Colony-Forming-Curve Analysis Method

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Sept. 1992, p /92/ $02.00/0 Copyright 1992, Amerian Soiety for Mirobiology Vol. 58, No. 9 Examination of Baterial Charateristis of Anaerobi Membrane Bioreators in Three Pilot-Sale Plants for Treating Low-Strength Wastewater by Appliation of the Colony-Forming-Curve Analysis Method NAOAKI KATAOKA,1 YUTAKA TOKIWA,2* YASUO TANAKA,1 KIICHI FUJIKI,3 HIROYUKI TARODA,4 AND KIYOSHI TAKEDA2 Development and Planning Department, Ebara Corporation, , Khonan, Minato-ku, Tokyo 108,1 Chemial Eology Division, Fermentation Researh Institute, Ageny of Industrial Siene and Tehnology, MITI, 1-1-3, Higashi Tsukuba, Ibaraki 305,2 Biotehnology and Environmental Engineering Department, Researh Institute, Ishikawajima-Harima Heavy Industries Co., Ltd., Isogo-ku, Yokohama-shi, Kanagawa 235,3 and Researh and Development Department, DIC-Degremont Co., Ltd., 631, Sakado, Sakura-shi, Chiba 285,4 Japan Reeived 2 January 1992/Aepted 10 June 1992 Charateristi sludge eosystems arising in anaerobi membrane bioreators of three pilot-sale plants treating low-strength (less than 1 g of biologial oxygen demand per liter) sewage or soybean-proessing wastewater were examined by analysis of the olony-forming-urves (CFC) obtained by ounting olonies at suitable intervals. The wastewaters, ontaining high amounts of suspended solids (SS) (SS/hemial oxygen demand ratio, 0.51 to 0.80), were treated by using two types of bioreators: (i) a hydrolyzation reator for solubilization and aidifiation of SS in wastewater and (ii) a methane fermentation reator for produing methane. The olony ounts for the two sewage treatment plants ontinued to inrease even after 3 weeks of inubation, whereas those for soybean-proessing wastewater reahed an approximately onstant level within 3 weeks of inubation. The CFCs were analyzed by orrelating the rate of olony appearane on roll tubes with the physiologial types of bateria present in the bioreators. It was found that there were large numbers of slow-olony-forming anaerobi bateria within the bioreators and that the viable populations onsisted of a few groups with different growth rates. It is onsidered that the slow-growing olonies appearing after 10 days of inubation were the dominant miroflora in the sewage treated by hydrolyzation reators. In partiular, highly onentrated sludge (30.0 g of mixed-liquor volatile SS per liter) retained by the membrane separation module ontained a large number of suh bateria. Slow-growing olonies of these bateria ould be ounted by using a sludge extrat medium prepared from only the supernatant of autolaved sludge. In addition, the highest olony ounts were almost always obtained with the sludge extrat medium, meaning that most of the anaerobi bateria in these sludges have omplex nutrient requirements for growth. This report also indiates the usefulness of appliation of the CFC analysis method to the study of baterial populations of anaerobi treatment systems. The anaerobi degradation of omplex organi matter to methane and arbon dioxide involves the interation of four metaboli groups of bateria, namely, hydrolyti, aetogeni, homoaetogeni, and methanogeni bateria (35). A onsiderable amount of researh into the mirobiology of anaerobi digesters, using tehniques suh as immunologi analysis (21-23), enzyme ativity measurement (2, 7, 8), lipid and poly-p-hydroxybutyri aid analysis (13), and mirosopi observation (10, 24, 27, 29), has reently been reported. Although the baterial olony-forming method using a roll tube (15) has frequently been applied to enumeration of the total viable number after a period of inubation and to isolation of the individual physiologial bateria groups in anaerobi digestion eosystems (9, 16, 31-33), little attention has so far been given to the rate of olony appearane on the roll tubes. It has been reported that the olony-forming-urves (CFC) of soil bateria obtained by ounting the number of visible olonies on a plate at suitable intervals give some information, suh as the ratio of the population size and the growth * Corresponding author rate of bateria by the superposition of the first-orderreation model proposed by Hattori (11, 12, 17, 18). The relationship between the number of visible olonies on a plate and the inubation time was formulated on the basis of the first-order-reation model. Ishikuri and Hattori (18) have previously reported that eah tier of the CFC was omposed of a group having a similar growth rate and that the retardation time until the appearane of visible olonies of the population on a plate diretly refleted the growth rate of bateria from natural environments. In methanogeni enrihment populations, the time required for formation of olonies by the pure isolates was similar to that required for the formation of a olony by that ulture during its initial isolation (19). Thus, the growth harateristis of anaerobi baterial populations an be represented by their CFCs. However, there are no reports of the CFC analysis method being applied to the analysis of baterial populations of anaerobi treatment systems. The present researh relates to obtaining the growth and physiologial harateristis of highly onentrated anaerobi baterial populations arising in membrane bioreators by the appliation of the CFC analysis method. The mirobiol-

2 2752 KATAOKA ET AL. Plant TABLE 1. Charateristis of wastewater fed to three plants Amt (mg/liter) of: BOD COD SS TOC A Na B C 852 1, a NT, not tested. ogy of these populations is investigated by orrelating the time required for formation of a visible olony on roll tubes with the physiologial types of anaerobi bateria present in the membrane bioreators. This paper also shows that the CFC analysis method is useful for studying the mirobiology of anaerobi sludge eosystems. MATERIALS AND METHODS Soure of samples. Anaerobi sludge samples were taken from the hydrolyzation and methane fermentation reators of three pilot-sale anaerobi wastewater treatment plants. Srew bottles (100 ml) were filled to the top with mixedliquor samples and immediately losed tightly with stoppers. Sludge samples transported to the laboratory were stored at 4 C for a maximum of 5 days until subjeted to hemial and baterial analysis. Operating onditions of the three plants. The analytial data on wastewaters fed to the three pilot-sale plants (A, B, and C) are shown in Table 1. The three wastewaters had low biologial oxygen demands (BOD) and high suspended-solid (SS) onentrations. The SS/hemial oxygen demand (COD) ratio was high, in the range of 0.51 to Plant A and B wastewaters were sewage with high levels of ellulosi materials. Plant C was soybean-proessing wastewater disharged from an edible-oil refinery plant and ontained relatively high amounts of SS omposed mainly of proteins (544 mg/liter) and arbohydrates (234 mg/liter) from soybean (34). Plant A was for domesti sewage and had a design apaity of 20 m3/day. The plant onsisted of pretreatment equipment (a sreen [pinhole opening, 20,um] for separating SS from the raw sewage), a hydrolyzation reator (total volume, 2.0 mi3) for solubilization of the separated SS, an upflow anaerobi sludge blanket (UASB)-type reator (total volume, 5.4 mi3; reation volume, 1.6 mi) for methane fermentation, and a membrane separation unit. The hydrolyzation reator was maintained at a temperature of 35 C and a ph of 6. Enzyme for ellulose hydrolysis was added at 3% per influent SS (9.5 g/liter). The hydrauli retention time was 3 days. The UASB reator, ombined with an externalpressure apillary-type ultrafiltration membrane module (moleular weight utoff, 15,000; membrane area, 100 mi2) made of polysulfone and polyvinylalohol, was operated at ordinary temperature; the effetive volumetri loading rate was 2.8 kg of BOD per m3/day. Plant B was for muniipal sewage and had a design apaity of 10 m3/day. The plant onsisted of pretreatment equipment (1 m3) (a polyurethane sponge for removing the SS), a hydrolyzation reator (total volume, 0.5 m3) for solubilization of the onentrated SS (more than 10 g/liter), a fluidized-bed-tyge reator (total volume, 1.0 mi3; reation volume, 0.66 m ) for methane fermentation, and membrane separation units. The hydrolyzation reator, ombined with an internal-pressure tubular-type ultrafiltration membrane TABLE 2. APPL. ENvIRON. MICROBIOL. Effiieny of treatment proesses in three plants Gas prodution Feed rate % Removal' of: (Normal liters of Plant (m3/day) gas/kg of COD) BOD COD TOC Total Methane A NT B C a (Total influent onentration - effluent onentration)/total influent onentration. NT, not tested. module (moleular weight utoff, 13,000; membrane area, 0.94 m2) made of polyarylonitrile, was maintained at a temperature of 30 C and a ph of 5.5 to 6.5. The hydrauli retention time