Characterization of Metabolic Performance of Methanogenic Granules Treating Brewery Wastewater: Role of Sulfate-Reducing Bacteria

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Dec. 1991, p /91/ $2./ Copyright 1991, American Society for Microbiology Vol. 57, No. 12 Characterization of Metabolic Performance of Methanogenic Granules Treating Brewery Wastewater: Role of Sulfate-Reducing Bacteria WEI-MIN WU,1"2 ROBERT F. HICKEY,.12* AND J. GREGORY ZEIKUS13.4 Michigan Biotechnology Institute, 39 Collins Road, Lansing, Michigan 4899,1* and Department of Civil and Environmental Engineering,2 Department of Biochemistry,3 and Department of Microbiology and Public Health,4 Michigan State University, East Lansing, Michigan Received 29 April 1991/Accepted 2 September 1991 Granules from an upflow anaerobic sludge blanket system treating a brewery wastewater that contained mainly ethanol, propionate, and acetate as carbon sources and sulfate (.6 to 1. mm) were characterized for their physical and chemical properties, metabolic performance on various substrates, and microbial composition. Transmission electron microscopic examination showed that at least three types of microcolonies existed inside the granules. One type consisted of Methanothrix-like rods with low levels of Methanobacterium-like rods; two other types appeared to be associations between syntrophic-like acetogens and Methanobacterium-like organisms. The granules were observed to be have numerous vents or channels on the surface that extended into the interior portions of the granules that may be involved in release of gas formed within the granules. The maximum substrate conversion rates (millimoles per gram of volatile suspended solids per day) at 35 C in the absence of sulfate were 45.1, 8.4, 4.14, and 5.75 for ethanol, acetate, propionate, and glucose, respectively. The maximum methane production rates (millimoles per gram of volatile suspended solids per day) from H2-CO2 and formate were essentially equal for intact granules (13.7 and 13.5) and for physically disrupted granules (42 and 37). During syntrophic ethanol conversion, both hydrogen and formate were formed by the granules. The concentrations of these two intermediates were maintained at a thermodynamic equilibrium, indicating that both are intermediate metabolites in degradation. Formate accumulated and was then consumed during methanogenesis from H2-C2. Higher concentrations of formate accumulated in the absence of sulfate than in the presence of sulfate. The addition of sulfate (8 to 9 mm) increased the maximum substrate degradation rates for propionate and ethanol by 27 and 12%, respectively. In the presence of this level of sulfate, sulfate-reducing bacteria did not play a significant role in the metabolism of H2, formate, and acetate, but ethanol and propionate were converted via sulfate reduction by approximately 28 and 6%, respectively. In the presence of 2. mm molybdate, syntrophic propionate and ethanol conversion by the granules was inhibited by 97 and 29%, respectively. The data show that in this granular microbial consortium, methanogens and sulfate-reducing bacteria did not compete for common substrates. Syntrophic propionate and ethanol conversion was likely performed primarily by sulfate-reducing bacteria, while H2, formate, and acetate were consumed primarily by methanogens. Methanogenic granules are self-immobilized consortia of methanogens, syntrophic acetogens, and hydrolytic-fermentative bacteria that convert soluble organic matter to methane and CO2. The published work on the granule formation, microbial composition, substrate conversion potentials, and microbial structures of granules has focused on methanogens and syntrophic acetogens (7-11, 15, 23, 32, 39). This is likely due to the observation that sulfate-reducing bacteria (SRB) are present at much lower levels than syntrophic acetogens in granules treating sugar wastewater (1). Considerable information as to the methanogenic degradation of acetate, formate, and H2-CO2 in granules has been reported (7, 8, 11). Little work, however, has been performed to evaluate the conversion of other common substrates such as ethanol, propionate, and butyrate. SRB are quite diverse in terms of metabolic activities, morphotypes, trophic properties, and substrate affinities. In the presence of sulfate, acetate can be oxidized to CO2 by some pure SRB cultures; propionate, butyrate, and other volatile fatty acids (VFAs) can be oxidized completely to * Corresponding author CO2 or converted to acetate or acetate plus propionate (in the case of odd long-chain acids with five or more carbon atoms); branched fatty acids such as isobutyrate, isovalerate, and 2-methylbutyrate can also be oxidized completely to CO2 or incompletely to acetate (35, 37). Hydrogen and formate can be utilized by many SRB as electron donors for sulfate reduction (35). Acetate and methanol are degraded via sulfate reduction by a coculture consisting of Desulfovibrio vulgaris and Methanosarcina barkeri (22). Methanol can be also degraded to CO2 via sulfate reduction by a coculture consisting of Desulfovibrio vulgaris and various homoacetogens (6, 12). In the absence of sulfate, certain SRB such as Desulfovibrio spp. may grow together with H2-utilizing methanogens to convert ethanol or lactate to acetate syntrophically (18, 34). The existence of syntrophic associations between H2-producing SRB and H2-consuming methanogens in lake sediments was suggested (5). No mention of syntrophic catabolism of VFAs by SRB in granular systems has been reported. VFAs such as propionate and butyrate are thought to be converted only by obligate syntrophic acetogens in concert with H2-utilizing methanogens (3, 19, 2, 28, 35). In this study, methanogenic granules grown on brewery

2 VOL. 57, 1991 METABOLIC ACTIVITIES OF METHANOGENIC GRANULES 3439 wastewater were characterized for their physical and chemical characteristics, substrate metabolism, and microbial structure. These granules were obtained from a full-scale upflow anaerobic sludge blanket (UASB) reactor treating brewery wastewater containing a moderate level of sulfate (.6 to 1.3 mm). The important role that SRB play in syntrophic methanogenesis was examined and is discussed. MATERIALS AND METHODS Chemicals and gases. All chemicals used were analytical grade and were obtained from Sigma Chemical Co. (St. Louis, Mo.) or Mallinckrodt, Inc. (Paris, Ky.), except for butyric acid (Eastman Kodak Co., Rochester, N.Y.). Nitrogen gas and gas mixtures of N2-CO2 (7:3) and H2-CO2 (8:2) were obtained from Union Carbide Corp., Linde Division (Warren, Mich.). Before use, all gases were passed over heated (37 C) copper filings to remove traces of 2. Analytical methods. Methane was analyzed with a Hewlett-Packard model 589A gas chromatograph (GC) equipped with a flame ionization detector. Separation was done by using Chromosorb 11 at 6 C with N2 as the carrier gas (1 ml/min). H2 was analyzed with the same GC using a thermal conductivity detector and Carbosphere column. CO2 was measured with a series 58 TCD GC (GOW-MAC Instrument Co., Bridgewater, N.J.), using Carbosphere for separation and He as a carrier. Gas samples were withdrawn from the headspace of test vials with a 1-ml glass syringe (Container Corp., Sioux City, Iowa) equipped a gas-tight MININERT syringe valve (Alltech Associates, Inc., Deerfield, Ill.) and a 25-gauge needle. Fatty acids were determined with a Hewlett-Packard model 589A GC using