38. Narihiro, T., and Y. Sekiguchi. 2007. Microbial communities in 54:201–208. anaerobic digestion processes for waste and wastewater treatment: a 51. Swofford, D.L. 2003. PAUP*, Phylogenetic Analysis Using microbiological update. Curr. Opin. Biotechnol. 18:273–278. Parsimony (*and Other Methods), version 4, Sinauer Associates, 39. Narihiro, T., T. Terada, A. Ohashi, J.H. Wu, W.T. Liu, N. Araki, Y. Sunderland, USA. Kamagata, K. Nakamura, and Y. Sekiguchi. 2009. Quantitative detec- 52. Vartoukian, S.R., R.M. Palmer, and W.G. Wade. 2007. The division tion of culturable methanogenic archaea abundance in anaerobic treat- “Synergistes”. Anaerobe 13:99–106. ment systems using the sequence-specific rRNA cleavage method. 53. Wu, J.H., W.T. Liu, I.C. Tseng, and S.S. Cheng. 2001. Characteriza- ISME J. doi:10.1038/ismej.2009.4. tion of microbial consortia in a terephthalate-degrading anaerobic 40. O’Flaherty, V., G. Collins, and T. Mahony. 2006. The microbiology granular sludge system. Microbiology 147:373–382. and biochemistry of anaerobic bioreactors with relevance to domestic 54. Yamada, T., Y. Sekiguchi, H. Imachi, Y. Kamagata, A. Ohashi, and sewage treatment. Rev. Environ. Sci. Biotechnol. 5:39–55. H. Harada. 2005. Diversity, localization, and physiological properties 41. Roest, K., H.G.H.J. Heilig, H. Smidt, W.M. de Vos, A.J.M. Stams, of filamentous microbes belonging to Chloroflexi subphylum I in and A.D.L. Akkermans. 2005. Community analysis of a full-scale mesophilic and thermophilic methanogenic sludge granules. Appl. anaerobic bioreactor treating paper mill wastewater. Syst. Appl. Environ. Microbiol. 71:7493–7503. Microbiol. 28:175–185. 55. Yamada, T., T. Yamauchi, K. Shiraishi, P. Hugenholtz, A. Ohashi, H. 42. Saitou, N., and M. Nei. 1987. The neighbor-joining method: A new Harada, Y. Kamagata, K. Nakamura, and Y. Sekiguchi. 2007. Charac- method for reconstructing phylogenetic trees. Mol. Biol. Evol. 4:406– terization of filamentous bacteria, belonging to candidate phylum 425. KSB3, that are associated with bulking in methanogenic granular 43. Schink, B. 1997. Energetics of syntrophic cooperation in methano- sludges. ISME J. 1:246–255. genic degradation. Microbiol. Mol. Biol. Rev. 61:262–280.