was 5 days. The fluidized-bed reator, ombined with an external-pressure hollow-fiber-type mirofiltration membrane module (pore size, 0.1,um; membrane area, 54 mi2) made of polyethylene, was operated at ordinary temperature; effetive volumetri loading rate was 1.1 kg of BOD per m3/day. Plant C was for soybean-proessing wastewater with a high soybean solid onentration and had a design apaity of 7.5 m3/day. The plant onsisted of a fixed-bed-type reator (total volume, 3 mi3), whih had a two-phase system omposed of aid fermentation (total volume, 1.0 mi3; reation volume, 0.54 mi3) and methane fermentation (total volume, 2.0 mi3; reation volume, 1.6 mi3) setions, and a membrane separation unit. Both setions were paked with nonwoven fabri arriers. The reator, ombined with an external-pressure apillary-type ultrafiltration membrane module (moleular weight utoff, 15,000; membrane area, 50 mi2) made of polysulfone and polyvinylalohol, was maintained at a temperature of 30 C, and the phs of the aid and methane fermentation setions were maintained at 6.0 and 7.5, respetively. The hydrauli retention times of the aid and methane fermentation setions were 3.3 to 3.5 and 6.7 to 7.0 h, respetively. The volumetri loading rate was 2.0 kg of BOD per m3/day. The treatment performanes of the three plants are shown in Table 2. The SS in the plant effluent was less than 1 mg/liter. Proedures for baterial enumeration. (i) Sample preparation. Strit anaerobi tehniques used for medium preparation were essentially the same as those of Hungate (15) as modified by Azuma and Suto to develop the gas jet method (3). Traes of 02 in the gases were removed by passing the gases through a redued-opper olumn. The omposition of medium for serial dilution was as follows: (NH4)2HP04, 1.0 g; KH2PO4, 0.2 g; K2HPO4, 1.6 g; MgSO4 7H20, 0.2 g; NaCl, 0.1 g; CaC12 2H20, 0.02 g; FeSO4-7H20, 0.01 g; Na2MoO4 2H20, 0.5 mg; Na2WO4. 2H20, 0.5 mg; MnSO4, 0.5 mg; ysteine. HCl1 H20, 0.25 g; Na2S. 9H20, 0.25 g; resazurin, 1 mg in 1 liter of distilled water. The ph was adjusted to 7.0. The mixed-liquor sludge samples were diluted 1:10 in the dilution medium and homogenized with a Nissei AM-77 homogenizer at 18,000 rpm for 1 min under ie-old onditions. The homogenates were then diluted 1:10 in the dilution medium and soniated with a Tomy Seiko UR-20R soniator at 20-W output for 2 min under ie-old onditions. The homogenization and soniation proedures were arried out in a Hirasawa ANX-1L anaerobi glove box under an atmosphere of 80% N2, 10% CO2, and 10% H2. Serial 10-fold dilutions were made with butyl rubber stopper-

3 VOL. 58, 1992 ANAEROBIC MEMBRANE BIOREACTORS 2753 TABLE 3. Charateristis of anaerobi sludges studied Plant andtypeh MLVSS TOC VFAb (mm) orp(g/liter) (mg/liter) Aetate Propionate Butyrate Valerate A HR MFR <0.1 ND ND B HR , MFR ND ND ND C (MFR) a HR, hydrolyzation reator; MFR, methane fermentation reator. b ND, not deteted. sealed tubes ontaining 9.0 ml of the dilution medium and prepared to a dilution of 10-' by using 1-ml gassed sterile plasti syringes. (ii) Culture media and olony ounts. Three kinds of media were used for the olony ounts by the roll tube method (15). Balh-1 medium was the same as medium number 1 of Balh et al. (4), whih ontained 0.25% sodium formate, 0.25% sodium aetate, 0.2% yeast extrat, 0.2% Tryptiase peptone, mineral salts, trae vitamins, trae metals, and reduing agents. m-vl medium, whih was modified from that desribed by Azuma and Suto (3) and Ueki et al. (32), was prepared by omitting the supematant of sewage digestor fluid from VLS agar (32). It ontained 0.025% eah of gluose, ellobiose, maltose, and soluble starh plus 0.5% Tryptiase peptone, 0.25% yeast extrat, 0.01% beef extrat, mineral salts, and reduing agents. The sludge extrat (SE) medium ontained 0.25 g eah of ysteine. HCl H20 and Na2S. 9H20 plus 1 mg of resazurin and 22 g of agar per liter of the supematant of autolaved sludge. SE for enumeration was made for eah sludge and prepared by autolaving at 121 C for 20 min and entrifuging at 12,000 x g for 30 min. Total organi arbon (TOC) values in the prepared SE were in the range of 600 to 1,050 mg/liter. The gas phases and final phs of these three media were N2 and 6.8 to 7.0, respetively. For olony ounts, samples (0.5 ml) diluted within the range of 10-5 to 10-8 were eah inoulated into dupliate 32-ml pressure glass tubes ontaining 9.0 ml of