Chromosorb 11 maintained at 175 C for separation and N2 as a carrier (2 mlmin). The same GC was used to determine methanol and ethanol, with separations performed at 12 and 15 C, respectively. Liquid samples were taken from anaerobic vials with 1-ml plastic syringes with 25-gauge needles. Samples were centrifuged at 2, x g for 5 min with an Epperndroft 5415 centrifuge (Brinkmann Instruments, Inc., Westbury, N.Y.). The supernatant was immediately acidified by using 1 N H3PO4 (1:1, vol sample/vol acid). Formate was determined by an enzymatic method with formate dehydrogenase by the method of Hopner and Knappe (14). The detection limit was.5,um. Glucose was determined based on an enzyme electrode sensor with a YSI model 27 glucose analyzer (Yellow Springs Instruments Co., Yellow Springs, Ohio). Sulfate was determined by using an Ion-Chromatography Module/SP (Dionex Corp., Itasca, Ill.) with a Dionex conductivity detector and lonpac A54A column. The granule bed density (grams of solids per milliliter of granule bed volume) was determined by using a volumetric displacement technique. The suspended solids (SS), volatile suspended solids (VSS), specific gravity, and chemical oxygen demand (COD) were determined in accordance with standard methods (1). The insoluble carbonate content in granules was determined by the method described previously (32). Source of methanogenic granules. Brewery granules were obtained from a full-scale UASB reactor operated at G. Heileman Brewery Co. (La Crosse, Wis.). This system was commissioned in 1982 using 2, kg of methanogenic granular sludge obtained from a UASB reactor treating sugar beet wastewater at the sugar factory CSM, Breda, The Netherlands, and 2, kg of municipal digestor sludge from the La Crosse treatment plant (26). The operational volu- TABLE 1. Composition of the influent brewery wastewater and effluent of the UASB reactor Parameter Influenta Effluent COD (mg/liter) 1,2-2, 17-3 Acetate (mm) Propionate (mm) Ethanol (mm) S42- (mm) Cl- (mm) ph a Trace amounts (<.2 mm) of glucose, butyrate, lactate, and succinate were also found in the influent. metric COD loading rate was about 1 kg of COD per m3 per day with a COD removal of 8 to 9%. The reactors were maintained at an operating temperature between 28 and 3 C. Influent sulfate concentrations in the raw wastewater ranged from.6 to 1.3 mm. Sulfate removal of between 2 and 6% was normally observed in this system. The major chemical compositions of the brewery wastewater fed to the UASB reactor and in the effluent are summarized in Table 1. The majority of COD in the influent was due to incoming ethanol, acetate, and propionate. Granules were anaerobically prepared, shipped to our laboratory anaerobically, and immediately transferred into 158-ml serum vials under an N2 atmosphere. The granules were stored at room temperature (about 2 C). The maximum substrate conversion rates and the maximum methane production rates were determined within 72 h. Medium and stock solutions. Strict anaerobic medium and chemical stock solutions were prepared in serum vials (158-ml volume; Wheaton Scientific, Millville, N.J.) sealed with butyl rubber stoppers (Bellco Biotechnology, Vineland, N.J.). The basal medium for substrate conversion and methane production assays and enrichments contained the following components (per liter): NaCl,.9 g; NH4Cl, 1. g; NaHCO3, 5. g; K2HPO4, 1.45 g; KH2PO4,.75 g; - MgCl2 6H2,.2 g; CaCl2 2H2,.1 g; resazurin,.1 g; and trace element solution, 1 ml (16). The medium ph was 7. to 7.1 with 1 atm (1 atm = kpa) of N2-CO2 (7:3) in the headspace. When H2-CO2 (8:2) gas mixture (1 atm) was used as the substrate, the ph was adjusted to 7. to 7.1 by adding a stock solution of NaHCO3 (1%) to the medium. - Just before an experiment was started, Na2S 9H2 solution (2.5%) and substrate stock solutions were added to the vials to obtain a starting sulfide concentration of.4 mm and respective desired substrate concentrations. Sulfate was supplied, when required, by the addition of a ferrous sulfate stock solution (1 M) except for enrichments. Acetate, propionate, and butyrate stock solutions contained 1 M of the respective acids, adjusted to ph 7. with NaOH. The concentrations of methanol, ethanol, sodium formate, and glucose stock solutions were 6 M, 1 M, 4 M, and 2 g/liter, respectively. Enrichment of trophic groups. Acetate (2 mm), propionate (2 mm), and ethanol (2 mm) were used as the substrates to enrich the prevalent species in the granules. One set of the enrichments was supplied with sodium sulfate (5 mm), while another was not. Disrupted granules (.5 ml) with a concentration of 3 g of VSS per liter were inoculated into the serum vials containing 5 ml of the respective enrichment medium. Disrupted granules were prepared from a cell suspension by using a cell homogenizer inside an anaerobic glovebag (Coy Laboratory Products, Inc., Ann

3 344 WU ET AL. Arbor, Mich.). These vials were then incubated at 35 C. The growth of different trophic groups was monitored by examining methane production and substrate consumption. The enrichments were transferred at least five times at a 1% (vol/vol) inoculum. The final enrichments were examined by microscopy and electron microscopy. Substrate degradation and maximum specific activity assays. Substrate degradation time curve assays were performed with ethanol and a mixture of acetate, propionate, and butyrate. Eight substrates (acetate, propionate, butyrate, methanol, ethanol, formate, H2-CO2, and glucose) were tested to determine maximum substrate conversion rates and/or maximum methane production rates of the granules at 35 C. The maximum substrate conversion rates were determined for acetate, propionate, butyrate, methanol, ethanol, and glucose, based on substrate consumption per gram of VSS per day by using the data within the initial several hours of the assay when the substrate conversion could be approximated by zero-order kinetics. The maximum methane production rates were calculated based on the methane production per gram of VSS per day. Both substrate conversion and methane production were conducted in the presence and in the absence of added FeSO4 (8 to 9 mm initial concentration) except for glucose conversion. For H2-CO2 and formate, both intact and disrupted granules were used as inocula to assay the effect of substrate diffusion. Before the media were inoculated with granules, the basal - medium (5 ml per vial) was reduced with an Na2S 9H2 stock solution (2.5%). Subsequently, the intact granules (about 2 ml) were transferred to each vial inside the anaerobic glovebag. The headspace in the vials was flushed with an N2-CO2 mixture (7:3) to remove the H2 gas from the glovebag. For inoculation of disrupted granules, the cell suspension (5 ml) was added by syringe directly to the vials. Inoculated vials were then supplied with a small amount (about 1 mm for fatty acids and ethanol and.1 g/liter for glucose) of the respective substrates and incubated in a 35 C shaking water bath for 2 to 3 h to activate the respective trophic bacteria in the granules and thereby eliminate any lag in substrate consumption. The serum vials were subsequently supplied with the respective substrates. The concentrations used were about 9 mm for acetate or propionate, 5 mm for butyrate, 6 mm for formate, 22 mm for methanol, 17 mm for ethanol, 2 g/liter for glucose, and 1. atm for H2-CO2 (8:2). These concentrations were much higher than the known values of Km for the