Microbial community structure and methanogenic activity duringstart-up of psychrophilic anaerobic digesters treating

synthetic industrial wastewaters

Gavin Collins, Adele Woods, Sharon McHugh, Micheal W. Carton, Vincent O’Flaherty �

Microbial Ecology Laboratory, Department of Microbiology, National University of Ireland, Galway, Ireland

Received 22 January 2003; received in revised form 31 May 2003; accepted 22 July 2003

First published online 11 September 2003

Abstract

Culture-independent, molecular techniques were applied to the characterization of microbial communities of an anaerobic granularsludge obtained from a full-scale digester. Procedures were optimised for total DNA recovery and polymerase chain reaction (PCR)amplification of 16S rDNA using archaea- and eubacteria-specific oligonucleotide primers. Cloned PCR products were subsequentlyscreened by amplified rDNA restriction analysis to identify operational taxonomic units (OTUs). Inserts from clones representing eachOTU were sequenced and phylogenetic trees were prepared. In addition, the microbial communities were characterised using terminalrestriction fragment length polymorphism (T-RFLP). The specific methanogenic activity of the biomass, against various substrates, wasalso ascertained. Two anaerobic bioreactors were seeded with granular and non-granular (i.e. crushed) aliquots of the characterisedsludge, respectively, and used to investigate the treatment of a volatile fatty acid (VFA)-based synthetic wastewater, at a loading rate of5 kg COD m33 day31 at low ambient temperatures (18‡C). DNA was isolated from sludge samples during the test period and shifts inarchaeal and eubacterial population structures were elucidated. The start-up period was successful with methane yields and COD removalefficiencies of 60^75% and 65^85%, respectively. Specific methanogenic activities of reactor biomass, obtained at the conclusion of thetrial, indicated the development of psychrotolerant biomass during the 90-day experiment. Furthermore, the efficacy of T-RFLP as amolecular tool for use in the surveyance of engineered ecosystems was confirmed.@ 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.

Keywords: 16S rDNA; T-RFLP; ARDRA; Speci¢c methanogenic activity; Psychrophilic anaerobic digestion; EGSB

1. Introduction

Anaerobic digestion of wastewater o¡ers several advan-tages over more conventional processes, including a reduc-tion in energy needed for aeration and the production ofmethane, a readily usable fuel. Currently, full-scale appli-cations of anaerobic treatment are almost exclusively man-aged at temperatures exceeding 18‡C [1], i.e. mesophilic(30^37‡C) and thermophilic (55^65‡C) operation. How-ever, under moderate climate conditions many waste-waters, including domestic and industrial e¥uents, are dis-

charged at a considerably lower temperature than the op-tima of biological wastewater treatment processes such asnitri¢cation, denitri¢cation and mesophilic methanogene-sis [2]. Moreover, the degradation of organic matter withsubsequent methane production occurs at low tempera-tures in most terrestrial ecosystems of boreal and northernclimate zones [3]. For example, methanogenesis has beendescribed in tundra soil, pond sediments and deep-lakeecosystems [4]. In addition, psychrophilic wastewatertreatment technologies o¡er an attractive potential alter-native to established systems, since the maintenance ofmesophilic anaerobic digestion facilities is expensive [5].To this end, signi¢cant advances have recently beenmade in anaerobic reactor designs, which may be applica-ble to low-temperature anaerobic digestion.Psychrophilic anaerobic treatment has been reported to

occur at lower rates and with less stability than undermesophilic conditions [6,7]. The principal di⁄culty is the

0168-6496 / 03 / $22.00 @ 2003 Federation of European Microbiological Societies. Published by Elsevier B.V. All rights reserved.doi :10.1016/S0168-6496(03)00217-4

* Corresponding author. Tel. : +353 (91) 52 44 11 Ext. 3734;Fax: +353 (91) 52 57 00.

E-mail address: vincent.o£aherty@nuigalway.ie (V. O’Flaherty).

FEMSEC 1578 7-10-03

FEMS Microbiology Ecology 46 (2003) 159^170

www.fems-microbiology.org

decreased level of mixing and £uidisation of reactor bio-mass, attributable to the reduced rate of biogas produc-tion during psychrophilic methanogenesis. This drawbackhas been overcome with a considerable degree of successby the introduction of the expanded granular sludge bed(EGSB) reactor [8]. Although growing, the number of re-ports of successful low-temperature digestion [1,2,5,9] ismodest and this research ¢eld is still in its infancy.Much scope exists for the application of further bioengi-neering innovations, which will allow for the implementa-tion of psychrophilic anaerobic digestion on a more globalbasis for the treatment of a wide variety of wastewaters.Despite the economic importance and widespread har-

nessing of anaerobic digestion in engineered systems andthe development of de novo reactor types, coupled with anincreased depth of understanding with regard to the com-plex microbial processes that occur in anaerobic digestion[10], relatively little knowledge has been acquired withregard to the structure and function of the microbial pop-ulations within the process. Understanding the dynamicsof such communities would be di⁄cult enough if eachpopulation could be isolated from the whole, cultivatedand characterised. Our inability, however, to cultivategreater than 90% of the members from many communities,makes the task considerably more daunting [11]. However,in recent years, nucleic acid-based, phylogenetic ap-proaches have proven useful to describe the microbialecology of such consortia, particularly when supportedby physiological, microscopic and biochemical data. Mo-lecular methods targeting the small subunit rRNA-encod-ing genes are currently used to investigate microbial com-munities in anaerobic digesters and polymerase chainreaction (PCR)-based molecular typing methods rapidlyprovide an image of the community structure of an eco-system (e.g. [12,13]). One of these rRNA gene-based ap-proaches is terminal restriction fragment length polymor-phism (T-RFLP) analysis of 16S rRNA genes, whichallows for the rapid identi¢cation of ribotypes from avariety of samples, including soils and sludges of environ-mental origin [14]. Due to the sensitivity and highthroughput of this method it is an ideal technique forcomparative community analyses [11].The goal of the present study was to investigate the

microbial assemblages within an anaerobic, mesophilicsludge, used to treat a synthetic wastewater under psy-chrophilic conditions. Molecular techniques, 16S rRNAgene sequencing and T-RFLP analysis, were used to char-acterise the microbial community structure of the sludgeand the population dynamics throughout the startingphase of reactor operation, respectively. Process dataand physiological pro¢ling analysis were used in conjunc-tion with T-RFLP, which is con¢rmed here as a useful ande¡ective tool for frequent biomonitoring. We examinedthe feasibility of anaerobic digestion of high-strength in-dustrial wastewaters at low temperatures and recorded asuccessful start-up period for EGSB lab-scale reactors at

18‡C, suggesting the prevalence of psychrotolerant organ-isms.