medium, and then the roll tubes were formed. Cultures were inubated at 30 C. Visible olonies on the roll tube were ounted at suessive intervals. By plotting the mean of olony ounts from dupliate tubes for eah inubation time, CFCs were obtained. The final ounts for roll tubes differed by less than 9% from the mean. Population numbers were expressed as CFU per milliliter of sludge or CFU per gram of mixedliquor volatile SS (MLVSS; desribed below). (iii) Counts of sulfate-reduing bateria. Sulfate-reduing bateria were enumerated by the roll tube method with m-butlin medium, whih was a modifiation of medium C of Butlin et al. (6) and ontained 0.35% sodium latate, 0.1% yeast extrat, mineral salts, and reduing agents. The gas phase was N2, and the final ph was 7.1 to 7.4. The method desribed by Ueki (31) was used to ount blak olonies, whih were onsidered sulfate-reduing bateria. (iv) Counts of methanogeni bateria. Aetate- or formateutilizing methanogeni bateria were enumerated by the most-probable-number (MPN) tehnique with three tubes per dilution. The enumeration was performed in the pressure glass tubes ontaining 9.0 ml of the Balh-1 medium desribed above. The gas phase was N2. Samples (1.0 ml) diluted within the range of 10-' to 10-' were inoulated into three MPN tubes. After 4 weeks of inubation at 30 C, the gas phase of eah tube was examined for the presene of methane, and aetate- or formate-utilizing methanogeni bateria were enumerated by the MPN tehnique on the basis of the numbers of tubes positive for methane prodution. Analytial proedures. The weight of the MLVSS in sludge was determined aording to Standard Methods (1). The weight of MLVSS indiates the amount of organi matter in the solid fration of the sludge. TOC was measured by a TOC analyzer (TOC-500; Shimadzu Co.), using potassium hydrogenphthalate as standard. The methane ontent of the gas phase in the ulture tube was analyzed by a Shimadzu GC-4C gas hromatograph with a stainless steel olumn (WG-100; 6.4 mm by 1.8 m) and a thermal ondutivity detetor. The olumn, inlet, and detetor temperatures were 50, 80, and 80 C, respetively. Argon was used as a arrier gas. Gas samples (0.2 ml) were olleted from the ulture tubes by inserting the needle of a pressure-lok syringe through the butyl rubber stoppers, and eah olleted gas was injeted into a gas hromatograph. Volatile fatty aids (VFA; aeti aid, propioni aid, butyri aid, and valeri aid) were analyzed by a Shimadzu GC-14A gas hromatograph with a paked glass olumn (FAL-M 10%; 3 mm by 2.1 m) with Shimalite TPA (60/80 mesh) support and a flame ionization detetor. The olumn, inlet, and detetor temperatures were 150, 190, and 190 C, respetively. N2 was used as a arrier gas. Samples (0.9 volume) were aidified with 0.1 volume of 1 N HCl, and then the aidified samples (1.0,u) were injeted into a gas hromatograph. RESULTS Chemial harateristis of sludge samples. The hemial harateristis of sludge samples used for analysis of baterial populations are shown in Table 3. Sludge from the UASB reator in plant A, whih was granular-type sludge, and the hydrolyzation reator in plant B had onsiderable amounts of MLVSS. Dissolved-TOC and -VFA values were high in both of the hydrolyzation reators for sewage beause of the inrease of produts by solubilization of sewage solids. On the other hand, VFA values in the methane fermentation reators were muh lower beause methanization of the solubilized solution and the pretreated filtrate proeeded without the aumulation of VFA. CFCs from hydrolyzation reators. The CFCs obtained with the Balh-1, m-vl, and SE media in two hydrolyzation reators (plants A and B) for sewage solids are shown in Fig. 1. The CFCs from both of the reators showed a very slow

4 2754 KATAOKA ET AL. APPL. ENVIRON. MICROBIOL C E I- 0C0o :. o U. : 'o '3 O _k Inubation time (day) Inubation time (day) FIG. 1. CFCs of anaerobi bateria in hydrolyzation reators. Colonies were ounted with three kinds of media: 0, SE; A, m-vl; and C, Balh-1. The means of olony ounts from dupliate roll tubes are plotted. rate of olony formation. Espeially in the ase of plant B, the onsiderable amount of highly onentrated sludge retained by the ombined ultrafiltration membrane module ontained large numbers of slow-growing olonies. Most of the olonies ounted with the SE and m-vl media began to appear after 1 week of inubation, and the ounts ontinued to inrease for up