respective substrates and allowed determination of maximum substrate conversion rates. The water bath was shaken at 125 strokes per min to reduce mass transfer resistance from liquid bulk to the surface of the granules. During the experiments, gas samples for methane and H2 and liquid samples for VFAs, methanol, ethanol, and glucose were taken every 1 to 3 h, depending on observed substrate conversion rates. After an experiment was concluded, the contents of serum vials were centrifuged at 2, x g for 6 to 1 min and the pellet was collected for solids analysis. Mass balance experiments. Mass balance experiments for H2-CO2, formate, ethanol, acetate, and propionate were performed in 158-ml serum vials as described above with the following exception. Intact granules (1 ml) were transferred into the vials and then incubated in a 35 C water bath for about 12 h to consume any residual substrate and, therefore, reduce potential errors in methane production calculation. Subsequently, gas samples were withdrawn from the vials for the determination of starting methane and H2 levels. The APPL. ENVIRON. MICROBIOL. respective substrate stock solutions or H2-CO2 gas mixture was then added to the vials. Vials without substrate added were used as controls to estimate the methane production from any endogenous metabolism. The vials containing the respective substrates were prepared as two groups: with and without addition of FeSO4. After 96 h of incubation at 35 C, substrate consumption (including sulfate) and methane formation were quantified for mass balance calculations. Molybdate addition experiments. The effects of molybdate on the conversion of acetate, propionate, and ethanol and related methane production rates were determined as follows. Intact granules were transferred into serum vials containing 5 ml of medium and then supplied with 1 mm of the respective substrate. An Na2MoO4 stock solution was then added to the vials to obtain the desired starting concentrations of molybdate. The vials were incubated in a 3 C shaking water bath for about 12 h to permit molybdateinduced inhibition levels to stabilize. The temperature of the water bath was then increased to 35 C, and the respective substrates were added to the vials. The experiments designed to measure the maximum substrate conversion rates and the maximum methane production rates were then begun. Microscopy. Phase-contrast and epifluorescence observations were made with an Olympus microscope (model BH2) equipped with a mercury lamp. The autofluorescence of methanogens was observed with a B (IF-49) excitation filter. Electron microscopic examination. The samples for scanning electron microscopy (SEM) were washed with.9% NaCl. The samples were then fixed in.1 M phosphate buffer (ph 7.) containing 4% glutaraldehyde for more than 3 min, washed with.1 M phosphate buffer for 1 min, and then dehydrated by using a graded series of ethanol solutions (25, 5, 75, 85, 95, 1% ethanol). The samples were dried by the critical-point method and coated with gold. Samples for X-ray analysis were prepared in essentially the same way as those for SEM except they were sputter coated with carbon instead of gold. SEM microphotographs were taken with a JSM 35C SEM (JEOL, Ltd., Tokyo, Japan) equipped with a Tracor Northern energy-dispersive X-ray spectrometer (model 55; EDX Instruments, Middletown, Wis.). The mineral composition of granules was analyzed by the X-ray spectrometer, based on the average values of different sites on the surface and cross section of granules. Granule samples prepared for transmission electron microscopy (TEM) were washed with.1 M phosphate buffer (ph 7.2) and fixed in.1 M phosphate buffer (ph 7.2) containing 2.5% glutaraldehyde for 12 to 16 h at 4 C. The fixed samples were rinsed three times at ambient temperature in.1 mm phosphate buffer (ph 7.2), postfixed with 1% OS4 in the same buffer, and dehydrated through a graded series of ethanol solutions and then propylene oxide. Samples were embedded in Polybed 812 (Polyscience, Inc., Warrington, Penn.). Thin sections were cut with a LKB Ultratome and then poststained with uranyl acetate and lead citrate. TEM was performed with a Philips CM-1 electron microscope. The microbial density in granules was determined by counting the number of cells with individual morphotypes in TEM microphotographs of thin cross sections. Total numbers of cells counted were more than 6,7. Calculations. The percentage of sulfate reduction was calculated according to the following equations on the basis of methane produced in the presence versus the absence of sulfate:

4 VOL. 57, 1991 METABOLIC ACTIVITIES OF METHANOGENIC GRANULES H2 + S42- + H+ 4H2 + HS- (1) 4H2 + HC3- + H+ 3H2 + CH4 (2) 4 formate- + So42- + H+ 4HC3- + HS- (3) 4 formate- + H2 + H+ 3HC3- + CH4 (4) acetate- + S42-2HC3- + HS- (5) acetate- + H2 - HC3- + CH4 (6) 2 ethanol + S acetate- + 2H2 + HS- + H+ (7) 2 ethanol + HC3-+ 2 acetate + CH4 + H+ + H2 (8) 4 propionate- + 3SO acetate- + 4HC3- + 3HS- + H+ (9) 4 propionate- + 3H2 -O 4 acetate- + HC3- + 3CH4 + H+ (1) The calculation of free energy (AG') available for interspecies electron transfer during syntrophic ethanol degradation and formate synthesis from H2 and HC3- was based on the data from Thauer et al. (31). Interspecies hydrogen transfer: ethanol + H2 -O acetate- + H+ + 2H2 AG"' = +9.6 kj per reaction (11) Interspecies formate transfer: ethanol + 2HC3-* acetate- + H+ + 2 formate + H2 AG"' = +6.9 kj per reaction (12) Formate synthesis from hydrogen-bicarbonate: H2 + HC3 -* formate + H2 AG"' =-1.3 kj per reaction (13) The AG"' is the increment of free energy for the reactions under standard conditions (273 K, 1 atm), ph 7., and 1 M concentration. The free energy at ph 7. under nonstandard conditions (35C or 38 K) is calculated as the following equations: Interspecies hydrogen transfer: AG' = log ([acetate-][h2]2/[ethanol]) (14) Interspecies formate transfer: AG' = log ([acetate j[formate -]2/ [ethanol][hco3 ]2) (15) Formate synthesis: AG' = log ([formate-]/[h2][hco3j) (16) RESULTS General characterization of granules. The granules were black, and their size varied from.3 to 4.5 mm in diameter (most less than 3 mm). The granules had a relatively high specific gravity (1.46) and granule bed density (92 g of SS per liter). The ash content in the granules (on the basis of dry weight) was 18%. Based on X-ray analysis, the major mineral composition was the following (%): silicon, 2.43; sulfur, 22; phosphorus, 31; potassium, 6.9; sodium, 13; calcium, 11; and iron, 13. The major mineral solids in the granules apparently were ferrous sulfide, calcium phosphate, and silicon compounds. Carbonate content was low (.24 mmol/g of SS). Assuming all the carbonate in the granules was present as calcium salts, a maximum of 2.4 mg of CaCO3/g of SS would have been present. Microbial composition analysis. Observation by phasecontrast-epifluorescence and electron microscopy of both granules and enrichment cultures derived from granules revealed that the granules contained a number of prevalent bacteria which were morphologically distinct (Fig. 1 to 3). In enrichments, Methanobacterium-like autofluorescent rods were identified as the prevalent H2-CO2- and formate-utilizing methanogens (Fig. 1A and B), while Methanothrix-like bamboo-shaped rods appeared to be the prevalent acetotrophic methanogenic bacteria (Fig. 1A). Methanospirillum-like spirilla were occasionally observed (Fig. 1E). In addition to methanogens, two