2. Materials and methods

2.1. Source of biomass

A granular, mesophilic, anaerobic sludge, obtained froma full-scale, 8000 m3 UASB bioreactor, operating at 37‡C,at Archer Daniels Midland (ADM, Co. Cork, Ireland),treating citric acid production wastewater, was used forthe present study.

2.2. Design and operation of psychrophilic anaerobicbioreactors

Two 3.5-l glass laboratory-scale EGSB reactors (R1 andR2) were used in the present study, as described by Col-leran and Pender [15]. A total mass of 70 g volatile sus-pended solids (VSS) of seed sludge were inoculated to eachbioreactor. In order to evaluate the signi¢cance of granu-lar aggregates for successful low-temperature digestion, R1and R2 were seeded with crushed (non-granular) andgranular sludge, respectively. The anaerobic reactorswere fed a synthetic wastewater (pH 7.5 R 0.2) consistingof ethanol, butyrate, propionate and acetate, in the chem-ical oxygen demand (COD) ratio of 1:1:1:1, to a total of10 g COD l31. The synthetic in£uent was bu¡ered withNaHCO3 and forti¢ed, as described by Shelton and Tiedje[16], with macro- (10 ml l31) and micro- (1 ml l31) nu-trients. Both reactors were maintained at 18R 1‡C and at a2-day hydraulic retention time (HRT) for the entire testrun of 90 days. E¥uent was recycled to give a liquid up-£ow velocity of 5 m h31 and samples of reactor e¥uentand biogas were routinely taken for volatile fatty acids(VFAs), COD, and CH4 determination as described pre-viously [17].

2.3. Determination of speci¢c methanogenic activity

The speci¢c methanogenic activity (SMA) of the seedsludge and reactor sludge samples on day 90 was ascer-tained at 37, 30, 22 and 15‡C. The tests were employedwith acetate, butyrate, propionate, ethanol and H2CO2 assubstrates as described previously [18]. The sludge sampleswere washed and a ¢nal biomass concentration of 2^5 gVSS l31 was added to anaerobic activity test medium, togive a ¢nal volume of 10 ml in 20 ml serum vials for liquidsubstrates and in 60 ml hypovials for gaseous substrates.Vials without any added substrate, or with the addition ofN2CO2 in the case of the 60 ml vials, served as controls.

2.4. Extraction and PCR ampli¢cation of 16S rDNA

An extraction protocol was ¢rst optimised for the re-

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170160

covery of total genomic DNA from sludge granules. Anumber of methods were examined for isolation of nucleicacids, the merits of which were assessed by determinationof the cell lysis e⁄ciency, i.e. the fewest unlysed cellswhile maintaining a high yield of non-degraded, highmolecular mass DNA. Microorganisms in sludge samplesbefore and after DNA extraction were visualised micro-scopically according to the method of Bitton and co-work-ers [19], with the exception that Sybr-Gold (MolecularProbes, USA) was used as a stain instead of acridineorange. A sample (0.1 g) of crushed sludge was suspendedin 500 Wl of ¢lter-sterilised water. An aliquot (100 Wl) ofthis was added to 10 ml of ¢lter-sterilised water. An 8 mlvolume of the diluted sample was ¢ltered onto blackIsopore0 (Whatman) membrane ¢lters and 200 Wl ofSybr-Gold (10U stock solution in TE bu¡er, pH 8) wasadded to the remaining 2 ml of diluted sample and left for5 min. The remainder of the sample was then drawn ontothe ¢lter that was mounted onto a glass slide and a mini-mum amount of mineral oil was added under a coverslip.The samples were viewed using a Nikon Optiphot-2UVmicroscope ¢tted with a 100 W mercury bulb, a B-2Aexcitation ¢lter for blue light, a 100U planar objectivelens and 10U eyepieces. Fluorescent microscopy of sludgegranules was carried out as described previously by Pender[20].Aggregates were initially disassociated by grinding in a

mortar and pestle, or by sonication (UltraSonik, Yucaipa,CA, USA) for 30 s, prior to microbial cell lysis using achemical lysis approach, as described by Zhou et al. [21].Alternatively, aggregates were disrupted and cells werelysed by bead beating combined with chemical lysis. Dif-ferent combinations of the above were tested and it wasfound that gently crushing the sludge granules with a pes-tle and mortar, before passing through a Soil DNA Kit(MoBio Laboratories) gave the highest yield of good-qual-ity DNA with little shearing, and provided overall the bestmicrobial cell lysis. Archaeal and eubacterial 16S rRNAgenes were ampli¢ed with forward primer 21F (5P-TT-CCGGTTGATCCYGCCGGA-3P) [22] and reverse primer958R (5P-YCCGGCGTTGAMTCCAATT-3P) [23], andforward primer 27F (5P-GAGTTTGATCCTGGCTCAG-3P) [23] and reverse primer 1392R (5P-ACGGGCGGTG-TGTRC-3P) [24], respectively. Reaction mixtures were pre-pared in a laminar air-£ow biological cabinet (Nuaire, Ply-mouth, UK) and PCR was performed with 10 ng of DNAas the template, the primer set (0.25 WM l31 of each prim-er), MgCl2 (0.125 mmol), 5 Wl 10U NH4 bu¡er and 1 U ofTaq DNA polymerase (Bioline). The cycle pro¢les usedwere denaturation at 95‡C for 1.5 min, annealing at55‡C (archaeal) or 52‡C (eubacterial) for 1.5 min and ex-tension at 72‡C for 1.5 min; the number of cycles was 30.PCR products were resolved by gel electrophoresis on 1%1U TAE agarose gels, containing ethidium bromide (1 Wgml31) and visualised by UV excitation. Primers were re-moved and the ampli¢ed products were concentrated, us-

ing a PCR Prep Puri¢cation kit (Promega) according tothe manufacturer’s protocol.

2.5. Cloning of 16S rDNA, ampli¢ed rDNA restrictionanalysis (ARDRA), sequencing and phylogeneticanalysis of seed sludge biomass

