Microbial community structure and population dynamics of granules developed in expanded granular sludge bed (EGSB) reactors for the anaerobic treatment of low-strength wastewater at

Page 1

PLEASE SCROLL DOWN FOR ARTICLE

This article was downloaded by: [Syutsubo, Kazuaki][National Inst for Environmental Studies.]On: 13 April 2010Access details: Access Details: [subscription number 918060784]Publisher Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Environmental Science and Health, Part APublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t713597268

Microbial community structure and population dynamics of granulesdeveloped in expanded granular sludge bed (EGSB) reactors for theanaerobic treatment of low-strength wastewater at low temperatureIkuo Tsushima a; Wilasinee Yoochatchaval a; Hiroki Yoshida a; Nobuo Araki b;Kazuaki Syutsubo a

a Water and Soil Environment Division, National Institute for Environmental Studies (NIES), Ibaraki,Japan b Department of Civil Engineering, Nagaoka National College of Technology, Niigata, Japan

Online publication date: 12 April 2010

To cite this Article Tsushima, Ikuo , Yoochatchaval, Wilasinee , Yoshida, Hiroki , Araki, Nobuo andSyutsubo,Kazuaki(2010) 'Microbial community structure and population dynamics of granules developed in expanded granularsludge bed (EGSB) reactors for the anaerobic treatment of low-strength wastewater at low temperature', Journal ofEnvironmental Science and Health, Part A, 45: 6, 754 — 766To link to this Article: DOI: 10.1080/10934521003651531URL: http://dx.doi.org/10.1080/10934521003651531

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.

Page 2

Journal of Environmental Science and Health Part A (2010) 45, 754–766Copyright C© Taylor & Francis Group, LLCISSN: 1093-4529 (Print); 1532-4117 (Online)DOI: 10.1080/10934521003651531

Microbial community structure and population dynamics ofgranules developed in expanded granular sludge bed (EGSB)reactors for the anaerobic treatment of low-strengthwastewater at low temperature

IKUO TSUSHIMA1, WILASINEE YOOCHATCHAVAL1, HIROKI YOSHIDA1,NOBUO ARAKI2 and KAZUAKI SYUTSUBO1

1Water and Soil Environment Division, National Institute for Environmental Studies (NIES), Ibaraki, Japan2Department of Civil Engineering, Nagaoka National College of Technology, Niigata, Japan

The anaerobic biological treatment of sucrose-based, low-strength wastewater was investigated in expanded granular sludge bed(EGSB) reactors at low temperatures over a 300-day trial period. During the trial, the operating temperature was lowered in astepwise manner from 20◦C to 5◦C. As a result, the reactors exhibited sufficient performances until 10◦C operation. The CODremoval rate was 3.1–3.8 kgCOD m−3 day−1 at 10◦C. In particular, the COD removal rate increased gradually through the low-temperature operation; indeed, the later stages of the 10◦C operation attained a rate similar to those achieved at 20◦C and 15◦C. Thisfinding is especially practical for applications of psychrophilic methane fermentation. Additionally, the structure of the microbialcommunity in the granular sludge was analyzed by clone analysis based on 16S rRNA genes and fluorescence in situ hybridization(FISH). As a result, the percentage of the phylum Firmicutes, which were assumed to be Anaerobivrio sp. and Lactococcus sp., greatlyincreased from 0.7% to 8.0% of the total cells, especially in the surface layer of the granular sludge. These bacteria would contribute tothe degradation of the sucrose substrate anaerobically at ambient temperatures. Moreover, the results suggest that a Methanospirillumspecies, which is a H2-utilizing methanogen, increased from 0.5% to 6.7% during the low-temperature incubation, with a significantincrease of methanogenic activity from H2/CO2 at 20◦C. Thus, the Methanospirillum species detected in this study may have a keyrole as hydrogen scavenger during hydrogen-metabolism in low-temperature conditions.

Keywords: Ambient temperature, anaerobic treatment, granular sludge bed reactor, low-strength wastewater, clone analysis, fluores-cence in situ hybridization (FISH).

Introduction

Psychrophilic (<20◦C) methane fermentation offers tech-nical and economical benefits over the more conventionaltreatment processes (except in necessary cases such as high-temperature effluent discharges), making it an ecologicalmethod for treating wastewater discharged at low or ambi-ent temperatures. Although psychrophilic methane fermen-tation was thought not to be viable due to the low biogasproduction rates and low microbial activity achieved underlow-temperature conditions,[1,2] it has recently been provenfeasible through the use of new or modified laboratory-scale

Address correspondence to Kazuaki Syutsubo, Water and SoilEnvironment Division, National Institue for EnvironmentalStudies (NIES), Ibaraki, Japan. E-mail: [email protected] October 19, 2009.

anaerobic treatment bioreactor designs that are based on abiofilm (granular sludge), such as internal circulation reac-tors, expanded granular sludge bed (EGSB) reactors, andhybrid granular sludge bed reactor systems combined withan anaerobic filter.[3−9]

Anaerobic treatment by granular sludge has someadvantages over conventional anaerobic digestion. First,it allows high biomass concentrations and a long sludge-retention time from the beginning of operation. Second, itis able to respond quickly to perturbations in the wastew-ater, including pH, ORP, and temperature. Third, thegranular structure provides ideal conditions for syntrophicassociations such as those between H2-producing aceto-genic bacteria and H2-utilizing methanogens. In addition,reactors using granular sludge have been applied to andexhibited sufficient performance in the treatment of low-strength wastewater, such as industrial wastewater from abrewery,[10] and malting wastewater from the batch steep

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 3

Expanded granular sludge bed for wastewater treatment 755

process of a malting factory.[11,12] In recent years, McHughet al.[13] reported a successful treatment of whey-containingwastewater in an EGSB reactor at 20–12◦C.

However, problems still remain with respect to psy-chophilic methane fermentation; for example, the processrequires a long start-up period, its performance may beunstable, and the kind of wastewater that can be treatedis limited due to the activity of methanogen. Moreover,wastewater treatment processes are still mostly operatedas “black boxes”, taking the effluent concentration as anoutput value that cannot be improved. In addition, theprocess control strategy, if applied at all, does not generallytake into account what happens at the microorganism level.In particular, only a few studies have described the struc-ture and function of microbial populations in psychrophilicanaerobic fermentation.[8,14] To gain knowledge of the op-erational range of psychrophilic anaerobic fermentation,more information is required on the nature, identity andcapacity of the microbial consortia.

Therefore, in the present study, the feasibility of psy-chrophilic anaerobic fermentation has been investigated bymonitoring the effect of decreasing temperatures and low-strength wastewater on reactor performances over a 300-day period. In addition, the microbial population struc-ture and dynamics of the granular sludge during the trialwere investigated using molecular biological techniques, in-cluding clone analysis and florescence in situ hybridization(FISH).