types of nonfluorescent, prevalent rod-shaped (syntrophic-like organisms) acetogens were observed. Compared with the morphotypes of pure cultures of Desulfovibrio vulgaris and Desulfobulbus propionicus (data not shown), an irregular Desulfovibrio-like rod was observed in ethanol enrichments plus and minus sulfate (Fig. 1B and D) and a regularly-shaped Desulfobulbus-like rod was found in propionate enrichments plus and minus sulfate (Fig. IC and E). The granules were generally spheroidal. A large number of small holes or vents were observed on the surface of the granules (Fig. 2A). These holes appeared to be randomly distributed on the granule surface with openings of 1 to 2 p,m (Fig. 2B). Some of the openings were surrounded by bamboo-shaped Methanothrix-like rods (Fig. 2C), while others were surrounded by Methanobacterium-like rods together with other fat rods (Fig. 2D). Thin-section analysis of the granules revealed that the distribution of different bacterial morphotypes was not homogeneous and that low- and high-cell-density areas existed within the granules. Although it is impossible to say with any certainty that the low-density areas were not simply artifacts of sample preparation, it appears that these areas may be internal gas channels. This is supported by the observation that these low-density areas were generally void of cells or dead cells and were approximately the same diameter as the vent holes on the granule surface. The high-cell-density areas were mainly composed of microcolonies or clusters which can be characterized into three distinct types. Type I microcolonies consisted almost entirely of Methanothrixlike bacteria. Methanobacterium-like rods were occasionally observed within these microcolonies (Fig. 3A). These microcolonies contributed about 3 to 4% of the total area of the thin sections. The remaining two types of microcolonies were more heterogeneous with respect to morphotypes. Cluster type II was homogeneously composed of regular fat rods, which was also observed in the propionate enrichments, and Methanobacterium-like rods (Fig. 3B). Microcolony type III was heterogeneously composed of irregular rods having the same morphotypes as those observed in ethanol enrichments and Methanobacterium-like rods (Fig. 3C). The two rods were the prevalent nonmethanogenic bacteria found in the thin sections of the granules. Other than the above morphotypes, Methanospirillum-like cells were occasionally observed within the granules (Fig. 3C). The approximate concentration of the different cell morphotypes observed in the TEM thin sections was as follows (17 cells per cm2): Methanothrix-like bacteria, 1.31; Methano-

5 3442 WU ET AL. APPL. ENVIRON. MICROBIOL. Downloaded from FIG. 1. Microbial species observed in acetate, ethanol, and propionate enrichments derived from disrupted granules. (A) Methanothrixlike long filaments (Mt) were the predominant species in the acetate enrichment minus sulfate. Methanobacterium-like (Mb) and Desulfovibrio-like (Dv) cells were also found. (B) Syntrophic growth of Desulfovibrio-like (Dv) and Methanobacterium-like (Mb) cells in ethanol enrichments minus sulfate. (C) Clump of Desulfobulbus-like (Db) regular fat rods in propionate enrichment (plus sulfate). (D) TEM thin section of Desulfovibrio-like (Dv) cells in ethanol enrichment showed the same morphology as the irregular rods in granules (see Fig. 3). (E) TEM micrograph of syntrophic growth on propionate of Desulfobulbus-like (Db) and Methanospirillum-like (Ms) cells in a propionate enrichment minus sulfate. on March 17, 219 by guest bacterium-like bacteria, 2.22; regular rods, irregular rods, and others, 1.49; and empty cells,.76. The granular microbial density of total cells observed in thin sections (see below) was 5.78 x 17 cells per cm2. Based on these data, the granules would have about 4.4 x 11 cells per cm3 or 4. x 112 cells per g of VSS. Substrate degradation experiments. Substrate degradation of some major intermediates was tested with essentially no sulfate (ca. 3.7 p.m of sulfate added with the trace element solution) to examine syntrophic methanogenic reactions in the granules. The results of syntrophic ethanol conversion are illustrated in Fig. 4 and summarized in Table 2. As ethanol was converted, acetate accumulated and then declined upon depletion of ethanol. As would be anticipated, H2 accumulated rapidly during ethanol conversion (up to 97 Pa) and was then consumed to low levels. Formate (up to 7,uM) was observed to accumulate and then subsequently to be consumed as ethanol levels decreased. Low levels of propionate (<.3 mm) accumulated during ethanol conversion and were then degraded. Simultaneous degradation of acetate, propionate, and butyrate was examined to determine the relative rates of conversion of these three VFAs (Fig. 5). Acetate was consumed much quicker than the other VFAs. Butyrate consumption was extremely slow, indicating that this acid is not a normal substrate for these granules as would be expected based on the composition of brewery wastewater. During butyrate degradation, considerable amounts of isobu-

6 VOL. 57, 1991 METABOLIC ACTIVITIES OF METHANOGENIC GRANULES 3443 FIG. 2. (A) Low-magnification SEM micrograph of a granule shows that small holes exist on the surface of granules. (B) Outline of vent holes on granule surface. (C) Opening of a vent hole surrounded by Methanothrix-like bacteria. (D) Another opening surrounded by Methanobacterium-like rods and other bacteria. Downloaded from tyrate (.7 mm) and trace amounts of 2-methylbutyrate (.5 mm) were observed (data not shown). The H2 partial pressure observed during this experiment was fourfold lower than that observed during ethanol consumption. Formate was not detectable (<5,uM). Substrate conversion rates and maximum methane production rates. Although the experiments conducted above provided some information as to the capacity of the granules to degrade common substrates, the assays were performed basically without added sulfate present. As indicated in the composition of brewery wastewater (Table 1), sulfate was normally present in both the influent and effluent wastewaters treated by these granules. The maximum substrate conversion rates and/or maximum methane production rates for key substrates, in the presence or absence of added sulfate, are compared in Table 3. In the absence of added sulfate, the granules demonstrated relatively high maximum specific conversion rates for ethanol (45.1 mmol/g of VSS per day), acetate (8. mmol/g of VSS per day), propionate (4.1 mmol/g of VSS per day), and glucose (5.7 mmol/g of VSS per day) and low rates for methanol (1.2 mmol/g of VSS per day) and butyrate (.85 mmol/g of VSS per day). The maximum specific methane production rates from formate and H2-CO2 were essentially the same. Granules physically disrupted to reduce mass transfer resistance had much higher specific methane production rates than the intact granules, and the methane production rates were somewhat higher for H2-CO2 than for formate. The addition of sulfate did not significantly affect methane production rates from H2-C2 or formate in intact or disrupted granules. When formate was used as the substrate to determine maximum methane production rates, significant concentrations of hydrogen (more than 15 Pa) accumulated at the beginning of the experiment. This H2 was subsequently consumed. The initial and maximum H2 accumulation rates from added formate were the same in