PCR amplicons were ligated into the plasmid vectorpCR0 2.1-TOPO (Invitrogen) and used to transformTOPO 10 (Invitrogen) competent Escherichia coli cells byfollowing the manufacturer’s instructions. Clone librarieswere generated, by growing 96 (A1 to H12) archaeal andeubacterial clones, originating from the R1 and R2 seedsludge in respective micro-well plates (NUNC), containing200 Wl LB broth and 50 Wg kanamycin ml31 per well.Ampli¢cation products, generated using the vector-speci¢cprimers M13F (5P-GTTTTCCCAGTCACGAC-3P) andM13R (5P-CAGGAAACAGCTATGAC-3P), were ob-tained from clones and digested using the tetrameric re-striction endonuclease HaeIII (Promega) at 37‡C over-night. Resultant DNA fragments were separatedelectrophoretically in 3.5% 1U TAE high-resolution aga-rose gels, containing ethidium bromide as before. Opera-tional taxonomic units (OTUs) [25] were identi¢ed, basedon restriction cleavage patterns and clones representingthe OTUs selected for sequencing. An alkaline-lysis mini-prep kit (Qiagen) was used, as per the manufacturer’s in-structions, to prepare plasmid DNA from overnight cul-tures of positive transformants, and sequencing wasachieved using vector-speci¢c primers on a Licor gel se-quencer (MWG Biotech, Milton Keynes, UK). Sequencedata were compared with the nucleotide database usingBLAST (Basic Local Alignment Search Tool [26]). Se-quences from this study were manually aligned with se-quences retrieved from the Ribosomal Database Project(RDP [27]), and the CHIMERA_CHECK software main-tained at the RDP was used to identify potential chimericsequences. Finally, phylogenetic trees were calculated withthe Paup*4.0b8 phylogenetic inference package [28], usingthe Kimura-2 parameter correction [29]. The partial 16SrRNA gene sequences determined in this study were de-posited in the GenBank database under accession numbersAY161236 to AY161261 with the generic name ofSSADM_E (Seed Sludge Archer Daniels Midland Eubac-teria) for eubacterial clones and SSADM_A for archaeons.

2.6. T-RFLP analysis of community structure

T-RFLP was used, in parallel with clone library analy-sis, to examine the microbial community structure of thesample biomass in seed sludge, and in sludge samples re-covered from the bioreactors on day 41 of operation andat the end of the test period (day 90). DNA was isolatedfrom the biomass as described above. PCR was per-formed, as above, but using archaeal (21F and 958R)and eubacterial (27F and 1392R) primer sets, where for-

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170 161

ward (21F and 27F) and reverse (958R and 1392R) prim-ers were labelled at the 5P end with the phosphoramiditedyes 6-FAM and HEX, respectively. All PCR productswere used directly for endonuclease restriction and sepa-rate digestions were processed, in the manufacturer’s rec-ommended bu¡ers, with the 4-bp cutting enzymes HhaIand AluI (separately) at 37‡C for 6 h. Terminal restrictionfragments (TRFs) were sized using an automated ABIPrism genetic analyzer by Oswel, Southampton, UK.Genescan0 3.1 software was used to quantify the electro-pherogram output and sample data consisted of the size(base pairs), peak height and peak area for each TRF peakin sample pro¢les. The abundance of individual TRFs in agiven sample was calculated, based on the total peak areaof the TRF pattern under investigation. Those TRFs,however, with a peak area not greater than 2.5% of thetotal area of the sample were excluded from further anal-ysis. TRF sizes were then rounded to the nearest bp, and1-bp bins constructed, as per Clement et al. [30]. Morethan 300 archaeal and 700 eubacterial in silico T-RFLPsimulations were carried out using the primer combina-tions and endonuclease restriction sites, and sequencesfrom both the RDP 16S rRNA database and those re-trieved from the seed sludge clone libraries. Predicted

TRF lengths were then used for comparison with actualTRFs obtained from sludge samples.

3. Results

3.1. Psychrophilic reactor performance

A rapid start-up of the laboratory-scale reactors wasachieved with less than 2500 mg COD l31 present in thee¥uents of both reactors after 60 days of operation. E⁄-cient and stable COD removal (generally between 70 and75%), at the applied loading rate of 5 kg COD m33 day31,was maintained for the remaining 30 days of the start-upexperiment and the two reactors were very similar in termsof performance during the 90 days of operation. The prin-cipal components of the e¥uents from both reactors werethe VFA propionate and acetate (Fig. 1). Methane yieldsfrom both R1 and R2 typically constituted throughout thetest period between 60 and 75% of total biogas produced(data not shown). The operational pH in the digesters wasmonitored during the trial and the extremities of this rangewere 6.8 and 7.7 (data not shown). There was no evidenceof granulation of the biomass in R1, which was seededwith a crushed inoculum and the biomass consisted of anon-granular £oculant sludge with no signi¢cant increasein the particle size distribution pattern to that determinedfor the seed sludge after the 90-day trial (data not shown).

3.2. Microbial population structure of the seed sludge asdetermined by 16S rRNA clone library and ARDRAanalysis

The DNA extraction protocols carried out in this studyprovided cell lysis e⁄ciencies of s 99% according to mi-croscopic analysis, which revealed fewer than 1% of allcells observed in seed sludge samples, prior to extractions,remained unlysed post-extraction (data not shown). ByARDRA, 13 di¡erent archaeal and 18 eubacterial OTUswere identi¢ed from the respective 96-clone libraries, andrepresentative clones from each were chosen for sequenc-ing. Chimeric clones, composed of 16S rRNA genes am-pli¢ed from di¡erent organisms, can arise during PCRampli¢cation of mixed DNA populations [31], and chime-ras found here, which included ¢ve from the eubacterialstudy, were removed from further analysis. BLAST searchresults and phylogenetic reconstruction (Fig. 2) revealedthat the majority (eight of 13) of the archaeal OTUs rep-resented euryarchaeotal clones closely related to the meth-anogens, and, in particular, the order Methanosarcinales.Five OTUs, accounting for 56% of the total library, how-ever, were found to represent close relatives of the Cren-archaeota. This indicates that, although greater apparentarchaeal diversity was present within the euryarchaeotalsegment of the community, most archaeons in the seedsludge were assigned to the terrestrial cluster of the Cren-

Fig. 1. Absolute values (mg l31) for VFA and ethanol present in R1(A) and R2 (B) reactor e¥uent; ethanol (R), acetate (F), propionate(b) and butyrate (a).

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170162

archaeota. The function of Crenarchaeota, with respect toanaerobic digestion, is unclear as high levels of these or-ganisms have not previously been reported in such sys-tems. Further study is required to investigate the role, ifany, of these organisms in anaerobic digestion.The bacterial community, as revealed by this inventory

(Fig. 3), included organisms closely related to the Syntro-phus group of the N-Proteobacteria (SSADM_ED8),the Flexibacter^Cytophaga^Bacteroides group (SSAD-

M_EF11) and the L-Proteobacteria (SSADM_EG9),which composed 24%, 12% and 7% of the eubacterialclone library, respectively. The remaining 57% of theclones were represented by the Gram-positive organisms,including clones related to the Clostridia (SSADM_ED5,SSAMD_EE9). The primary functions of the major con-stituents of the eubacteria in anaerobic digester ecosystemsare hydrolysis, fermentation and, as in this case, the syn-trophic metabolism of organic acids, ketones and alcohols.

Fig. 2. Phylogenetic relationships of archaeal 16S rDNA clones from the seed sludge biomass based on the Kimura two-parameter algorithm. Scale bar,0.1 estimated substitutions/nucleotide position. Numerical values at nodes represent percentages of 100 bootstrap replications that support branching or-der.