Materials and methods

Reactor operation and seeding sludge

An EGSB reactor (Run 1) with a working volume of16.8 L (Column: 11.7 L, Gas-Solid separatior: 5.1 L) wasoperated for 104 days at 20◦C. The liquid-upflow veloc-ity was set to 5 m/h, and effluent recirculation was setto maintain this required upflow velocity. The reactorwas inoculated with mesophilic granular sludge obtainedfrom a full-scale UASB reactor receiving sugar-containingwastewater, and started with 360 gVSS, giving the reactora concentration of 45 g VSS L−1. The synthetic wastew-ater ranged from 0.6 to 0.8 g COD L−1 during the re-actor operation. This wastewater was composed of su-crose, acetate, propionate, and yeast extract as a carbonsource in the COD ratio of 4.5:2.25:2.25:1. The compo-sitions of basal minerals and trace elements were as fol-lows (mg L−1): NH4Cl, 37; KH2PO4, 33; MgCl2·6H2O, 13:CaCl2·2H2O, 33; KCl, 10; FeSO4·7H2O, 7; CoCl2·6H2O,0.17; ZnSO4·7H2O, 0.15; H3BO3, 0.06; MnCl2·4H2O, 0.42;NiCl2·6H2O, 0.04; CuCl2·2H2O, 0.027; Na2MoO4·2H2O,0.025; NaHCO3, 800. The average concentration of sulfatewas 46 mg SO2−

4 L−1, which originated from tap water. Af-ter Run 1, which operated for 104 days, the granule sludgewas inoculated into another EGSB reactor (Run 2) with a

working volume of 2 L (Column: 1.3 L, Gas-Solid separa-tior: 0.7 L). Run 2 was operated at 15◦C for the first 140days (day 104 to day 243) with the same operational con-ditions and wastewater as Run 1. The process temperaturewas then reduced to 10◦C at day 244 and to 5◦C at day 346.

Analysis procedure

The COD and sulfate concentration of the influent and ef-fluent were analyzed by DR-2500 (Hach Company, USA) inaccordance with the manufacturer’s manual. A raw samplewas used to estimate total the COD, and a 0.45-µm-filteredsample was used to estimate the dissolved COD. After sam-pling, the effluent was homogenized for suspended solids(SS) and total COD measurement. For the measurement ofeffluent COD, a small amount of sulfuric acid was added tothe samples, which were then purged with nitrogen gas toremove sulfide. The COD removal efficiency was calculatedby the difference between the influent total COD and theeffluent dissolved COD.

Microbial community analysis

To monitor changes in the bacterial and archaeal commu-nity structures in the reactor during long-term operation,granular sludge samples were harvested from the bottompart of the reactor. These sludge samples were used for anal-ysis of the microbial community structure by 16S rDNA-targeted clone analysis and FISH.

DNA extraction and PCR amplification for cloning

Total DNA was extracted from granular sludge samples(approximately 0.5 mL) with a DNA Isolation kit (ISOILfor Beads Beating, NIPPON GENE CO., LTD., Tokyo,Japan), as described in the manufacturer’s instructions.16S rRNA gene fragments from the extracted total DNAwere amplified with Taq DNA polymerase (TaKaRa BioInc., Ohtsu, Japan) and Eubacteria-targeted primer setBac27F and Univ1500R,[15] or Archaea-targeted primerset A109Fm and Univ1500R. The PCR conditions were asfollows: an initial 9 min denaturation at 95◦C, followed by25 cycles of 40 s at 94◦C, 30 s at 50◦C, and 2 min at 72◦C.The final extension was carried out for 10 min at 72◦C. ThePCR products were electrophoresed on a 1% (w/v) agarosegel.

Phylogenetic Analysis

PCR products were ligated into a pCR-XL-TOPO vectorand transformed into ONE SHOT Escherichia coli cellsby following the manufacturer’s instructions (TOPO XLPCR cloning; Invitrogen, CA, USA). Recombinant trans-formants were screened by blue and white selection, andcolonies were randomly selected. Plasmid DNA was ex-tracted and purified from clones with SephacrylTM S-300

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 4

756 Tsushima et al.

(GE Healthcare Bio-Sciences) according to the manufac-turer’s instruction. Almost complete sequencing of the16S rRNA gene inserts (about 1,350 bp) were performedby an automatic sequencer (ABI Prism 3100 Avant Ge-netic Analyzer; Applied Biosystems, CA., USA) with aBigDye terminator Ready Reaction kit (Applied Biosys-tems). All sequences were checked for chimeric artifactsby the CHECK CHIMERA program in the RibosomalDatabase Project[16] and compared with similar sequencesof the reference organisms by a BLAST search.[17] Sequencedata were aligned with the CLUSTAL W package.[18]

Clones with more than 97% sequence similarity weregrouped into the same operational taxonomic unit (OTU),and their representative sequences were used for phyloge-netic analysis. Phylogenetic trees were constructed with theneighbor-joining method.[19] Tree topology was also testedby using the maximum-parsimony method. Bootstrap re-sampling analysis for 1,000 replicates was performed toestimate the confidence of the tree topologies.

Sample fixation

For FISH analysis, intact granular sludge samples werefixed in 4% paraformaldehyde solution for 8 h at 4◦C,washed three times with phosphate-buffered saline (10 mMsodium phosphate buffer, 130 mM sodium chloride [pH7.2]), and embedded in Tissue-Tek OCT (optimal cuttingtemperature) compound (Sakura Finetek, Torrance, CA,USA) overnight to infiltrate the OCT compound into thegranules, as described previously.[20] After rapid freezing at -21◦C, 10-µm-thick slices were prepared by a cryostat. For insitu hybridization, dispersed cells and thin sections of gran-ules were immobilized on glass slides coated with gelatin.[20]

Hybridization was carried out as described previously.[20,21]

Samples hybridized with probes were mounted with theSlow Fade Light antifading kit (Molecular Probes).

Fluorescent images were recorded by a fluorescence mi-croscope (BX51, Olympus, Japan) equipped with a CCDcamera (Diagnostic instruments inc., MI, USA). Imagecombining, processing, and analysis were performed withstandard software packages: IPLab provided by BD Bio-sciences Bioimaging (MD, USA), and Image-PRO pro-vided by Media Cybernetics (MD, USA).

Probe design and in situ hybridization

The 16S rRNA-targeted oligonucleotide probes spe-cific for Methanospirillum, Mspi614 (for cluster 2a) (5′-CTGAACGCCYATCAGTTRAGCCG-3′) and Mspi1425(for cluster 1) (5′-GATACATCCTCATCAAGACCTC-3′),were designed with the ARB program in this study.In addition, the following oligonucleotide probes wereused: (i) EUB338,[22] EUB338II, and EUB338III[23] forthe domain Bacteria; (ii) ARC915[24] for the domain Ar-chaea; (iii) MX825[24] for the family Methanosaetaceae;(iv) MB1174[24] for the family Methanobacteriaceae;

(v) Synbac824[25] for Syntrophobacter; (vi) LGC354A,LGC354B, LGC354C[26] for the phylum Firmicutes(low-G+C Gram-positive bacteria) not including Syn-trophomonas; (vii) GNSB941 (Gich et al.[27]) for the phylumChloroflexi; (viii) ALF1B[28] for the class Alphaproteobac-teria; (ix) BET42a[28] for the class Betaproteobacteria; and(x) GAM42a[28] for the class Gammaproteobacteria.