the absence and presence of added sulfate (Table 3). In the presence of sulfate, the maximum substrate conversion rates for methanol, ethanol, acetate, and butyrate were essentially equal to those of samples without added sulfate. The addition of sulfate did not change the maximum methane production rates from acetate, but it did reduce CH4 production rates from ethanol. The maximum propionate consumption rate increased by 3% because of the addition of sulfate. The maximum methane production rate from propionate was also observed to increase as a consequence of sulfate addition. Formate production from H2-CO2. Formate production was observed during methanogenesis from H2-CO2 by the intact granules in the absence of sulfate (Fig. 6B). The level of formate produced was influenced by the addition of sulfate and the hydrogen pressure applied. When hydrogen (.34 atm) was added, formate accumulated up to.6 mm in the liquid phase within 1 h and was then consumed concurrent with hydrogen consumption. This indicates that a reversible reaction between formate synthesis and hydrogen production can be performed by the granules. When sulfate (7.5 mm) was present, formate accumulation was close to or lower than the detectable level (<5,uM) during methanogenesis from H2-CO2. Sulfate was reduced only to a small extent (Fig. 6B). When a higher partial pressure of hydrogen (.9 on March 17, 219 by guest

7 3444 WU ET AL. APPL. ENVIRON. MICROBIOL i E 1- IL 8- z 4 1 z 8 LU 6 4 >' A - A;A l FIG. 3. TEM micrographs of thin cross section of granules show the presence of different microcolonies. (A) Type I microcolony of primarily Methanothrix-like bacteria (Mt) and low levels of Methanobacterium-like bacteria (Mb). (B) Microcolony type II is homogeneously composed of Methanobacterium-like rods (Mb) and Desulfobulbus-like rods (Db). (C) Desulfovibrio-like bacteria (Dv) and Methanobacterium-like rods (Mb). Methanospirillum-like cells (Ms) are also observed outside the colony E 1 z 1 ui LU 1 4 ot E 15 F-O z Lu A '2 4 3 LU 2, TIME (hrs) FIG. 4. Syntrophic ethanol degradation by brewery granules in the absence of sulfate at 35 C and major intermediate product formation. atm) was supplied, formate accumulated to.8 and.2 mm in the absence and the presence of sulfate, respectively, and was then consumed. Effect of sulfate reduction on methane yield. Mass balances were conducted with various substrates (H2, formate, acetate, propionate, and ethanol) in the absence or presence of added sulfate for determination of substrate consumption, methane production, and sulfate consumption (Table 4). Sulfate was added to obtain initial S42- concentrations ranging from 8.2 to 9.4 mm. In the experiments conducted with added sulfate, a minimum of 1.6 mm and, in general, much higher levels were present at the conclusion of the experiments. The amount of methane formed from these substrates in the presence versus the absence of added sulfate indicated a limited role for sulfate reduction in the complete conversion of these substrates except for ethanol and propionate. Only a small amount of H2, formate, and acetate was converted by sulfate reduction. The methane yield from propionate and ethanol in the presence of sulfate was reduced by 28 and 12%, respectively, compared with samples converting propionate and ethanol without added sulfate. Based on equations 7 and 9, ca. 6% of propionate and 28% of ethanol were converted to acetate by sulfate reduction when a high level of sulfate was supplied. Effect of molybdate. Molybdate is an effective and relatively selective inhibitor of SRB (2, 27, 3). Experiments using different concentrations of sodium molybdate were conducted to further support the role of SRB in the conversion of propionate, acetate, and ethanol (Table 5). The addition of 2 mm MoO42- reduced the maximum conversion rate of acetate by only 13%. Syntrophic propionate conversion, however, was extremely sensitive to molybdate addition. Only.1 mm molybdate was required to decrease the maximum propionate degradation rate by 67%. Addition of 2. mm molybdate inhibited propionate degradation by 97%. 1o IL

8 VOL. 57, 1991 METABOLIC ACTIVITIES OF METHANOGENIC GRANULES 3445 TABLE 2. Intermediates formed during syntrophic ethanol degradation by brewery granules and energetic analysis of interspecies electron transfer and formate synthesisa AG' (kj/reaction)b Time Ethanol Acetate Propionate H2 Formate Interspecies electron (h) (mm) (mm) (mm) (Pa) (>LM) transfer Synthesis (H2f-ormate) H2 Formate <.2 1 <5-62 <-54 < < < NCc NC < <5 NC NC NC a Test conditions: initial ph, 7.; HC3-, 59 mm; 35 C. b AG' values were estimated using equations 14, 15, and 16 for interspecies hydrogen transfer, interspecies formate transfer, and formate synthesis from hydrogen plus bicarbonate, respectively. I NC, not calculated. Detection limits were.2 mm,.5 mm, and 5 F.M for propionate, ethanol, and formate, respectively. Syntrophic conversion of ethanol was moderately sensitive to molybdate addition. In the presence of 2. mm molybdate, ethanol conversion was reduced by 29%. DISCUSSION Substrate composition has a strong effect on determination of the microbial composition in an ecosystem. Characterization of the sulfate-containing brewery waste influent revealed that the major electron donors consistently availis EO 15 z 4c I- 5 1 f8 cn LI TIME (hrs) FIG. 5. VFA (acetate, propionate, and butyrate) degradation in the absence of sulfate by brewery granules at 35 C.,1 4 2 able to the granules used in this study were ethanol, acetate, and propionate (65 to 9% of total COD). Therefore, the prevalent microbial trophic groups in the granules would be expected to consist of ethanol-, propionate-, acetate-, and H2-formate-utilizing bacteria. This was, in fact, borne out by subsequent substrate degradation experiments. Hydrolyticfermentative bacteria constituted a limited portion of the overall population in the granules, based on the relatively low glucose conversion rate. Methanol-utilizing bacteria and butyrate degraders also did not appear to be present in great numbers since poor substrate conversion and methane production rates were observed from these substrates. The level of sulfate (.6 to 1.3 mm) in the influent provided favorable conditions for SRB to proliferate. Ethanol was the major carbon and electron source present in the brewery wastewater. Ethanol degradation studies showed that acetate, formate, and H2 were formed as intermediates and were subsequently converted to methane. Acetate accumulation is a consequence of the maximum rate of acetate consumption being much lower than the maximum rate of ethanol consumption (Table 3). Formate was sug- TABLE 3. Comparison of substrate conversion rates and methane production rates by intact versus disrupted brewery granules in the presence or absence of added sulfatea Maximum substrate Maximum CH4 conversion production rate Substrate Granule (mmol/g of (mmollg of form VSS/day) VSS/day) -SO42- +SO42- -So42- +SO42- Glucose Intact 5.7 ND 8.8 ND Methanol Intact Ethanol Intact H2-CO2 Intact NDb ND H2-C2 Disrupted ND ND Formatec Intact ND ND 13.5c 13.4c Formated Disrupted ND ND 36.8d 37.3d Acetate Intact Propionate Intact Butyrate Intact a FeSO4. 5H2 was used as the sulfate source. b ND, not determined. c The initial H2 accumulation rates from formate were 1.4 and 1.8 mmollg of VSS per day, respectively. d The initial H2 accumulation rates from formate were 1.61 and 1.41 mmol/g of VSS per day, respectively.