to 5 weeks of inubation. However, in Balh-1 medium, late-appearing olonies were very few. SE medium gave the highest olony ounts, at 2.1 x 108 CFU/ml (3.2 x 1010 CFU/g of MLVSS) and 5.4 x 108 CFU/ml (1.8 x 1010 CFU/g of MLVSS) after 5 weeks of inubation in reators of plants A and B, respetively. This suggests that slow-growing bateria are the dominant miroflora in hydrolyzation reators. CFC from methane fermentation reators. Figure 2 shows the CFCs from three kinds of methane fermentation reators, namely, UASB (plant A), fluidized-bed (plant B), and fixed-bed (plant C) types of reators. Plant C (for soybeanproessing wastewater) showed the highest olony ounts, whereas plants A and B (for sewage) showed muh lower C 0I -.E_ E- -. CU. o C) o 10, _ C o E -. bo E U. o - 0 ounts. The highest ounts after 5 weeks of inubation were 5.4 x 107 CFU/ml (1.4 x 109 CFU/g of MLVSS), 7.9 x 107 CFU/ml (9.3 x 109 CFU/g of MLVSS), and 4.8 x 108 CFU/ml (2.4 x 1011 CFU/g of MLVSS) in plants A, B, and C, respetively. SE medium had the highest ounts among the three media (not inluding plant A, whih was not tested with SE medium). The CFC pattern for plant C differed greatly from those of plants A and B. The ounts with the three media for plant C reahed an approximately onstant level after 3 weeks of inubation; that is, 95% of the total ounts appeared within 3 weeks of inubation (Fig. 2C). On the other hand, the CFCs from plants A and B resembled those from the hydrolyzation reators exept for the minor ounts of late-appearing olonies. The ounts with SE and m-vl media inreased smoothly until the seond to third week of inubation, whereas ounts with Balh-1 medium ontinued to inrease for up to 5 weeks of inubation (Fig. 2A and B). The ounts with m-vl medium for enumerating mainly aidogeni bateria were low in plants A and B (4.8 x 107 and 2.0 x 107 CFU/ml at 5 weeks of inubation) 5 E '0 E U. C o o C) Inubation time (day) Inubation time (day) Inubation time (day) FIG. 2. CFCs of anaerobi bateria in methane fermentation reators. Colonies were ounted with three kinds of media: 0, SE; A, m-vl; and 0, Balh-1. The means of olony ounts from dupliate roll tubes are plotted.

5 VOL. 58, 1992 ANAEROBIC MEMBRANE BIOREACTORS 2755 TABLE 4. Enumeration of sulfate reduers and methanogens in three plants No. of bateria/ml ina: Bateria Plant A Plant B Plant C HR MFR HR MFR (MFR) Sulfate-reduersb 3.6 x x x x x 107 Methanogens 4.0 x X x X 107 NT a HR, hydrolyzation reator; MFR, methane fermentation reator; NT, not tested. b Blak olonies were ounted on the roll tube with m-butlin medium after 4 weeks of inubation. Aetate- or formate-utilizing methanogens were enumerated by the MPN tehnique with three tubes per dilution; they were ultured with Balh-1 medium. Values were determined from the presene of methane in sets of three tubes after 4 weeks of inubation. ompared with those for hydrolyzation reators. The superiority of aidogeni bateria in hydrolyzation reators is reasonable enough, onsidering the funtion of these reators. However, the methane fermentation reator of plant C ontained large numbers of aidogeni bateria (3.4 x 108 CFU/ml). In plants A and B, ounts with Balh-1 medium ontaining formate and aetate as a arbon soure were almost the same (5.6 x 107 to 7.6 x 107 CFU/ml) after 5 weeks of inubation. The CFCs for the three plants show a pattern of relatively slow development. The olony-forming rates of anaerobi bateria in the reators of plants A and B differ from that in the methane fermentation reator of plant C. Most of the olonies from the hydrolyzation reators showed a muh slower growth than those from the methane fermentation reator of plant C. The CFCs also indiate that the baterial populations of the hydrolyzation and methane fermentation reators differ in their nutritional requirements. Enumeration of sulfate-reduing and methanogeni bateria. The ounts of sulfate-reduing bateria (sulfate reduers) and methanogeni bateria (methanogens) in the three plants are shown in Table 4. Sulfate reduers were enumerated by ounting blak olonies on roll tubes ontaining latate as an eletron donor. The blak olonies began to appear on or 10-6-dilution roll tubes after 1 to 2 weeks of inubation, and their numbers reahed a maximum after 3 weeks of inubation. The proportion of sulfate reduer ounts ompared with total olony ounts on the same roll tube ranged from 0.5 to 28%. The hydrolyzation reator of plant A ontained onsiderable numbers of sulfate reduers (3.6 x 107 CFU/ml [5.4 x 109 CFU/g of MLVSS]). Of the three methane fermentation reators, that of plant C had the highest sulfate reduer ount (2.7 x 107 CFU/ml [1.3 x 1010 CFU/g of MLVSSJ). Aetate- or formate-utilizing methanogens were enumerated by the MPN method. The numbers of aetate-utilizing methanogens in the methane fermentation reators were high (1.1 x 107 ells per ml), whereas those in the hydrolyzation reators were low (4.0 x 104 ells per ml). Using a method studied previously (19), we observed the roll tubes with Balh-1 medium from the methane fermentation reator of plant A diretly under an epifluoresene mirosope (Olympus model BHS-RFK) with a 2x objetive and a V (BP-405) exitation filter. However, no fluoresent olony of methanogens was found, although the lowerdilution tubes (10-4 and 10-5) produed trae amounts of methane (less than 2.6% in volume). DISCUSSION This investigation was an attempt to analyze the baterial populations in harateristi sludge eosystems arising in anaerobi membrane bioreators for treating low-strength wastewater. We observed the time required for baterial olonies to appear on roll tubes and tried to orrelate that with the physiologial harateristis of bateria present in the bioreators. The results of CFC analysis show that the type of wastewater and the operating onditions of bioreators signifiantly influene baterial biomass, baterial flora (as regards, for example, the proportion of slower-growing bateria), and nutritional requirements. The present work also shows that the CFC analysis method is an effetive way of studying the mirobiology of anaerobi treatment systems. The CFCs showed a harateristi pattern that differed depending on the medium. SE medium ontaining only the supematant of autolaved sludge gave the highest ounts, with many slower-growing olonies appearing after 10 days of inubation. Large numbers of bateria in the sludges, espeially the slower-growing bateria, probably have omplex requirements for growth. Iannotti et al. (16) stated that the long time for development of visible olonies appeared to be harateristi of anaerobi digesters and that the improvements in reovery (viable ount per total mirosopi ount) were due to slower-growing organisms not found in other media. The results of our researh support this idea. Our findings show that the anaerobi baterial populations for treating sewage (plants A and B) are mainly slow-growing ompared with those for treating soybean-proessing wastewater (plant C). The large numbers of slower-growing bateria in hydrolyzation reators seem to arise from the diffiulty of effeting hydrolysis of sewage solids ontaining large amounts of ellulosi materials. Pfeffer (28) and Klass (20) have reported that in domesti wastewater sewage digestors, the first-step reation, namely, hydrolysis of organi solids, is the rate-limiting step in the overall proess beause of the long retention times required for hydrolysis of these solids. It has been reported that in sewage sludge digestors, the number of hydrolyti and fermentative bateria was large (108 to 109/ml of mesophili sewage sludge) (25, 30), whereas that of the ellulolyti bateria was very small (1.6 x 104 to 9.7 x 105/ml of sewage sludge) (26). Likewise, Hobson and Shaw (14) reported that the total numbers of anaerobi bateria and ellulolyti bateria in domesti digestor sludge were 2.4 x 107 and 4 x 103/ml, respetively. It is reasonable to onsider that the population size of ellulolyti bateria is very small in hydrolyzation reators for treating sewage solids. The hydrolyzation reators employed were operated with a 3- or 5-day hydrauli retention time for obtaining suffiient solubilization of sewage solids. The long retention time probably influenes aspets of the baterial flora suh as the proportion of slower-growing bateria. We an further explain the predominane of the slower-growing bateria in the hydrolyzation reators by the

6 2756 KATAOKA ET AL. interations between hydrolyti bateria and their dependent assoiates regarding substrate inorporation. That is to say, the hydrolyti and fermentative bateria serve as a primary onsumer of various substrates in influent solids, whereas nonhydrolyti bateria suh as the slower-growing bateria grow by utilizing organi ompounds and growth fators, for example, vitamins and amino aids, derived from ellular omponents produed by death and lysis of ells and metabolites produed by viable ells. The higher ounts obtained with SE medium also suggest the presene of syntrophi bateria. Both the long retention time and the high reator biomass onentration by membrane separation seem to be related to the advane of syntrophi interation among the speies. For effiient methane fermentation of wastewaters, it is essential to enhane the solubilization and methanization of high amounts of solids. In domesti sewage, approximately 75% of the organi materials are in the suspended fration (5). The present investigation employed two different hydrolyzation proesses: (i) a reator adding enzyme for hydrolysis of ellulose (plant A) and (ii) a reator retaining highly onentrated sewage solids and bateria by a membrane separation unit (plant B). The solubilization rates of the two proesses were low, at 40.9% for the volatile SS standard (plant A) and 25% for the TOC standard (plant B). Eah reator had the following baterial properties: (i) large number of sulfate reduers (3.6 x 10 /ml; plant A); (ii) large numbers of slower-growing bateria (plant B). Muniipal sewage usually ontains 50 to 200 mg of sulfate per liter, and sulfate redution an thus proeed during the solid solubilization proess. In sewage digestor fluids, sulfate reduers are present at a onentration of 107 to 108/ml (31). We an say from the sulfate reduer ounts that sulfate redution ours simultaneously with hydrolysis of sewage solids and probably favors anaerobi deomposition of the solids. In order to improve the low solubilization rate of sewage solids, the introdution of not only solubilization and aidifiation but also of an effiient methanization step with sulfate redution into the hydrolyzation reators, like anaerobi digestion of sewage sludge, is probably effetive. In pratie, the methane fermentation of highly onentrated sewage solids within the hydrolyzation reators ombined with a membrane separation module inreased the solubilization rate from 60 to 77% and from 45 to 62% as the volatile SS standard in plants A and B, respetively (data not shown). We found that the bioreators studied make harateristi anaerobi sludge eosystems ontaining large numbers of slow-growing bateria. The physiologial and eologial understanding of slow-growing bateria within the highly onentrated sludge will likely be inreasingly important for reator effiieny. As for the CFC analysis method, we an say that it is a useful tehnique for asertaining the physiologial harateristis of populations in anaerobi treatment systems, although there are some restritions due to destrution from omplex interations among speies in baterial aggregates, the presene of a number of nonultivable bateria in artifiial ulture media, and the diffiulty of ahieving optimum onditions for diverse bateria. This tehnique requires further testing in other anaerobi eosystems to onfirm our proposal. ACKNOWLEDGMENTS This researh was arried out as part of the National Researh and Development Program for New Wastewater Treatment System (Aqua Renaissane '90 Projet) being onduted by the Ministry of International Trade and Industry. This study was also supported by APPL. ENvIRON. MICROBIOL. the New Energy and Industrial Tehnology Development Organization. REFERENCES 1. Amerian Publi Health Assoiation Standard methods for the examination of water and wastewater, 16th ed. Amerian Publi Health Assoiation, Washington, D.C. 2. Ashley, N. V., and T. J. Hurst Aid and alkaline phosphatase ativity in anaerobi digested sludge: a biohemial preditor of digester failure. Water Res. 15: Azuma, R., and T. Suto Validity of transfer of the taxonomial position of Corynebaterium pseudopyogenes from genus Corynebaterium to genus Atinomyes, p In H. Iizuka and T. Hasegawa (ed.), Proeedings of the First International Conferene on Culture Colletions. University of Tokyo Press, Tokyo. 4. Balh, W. E., G. E. Fox, L. J. Magrum, C. R. Woese, and R. S. Wolfe Methanogens: reevaluation of a unique biologial group. Mirobiol. Rev. 43: Barbosa, R. A., and G. L. Sant'Anna, Jr Treatment of raw domesti sewage in an UASB reator. Water Res. 23: Butlin, K. R., M. E. Adams, and M. Thomas The isolation and ultivation of sulfate-reduing bateria. J. Gen. Mirobiol. 3: Chung, Y.-C., and J. B. Neethling ATP as a measure of anaerobi sludge digester ativity. J. Water Pollut. Control Fed. 60: Chung, Y.-C., and J. B. Neethling Mirobial ativity measurements for anaerobi sludge digestion. J. Water Pollut. Control Fed. 61: De Haast, J., and T. J. Britz Charaterization of some anaerobi bateria from the liquid phase of a mesophili anaerobi digester fed with a prefermented heese whey substrate. Mirob. Eol. 14: Dolfing, J., A. Griffioen, A. R. W. van Neerven, and L. P. T. M. Zevenhuizen Chemial and bateriologial omposition of granular methanogeni sludge. Can. J. Mirobiol. 31: Hattori, T Analysis of plate ount data of bateria in natural environments. J. Gen. Appl. Mirobiol. 28: Hattori, T The analysis of olony-forming urves, p In T. D. Brok (ed.), The viable ount: quantitative and environmental aspets. Siene Teh Publishers, Madison, Wis. 13. Henson, J. M., P. H. Smith, and D. C. White Examination of thermophili methane-produing digesters by analysis of baterial lipids. Appl. Environ. Mirobiol. 50: Hobson, P. N., and B. G. Shaw The baterial population of piggery-waste anaerobi digesters. Water Res. 8: Hungate, R. E A roll tube method for ultivation of strit anaerobes, p In J. R. Norris and D. W. Ribbons (ed.), Methods in mirobiology, vol. 3B. Aademi Press, In., New York. 16. Iannotti, E. L., J. R. Fisher, and D. M. Sievers Medium for the enumeration and isolation of bateria from a swine waste digester. Appl. Environ. Mirobiol. 36: Ishikuri, S., and T. Hattori Formation of baterial olonies in suessive time intervals. Appl. Environ. Mirobiol. 49: Ishikuri, S., and T. Hattori Analysis of olony forming urves of soil bateria. Soil Si. Plant Nutr. 33: Kataoka, N., Y. Tokiwa, and K. Takeda Improved tehnique for identifiation and enumeration of methanogeni baterial olonies on roll tubes by epifluoresene mirosopy. Appl. Environ. Mirobiol. 57: Klass, D. L Methane from anaerobi fermentation. Siene 223: Kobayashi, H. A., E. Conway de Maario, R. S. Williams, and A. J. L. Maario Diret haraterization of methanogens in two high-rate anaerobi biologial reators. Appl. Environ. Mirobiol. 54: Maario, A. J. L., and E. Conway de Maario Quantitative immunologi analysis of the methanogeni flora of digestors

7 VOL. 58, 1992 ANAEROBIC MEMBRANE BIOREACTORS 2757 reveals a onsiderable diversity. Appl. Environ. Mirobiol. 54: Maario, A. J. L., J. F. K. Earle, D. P. Chynoweth, and E. Conway de Maario Distintive patterns of methanogeni flora determined with antibody probes in anaerobi digestors of different harateristis operated under ontrolled onditions. Syst. Appl. Mirobiol. 12: MaLeod, F. A., S. R. Guiot, and J. W. Costerton Layered struture of baterial aggregates produed in an upflow anaerobi sludge bed and filter reator. Appl. Environ. Mirobiol. 56: Mah, R. A., and C. Sussman Mirobiology of anaerobi sludge fermentation. I. Enumeration of the nonmethanogeni anaerobi bateria. Appl. Mirobiol. 16: Maki, L. R Experiments on the mirobiology of ellulose deomposition in a muniipal sewage plant. Antonie van Leeuwenhoek 20: Morgan, J. W., L. M. Evison, and C. F. Forster The internal arhiteture of anaerobi sludge granules. J. Chem. Teh. Biotehnol. 50: Pfeffer, J. T Inreased loadings on digesters with reyle of digested solids. J. Water Pollut. Control Fed. 40: Robinson, R. W., D. E. Aidn, R. A. Nordstedt, M. V. Thomas, and H. C. Aldrih Light and eletron mirosopi examinations of methane-produing biofilms from anaerobi fixedbed reators. Appl. Environ. Mirobiol. 48: Toerien, D. F., and M. L. Siebert A method for the enumeration and ultivation of anaerobi aid-forming bateria present in digesting sludge. Water Res. 1: Ueki, A., H. Minato, R. Azuma, and T. Suto Enumeration and isolation of anaerobi bateria in sewage digestor fluids: enumeration of sulfate-reduers by the anaerobi roll tube method. J. Gen. Appl. Mirobiol. 26: Ueki, A., E. Miyagawa, H. Minato, R. Azuma, and T. Suto Enumeration and isolation of anaerobi bateria in sewage digestor fluids. J. Gen. Appl. Mirobiol. 24: Varel, V. H Charateristis of some fermentative bateria from a thermophili methane-produing fermenter. Mirob. Eol. 10: Yushina, Y., J. Hasegawa, and H. Satoh Comparison studies on two-phase methane fermentation proesses ombined with membrane separation. Suishitsu Odaku Kenkyuu 14: (In Japanese.) 35. Zeikus, J. G Mirobial populations in digesters, p In D. A. Stafford, B. I. Wheatley, and D. E. Hughes (ed.), Anaerobi digestion. Applied Siene Publishers, Barking, Essex, England. Downloaded from on January 27, 2019 by guest