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170 163

3.3. Diverisity of seed sludge and microbial populationdynamics in the laboratory-scale reactors as determinedby T-RFLP analysis

Archaeal TRF patterns derived from both endonu-cleases used included TRF peaks, which, when comparedto predicted peaks from in silico experiments, suggestedthe presence of both euryarchaeotal and crenarchaeotalmembers. More speci¢cally, T-RFLP revealed the domi-

nance of Methanosarcina within the methanogenic popu-lation and the signi¢cant proportion of Crenarchaeota-likeorganisms within the sludge, thus supporting the conclu-sions of the clone libraries and sequence analysis. To beginwith, the archaeal TRF pro¢le of the seed biomass, gen-erated using HhaI as restriction endonuclease and the uni-versal (reverse) primer, revealed the dominance of peaks at579 and 585 bp representing members of the Methanomi-crobiales and marine Crenarchaeota, respectively (data not

Fig. 3. Phylogram illustrating the eubacterial diversity from this study. Scale bar, 0.1 estimated substitutions/nucleotide position. Archaeal 16S rDNAclones were used as outgroups and 100 bootstrap replications were performed.

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170164

shown). However, the use of the forward primer for thesame sample provided a more in-depth view of the struc-ture of this community, with peaks representing the hal-ophilic archaea, Methanosarcinales and Methanobacteria(Fig. 4). Secondly, archaeal T-RFLPs from the seedsludge, which were digested with AluI and generated usingthe reverse primer, described the presence of a peak at 171bp, representing SSADM_AC4 and SSADM_AD3 fromthe clone library in this study, and larger peaks at 439and 631 bp for which no matches were found in in silicoexperiments (data not shown). Again, results from experi-ments using the forward primer o¡ered more clarity withrespect to the composition of the community under inves-tigation and suggested the presence of members of theMethanosarcinales and a peak at 92 bp, for which nomatch was found in the in silico experiments (data notshown).T-RFLP pro¢les of the archaeal community within the

seed sludge and the reactor biomass, at days 41 and 90, forthe R1 and R2 experiments, using the reverse primer sug-gest that a stable archaeal community composition wasmaintained for the duration of both reactor trials. A dom-inant peak at 439 bp, for example, in the AluI-digestedpro¢les was clear for both digesters (data not shown).The pro¢les for the R1 (Fig. 4(i)) and R2 (Fig. 4(ii)) ex-periments, using the forward primer and HhaI, illustratethe persistence of a peak throughout the trial period at 198bp representing the Methanosarcinales population. Peaksat 328 and 330 bp represent organisms closely related toMethanosarcina vacuolata and Methanobacterium palustre,respectively (Fig. 4). Despite, di¡erences in the levels ofdiscrimination between forward and reverse primer-based

T-RFLP, there was good agreement in the general pat-terns of community structure observed for both thearchaeal and bacterial populations of the reactors.Patterns obtained from eubacterial samples similarly

supported the evidence as per ARDRA and clone libraryanalysis, demonstrating the bacterial diversity of the bio-mass. Reverse primer-generated patterns for the seed bio-mass, which were restricted with HhaI, suggest a dominantProteobacteria segment in the community (300 bp), withsmaller peaks at 287 and 302 bp, representing green non-sulphur bacteria, and Fibrobacter and Gram-positive bac-teria, respectively. The corresponding forward primer pro-¢le (Fig. 5), however, discriminates the composition of theproteobacterial community, revealing the presence of aribotype similar to Helicobacter pylori at 98 bp, Iosphaeraand Wollinea at 376 bp and also records a peak at 562bp representing a ribotype similar to Arhodomonas. Theeubacterial structure as described by the reverse primerpro¢le, and using AluI as digestion enzyme, revealed thepresence of members of the Flexibacter^Cytophaga^Bacteroides group by a peak at 73 bp and of clonesfrom this study (SSADM_ED5, SSADM_EE3, SSAD-M_EE5 and SSADM_EG12) by a peak at 130 bp (datanot shown). When the pro¢le was obtained using the for-ward primer TRF, organisms related to Bacillus sp. wereidenti¢ed at the 73 bp peak and peaks were recorded at175, 189 and 248 bp, representing no match, and organ-isms related to Haloanaerobium and Paracoccus, respec-tively (data not shown).TRF pro¢les of the bacterial assemblages throughout

the experiments for both bioreactors using the reverseprimer show the persistent dominance of the Proteobacte-

Fig. 4. Electropherograms illustrating forward primer-generated archaeal TRF peaks, after HhaI digestion, throughout R1 (i) and R2 (ii) operation;seed sludge (A), day 41 (B) and day 90 (C).

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170 165

ria at 300 bp. This suggests an inherent stability within thebacterial populations of the two digesters. The TRF pro-¢les, however, of bacterial communities within the seedsludge and the digester biomass, at days 41 and 90, forR1 (Fig. 5(i)) and R2 (Fig. 5(ii)), using the forward primerand HhaI, showed very di¡erent dynamics within the di-gester ecosystems. The peak at 98 bp in the seed sludge,representing members of the Proteobacteria, is not domi-nant in the R1 biomass by the trial conclusion. This peakalmost disappeared by day 41 in the R2 biomass but re-emerges as a peak of interest by day 90. A peak at about1125 bp, which could not be identi¢ed by in silico restric-

tions, is evident in the biomass of both bioreactors (Fig.5). Overall, a more diverse bacterial population is repre-sented by the pro¢les from the R1 biomass, which alsoappears to exhibit the least compositional stability.It should be noted that the electronic simulations, from

which predicted TRF peak lengths were obtained, werenot exhaustive. Some peaks, therefore, were found inTRF pro¢les for which no matches were available throughin silico experiments. Furthermore, some TRFs could rep-resent microorganisms not yet added to the database, andwere therefore used only to identify microbial groupsrather than species. Despite this, T-RFLP does allow bio-

Table 1SMA values (ml CH4 g VSS31 day31 at standard temperature and pressure (STP) for the seed (day 0), R1 and R2 (day 90) biomass at various temper-atures

Substrate/Test Seed (day 0) R1 (day 90) R2 (day 90)

Temperature 15‡C 37‡C 15‡C 22‡C 30‡C 37‡C 15‡C 22‡C 30‡C 37‡C

Ethanol 24.69 165.33 186.35 260.18 261.87 950.66 88.92 307.74 168.04 1052.78Butyrate 10.42a 58.12b 79.41 172.18 270.80 445.33 83.09 217.46 125.88 527.16Acetate 24.81 59.77 81.96 131.23 297.94 784.64 66.39 110.90 102.11 461.23Propionate N.D.c 1.20 7.34d 48.63e 70.21f 81.92g 12.42h 15.35i 146.29j 86.70H2/CO2 25.87 117.46 22.61 125.28 425.64 652.04 31.43 60.79 259.79 565.13

aLag 290 h.bLag 78 h.cN.D., non-dectable.dLag 780 h.eLag 590 h.fLag 214 h.gLag 215 h.hLag 830 h.iLag 430 h.jLag 200 h.