9 3446 WU ET AL. APPL. ENVIRON. MICROBIOL. E z w a C.4 1 (A) m"/ Hydrogen A..<..\.A ".. ' thane * Formate s b:. -~@~ ~*g._ m (B) L~~.ksuhesue Hydrogen '- A A SI.. than A, * " ' E E UJ LL ZC X -6 -L TIME (hrs) FIG. 6. Methanogenesis from H2-CO2 by the granules (initial ph 7.1). (A) Initial H2 partial pressure was.34 atm. Formate accumulated and then was consumed in the absence of sulfate. (B) Initial H2 pressure was.38 atm, and sulfate was 7.5 mm. Formate did not accumulate to a detectable level (>5 FiM), and sulfate was converted to a small extent. gested to be a prevalent intermediate during syntrophic ethanol degradation by anaerobic whey digestor flocs (33). In this study, accumulation of both H2 and formate to relatively high concentrations during syntrophic ethanol conversion by the granules suggests that both were involved in interspecies electron transfer. Energetic analysis indicated that the free energy available for both interspecies formate and hydrogen transfer during syntrophic ethanol degradation was basically the same and that the concentration of formate and the partial pressure of hydrogen were at an equilibrium level (AG' for formate synthesis from H2 plus HC3- was near zero) (Table 2). The levels of formate and hydrogen suggested that a reversible conversion (formate- + H2 H2 + HCO3) was at equilibrium during syntrophic ethanol degradation by the granules. The presence of the reversible reaction was confirmed by the production of hydrogen TABLE LL TABLE 5. Effect of molybdate addition on the maximum substrate conversion rates of acetate, propionate, or ethanol by brewery granules in the absence of added sulfate Molybdate Substrate conversion rate (mmol/g of VSS/day) concn (% inhibition) (mm) Acetate Propionate Ethanol 8.16 () 4.17 () 39. () (6.9) 1.37 (67.1) 36.6 (6.3) (12.6).14 (96.6) 27.7 (29.1) during methanogenesis from formate and formate accumulation during methanogenesis from H2-C2. Our analysis suggests that reactions involved in syntrophic ethanol or VFA degradation are more complicated than other models have suggested to date (4, 21, 32). The accumulation of propionate during syntrophic ethanol degradation is probably due to the action of some SRB. For example, Desulfobulbus propionicus can ferment ethanol to produce propionate and acetate in the absence of sulfate (17, 35). This reaction, however, may not be important for ethanol degradation since the molar ratio between propionate formed versus ethanol converted was less than 2% (Table 2) and the propionate conversion rate was much slower than the ethanol conversion rate. The observation that the methane production rates from both formate and H2-CO2 increased over threefold when the granules were disrupted reveals the presence of mass transfer limitation for both formate and H2-CO2 within intact granules. The observation that methane production rates from these two substrates were essentially identical for intact granules is somewhat perplexing and contrary to what would be anticipated based on mass transport calculation for microbial flocs (21) and individual cells (6). The catabolism of formate should be more favorable as mass transfer limitations become greater because of the higher diffusivity of formate compared with hydrogen in the intact granules. The production of methane from formate would be expected to be greater than that from hydrogen. The reason this was not observed is not clear. There are several plausible explanations. It is possible that the gas channels or vents observed in the granules not only serve as conduits for release of gas formed within the granules but also exchange dissolved and gaseous substrates into the interior portions of the granules, effectively reducing mass transfer limitations. The observation of such channels or vents is not new. Similar, although Comparison of brewery granule substrate-product conversion in the absence versus the presence of added sulfatea Substrate consumed S42- CH4 formed Substrate (mmol) consumed (mmol) + Sof4 (S4 reduction Granule (mmol) Momml slae - sulfate + sulfate (+ sulfate) - sulfate + sulfate (mmol:mmol) H Intact H Disrupted Formate Intact Formate Disrupted Acetate Intact Propionate Intact Ethanol Intact a The percentages of substrates converted via sulfate reduction for hydrogen, formate, and acetate were calculated from the ratio of methane produced in the presence versus the absence of sulfate, based on equations 1 to 6. The percentages for propionate and ethanol were estimated from the ratio of methane produced, based on equations 7 to 1 and considering the fraction of conversion of acetate, which was produced during degradation of propionate or ethanol, via sulfate reduction.