The oligonucleotide probes were 5′-labeled with indocar-bocyanine (Cy3), tetramethelrhodamine-5-isothiocyanate(TRITC), and flourescein isothiocyanate (FITC) (TaKaRa,Shiga, Japan). The stringency of hybridization was ad-justed by adding formamide to the hybridization buffer(15% for Mspi614 and Mspi1425). To determine the speci-ficity of Mspi614 and Mspi1425, Methanospirillum hungatei(NBRC100397) and an enriched culture of Methanospiril-lum were used. For each probe, the number of hybridizedcells (cells cm−2) was determined by multiplying the totalnumber of cells stained with 4′,6-diamidino-2-phenylindole(DAPI) by the ratio of probe-hybridized cells to total DAPI-stained cells as described previously[29]. This measurementwas performed in duplicate, and the average cell num-bers were reported in this study. Total DAPI-stained cellswere enumerated by the direct-counting method of Hobbieet al.[30] after staining with DAPI. For each probe, the ra-tio of hybridized cells to total DAPI-stained cells was de-termined by direct counting at least 20 randomly chosenmicroscopic fields for each sample, which corresponded to600–900 DAPI-stained cells.

Methanogenic activity determination

The methanogenic activities of retained sludge were de-termined in duplicate at days 0 (seed), 104, 146, 236, and300, with 122 mL serum vial bottles, according to Syut-subo in 1997.[31] The sludge samples for the measurementof activity were washed with a 25 mM phosphate bufferto remove extra substrates and disintegrated by a homog-enizer (anaerobic condition maintained by purging withnitrogen gas). The test substrates were acetate, propionatesucrose, and H2/CO2 (80%:20%, v/v). The initial concen-tration of acetate was 2 g COD/L, and the initial concen-trations of propionate and sucrose were 1 g COD/L. Thevial headspace was filled with N2 gas at 1 atm (101 kPa).For the measurement of hydrogenotrophic activity, the vialheadspace was filled with H2/CO2 gas at 1.4 atm (142 kPa).All vials were incubated on a reciprocal-shaker (120 rpm)at 20◦C and 35◦C.

Nucleotide sequence accession numbers

The GenBank/EMBL/DDBJ accession numbers for the16S rRNA genes used for the phylogenetic tree analysis areAB447621-AB447879.

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 5

Expanded granular sludge bed for wastewater treatment 757

Results

Reactor performance

An EGSB reactor was operated with low-strength wastew-ater (0.6-0.8 gCOD L−1) (see Table 1). Run 1 was operatedfor 104 days at 20◦C seeding with mesophilically (30–35◦C)grown granular sludge. The COD loading was increasedstepwise up to 12 kgCOD m−3 day−1 (0.28 gCOD gVSS−1

day−1) by reducing the hydraulic retention time (HRT) from12 hours to 1.5 hours. After Run 1 had operated for 104days, the retained granular sludge was inoculated into an-other EGSB reactor of 2.0L (Run 2), which was operatedat 15◦C for the first 140 days (from day 104 to day 243),and then at 10◦C until day 346. The COD loading was setat 6 kgCOD m−3 day−1 (0.22 gCOD gVSS−1 day−1) withan HRT of 3 hours except for the first 25 days.

During Run 1, the COD removal efficiency was morethan 80% until day 76, when the organic loading rate (OLR)reached 6.4 kgCOD m−3 day−1. Subsequently, the COD re-moval efficiency decreased as the OLR increased to 11.0 kg-COD m−3 day−1, although the COD removal rate increasedto 7.6 kgCOD m−3 day−1 (0.17 gCOD gVSS−1 day−1).

During Run 2, the COD removal rate dropped after thedecrease in operation temperature, whereas the COD re-moval rate gradually increased over the course of time andexhibited sufficient performance until the temperature wasdropped to 10◦C. In the later part of each operational tem-perature, the average COD removal rate recovered from 3.0kgCOD m−3 day−1to 4.3 kgCOD m−3 day−1 at 15◦C andfrom 2.4 kgCOD m−3 day−1to 3.8 kgCOD m−3 day−1 at10◦C.

At the end of Run 2, the process temperature was reducedto 5◦C. Due to the 5◦C-operation, the methanogenic activ-ity of the retained sludge was reduced significantly, and theresidual VFA concentration in the effluent was increased.In addition, the physical properties of the granular sludge,such as sludge settleability and sludge concentration, weredeteriorated (data not shown), whereas no sucrose was de-tected in the effluent even in 5◦C-operation.

Microbial diversity of granular sludge as determined byphylogenetic analysis

Phylogenetic analysis of the microbial community in thegranular sludge indicated that there were major differ-ences in the bacterial and archaeal populations betweenday 0 (seed granular sludge) and day 346 (low tempera-ture adapted granular sludge) (Table 2). We obtained 139bacterial clones and 120 archaeal clones from the granularsludge: 66 bacterial clones and 33 archaeal clones on day 0,and 73 bacterial clones and 87 archaeal clones on day 346.

Regarding the result of the bacterial analysis of thesludge on day 0, the 66 clones were affiliated with a ma-jor family: namely, Syntrophobacter, Propionibacterineae,Treponemaceae, and Alphaproteobacteria, with a distribu-

tion of 51.5%, 13.6%, 10.6%, and 7.6% of the total bacterialclones, respectively. On the other hand, among the 73 clonesin the day-346 sludge, the most frequent phylum detectedwas the Firmicutes (45.2%), including the families Strepto-coccaceae (20.5%) and Acidaminococcaceae (15.0%). Theother minor families detected were Geobacteraceae (9.6%),Bacteroidales (9.6%) and Syntrophobacteraceae (6.8%).Small percentages of other candidate phyla were detected,i.e., WCHB1-32, b-17BO, WD272 CD2, CCM11b andOP11. Specifically, the clones that belonged to Streptococ-caceae were related to Lactococcus sp. F116 (EF204365)with 99% sequence similarity, and the clones that belongedto Acidaminococcaceae were related to Anaerovibrio burk-inabensis DSM 6283(T) (AJ010961) with 99% sequencesimilarity.

In the archaeal clone analysis (Table 2 and Fig. 1), the33 clones from the day-0 sludge and the 87 clones fromthe day-346 sludge were affiliated with 12 operational taxo-nomic units (OTUs) belonging to the families Methanosae-taceae and Methanobacteriaceae, with a clone distributionof 69.7% and 27.3% on day 0, and 64.4% and 16.1% on day346, respectively. It is striking that the clones in OTU 8 andOTU 9 that belong to the Methanospirillacea (in cluster 1and 2a, respectively in Fig. 2) were detected only on day 346with a distribution of 2.3% and 10.3% of the total 87 clones,respectively. According to previous studies, Methanospiril-lacea are divided into three clusters (cluster 1, 2a, and 2b)according to their growth temperatures. The clones in OTU8 (in cluster 1) were related to Uncultured Methanospir-illaceae clone AR-Eth-B (AB236073) with 99% sequencesimilarity and to Methanospirillum hungatei (M60880) with98% sequence similarity, known as a mesophilic H2/CO2-utilizing methanogen. The clones in OTU 9 were related toUncultured Methanospirillaceae archaeon gene LF-H2-A(AB236096), obtained from lotus field sediment, with 97%sequence similarity and Methanospirillum sp. strain TM20-1, isolated from paddy field soil by 20◦C enrichment withH2/CO2,

[32] with 97% sequence similarity.