Fig. 5. Electropherograms illustrating forward primer-generated eubacterial TRF peaks, after HhaI digestion, throughout R1 (i) and R2 (ii) operation;seed sludge (A), day 41 (B) and day 90 (C).

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170166

monitoring of the fate of these peaks during the samplingand experimental period, maintaining the integrity of thetechnology as a suitable biomonitoring tool.

3.4. SMA pro¢les of seed sludge and reactor biomasssampled on day 90

The mesophilic seed sludge displayed relatively highSMA values for ethanol, acetate and H2CO2 when mea-sured at 37‡C (Table 1). The activity with butyrate waslow and long lag phases were observed before the onset ofsubstrate conversion to methane. No propionate degrad-ing activity was detectable. Considerably lower activitieswere observed under psychrophilic conditions, i.e. 15‡C,than at 37‡C (Table 1).The SMA activities of sludge biomass from reactor 1 at

37‡C for ethanol, butyrate, acetate, propionate andH2CO2 increased by a factor of 5, 7, 13, 81 and 5, respec-tively, when measured after 90 days of operation (Table1). A substantial increase was also apparent for activitywith butyrate. The activity with propionate, however, re-mained low after 90 days of operation, despite the pres-ence of propionate as a wastewater constituent. Stabledegradation of acetate was achieved, thereby indicatinggreater reactor stability, as 70% of all methane formedduring anaerobic digestion originates from acetate [32].Although lesser activities were observed at 30, 22 and15‡C, greater methanogenic activity than that of the seedsludge was evident with all of the above substrates. Asimilar SMA pro¢le was established for R2 biomass (Ta-ble 1), with high rates of substrate conversion to biogas.This was represented by extremely high activities when thetests were carried out at 37‡C. Low propionate activitieswere again observed.

4. Discussion

The level of performance achieved by the two reactorsinvestigated here was somewhat less than that reportedpreviously for mesophilic reactors treating VFA waste-waters [33] and also of one set of psychrophilic systemsseeded with a di¡erent sludge [34], but is comparable withother reported data for psychrophilic systems and a⁄rmsthe potential of this approach [9]. The methane yield ofthe biogas compares favourably with previous reports ofboth mesophilic [35] and psychrophilic [9] anaerobic di-gesters. One of the major concerns with regard to sub-ambient anaerobic reactors is the low biogas productionrate [2]. High methane production not only re£ects in-creased methanogenic activity, but also promotes mixingand £uidisation of reactor biomass, thus inferring greaterCOD removal, due to increased sludge^substrate contact.Propionate was the principal VFA present in the e¥u-

ents of R1 and R2 throughout the trial. Rebac [9] reportedthat propionate oxidation is most sensitive in a psychro-

philic anaerobic environment and may thus be the rate-limiting step for reactor operation under low-temperatureconditions. The poor performance of the reactor biomassin terms of propionate degradation is re£ected in the ex-tremely low to non-existent metabolic activity of the reac-tor biomass with this substrate (Table 1). Furthermore,the high acetate levels observed in the reactor e¥uentsmay contribute to inhibition of the development of a pro-pionate-degrading biomass, and the long lag phases ob-served before the onset of propionate degradation in batchtests, as previous studies have demonstrated that acetatehas a strong inhibitory e¡ect on propionate degradation inanaerobic bioreactors [31]. Although the low levels of pro-pionate degrading activity in the seed sludge were notdesirable for the successful start-up of the digesters, it isclear that no population capable of occupying this nichehad emerged after the 90-day trial. Longer experimentswill determine whether the development of such a popu-lation is possible or whether high propionate degradingactivity will be the key prerequisite for seeding low-tem-perature anaerobic digesters. Rebac [9] suggested that bu-tyrate degradation was the most stable with respect totemperature £uctuations and it was seen here that the in-crease in the butyrate activity was the most pronounced ofall the VFA substrates (seven-fold), and that butyrate de-grading capacity could be developed during low-temper-ature reactor operation.Crushing of the sludge granules allows for the penetra-

tion of the substrate to the core of the aggregates [36] andwe suggest here that this may account for the greateracetoclastic methanogenic activities recorded for R1 bio-mass. The lack of granulation observed in the reactorseeded with crushed anaerobic sludge may be the resultof unfavourable substrate conditions, e.g. the absence ofcarbohydrates in the in£uent [37]. Alternatively, the ap-plied loading rates and applied liquid up£ow velocitymay have been unsuitable for granulation at low temper-atures, although granulation was observed under identicalsubstrate, up£ow velocity and loading rate conditionsunder mesophilic conditions [38]. It is also unlikely to bea particular psychrophilic issue as granulation has beenobserved in low-temperature anaerobic reactors [G. Col-lins, unpublished data]. The exact liquid up£ow velocityregime and loading rates required during start-up for theinduction of granulation at low temperatures remain un-clear and are an important area for further research beforefull-scale psychrophilic applications are contemplated.The microbial populations involved in the metabolism

of the substrates studied in batch tests remained mesophil-ic (temperature optima of 37‡C) after the 90-day start-upperiod (Table 1). All SMA pro¢les, except those forH2CO2 and ethanol (in the case of R2) collated at 15‡Cfor the reactor sludges were higher than those recorded at37‡C for the seed sludge, thus revealing a satisfactory de-velopment of the methanogenic activity of the community.We posit that the euryarchaeotal or methanogenic segment

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170 167

of the microbial community, therefore, may have becomepsychrotolerant and thus displayed elevated activities at15‡C. Chin and Conrad [39] demonstrated that H2CO2was consumed mainly by methanogenesis at 30‡C but byhomoacetogenesis at 15‡C, and comparisons of the rela-tive amounts of accumulated intermediates, in that study,indicated that H2 consumption by methanogenesis wasmore sensitive to low temperatures than homoacetogenichydrogen consumption. This may explain the absence ofincreased methanogenesis at 15‡C of H2CO2. However, noevidence of homoacteogenic activity (acetate accumula-tion) was observed in batch activity tests (data not shown),and no signi¢cant development of new populations wasdetected by T-RFLP analysis during the 90-day trial.The relatively poor hydrogenophilic activity suggests thatelevated hydrogen partial pressures may also have contrib-uted to the poor conversion of propionate in the labora-tory-scale reactors [9].The high levels of acetate observed in the reactor e¥u-