10 VOL. 57, 1991 METABOLIC ACTIVITIES OF METHANOGENIC GRANULES 3447 considerably larger, vents (1 to 5,um) with a smooth microbial lining were observed in thick biofilms from anaerobic filters, while small vents (1 to 2,um) were observed in thinner films from the walls of the reactors by Robinson et al. (24). Holes and channels were also observed in biofilms and granules from anaerobic filters, UASB reactors, and fluidized bed reactors (29). The presence of these channels could be, in part, responsible for the high effectiveness factors observed for granules degrading acetate and other fatty acids. Although these channels may reduce mass transfer resistance, they do not altogether eliminate it as demonstrated by the differences in formate and H2 conversion rates by intact and disrupted granules. An alternate explanation for the same reaction rates of formate and H2 could be that a reversible conversion between formate and hydrogen occurred in the granules. If formate is rapidly converted to H2 and vice versa in granules to maintain close to a thermodynamical equilibrium, little difference in methane production rates via formate and H2 would be expected. These two explanations are not mutually exclusive and, in fact, both could contribute to the observed phenomena. Based on the relative amounts of methane formed in the presence and in the absence of sulfate during the conversion of H2-CO2 and formate (Table 4), SRB played a minor role in the consumption of these electron transfer intermediates. Only 5 and 6% of the total amount of hydrogen and formate were respectively used for sulfate reduction. Similarly, only about 4% of available acetate was converted via sulfate reduction. This is consistent with the observation that the majority of the morphotypes observed in the thin sections of the granules were acetate-utilizing Methanothrix-like and hydrogen-formate-utilizing Methanobacterium-like cells. However, consumption of sulfate during degradation of ethanol and propionate indicates involvement of SRB. In this study, the role of SRB in syntrophic acetogenesis of ethanol and propionate was investigated by adding molybdate in the absence of sulfate. Molybdate is well established as a inhibitor of SRB (2, 27, 3, 4). The concentration of molybdate required for essentially complete inhibition of sulfate reduction was found to be.2 mm for a lake sediment (27). A molybdate concentration of.1 mm completely inhibits the growth of Desulfovibrio desulfuricans on ethanol in a syntrophic coculture with Methanobacterium formicicum MF (4). The inhibition level of molybdate for methanogens appeared to be much higher than that for SRB. Molybdate at a concentration of 1 mm did not inhibit the growth of M. formicicum (4). In lake sediments, a high level (2 mm) of molybdate inhibited methanogenesis (27). A decrease of methanogenesis and the accumulation of acetate in an anaerobic filter was reported to be due to a concentration of 1 mm molybdate (13). Our data indicated that in the absence of sulfate, the maximum acetate conversion rate was only slightly reduced by the addition of molybdate (.1 to 2. mm). Mass balance experiments revealed that ca. 29% of ethanol was converted via sulfate reduction when sufficient sulfate was present. The molybdate inhibition experiments showed that in the absence of sulfate, syntrophic ethanol conversion was inhibited by 29% by 2. mm of molybdate. These results strongly suggested that SRB were involved in ethanol conversion whether sulfate was available or not. It is well known that in the absence of sulfate, SRB can syntrophically catabolize ethanol together with H2-utilizing methanogens (18, 34). Molybdate severely inhibited the syntrophic conversion of propionate at low levels (.1 mm) with essentially complete inhibition (97%) at higher molybdate concentrations (2. mm). In an experiment performed by using another type of granule developed on a VFA mixture (acetate, propionate, and butyrate) medium containing low sulfate (.15 mm), the same level of molybdate (.1 to 2. mm) inhibited syntrophic propionate conversion only by 22% (38a). These results suggest that syntrophic conversion of propionate in the brewery granules was mainly performed by SRB. To date, several strains of propionate-degrading SRB have been reported (25, 35-38), but the syntrophic conversion of propionate by SRB associated with methanogens has not been previously reported. These results suggest that certain propionate-degrading SRB are able to couple propionate oxidation with sulfate reduction if sulfate is available or to complete propionate oxidation syntrophically with another partner that is able to remove H2 or formate formed during the propionate oxidation if sulfate is not available. This represents a considerable ecological advantage of this type of SRB over obligate syntrophic propionate degraders in ecosystems where sulfate is continuously or intermittently available. These results also suggest that certain propionate- and ethanol-degrading SRB cannot effectively compete with methanogens for hydrogen or formate. Otherwise, a considerably greater portion of the hydrogen and formate would have been used for sulfate reduction during the mass balance experiments. One explanation for this may be that H2- formate-using methanogens are in close association with the SRB that function as syntrophic acetogenic bacteria within the microcolonies or clusters observed by using TEM. This close association would allow the methanogens greater access to these intermediate products and allow them to outcompete other H2-formate-using organisms via physical proximity or access to the substrates. In freshwater lake sediments, sulfate is reported to inhibit H2-dependent methanogenesis as a consequence of changing the metabolism of SRB from producing hydrogen in the absence of sulfate to consuming hydrogen in the presence of sulfate (5). In this study, the lower concentration of formate accumulated in the presence of sulfate compared with that accumulated in the absence of sulfate indicates that SRB may switch their metabolism from sulfate reduction to formate synthesis using H2 or some other electron donor when sulfate is limited. Consequently, the ability of certain methanogenic and sulfate-reducing bacteria which can bind to each other in brewery granules establishes a unique microniche where methanogens and sulfate reducers do not compete for common electron donors (i.e., H2, formate, or acetate) but coexist on different substrates in the presence of high or low sulfate levels. Microcolonies or clusters composed of the syntrophic acetogens Syntrophomonas sp. or Syntrophobacter sp. and H2-utilizing Methanobrevibacter sp. were reported to be found inside granules examined by Dubourguier et al. (1). Our TEM examination revealed that syntrophic microcolonies also existed in the brewery granules but that these microcolonies are composed of the regular and irregular rods as the syntrophic VFA degraders and Methanobacteriumlike bacteria as the H2- or formate-utilizing methanogen. In addition, it was observed that Methanothrix-like bacteria formed their own microcolonies with low levels of Methanobacterium-like bacteria. The irregular rods observed in the granules and ethanol enrichments had morphotypes similar to that of Desulfovibrio vulgaris, which can degrade ethanol syntrophically. The regular rods observed in the granules and propionate enrichments appear novel. Morpho-