Population dynamics and spatial structure as determined byFISH

The composition of the microbial community determinedby FISH relative to the total number of cells stained byDAPI (4′,6-diamidino-2-phenylindole) in sludge from day0 and day 346 is shown in Figure 2. The abundance of totalDAPI-stained cells was 3.9 (± 1.3) × 1014 cell g-VSS−1 (±SD) and 2.0 (± 0.2) × 1016 cell g-VSS−1 in the day-0 sludgeand day-346 sludge, respectively.

The archaeal cells (ARC915) are mainly comprised by thefamily Methanobacteriaceae (MB1174), the Methanosaetaspecies (MX825), the Methanospirillum species belongingto cluster 2a (Mspi614), and the mesophilic Methanospir-illum species (Mspi1425) in the day-0 sludge and day-346 sludge. In order to detect the genes of Methanospir-illum cells, the probes Mspi614 and Mspi1425 were newly

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 6

Tab

le1.

EG

SBre

acto

rpe

rfor

man

cean

dop

erat

ing

cond

itio

nsdu

ring

the

cont

inuo

usflo

wex

peri

men

t.

Tot

also

lubl

eC

OD

Loa

din

efflu

ent

(Res

idua

lVFA

rati

o)R

eact

orte

mp.

Tim

eH

RT

OL

RC

OD

slud

gelo

adin

gS

peci

ficC

OD

rem

oval

rate

Bio

mas

sco

nc.

CO

Dre

mov

alra

teC

OD

T-C

Ha 4

S-C

Hb 4

(gm

−3da

y−1)

(◦ C)

(Day

s)(H

ours

)(k

gm

−3da

y−1)

(g-C

OD

g-V

SS

−1da

y−1)

(g-C

OD

g-V

SS

−1da

y−1)

(g-V

SS

L−1

-rea

ctor

)(k

gm

−3da

y−1)

rem

oval

(~E

)co

nv.(

%)

conv

.(%

)(%

)

20◦ C

0–6

11.9

1.6

0.05

0.04

34.4

1.4

83.8

(±9.

1)58

.5(±

18.4

)10

.1(±

2.1)

149

(79%

)(R

un1)

7–34

6.0

3.1

0.08

0.07

36.5

2.7

87.3

(±4.

0)65

.1(±

4.3)

10.9

(±0.

6)99

(90%

)35

–50

4.0

4.4

0.11

0.09

41.9

3.9

87.5

(±1.

8)66

.2(±

4.5)

11.5

(±0.

6)92

(90%

)51

–76

2.5

6.4

0.14

0.11

46.4

5.3

82.3

(±1.

6)60

.0(±

6.2)

13.8

(±1.

8)11

8(9

4%)

77–1

041.

511

0.25

0.17

43.6

7.6

69.3

(±2.

0)47

.2(±

4.8)

15.3

(±2.

0)20

6(8

1%)

15◦ C

104–

124

4.2

4.0

0.12

0.09

32.6

3.0

73.5

(±2.

5)54

.5(±

4.5)

12.1

(±0.

7)18

5(1

20%

)(R

un2)

125–

154

2.9

5.4

0.17

0.12

31.6

3.7

68.9

(±2.

5)48

.1(±

3.2)

12.8

(±1.

0)21

2(9

0%)

155–

204

2.9

5.6

0.18

0.12

31.5

3.9

69.3

(±5.

6)50

.3(±

5.0)

12.4

(±1.

1)21

2(6

5%)

205–

243

2.9

5.4

0.17

0.14

31.0

4.3

78.5

(±3.

7)52

.9(±

6.3)

12.3

(±1.

1)14

6(8

4%)

10◦ C

244–

249

5.9

3.0

0.10

0.08

29.8

2.4

78.0

(±1.

9)62

.5(±

9.4)

11.8

(±1.

3)15

5(7

9%)

(Run

2)25

0–27

04.

04.

30.

150.

1128

.33.

172

.0(±

1.7)

48.1

(±5.

8)11

.8(±

1.4)

207

(102

%)

271–

304

3.0

5.8

0.21

0.13

27.8

3.6

62.3

(±4.

1)38

.3(±

2.8)

11.6

(±1.

0)26

9(9

4%)

305–

346

2.9

60.

240.

1525

.33.

863

.0(±

2.4)

42.8

(±5.

8)12

.6(±

0.9)

270

(94%

)5◦ C

347–

354

5.9

3.3

0.12

0.08

28.9

2.4

70.4

(±2.

9)47

.5(±

7.6)

13.8

(±1.

8)18

9(7

9%)

(Run

2)35

5–39

53.

94.

30.

160.

0927

.12.

557

.6(±

1.4)

32.4

(±4.

3)12

.7(±

1.3)

309

(90%

)39

6–50

26.

02.

80.

100.

0727

.11.

963

.2(±

5.4)

32.8

(±6.

6)13

.0(±

1.8)

265

(77%

)

aTo

talC

H4.

bSo

lubl

eC

H4.

758

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 7

Expanded granular sludge bed for wastewater treatment 759

Table 2. Distribution of 16S rRNA clones detected in the anaerobic granular sludge.

Day 0 Day 346Clone number (Percentage of total clones)

GammaproteobacteriaPseudomonadaceae 1 (1.4)

AlphaproteobacteriaVerorhodospirilla 5 (7.6) 1 (1.4)

DeltaproteobacteriaSyntrophaceae 2 (2.7)Syntrophobacteraceae 34 (51.5) 5 (6.8)Geobacteraceae 7 (9.6)Desulfovibrionaceae 2 (2.7)Desulfomicrobiaceae 1 (1.4)Nitrospira 2 (3.0)

ActinobacteriaPropionibacterineae 9 (13.6)

BacteroidetesBacteroidales 5 (7.6) 7 (9.6)

WCHB1-32 3 (4.1)b-17BO 1 (1.4)WD272 CD2 2 (2.7)CCM11b 1 (1.4)Sprirochaetes

Treponemaceae 7 (10.6)Leptospirales 1 (1.4)

ChloroflexiT78 4 (6.1) 2 (2.7)Anaerobic filamentous bacterium YMTK-2 2 (2.7)

FirmicutesLachnospiraceae(Sporobacter) 2 (2.7)Clostridium botulinum et al. 1 (1.4)RC6 (4C0d-14) 1 (1.4)RC6 1 (1.5) 1 (1.4)Eubacterium 1 (1.4)Streptococcaceae (Lactococcus) 15 (20.5)Acidaminococcaceae 11 (15.0)Syntrophomonadaceae 1 (1.4)

AminanaerobiaTTA B6 1 (1.4)

OP11-5 (ABY1) 1 (1.4)Total Clone number 66 73

MethanosarcinalesMethanosarcinaceae 2 (2.3)Methanosaetaceae 23 (69.7) 56 (64.4)

MethanomicrobialesMethanospirillaceae 12 (13.8)Methanomicrobiaceae 3 (3.4)Uncultured Methanomicrobiales 1 (3.0)

MethanobacteialesMethanobacteriaceae 9 (27.3) 14 (16.1)

Total Clone number 33 87

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 8

760 Tsushima et al.

Fig. 1. Phylogenetic tree of Archaea showing the positions of the clones obtained from the granular sludge on day 0 and day 346.The tree was generated by using 1391-bp of the 16S rRNA sequence and the neighbor-joining method. The scale bar represents 5%sequence divergence. The filled and empty circles at the nodes represent bootstrap values higher than 95% and 80%, respectively (1000times resampling analysis). The E.coli gene sequence served as the outgroup for rooting the tree. Numbers in parentheses indicate thefrequency of appearance of identical clones among the total clones analyzed with a specific primer set.

designed in this study. The specificities of these probeswere tested using an enriched culture of Methanospirillumspecies belonging to cluster 2a and Methanospirillum hun-gatei (NBRC100397) (data not shown).

Cells of the Bacteria (EUB338mix) domains, thephyla Alphaproteobacteria (ALF1B), Betaproteobacteria(BET42a), Gammaproteobacteria (GAM42a) and Firmi-

cutes (LGC354mix; LGC354A + LGC354B + LGC354C),the Syntrophobacter group (Synbac824) and Chloroflexi(GNSB941) were present in the day-0 and day-346 sludge.

Less than 0.5% of the cells were hybridized byHGC69a (Actinobacteria), Synm700 (Syntrophomon-adaceae), and MS1414 (Methanosarcina, Methanococ-coides, and Methanolobus) (data not shown).

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 9

Expanded granular sludge bed for wastewater treatment 761

Fig. 2. Relative abundance of microbial community components in anaerobic granular sludge on day 0 and day 346 analyzedby FISH using specific probes and DAPI staining as a reference. Specific probes were used for Methanobacteriaceae (MB1174),Methanosaeta of Methanosarcinaceae (MX825), Methanospirillum species that belong to cluster 2a (Mspi614), Methanospirillumspecies that belong to cludter 1 (Mspi1425), Alphaproteobacteria (ALF1B), Betaproteobacteria (BET42a), Gammaproteobacteria(GAM42a), Firmicutes/low-G+C Gram-positive bacteria (LGC354mix; LGC354A + LGC354B + LGC354C), Syntrophobactergroup (Synbac824), and Chloroflexi (GNSB941). In Archaea (upper) and Bacteria (bottom), “Others” refers to the amount of cellshybridized by each specific probe relative to the amount of cells hybridized by the probe ARC915 and EUB338mix (EUB338 +EUB338II + EUB338III), respectively. Error bars indicate the standard deviation of duplicate measurements.

The relative abundance was determined by calculatingthe percentage of one specific group of cells in relationto the total number of DAPI-stained cells. This analysisshowed that the ARC915 probe detected the entire major-ity of archaeal cells in day-0 sludge and day-346 sludgeas compared with domain bacterial cells. The archaeal do-main corresponded to 50.9% in the day-0 sludge and 61.2%in the day-346 sludge. The acetate-utilizing Methanosaetacells (MX825) predominated in the archaeal population onboth day 0 and day 346. The percentage of Methanosaetacells increased slightly from 27.1% (day 0) to 29.4% (day346). The percentage of Methanospirillum cells belonging tocluster 2a (Mspi614) clearly increased from 0.5% (day 0) to6.7% (day 346), whereas the percentage of Methanospir-illum cells belonging to cluster 1 (Mspi1425) decreasedfrom 5.4% (day 0) to 3.1% (day 346), although the cellnumber slightly increased from 5.9 × 1010 cell g-VSS−1 to2.2 × 1011 cell g-VSS−1. The percentage of Methanobac-terium cells (MB1174), another H2-utilizing methanogen,increased from 2.0% (day 0) to 4.5% (day 346). Meanwhile,bacterial cells detected with the EUB338mix probe werealso present in the samples, accounting for 31.8% (day 0)and 25.8% (day 346). The percentage of members of thephylum Firmicutes (detected with LGC354mix) increasedfrom 0.7% (day 0) to 8.0% (day 346).

Cross-sectional images showed that the granules had amultilayered structure consisting of a biomass and intersti-tial voids (Fig. 3A). FISH images of vertical cross-sectionsof the granular sludge samples indicated that bacterial cellshybridized by the probe EUB338mix (green fluorescence)were abundant in the outermost thin layer (10-150 µm) and

in the center of granules with high fluorescence intensity.Archaeal cells hybridized by the probe ARC915 (red fluo-rescence) were frequently detected in the inner layer of thegranules (at a depth of ca. 350 µm below the surface), ata position deeper than the bacterial cells found on the sur-face. This microbial structure was observed repeatedly inall of the granular sections analyzed. Long rod-shaped cellsand spherical bacteria were observed on the surface of theoutermost thin layer, and these cells were hybridized by theprobe LGC354mix (yellow-fluorescence), specific for Fir-micutes (Fig. 3B). A few filamentous cells were observedin the outermost thin layer and these cells hybridized bythe probe GNSB941, which is specific for almost all mem-bers of the phylum Chloroflexi (data not shown). Figure3C showed that Methanospirillum belonging to cluster 2aand Methanobacteriaceae cells, hybridized by the probesMspi614 (green fluorescence) and MB1174 (red fluores-cence), respectively, which are H2-utilizing methanogens,were present in the archaeal layer. Methanobacteriaceaewere detected much less frequently than Methanospirillumbelonging to cluster 2a. Methanobacateriaceae tended tobe observed inside the granules, although they were dis-tributed unevenly. Methanospirillum belonging to cluster2a were detected, along with saccharolytic bacteria, hy-bridized by the probe LGC354mix (red fluorescence) in thesurface and second layers (Fig. 3D).

A number of Methanobacteriaceae cells were detectedin the central part of the granular samples and co-existed with propionate-oxidizing Syntrophobacteraceaecells hybridized by the probe Synbac824 (green fluores-cence) (Fig. 3E). A small amount of Syntrophobacteraceae

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 10

762 Tsushima et al.

Fig. 3. Microscope images of thin sections of anaerobic granules obtained on day 346 showing the in situ spatial organization ofbacteria and archaea. (A) FISH with the Cy3-labeled probe ARC915 (red) and the FITC-labeled EUB338mix probe (green). (B) FISHwith the TRITC-labeled LGC354mix probe (red) and the FITC-labeled EUB338-mixed probe (green). (C) FISH with the TRITC-labeled MB1174 probe (red) and the FITC-labeled Mspi614 probe (green). (D) DAPI staining and FISH with the Cy3-labeled probeLGC354mix (red) and the FITC-labeled Mspi614 probe (green). (E) FISH with the TRITC-labeled MB1174 probe (red) and theFITC-labeled Synbac824 probe (green). (F) DAPI staining and FISH with the Cy3-labeled probe Mspi614 (red) and the FITC-labeledSynbac824 probe (green). (G) DAPI staining and FISH with the Cy3-labeled probe Mspi1425 (red) and the FITC-labeled Mspi614probe (green).

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 11

Expanded granular sludge bed for wastewater treatment 763

Fig. 4. Methanogenic activities of retained sludge with H2/CO2,acetate and propionate at 20◦C on day 0, 104, 146, 236, and 300with 122 mL serum vial bottles.

(green fluorescence) co-existed in the second layer withMethanospirillum belonging to cluster 2a (red fluorescence)(Fig. 3F). However, a significant symbiotic relationshipsuch as Figure 3F was not observed. Figure 3G shows thatMethanospirillum belonging to cluster 2a (hybridized byprobe Mspi614 (green)) were present on the surface of thearchaeal layer, a small amount of Methanospirillum that be-long to cluster 1 (hybridized by probe Mspi1425 (red)) werelocated in the deep part of the layer where Methanospiril-lum that belong to cluster 2a existed. The morphology ofthe Methanospirillum cells was a curved rod in the granularsludge sample in the UASB reactor; by contrast, these cellsoccasionally formed wavy filaments over 100 µm in lengthin a dilution culture (data not shown).

Methanogenic activities of retained sludge

The methanogenic activities of the retained sludge wereinvestigated at 20◦C and 35◦C. The test substrates usedwere H2/CO2, acetate, and propionate. Figure 4 shows thatmethanogenic activities at both 20◦C and 35◦C increasedas the reactor operation progressed. At 20◦C, the activitiesof seed sludge (day 0) fed with H2/CO2, acetate, and propi-

onate were 0.056, 0.014, and 0.021 g-COD g-VSS−1 day−1,respectively. These activities increased to 0.157 (H2/CO2),0.256 (acetate), and 0.122 (propionate) g-COD g-VSS−1

day−1 on day 104, and to 0.995 (H2/CO2), 0.598 (acetate),and 0.245 (propionate) g-COD g-VSS−1 day−1 on day 300.However, at 35◦C, the activities of seed sludge fed withH2/CO2, acetate, and propionate were 0.402, 0.115, and0.057 g-COD g-VSS−1 day−1, respectively, on day 0. Theseactivities increased to 0.627 (H2/CO2), 0.864 (acetate), and0.303 (propionate) g-COD g-VSS−1 day−1 on day 104, andto 1.732 (H2/CO2), 1.942 (acetate), and 0.533 (propionate)g-COD g-VSS−1 day−1 on day 300.

From the 104th day until the 146th day (20◦C), themethanogenic activity increased 6.8 times with H2/CO2 fedseed sludge, while the activity with acetate and propionatefed seed sludge increased 1.7 times. After the 146th day, theactivities with H2/CO2 fed seed sludge were higher thanthose with acetate fed seed sludge. Methanogenic activitiesat 35◦C were higher than those at 20◦C. Additionally, themethanogenic activities with acetate fed seed sludge werehigher than those with H2/CO2 fed seed sludge, except forday 236 (35◦C).

Discussion

In this study, we operated the EGSB reactors fed withlow-strength wastewater under low-temperature conditionsover the course of 300 days. As a result, the reactors ex-hibited sufficient performance, and we were able to char-acterize the microbial structure and population dynam-ics of the granular sludge and confirm the increment ofmethanogenic activity during the low-temperature opera-tion. In this study, the COD removal efficiency was 60–65% (at COD loading of 0.21–0.24 kgCOD m−3 day−1),the COD removal rate was 3.1–3.8 kgCOD m−3 day−1, thespecific COD removal rate was 0.11–0.15 gCOD gVSS−1

day−1, and the total methane conversion efficiency was 40–50% at 10◦C. In particular, the specific COD removal rateincreased gradually through the low-temperature opera-tion, and a level similar to that achieved at 20◦C and 15◦Cwas attained in the later stage of the 10◦C operation (days305–346). This finding is especially practical for applica-tions of psychrophilic methane fermentation. Previously,the sludge retention time of the Run 1 reactor operated at20◦C had been estimated to be around 40 days at an HRTof 1.5 h.[33] This sufficient maintenance of sludge retentionis effective for the accumulation of bacteria present in thegranular sludge at low temperatures.

The microbial structure of granular sludge in a reactorthat had operated for 346 days at low temperatures (10–20◦C) was, for the first time to our knowledge, revealedusing clone analysis and FISH analysis. These analysesshowed that Bacteria dominated the surface layer and Ar-chaea dominated the interior layer in the granular sludge,

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 12

764 Tsushima et al.

in common with the mesophilic and thermophilic granu-lar sludges previously reported by Sekiguchi et al.[34] andSatoh et al.[35] However, we observed that several fea-tures differed between our low-temperature granule andthe meso/thermophilic granules. In this study, the surfacelayer dominated by Bacteria was thinner than that in themeso/thermophilic granular sludge. This is likely becausethe number of Bacteria decreased with the lower temper-atures or lower organic concentrations. The percentage ofBacteria in relation to the total number of cells stained withDAPI decreased from 31.8% (day 0) to 25.8% (day 346),whereas the relative percentage of Archaea increased from50.9% to 61.2%. This value is approximately 10% higherthan that in the meso/thermophilic granular sludge, whichwas fed with the same substrate (sucrose and VFA).[36]

The sufficient performance achieved at low temperaturesis likely due to the high concentration of Archaea enrichedin the granules. In addition, Figure 2 indicates that the ra-tio of the dominant Alphaproteobacteria, Betaproteobac-teria, Gammaproteobacteria and Syntrophobacter bacteriadecreased during low-temperature operation, whereas thatof Firmicutes greatly increased during this period (Table 2),especially in the surface layer (Fig. 3B). According to thephylogenetic analysis, bacteria belonging to the phylumFirmicutes were assumed to be Anaerobivrio sp. and Lac-tococcus sp. Much of the population of Bacteria exceptFirmicutes was reduced in this study, whereas the popula-tion of Anaerobivrio sp. and Lactococcus sp. were increased.It is suggested that these bacteria would contribute to thedegradation of the sucrose substrate anaerobically at am-bient temperatures as described in other studies.

Anaerovibrio burkinabensis isolated from an ambient en-vironmental sample grows on a wide range of saccha-ride substrates.[37] Lactococcus sp. is a characteristic fac-ultative anaerobe that is able to degrade sucrose duringpsychrophilic fermentation. McHugh et al.[8] reported pre-viously that Streptococcus, which is related to Lactococcus,was detected in a psychrophilic anaerobic digestion reactorfed with only a sucrose substrate. In addition, the FISHresults of this study revealed that bacteria hybridized bythe probe EUB338mix existed at a relatively high intensityin the center of the granule. This finding differs from re-ports of other meso/thermophilic granules by Sekiguchi etal.[34] and Satoh et al.[35] Generally, in meso/thermophilicfermentation, organisms in the seeding sludge are digestedthemselves and a new microbial community develops im-mediately. At low temperatures, however, the structure ofthe seeding granular sludge remained for a while, indicatingthat no significant self-digestion occurred. Moreover, newbiofilms formed on the surface of the seed granular sludge.In this study, as shown Figure 3A, the granular sludge thatdeveloped at low temperatures had a double structure, con-sisting of a bacterial layer in the center and an archaeal layerin the middle divided by large cavities. This is probably be-cause the bacteria that were present originally in the seedsludge still remained in the center of granule, although they

were digested gradually due to the low temperatures. Someof these bacteria have been identified as Syntrophobacterin this study. Further studies are needed to clarify the mi-crobial function in the core part of the granule in order todetermine whether other bacteria have some effect.

Collins et al.[6] reported that Crenarchaeota were de-tected by phylogenetic analysis in granular sludge from apsychrophilic anaerobic digestion reactor, and McHughet al.[8] reported that Methanocorpusculum parvum was de-tected, whereas in the present study these archaea were notdetected. Interestingly, Methanospirillum species related toMethanospirillum sp. strain TM20-1 with 97% sequencesimilarity were detected in granular sludge from our reac-tors, indicating that this species could be most likely clas-sified as a species of psychro-tolerant methanogens. In ad-dition, an increment in this Methanospirillum species wasobserved by FISH analysis; that is, the percentage of cellshybridized by the probe Mspi614 in relation to the totalcells stained with DAPI increased drastically as comparedwith other species during operation, even allowing for therelatively large error bars as shown in Figure 2.

Lettinga et al.[2] reported that Methanospirillum-like cellswere observed in biomass incubated at 10◦C with H2/CO2inoculated from the sludge from EGSB reactor under low-temperature conditions fed with VFA (propionate, acetateand butyrate.). We observed that these cells co-existedwith saccharolytic bacteria in the outer layer in the granu-lar sample and Methanobacteriaceae co-existed with Syn-trophobacter in the deeper part of the granular sample,whereas a low number of Syntrophobacter co-existed in theouter layers with Methanospirillum belonging to cluster 2a.Methanospirillum belonging to cluster 1 and Methanobac-teriaceae were present in the deeper part of the granularsample as compared with the dominant Methanospirillumspecies belonging to cluster 2a. This finding indicates thatMethanospirillum species belonging to cluster 2a grow byusing H2 produced by saccharolytic bacteria in preferenceof the propionate-oxidizing bacteria in the outer layer dur-ing the low-temperature operations. Similarly, it is possiblethat Methanobacteriaceae consumed the H2that was pro-duced by propionate-oxidizing bacteria located near thecenter of the granular sludge. However, the bacteria de-tected in this part of the granule would have disappearedover time during the operation.

Additionally, the amount of acetate-utilizingMethanosaeta hybridized with probe MX825 (29.5%of total DAPI-stained cells) was approximately twice thetotal amount of H2/CO2-utilizing methanogens hybridizedwith MB1174, Mspi614 and Mspi1425 (Fig. 2). As a resultof the methanogenic activity tests at 20◦C, the activity fedwith acetate was higher than the activity fed with H2/CO2during the initial stage, whereas from the 146th day, the ac-tivity fed with H2/CO2 was higher than the activity fed withacetate. However, as a result of the methanogenic activitytests at 35◦C, the activities fed with acetate were higherthan those fed with H2/CO2, suggesting that H2/CO2

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 13

Expanded granular sludge bed for wastewater treatment 765

utilizing psychro-tolerant mesophilic methanogens wereincreased (Fig. 4). Therefore, the increase of the activityfed with H2/CO2 was likely due to the increment ofMethanospirillum species belonging to cluster 2a duringlow-temperature operations, which was observed by FISHanalysis, indicating that these Methanospirillum species arelikely able to proliferate under psychrophilic conditions.

The Methanospirillum species (cluster 2a) and Firmicutes(Anaerobivrio sp. and Lactococcus sp.) detected in this studymay have a key role to contribute to the treatment of sugar-containing wastewater at low temperatures. However,further studies on the isolation and/or the granulation ofthese enriched Methanospirillum species are needed.

Conclusion

In conclusion, this study demonstrated that the EGSB re-actors could operate steadily while fed with low-strengthwastewater under low temperatures (< 20◦C). The commu-nity and special structures in the granule were also ana-lyzed. The increment of methanogenic activity during low-temperature operations was confirmed. Anaerobivrio sp.and Lactococcus sp. belonging to the phylum Firmicutesand Methanospirillum species related to Methanospirillumsp. strain TM20-1 were suggested to be increased in thegranular sludge. These results will provide profound in-sights into the optimal operation of psychrophilic anaero-bic fermentation reactors.

Acknowledgments

A part of this study was supported by New Energy and In-dustrial Technology Development Organization (NEDO).This study was also supported by a special-research pro-gram of NIES (National Institute for Environmental Stud-ies). We are grateful to Ms. Kaori Oki and Ms. Yue-QinFang for their technical assistance.

References

[1] Lin, C.Y.; Noike, T.; Sato, K.; Matsumoto, J. Temperature char-acteristics of the methanogenesis process in anaerobic digestion.Water. Sci. Technol. 1987, 19, 299–310.

[2] Lettinga, G.; Rebac, S.; Parshina, S.; Nozhevnikova, A.; van Lier,J. B.; Stams, A.J.M. High-rate anaerobic treatment of wastewater atlow temperature. Appl. Environ. Microbiol. 1999, 65,1696–1702.

[3] Rebac, S.; Ruskova, J.; Gerbens, S.; van Lier, J.B.; Stams, A.J.M.;Lettinga, G. High-rate anaerobic treatment of wastewater underpsychrophilic conditions. J. ferment. Bioeng. 1995, 80, 499–506.

[4] Lettinga, G.; Rebac, S.; Zeeman, G. Challenge of psychrophilicanaerobic wastewater treatment. Trends Biotechnol. 2001, 19, 363–370.

[5] Elmitwalli, T.A.; Sklyar, V.; Zeeman, G.; Lettinga, G. Low temper-ature pre-treatment of domestic sewage in an anaerobic hybrid oran anaerobic filter reactor. Bioresour. Technol. 2002, 82, 233–239.

[6] Collins, G.; Woods, A.; McHugh, S.; Carton, M.; O’Flaherty, V.Microbial community structure and methanogenic activity dur-ing start-up of psychrophilic anaerobic digesters treating syntheticindustrial wastewaters. FEMS Microbiol. Ecol. 2003, 46, 159–170.

[7] Collins, G.; Foy, C.; McHugh, S.; O’Flaherty V. Anarobic treat-ment of 2,4,6-trichlorophenol in an expanded granular sludge bed-anarobic filter (EGSB-AF) reactor at 15◦C. FEMS Microbiol. Ecol.2005, 53, 167–178.

[8] McHugh, S.; Carton, M.; Collins, G.; O’Flagerty V. Reactor perfor-mance and icrobial community dynamics during anaerobic biolog-ical treatment of wastewaters at 16-37◦C. FEMS Microbiol. Ecol.2004, 48, 369–378.

[9] Yoochatchaval, W.; Nishiyama, K.; Okawara, M.; Ohashi., A.;Harada, H.; Syutsubo, K. Influence of effluent-recirculation con-dition on the process performance of expanded granular sludge bedreactor for treating low strength wastewater. Water Sci. Technol.2008, 57 (6), 869–873.

[10] Kato, M. T.; Rebac, S.; Lettinga, G. Anaerobic treatment of low-strength brewery wastewater in expanded granular sludge bed reac-tor. Appl. Biochem. Biotechnol. 1999, 76(1), 15–32.

[11] Rebac, S.; van Lier, J. B.; Janssen, M.G.J.; Dekkers, F.; Swinkels,K. Th. M.; Lettinga, G. High-rate anaerobic treatment of malt-ing wastewater in a pilot-scale EGSB system under psychrophilicconditions. J. Chem. Tech. Biochnol. 1997, 68, 135–146.

[12] Rebac, S.; van Lier, J. B.; Lens, P.; van Cappellen, J.; Vermeulen,M.; Stams, A.J.M.; Dekkers, F.; Swinkels, K. Th. M.; Lettinga, G.Psychrophilic (6–15◦C) high-rate anaerobic treatment of maltingwastewater in a two-module expanded granular sludge bed system.Biotechnol. Prog. 1998, 14, 856–864.

[13] McHugh, S.; Collins, G.; O’Flagerty, V. Long-term, high rate anaer-obic biological treatment wastewaters at psychrophilic tempera-tures. Bioresour. Technol. 2006, 97, 1669–1678.

[14] Connaughton, S.; Collins, G.; O’Flaherty, V. Development of mi-crobial community structure and activity in a high-rate anaerobicbioreactor at 18◦C. Water Res. 2006, 40, 1009–1017.

[15] Weisberg, W.; Barns, S.; Pelletier, D.; Lane, D. 16S ribosomal DNAamplification for phylogenetic study. J. Bacteriol. 1991, 173, 697–703.

[16] Maidak, B.L.; Olsen, G.L.; Larsen, N.; Overbeek, R.; McCaughey,M. J.; Woese, C. R. The RDP (Ribosomal Database Project). Nuc.Acids Res. 1997, 25, 109–110.

[17] Altschul, S.F.; Gish, W., Miller, W.; Myers, E.W.; Lipman, D.J. Basiclocal alignment search tool. J. Mol. Biol. 1990, 215, 403–410.

[18] Thompson, J.D.; Higgins, D.G.; Gibson, T.J. CLUSTAL W: im-proving the sensitivity of progressive multiple sequence alignmentthrough sequence weighting, position-specific gap penalties andweight matrix choice. Nucl. Acids Res. 1994, 22, 4673–4680.

[19] Saito, N.; Nei, M. The neighbor-joining method: a new methodfor constructing phylogenetic trees. Mol. Biol. Evol. 1987, 4, 406–425.

[20] Okabe, S.; Satoh, H.; Watanabe, Y. In situ analysis of nitrify-ing biofilms as determined by in situ hybridization and the useof microelectrodes. Appl. Environ. Microbiol. 1999, 65, 3182–3191.

[21] Wagner, M.; Amann, R. I.; Kampfer, P.; Assmus, B.; Hartmann, A.;Hutzler, P.; Springer, N.; Schleifer, K.-H. Identification and in situdetection of gram-negative filamentous bacteria in activated sludge.Syst. Appl. Microbiol. 1994, 17, 405–417.

[22] Amann, R.I.; Binder, B.J.; Olson, R.J.; Chisholm, S.W.; Devereux,R.; Stahl, D.A. Combination of 16S rRNA-targeted oligonucleotideprobes with flow cytometry for analyzing mixed microbial popula-tions. Appl. Environ. Microbiol. 1990, 56, 1919–1925.

[23] Daims, H.; Bruhl, A.; Amann, R.; Schleifer, K.-H.; Wagner, M. Thedomain-specific probe EUB338 is insufficient for the detection ofall bacteria: Development and evaluation of a more comprehensiveprobe set. Syst. Appl. Microbiol. 1999, 22, 434–444.

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010

Page 14

766 Tsushima et al.

[24] Raskin, L.; Stromley, J.M.; Rittmann, B.E.; Stahl, D.A. Group-specific 16S rRNA hybridization probes to describe natural com-munities of methanogens. Appl. Environ. Microbiol. 1994, 60,1232–1240.

[25] Ariesyady, H.D; Ito, T.; Okabe, S. Functional bacterial and archaealcommunity structures of major trophic groups in a full-scale anaer-obic sludge digester. Water Res. 2007, 41, 1554–1568.

[26] Meier, H.; Amann, R.; Ludwig, W.; Schleifer, K.-H. Specificoligonucleotide probes for in situ detection of a major group ofgram-positive bacteria with low DNA G+C content. Syst. Appl.Microbiol. 1999, 22, 186–196.

[27] Gich, F.; Garcia-Gil, J.; Overmann, J. Previously unknown andphylogenetically diverse members of the green nonsulfur bacteriaare indigenous to freshwater lakes. Arch. Microbiol. 2001, 177, 1–10.

[28] Manz, W.; Amann, R.; Ludwig, W.; Wagner, M.; Schleifer, K.-H.Phylogenetic oligodeoxynucleotide probes for the major subclassesof Proteobacteria: problems and solutions. Syst. Appl. Microbiol.1992, 15, 593–600.

[29] Savichtcheva, O.; Okayama, N.; Ito, T.; Okabe, S. Application ofa direct fluorescence-based live/dead staining combined with flu-orescence in situ hybridization for assessment of survival rate ofBacteroides spp. in drinking water. Biotechnol. Bioeng. 2005, 92,356–363.

[30] Hobbie, J. E.; Daley, R. J.; Jasper, S. Use of nuclepore filters forcounting bacteria by fluorescence microscopy. Appl. Environ. Mi-crobiol. 1977,33, 1225–1228.

[31] Syutsubo, K.; Harada, H.; Ohashi, A.; Suzuki, H. An effective start-up of thermophilic UASB reactor by seeding mesophilically-growngranular sludge. Wat. Sci. Tech. 1997, 36 (6–7), 391–398.

[32] Tonouchi, A. Isolation and characterization of a motile hy-drogenotrophic methnogen from rice paddy field soil in Japan.FEMS Microbiol. Lett. 2002, 208, 239–243.

[33] Yoochatchaval, W.; Ohashi, A.; Harada, H.; Yamaguchi, T.; Syut-subo, K. Characteristics of granular sludge in an EGSB reactor fortreating low strength wastewater at 20◦C. Int. J. Environ. Res. 2008,2(4), 319–328.

[34] Sekiguchi, Y.; Kamagata, Y.; Nakamura, K.; Ohashi, A.; Harada,H. Fluorescence in situ hybridization using 16S rRNA-targetedoligonucleorides reveals localization of methanogens and selecteduncultured bacteria in mesophilic and thermophilic sludge granules.Appl. Environ. Microbiol. 1999, 65, 1280–1288.

[35] Satoh, H.; Miura, Y.; Tsushima, I.; Okabe, S. Layered structureof bacterial and archaeal communities and their in situ activi-ties in anaerobic granules. Appl. Environ. Microbiol. 2007, 73,7300–7307.

[36] Tagawa, T.; Syutsubo, K.; Sekiguchi, Y.; Ohashi, A.; Harada, H.Quantification of methanogen cell density in anaerobic granularsludge consortia by fluorescence in-situ hybridization. Wat. Sci.Technol. 2000, 42, 77–82.

[37] Ouattara, A.S.; Traore, A.S.; Garcia, J.L. Characterization ofAnaerovibrio burkinabensis sp. nov., a lactate-fermenting bacteriumisolated from rice field soils Int. J. Syst. Bacteriol. 1992, 42, 390–397.

Downloaded By: [Syutsubo, Kazuaki][National Inst for Environmental Studies.] At: 00:51 13 April 2010