ents and the poor granulation of R2 biomass are alsoconsistent with the observed dominance ofMethanosarcinasp. in reactor biomass, as determined by both T-RFLPand clone library analysis, and although this organismhas been shown to be present in granules and also to becapable of ‘spontaneous granulation’ [40], it is generallyaccepted that the dominance of the ¢lamentous acetoclas-ticMethanosaeta sp. is required for practically useful gran-ules. Methanosaeta sp. are rod-shaped organisms thatgrow as ¢laments and are regarded as being importantin the onset of granulation and the maintenance of stablegranules during system perturbations [41,42]. In situ hy-bridisation studies carried out by Merkel et al. [43] onsludge from a mesophilic digester fed lactate, propionate,butyrate and acetate revealed that members of the genusMethanosaeta made up more than 90% of the archaealcommunity in the reactor. Other researchers have alsoidenti¢ed Methanosaeta as the predominant acetoclasticmethanogen in anaerobic reactors when using microscopic[44] and MPN methodologies [45]. This organism is se-lected over the faster-growing Methanosarcina throughits greater substrate a⁄nity for acetate and is selectedfor in well-functioning anaerobic digesters with low levelsof acetate [46]. Consequently, the choice of inoculum andthe high levels of acetate recorded in the reactors duringthis trial may well have inhibited granulation.While the cell envelopes of some archaea are very fragile

and easy to lyse in water, other members are almost un-breakable by conventional methods [47]. Some have a par-ticularly thick and rigid outer layer named the methano-chondroitin [48], which shows a high resistance towardschemical lysing agents such as sodium dodecyl sulphate(SDS) or Triton X-100, and to physical disruption [49].In order to retrieve representative DNA yields for com-munity structure analyses, therefore, a robust and e⁄cientcell lysis procedure must be employed and the methodused in this study ful¢lled these criteria and allowed the

generation of community pro¢les from the reactors duringstart-up. Although some researchers have reported a biastowards Methanosaeta-like species using cloning-basedtechniques [50], this was not observed in this case. Inthis study, TRF patterns obtained using the universal re-verse primers 958R (Archaea) and 1392R (Bacteria) pro-vided a ‘macro-view’ of the community structure withinsamples, where the conserved nature of the target regionresulted in a summary of community structure only dis-criminated to the level of group, e.g. Methanobacteriales,based on the peak pro¢les. Conversely, T-RFLP pro¢lesderived from £uorescently-labelled forward primers 21F(archaeal) and 27F (eubacterial) o¡ered a more discrimi-natory and detailed description of population structure,down to genus level in some cases, as a consequence ofthe length heterogeneities at the 5P end of the gene, withinthe V1, V2 and V3 regions [51]. Both types of primer,however, are complementary and, when used together,provided a comprehensive snapshot of the status of micro-bial communities within environmental samples. The re-sults described here, illustrate the use of these primersfor T-RFLP, to provide high resolution and sensitivityfor the detection of OTUs from a complex community,notwithstanding the potential technical problems of themethod [14,52], including potential PCR bias [53,54].The T-RFLP results demonstrated the relatively well re-plicated nature of the archaeal population structures anddynamics of both reactor systems and suggest that a re-producible and stable response to the environmental pa-rameters occurred in both reactors. This is in agreementwith other studies suggesting that stable, replicated archae-al populations develop in replicated reactor experiments[55]. A peak at 630 bp is present in the archaeal TRFpattern for the seed biomass, obtained using AluI andthe reverse primer, but is almost absent in R1 at day 90.The apparent decrease in the relative abundance of organ-isms represented by this peak may be due to the non-gran-ular nature of the R1 biomass, which suggests the impor-tance of granular sludge for the retention of importantcommunity members within such engineered systems. Insummary, however, the archaeal communities showed nomajor shift in general community structure and T-RFLPdata obtained suggest that the overall archaeal diversity ofthe sludge did not change throughout the trial. This sup-ports the evidence of a psychrotolerant population withinthe biomass, as also suggested by the SMA measurements.The apparent diversity within the bacterial community,

based on the number of TRF peaks and ARDRA-derivedOTUs, was apparently greater than that among the Ar-chaea in the seed sludge and the population structuresobserved here are in agreement with the ¢ndings of pre-vious authors who studied mesophilic sludges [56,57].However, the limitations of the techniques used hereshould be noted since more than one representative TRFmay be located under a T-RFLP peak and the ARDRAapproach may underestimate the diversity of microbial

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170168

populations in environmental samples [55]. Furthermore,the bacterial populations in the reactors were more dy-namic than the archaeal communities. It has been sug-gested, however, that an extremely dynamic communitysustains a functionally stable ecosystem [56] and that prev-alent members and diversity within the microbial com-munities of such ecosystems can change dramaticallyeven over short periods.The T-RFLP and clone library analyses provided corre-

lating data on the archaeal and bacterial population struc-ture of the sludge samples. The population structure re-corded for the reactors also corresponded well with theSMA pro¢les and the reactor performance, for example,the elevated acetate concentrations observed in the reactore¥uent and the poor granulation of R2 biomass was con-sistent with the observed dominance ofMethanosarcina sp.in the reactor biomass, as determined by both T-RFLPand clone library analysis. In general, the polyphasic ex-perimental approach employed here provided evidencethat links between low-temperature metabolic activity pro-¢les, microbial population structure and reactor perfor-mance may be established. However, the value of the ap-proach will be more comprehensively tested during long-term reactor trials and by monitoring the performance andpopulation dynamics of reactor biomass in response tochanging environmental conditions.

5. Conclusions

In this study, T-RFLP was successfully used as a mo-lecular approach for biomonitoring of the microbial com-munities in anaerobic digesters and presents a sensitive,high-throughput approach. Molecular techniques arenow an invaluable tool for elucidating members of com-plex microbial communities, and when used in conjunctionwith process engineering and physiological measurementsthey will allow a more extensive knowledge and under-standing of the physiology and biochemistry of the micro-bial populations involved in psychrophilic anaerobic diges-tion. In particular, long-term low-temperature reactortrials will be required to develop a comprehensive under-standing of the microbial interactions which occur in thesesystems and also the ecological factors which will deter-mine reactor performance, such as, whether new, psychro-philic, microbial populations would emerge over time.

Acknowledgements

The receipt of ¢nancial support from the Higher Edu-cation Authority (HEA) Programme for Research in ThirdLevel Institutes (PRTLI)-Cycle II, through the Environ-mental Change Institute, NUI, Galway, and an EnterpriseIreland Research Scholarship to G.C. is gratefully ac-knowledged.

References

[1] Lettinga, G., Rebac, S., Parshina, S., Nozhevnikova, A., van Lier,J.B. and Stams, A.J.M. (1999) High-rate anaerobic treatment ofwastewater at low temperatures. Appl. Environ. Microbiol. 65,1696^1702.

[2] Rebac, S., van Lier, J.B., Lens, P.N., Stams, A.J.M., Dekkers, F.,Swinkels, K.Th.M. and Lettinga, G. (1999) Psychrophilic anerobictreatment of low strength wastewaters. Water Sci. Technol. 39,203^210.

[3] Nozhevnikova, A.N., Simankova, M.V., Parshina, S.N. and Kotsyur-benko, O.R. (2001) Temperature characteristics of methanogenic ar-chaea and acetogenic bacteria isolated from cold environments.Water Sci. Technol. 44, 41^48.

[4] Nozhevnikova, A., Holliger, C., Ammann, A. and Zehnder, A.J.B.(1997) Psychrophilic methanogenesis in sediments of deep lakes.Water Sci. Tecnol. 36, 57^64.

[5] Lettinga, G., Rebac, S. and Zeeman, G. (2001) Challenge of psychro-philic anaerobic wastewater treatment. Trends Biotechnol. 19, 363^370.

[6] Matasushige, K., Inamori, Y., Mizuochi, M., Hosomi, M. and Sudo,R. (1990) The e¡ects of temperature on anaerobic ¢lter treatment forlow strength organic wastewater. Environ. Technol. 11, 899^910.

[7] Zeeman, G. and Lettinga, G. (1999) The role of anaerobic digestionof domestic sewage in closing the water and nutrient cycle at com-munity level. Water Sci. Technol. 39, 187^194.

[8] Lettinga, G. (1995) Anaerobic digestion and wastewater treatmentsystems. Antonie van Leeuwenhoek 67, 3^28.

[9] Rebac, S., Gerbens, S., Lens, P.N., van Lier, J.B., Stams, A.J.M.,Keesman, K.J. and Lettinga, G. (1999) Kinetics of fatty acid degra-dation by psychrophilically grown anaerobic granular sludge. Biore-sour. Technol. 69, 241^248.

[10] Speece, R.E. (1996) Anaerobic Biotechnology for Industrial Waste-water. Archae Press, Nashville, TN.

[11] Marsh, T.L. (1999) Terminal restriction fragment length polymor-phism (T-RFLP): an emerging method for characterizing diversityamong homologous populations of ampli¢cation products. Curr.Opin. Microbiol. 2, 323^327.

[12] Leclerc, M., Delbes, C., Moletta, R. and Godon, J.-J. (2001) Singlestrand conformation polymorphism monitoring of 16S rDNA Ar-chaea during start-up of an anaerobic digester. FEMS Microbiol.Ecol. 34, 213^220.

[13] McHugh, S., Carton, M., Mahony, T. and O’Flaherty, V. (2003)Methanogenic population structure of anaerobic granular sludgesfrom pscyhrophilic, mesophilic and thermophilic bioreactors. FEMSMicrobiol. Lett. 219, 297^304.

[14] Hiraishi, A., Mitsuru, I. and Hisashi, S. (2000) Terminal restrictionpattern analysis of 16S rRNA genes for the characterization of bac-terial communities of activated sludge. J. Biosci. Bioeng. 90, 148^156.

[15] Colleran, E. and Pender, S. (2002) Mesophilic and thermophilicamaerobic digestion of sulphate-containing wastewaters. Water Sci.Technol. 45, 231^235.

[16] Shelton, D.R. and Tiedje, J.M. (1984) General method for determin-ing anaerobic biodegradation potential. Appl. Environ. Microbiol.47, 850^857.

[17] APHA (1998) Standard Methods for the Examination of Water andWastewater, 20th Edn. L.S. Clesceri, A.E. Greenberg and A.D.Eaton (Eds.). American Public Health Association, Washington,D.C.

[18] Colleran, E., Concannon, F., Golden, T., Geoghan, F., Crumlish, B.,Killilea, J.E., Henry, M. and Coates, J. (1992) Use of methanogenicactivity tests to characterize anaerobic sludges, screen for anaerobicbiodegradability and determine toxicity thresholds against individualanaerobic trophic groups and species. Water Sci. Technol. 25, 31^40.

[19] Bitton, G., Koopman, B., Jung, K.H., Voiland, G. and Kotob, M.(1993) Modi¢cation of the standard epi£uorescence microscopic

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170 169

method for total bacterial counts in environmental-samples. WaterRes. 27, 1109^1112.

[20] Pender, S. (2000) Anaerobic treatment of molasses wastewater. Ph.D.Thesis, National University of Ireland, Galway.

[21] Zhou, J., Bruns, M.A. and Tiedje, J.M. (1996) DNA recovery fromsoils of diverse composition. Appl. Environ. Microbiol. 62, 316^332.

[22] Stackebrandt, E. and Goodfellow, M. (1991) Nucleic Acid Tech-niques in Bacterial Systematics. Wiley.

[23] DeLong, E.F. (1992) Archaea in coastal marine environments. Proc.Natl. Acad. Sci. USA 89, 5685^5689.

[24] Lane, D.J., Pace, B., Olsen, G.J., Stahl, D.A., Sogin, M.L. and Pace,N.R. (1985) Rapid determination of 16S ribosomal RNA sequencesfor phylogenetic analyses. Proc. Natl. Acad. Sci. USA 82, 6955^6959.

[25] Moyer, C.L., Dobbs, F.C. and Karl, D.M. (1994) Estimation ofdiversity and community structure through restriction fragmentlength polymorphism distribution analysis of bacterial 16S rRNAgenes from a microbial mat at an active, hydrothermal vent system,Loihi Seamount, Hawaii. Appl. Environ. Microbiol. 60, 871^879.

[26] Altschul, S.F., Madden, T.L., Scha«¡er, A.A., Zhang, J., Zhang, Z.,Miller, W. and Lipman, D.J. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nu-cleic Acids Res. 25, 3389^3402.

[27] Maidak, B.L., Cole, J.R., Lilburn, T.G., Parker, C.T.Jr., Saxman,P.R., Stredwick, J.M., Garrity, G.M., Li, B., Olsen, G.J., Pramanik,S., Schmidt, T.M. and Tiedje, J.M. (2000) The RDP (RibosomalDatabase Project) continues. Nucleic Acids Res. 28, 173^174.

[28] Swo¡ord, D.L. (2001) PAUP*. Phylogenetic Analysis Using Parsi-mony (* and Other Methods). Version 4. Sinauer Associates, Sunder-land, MA.

[29] Kimura, M. (1980) A simple method for estimating evolutionaryrates of base substitutions through comparative studies of nucleotidesequences. J. Mol. Evol. 16, 111^120.

[30] Clement, B.G., Kehl, L.E., DeBord, K.L. and Kitts, C.L. (1998)Terminal restriction fragment patterns (TRFPs), a rapid, PCR-basedmethod for the comparison of complex bacterial communities.J. Microbiol. Methods 31, 135^142.

[31] Kopckynski, E.D., Bateson, M.M. and Ward, D.M. (1994) Recog-nition of chimeric small-subunit ribosomal DNAs composed of genesfrom uncultivated microorganisms. Appl. Environ. Microbiol. 60,746^748.

[32] Zinder, S.H. (1993) Physiological ecology of methanogens. In: Meth-anogenesis: Ecology, Physiology, Biochemistry and Genetics (Ferry,J.G., Ed.), pp. 128^206. Chapman and Hall, New York.

[33] O’Flaherty, V., Colohan, S., Mulkerrins, D. and Colleran, E. (1999)E¡ect of sulphate addition on volatile fatty acid and ethanol degra-dation in an anaerobic hybrid reactor II: Microbial interactions andtoxic e¡ects. Biores. Technol. 68, 109^120.

[34] McHugh, S., Carton, M., Mahony, T. and O’Flaherty, V. (2001)Methanogenic population structure of anaerobic granular sludgesfrom psychrophilic, mesophilic and thermophilic bioreactors. In:Proc. 9th World Congress Anaerobic Digestion 2001 (van Velsen,A.F.M. and Verstraete, W.H., Eds.), Part 2, pp. 93^98. TechnologicalInstitute.

[35] O’Flaherty, V., Lens, P.N. and Colleran, E. (1997) E¡ect of feedcomposition and up£ow velocity on aggregate characteristics in an-aerobic up£ow velocity. Appl. Microbiol. Biotechnol. 47, 102^107.

[36] Van Lier, J.B., Groeneveld, N. and Lettinga, G. (1994) Character-ization and development of thermophilic methanogenic sludge incompartmentalized up£ow reactors. Biotechnol. Bioeng. 30, 251^261.

[37] Jhung, J.K. and Choi, E. (1994) A comparative study of UASB andanaerobic ¢xed ¢lm reactors with development of sludge granulation.Water Res. 29, 271^277.

[38] O’Flaherty, V. and Colleran, E. (1999) E¡ect of sulphate addition onvolatile fatty acid and ethanol degradation in an anaerobic hybridreactor I: Process disturbance and remediation. Biores. Technol. 68,101^107.

[39] Chin, K.-J. and Conrad, R. (1995) Intermediary metabolism in meth-anogenic paddy soil and the in£uence of temperature. FEMS Micro-biol. Ecol. 18, 85^102.

[40] Wiegant, W.M. and De Man, A.W.A. (1986) Granulation of biomassin thermophilic up£ow anaerobic sludge blanket reactors treatingacidi¢ed wastewaters. Biotechnol. Bioeng. 28, 718^727.

[41] Grotenhuis, J.T.C., Smit, M., Plugge, C.M., Yuansheng, X, vanLammeren, A.A.M., Stams, A.J.M. and Zehnder, A.J.B. (1991) Bac-teriological composition and structure of granular sludge adapted todi¡erent substrate. Appl. Environ. Microbiol. 57, 1942^1949.

[42] Gonzales-Gil, G., Lens, P.N.L., Van Aelst, A., Van As, H., Ver-sprille, A.I. and Lettinga, G. (2001) Cluster structure of anaerobicaggregates of an expanded granular sludge bed reactor. Appl. Envi-ron. Microbiol. 67, 3683^3692.

[43] Merkel, W., Manz, W., Szewzyk, U. and Krauth, K. (1999) Popula-tion dynamics in anaerobic wastewater reactors: modeling and in situcharacterization. Water Res. 33, 2392^2402.

[44] Macario, A.J.L. and Conway de Macario, E. (1988) Quantative im-munologic analysis of the methanogenic £ora of digesters reveals aconsiderable diversity. Appl. Environ. Microbiol. 54, 79^86.

[45] Oude Elferink, S.J.W.H., Visser, A., Hulsho¡ Pol, L.W. and Stams,A.J.M. (1994) Sulphate reduction in methanogenic bioreactors.FEMS Microbiol. Rev. 15, 119^136.

[46] Jetten, M.S.M., Stams, A.J.M. and Zehnder, A.J.B. (1992) Methano-genesis from acetate: a comparison of the acetate metabolism inMethanothrix soehngenii and Methanosarcina spp. FEMS Microbiol.Rev. 88, 181^198.

[47] Kreisl, P. and Kandler, O. (1986) Chemical structure of the cell wallpolymer of Methanosarcina. Syst. Appl. Microbiol. 7, 293^299.

[48] Lange, M. and Ahring, B.K. (2001) A comprehensive study into themolecular methodology and molecular biology of methanogenic Ar-chaea. FEMS Microbiol. Rev. 25, 553^555.

[49] Boone, D.R. and Whitman, W.B. (1998) Proposal for minimal stan-dards for describing new taxa of methanogenic bacteria. Int. J. Syst.Bacteriol. 38, 212^219.

[50] Liu, W.T., Chan, O.C. and Fang, H.P. (2002) Characterization ofmicrobial community in granular sludge treating brewery wastewater.Water Res. 36, 1767^1775.

[51] Suzuki, M.T., Rappe¤, M.S. and Giovannoni, S.J. (1998) Kinetic biasin estimates of coastal picoplankton community strucyure obtainedby measurements of small-subunit rRNA gene PCR amplicon lengthheterogeneity. Appl. Env. Microbiol. 64, 4522^4529.

[52] Osborne, A.M., Moore, E.R. and Timmis, K.N. (2000) An evalua-tion of terminal-restriction fragment length polymorphism (T-RFLP)analysis for the study of microbial community structure and dynam-ics. Environ. Microbiol. 2, 39^50.

[53] Reysenbach, A.L., Giver, L.J., Wickham, G.S. and Pace, N.R. (1992)Di¡erential ampli¢cation of rRNA genes by polymerase chain reac-tion. Appl. Environ. Microbiol. 58, 3417^3418.

[54] Suzuki, M.T. and Giovannoni, S.J. (1996) Bias caused by templateannealing in the ampli¢cation of mixtures of 16S rRNA genes byPCR. Appl. Environ. Microbiol. 62, 625^630.

[55] Pender, S., Toomey, M., Carton, M., Eardly, D., Patching, J.W.P.,Colleran, E. and O’Flaherty, V. (2003) E¡ect of sulphate on meth-anogenic population dynamics in anaerobic bioreactors during long-term operation under mesophilic and thermophilic conditions. WaterRes. (in press).

[56] Fernandez, A., Huang, S., Seston, S., Xing, J., Hickey, R., Criddle,C. and Tiedje, J. (1999) How stable is stable? Function versus com-munity composition. Appl. Environ. Microbiol. 65, 3697^3704.

[57] Godon, J.-J., Sumstein, E., Dabert, P., Habouzit, F. and Moletta, R.(1997) Molecular microbial diversity of an anaerobic digester as de-termined by small-subunit rDNA sequence analysis. Appl. Environ.Microbiol. 63, 2802^2813.

FEMSEC 1578 7-10-03

G. Collins et al. / FEMS Microbiology Ecology 46 (2003) 159^170170