11 3448 WU ET AL. typically, these rods are similar to Desulfobulbus propionicus reported by Widdel and Pfennig (38). However, no report has shown that Desulfobulbus propionicus can grow syntrophically with an H2-formate-utilizing methanogen (35). ACKNOWLEDGMENTS We thank G. Heilemen Brewery Co., in particular Brad Thornton and Mike Biedron, for the gift of the granules. Special thanks are due to H. Stuart Pankratz for his help with the TEM work and valuable discussions during the preparation of the manuscript. Thanks are also due to Stanley Flegler for his help with SEM and April Sunday for her help in the laboratory. REFERENCES 1. American Public Health Association Standard methods for the examination of water and wastewater, 16th ed. American Public Health Association, Washington, D.C. 2. Banat, I. M., E. B. Lindstrom, D. B. Nedwell, and M. T. Balba Evidence for coexistence of two distinct functional groups of sulfate-reducing bacteria in salt marsh sediment. Appl. Environ. Microbiol. 42: Boone, D. R., and M. P. Bryant Propionate-degrading bacterium, Syntrophobacter wolinii sp. nov. gen. nov., from methanogenic ecosystems. Appl. Environ. Microbiol. 4: Boone, D. R., R. L. Johnson, and Y. Liu Diffusion of the interspecies electron carriers H2 and formate in methanogenic ecosystems and its implications in the measurement of Km for H2 or formate uptake. Appl. Environ. Microbiol. 55: Conrad, R., F. L. Lupton, and J. G. Zeikus Hydrogen metabolism and sulfate-dependent inhibition of methanogenesis in a eutrophic lake sediment. FEMS Microbiol. Ecol. 45: Cord-Ruwisch, R., and B. Ollivier Interspecies hydrogen transfer during methanol degradation by Sporomusa acidovorans and hydrogenophilic anaerobes. Arch. Microbiol. 144: Dolfing, J Kinetics of methane formation by granular sludge at low substrate concentrations, the influence of mass transfer limitation. Appl. Microbiol. Biotechnol. 22: Dolfing, J., and W. G. B. M. Blomen Activity measurements as a tool to characterize the microbial composition of methanogenic environments. J. Microbiol. Methods 4: Dolfing, J., A. Griffloen, A. R. W. van Neerven, and L. P. T. M. Zevenhuizen Chemical and bacteriological composition of granular methanogenic sludge. Can. J. Microbiol. 31: Dubourguier, H. C., D. B. Archer, G. Albagnac, and G. Prensier Structure and metabolism of methanogenic microbial conglomerates, p In E. R. Hall and P. N. Hobson (ed.), Anaerobic digestion Pergamon Press, Oxford. 11. Harada, H., G. Endo, Y. Tohya, and K. Momonoi High rate performance and its related characteristics of granulated sludges in UASB reactors treating various wastewaters, p In A. Tiche and A. Rozzi (ed.), Poster. Fifth International Symposium on Anaerobic Digestion. Bologna, Italy, May Monduzzi Editore S.p.A., Bologna, Italy. 12. Heijthuijsen, J. H. F. G., and T. A. Hansen Interspecies hydrogen transfer in co-cultures of methanol-utilizing acidogens and sulfate reducing on methanogenic bacteria. FEMS Microbiol. Ecol. 38: Hilton, M. G., and D. B. Archer Anaerobic digestion of a sulfate-rich molasses wastewater: inhibition of hydrogen sulfide production. Biotechnol. Bioeng. 31: Hopner, T., and J. Knappe Determination with formate dehydrogenase, p In H. U. Bergmeyer (ed.), Methods of enzymatic analysis. Academic Press, Inc., New York. 15. HulshoffPol, L. W., W. J. de Zeeuw, C. T. M. Velzeboer, and G. Lettinga Granulation in UASB reactor. Water Sci. Technol. 15: Kenealy, W., and J. G. Zeikus Influence of corrinoid APPL. ENVIRON. MICROBIOL. antagonists on methanogen metabolism. J. Bacteriol. 146: Laanbroek, H. J., T. Abee, and I. M. Voogd Alcohol conversions by Desulfobulbus propionicus Lindhorst in the presence and absence of sulfate and hydrogen. Arch. Microbiol. 133: McInerney, M. J., and M. P. Bryant Anaerobic degradation of lactate by syntrophic associations of Methanosarcina barkeri and Desulfovibrio species and effect of H2 on acetate degradation. Appl. Environ. Microbiol. 41: McInerney, M. J., M. P. Bryant, R. B. Hespell, and J. W. Costerton Syntrophomonas wolfei gen. nov., an anaerobic, syntrophic, fatty acid-oxidizing bacterium. Appl. Environ. Microbiol. 41: McInerney, M. J., M. P. Bryant, and N. Pfenning Anaerobic bacterium that degrades fatty acids in syntrophic association with methanogens. Arch. Microbiol. 122: Ozturk, S. S., B.. Palsson, and J. H. Thiele Control of interspecies electron transfer flow during anaerobic digestion: dynamic diffusion reaction models for hydrogen gas transfer in microbial flocs. Biotechnol. Bioeng. 33: Phelps, T. J., R. Conrad, and J. G. Zeikus Sulfatedependent interspecies H2 transfer between Methanosarcina barkeri and Desulfovibrio vulgaris during coculture metabolism of acetate or methanol. Appl. Environ. Microbiol. 5: Prensier, G., H. C. Dubourguier, I. Thomas, G. Albagnac, and M. N. Buisson Specific immunological probes for studying the bacterial associations in granules and biofilms, p In G. Lettinga, A. J. B. Zehnder, J. T. C. Grotenhuis, and L. W. HulshoffPol (ed.), Granular anaerobic sludge; microbiology and technology. Pudoc, Wageningen, Netherlands. 24. Robinson, R. W., D. E. Akin, R. A. Nordstedt, M. V. Thomas, and H. C. Aldrich Light and electron microscopic examinations of methane-producing biofilms from anaerobic fixedbed reactors. Appl. Environ. Microbiol. 48: Samain, E., H. C. Dubourguier, and G. Albagnac Isolation and characterization of Desulfobulbus elongatus sp. nov. from a mesophilic industrial digester. Syst. Appl. Microbiol. 5: Sax, R. I Anaerobic pretreatment of brewery wastewater on the industrial scale, p In Proceedings of the 4th Industrial Waste Conference, May, 1985, Purdue University. Butterworths, Boston. 27. Smith, R. L., and M. J. Klug Electron donors utilized by sulfate-reducing bacteria in eutrophic lake sediments. Appl. Environ. Microbiol. 42: Stieb, M., and B. Schink Anaerobic oxidation of fatty acids by Clostridium bryantii sp. nov., a sporeforming, obligately syntrophic bacterium. Arch. Microbiol. 14: Switzenbaum, M. S., and R. B. Einstad Analysis of anaerobic biofilms. Environ. Technol. Lett. 8: Taylor, B. F., and R. S. Oremland Depletion of adenosine triphosphate in Desulfovibrio by oxyanions of group VI elements. Curr. Microbiol. 3: Thauer, R. K., K. Jungermann, and K. Decker Energy conversion in chemotrophic anaerobic bacteria. Bacteriol. Rev. 41: Thiele, J. H., W.-M. Wu, M. K. Jain, and J. G. Zeikus Ecoengineering high rate anaerobic digestion systems: analysis of improved syntrophic biomethanation catalysts. Biotechnol. Bioeng. 35: Thiele, J. H., and J. G. Zeikus Control of interspecies electron flow during anaerobic digestion: the role of formate versus hydrogen transfers during syntrophic methanogenesis in flocs. Appl. Environ. Microbiol. 54: Traore, A. S., M. L. Fardeau, E. C. Hatchikan, J. LeGall, and J. P. Belaich Energetics of growth of a defined mixed culture of Desulfovibrio vulgaris and Methanosarcina barkeri: interspecies hydrogen transfer in batch and continuous culture. Appl. Environ. Microbiol. 46: Widdel, F Microbiology and ecology of sulfate- and sulfur-reducing bacteria, p In A. J. B. Zehnder (ed.), Biology of anaerobic microorganisms. John Wiley & Sons, Inc., New York.

12 VOL. 57, 1991 METABOLIC ACTIVITIES OF METHANOGENIC GRANULES Widdel, F., G.-W. Kohring, and F. Mayer Studies on dissimilatory sulfate-reducing bacteria that decompose fatty acids. III. Characterization of the filamentous gliding Desulfonema limicola gen. nov. sp. nov., and Desulfonema magnum sp. nov. Arch. Microbiol. 134: Widdel, F., and N. Pfennig Dissimilatory sulfate- or sulfur-oxidizing bacteria, p In N. R. Krieg and J. G. Holt (ed.), Bergey's manual of systematic bacteriology, vol. 1. The Williams & Wilkins Co., Baltimore. 38. Widdel, F., and N. Pfennig Studies on dissimilatory sulfate-reducing bacteria that decompose fatty acids. II. Incomplete oxidation of propionate by Desulfobulbus propionicus gen. nov., sp. nov. Arch. Microbiol. 131: a.Wu, W.-M Ph.D. dissertation. Michigan State University, East Lansing. 39. Wu, W.-M., J.-C. Hu, X.-S. Gu, Y.-Z. Zhao, H. Zhang, and G.-G. Gu Cultivation of anaerobic granular sludge in UASB reactors with aerobic activated sludge as seed. Water Res. 21: Yadav, V. K., and D. B. Archer Specific inhibition of sulphate-reducing bacteria in methanogenic co-culture. Lett. Appl. Microbiol. 7: