Isolation and Identification of Fecal Bacteria from Adult

Transcription

1 APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1977, p Copyright 1977 American Society for Microbiology Vol. 33, No. 1 Printed in U.S.A. Isolation and Identification of Fecal Bacteria from Adult Swine J. P. SALANITRO,* I. G. BLAKE, AND P. A. MUIRHEAD Department ofanimal Physiology and Growth, Biological Sciences Research Center, Shell Development Company, Modesto, California Received for publication 7 July 1976 An examination ofthe fecal microflora of adult swine was made with regard to the efficiency of several roll tube media in enumeration and recovery of anaerobes, the effects of medium constituents on recovery, and the isolation and identification of the predominant kinds of bacteria. Total number of organisms by microscopic bacterial counts varied among fecal samples from 4.48 x 101 to 7.40 x 10"0 bacteria/g (wet weight). Comparison of different nonselective roll tube media indicated that about 30% of the fecal bacteria could be recovered with a rumen fluid (40%, vol/vol) medium (M98-5). Recoveries of 21 and 15%, respectively, were obtained with M10 and rumen fluid-glucose-cellobiose agar (RGCA) media. Rumen fluid, Trypticase, sugars, and CO2 gas phase were important components required for maximum recovery with this medium. Similar high recoveries of anaerobes were also obtained with M98-5 containing swine cecal extract of place in rumen fluid or M10 plus swine cecal extract. Significantly lower recoveries were observed with RCGA, media supplemented with swine fecal extracts, reinforced clostridial medium, brain heart infusion agar, and prereduced blood agar. Ninety percent of the bacteria isolated from roll tube media were gram positive and consisted of facultatively anaerobic streptococci, Eubacterium sp., Clostridium sp., and Propionibacterium acnes. The remainder of the flora (8%) included several other species of anaerobes and Escherichia coli. Rumen fluid (or volatile fatty acids), Trypticase, and yeast extract additions to basal media stimulated the growth of anaerobic strains. Variation in the relative proportions of the predominant fecal microflora was observed. This work indicates that satisfactory enumeration, isolation, and cultivation of the predominant microflora in swine feces can be obtained when strict anaerobic culture methods and a rumen fluid medium are used. There have been several studies made to determine the types of bacteria that can be isolated from intestinal or fecal material of swine (6, 9, 13, 14, 20, 26). However, such efforts, using selective plate media, have led to the enumeration of only particular bacterial groups, namely, lactobacilli, streptococci, Bacteroides, Escherichia coli, and Clostridium perfringens. In other work on swine microflora, gram-negative anaerobes comprising species of Bacteroides, Veillonella, Fusobacterium, and Peptostreptococcus elsdenii have been isolated from different segments along the intestinal tract (1, 7). Also, a new species of Bacteroides, B. multiacidus, has recently been isolated from pig feces and described (17). Of the gram-positive bacteria isolated from swine feces, Bifidobacterium suis and Lactobacillus acidophilus have been identified (8, 16, 28). In addition, Vervaeke and van Nevel (27) presented results showing that the anaerobic roll tube method (compared with plate techniques) was best for obtaining high culture counts of bacteria isolated from intestinal contents of pigs. Studies comparing different nonselective roll tube media for efficiency of isolation of predominant anaerobes, however, are lacking. In this paper we investigated several roll tube media for optimum recovery and cultivation of anaerobes from feces of adult swine. A rumen fluid roll tube medium, M98-5, previously used for isolating the more numerous chicken cecal organisms (22), also proved to be a better culture medium for recovering fecal bacteria. Isolates were identified, and their distribution among animals on the same diet was determined. MATERIALS AND METHODS Animals. The swine used in these experiments were Yorkshire or Yorkshire-Hampshire crossbred sows or gilts. They were fed a commercial gestation ration (15% protein) and raised on slatted floors. Fecal sampling and processing. Rectal samples of fecal material (six individual animals) were taken 79

2 80 SALANITRO, BLAKE, AND MUIRHEAD manually with a glove moistened with glycerol. Collected fecal material (ca. 100 g) was placed in a plastic bag, closed tightly to exclude air, and kneaded for a few minutes to mix the digesta. Samples were brought to the laboratory and immediately processed in a dilution series for inoculating roll tubes and plates. Ten-gram (wet weight) samples of feces were placed in a rubber-stoppered Erlenmeyer flask (250 ml) containing 190 ml of prereduced sterile anaerobic dilution solution (ADS; 2). The vessel was gassed with C02 and stoppered, and homogenization of the fecal material was done on a magnetic stirrer for 20 min. A subsample of the blend was serially diluted further in anaerobic dilution solution, and aliquots (usually 0.5 to 1.0 ml of 10-8 and 10-9 dilutions) were used to inoculate various media. Other portions of a fecal sample were used for ph, dry weight, and volatile fatty acid (VFA) determinations. Diluted (10-') fecal material was centrifuged (10,000 rpm, 10 min), and 1 ml of the supernatant was processed for fatty acid analysis as described (23). Microscopic and colony counts. Microscopic bacteria counts and roll tube colony counts were made as described previously (22). Mean colony counts (usually from six samples) were determined after 7 days of incubation (37 C) and converted to percent recovery of the microscopic bacterial counts. Separation of the mean recoveries was made using a standard t test at the 5% level of significance (25). Culture media and methods. Strictly anaerobic techniques (2, 12) were used for preparation of media, dilution and inoculation of samples, and isolation and identification of strains. A comparison of several roll tube media was made to determine the optimum culture requirements for recovery of the predominant anaerobic bacteria. Preparation of M98-5 and M10 media was previously described (22), and preparation of RGCA (rumen fluid-glucose-cellobiose agar) and BHIAS (brain heart infusion agar supplemented) was as indicated elsewhere (10). Reinforced clostridial medium (RCM) was used at half strength (1.9%, wt/vol) since higher levels of RCM frequently resulted in many large and spreading colonies, making colony enumeration difflcult in roll tubes. In some cases, roll tubes of BHIAS and RCM were supplemented with defibrinated sheep blood (5%, vol/vol). Cecal or fecal extracts were added to M10 and M98-5 media. All roll tube media contained Na2CO3, 0.40% (wt/vol) (0.16% for media equilibrated with 10% C02), and were equilibrated and tubed (18 by 150 mm, Bellco Glass) under CO2 (100%) gas phase. Sodium carbonate was deleted from roll tube media equilibrated with 100% N2 gas phase (the ph of these media was 6.7 after autoclaving). Four replicate roll tubes were inoculated with aliquots (0.2 to 1.0 ml) of each dilution used. Aerobic counts (pour plate method) were made on BHIA (brain heart infusion agar) or EA (Eugonagar) plate medium. For preparation of cecal extract, cecal fluid was collected from several slaughtered adult swine. Ceca were removed from the digestive tract, and cecal contents were expelled into flasks. For preparation of fecal extract, fecal material was collected from swine as described above. Cecal and fecal extracts APPL. ENVIRON. MICROBIOL. were prepared by taking cecal fluid or fecal material from adult pigs, diluting 1:1 (cecal extract) or 1:4 (fecal extract) with distilled water, and homogenizing in a Waring blender. The slurry was autoclaved (15 lb/in2, 15 min), cooled, and then centrifuged at 12,000 x g 15 min. The supernatant extracts were adjusted to ph 6.8, autoclaved again, and stored at 4 C. Before use in media, cecal and fecal extracts were recentrifuged to remove denatured material and debris. Isolation and identification of fecal bacteria. About 60 to 70 colonies were picked from roll tubes of M98-5 medium with cecal extract (in place of rumen fluid) from each of six fecal samples (individual animals). Colonies isolated from tubes containing the highest dilution (usually 10-8 to 10-9 and 100 to 200 colonies) were picked to slants of the same medium and examined for purity, morphology, and Gram stain reaction (24- to 72-h cultures) and then subcultured in peptone-yeast extract-glucose (PYG) medium (10) for 7 days. The following additional features obtained from PYG medium were used to presumptively identify and group strains: glucose fermentation (terminal ph), catalase, H2 gas production, and fermentation products from glucose. Representative strains from each group were then processed through several carbohydrate fermentation and biochemical tests using the media and methods described elsewhere (10). All media (except for roll tubes and slants) for these tests were prepared and dispensed in 3-ml amounts in 5-ml serum bottles fitted with butyl rubber septa (Wheaton Scientific) according to the method of Miller and Wolin (15). Bottles of media were sterilized by autoclaving at 15 lb/in2 for 15 min. The inoculum (0.05 ml) used in these tests was from cultures grown in liquid media of the same composition as that of the roll tubes. All fermentation and physiological tests were incubated at 37 C and analyzed after 7 days. Crude nutritional studies on representative anaerobic isolates were performed as described previously (21). In these experiments, single additions of rumen fluid (20%, vol/vol), VFA, Trypticase (0.2%, wt/vol), or yeast extract (0.05%, wt/vol) were made to a basal medium consisting of minerals 1 and 2 (each 7.5%, vol/vol), glucose (0.4%, wt/vol), Na2CO3 (0.16%, wt/vol), cysteine hydrochloride (0.05%, wt/ vol), and CO2 gas phase. Analysis of fermentation products. Fermentation acids (formic, acetic, propionic, butyric, lactic, and succinic) were identified from culture media according to a previously published gas chromatographic method (23). Hydrogen (percentage, vol/vol) in the gas phase of serum bottle cultures was determined by a thermal conductivity gas chromatographic method (3). The amount of ethanol in fermentation media was analyzed enzymatically using alcohol dehydrogenase (Sigma kit no. 331-UV). With this method, the enzyme reacts predominantly with ethanol and to a lesser degree with other shortchain aliphatic alcohols. RESULTS Effects of different roll tube tube media on recovery of fecal anaerobes. A comparison of the various roll tube media on recovery of an-

3 VOL. 33, 1977 aerobes and facultative anaerobes is given in Table 1. Highest percent recoveries (29%) per gram (wet weight) of feces were obtained with M98-5 and M10 supplemented with cecal extract. M10 or M10 plus fecal extract recovered 20 to 22% of the bacteria in samples, and RGCA, RCM, and BHIAS recovered only 12 to 18% of the anaerobic bacteria. Addition of sheep blood slightly improved recovery in BHIAS medium but not in RCM medium. If incubation was extended from 7 to 14 days, 4% higher recoveries were obtained with most media. Since recoveries of fecal anaerobes were highest in M98-5, it was of interest to determine the effects of other components in this medium (Table 2). When culture gas phases contained either CO2 (100%), mixtures of N2/CO2 (90%/ 10%, vol/vol), or C02/H2 (90%/10%), recovery of anaerobes in M98-5 was not significantly different (data not shown). However, replacement of CO2 with an N2 gas phase in M98-5 reduced recovery by 9%. Similarly, deletion of either Trypticase, glucose, or maltose reduced recoveries by 6 to 9% (P = 0.05). Lowest recoveries (12% less) in M98-5 were noted upon deletion of rumen fluid, suggesting that many fecal organisms required factors in rumen fluid for growth. Roll tube counts were always higher in the presence of increasing concentrations of rumen fluid in the medium. Rumen fluid at concentrations of 20 or 30% (vol/vol) resulted in recoveries of 24.6 and 26.2%, respectively (data not shown), but these were not significantly different from recoveries obtained in media with 40% rumen fluid. Replacing rumen fluid with fecal extract (40%, vol/vol) did not improve recoveries (23%), but in fact colony counts were TABLE 1. Comparison of various roll tube media on percent recovery of fecal anaerobes Mediuma Percent recov- Range Statistical sigeryb nificancec M W M1O X M % fecal extract X M % cecal extract W RGCA YZ RCM XY RCM + 5% sheep blood Z BHIAS Z BHIAS + 5% sheep blood YZ a All media were equilibrated and tubed under 100% C02 gas phase. b Means of single fecal samples from six animals. Percent recovery (wet-weight basis) = (roll tube counts/microscopic bacterial count) x 100. Microscopic bacterial counts from six pigs varied from 2.70 x 101" to 5.95 x 101' organisms per g (wet weight). c Means not followed by the same letter are significantly different at the 5% level of probability. FECAL BACTERIA OF SWINE 81 TABLE 2. Effects ofdifferent medium components on percent recovery in M98-5 roll tube medium Per- Statis- Medium deletion or addi- cent tical tion recov- Range signifierya canceb Complete WX N2 gas phasec YZ - Trypticase YX -Glucose or maltose YX - Maltose plus Starch XY - RFd Z - RF + fecal extracte XY - RF + cecal extracte W a Means of single fecal samples from six animals. Percent recovery (wet-weight basis) = (roll tube counts/microscopic bacterial count) x 100. Microscopic bacteria counts from six pigs varied from 4.48 x 1010 to 7.40 x 1010 organisms per g (wet weight). b Means not followed by the same letter are significantly different at the 5% level of probability. c In place of CO2 and medium without added Na2CO3. d RF, Rumen fluid. ' Fecal extract or cecal extract, 40% (vol/vol). 6% lower than in complete M98-5; recoveries with 10 or 20% (vol/vol) fecal extract in place of rumen fluid also were similar (23%). Substituting cecal extract (40%, vol/vol) for rumen fluid gave recoveries that were 4.5% higher but not significantly different from those obtained with M98-5. Finally, addition of yeast extract (0.20%, wt/vol) to M98-5 did not change the percent recovery. Isolation and identification of fecal bacteria. The ph of fecal samples was usually 6.8, the mean dry weight was 38.3% (range, 29.7 to 40.4), and some of the major VFA (molar percent basis) were acetic (53%; range, 48 to 59%), propionic (15%; range, 10 to 20%), and butyric (8%; range, 4 to 13%). Strains were tentatively grouped and presumptively identified on the basis of morphology and Gram stain, glucose fermentation, and products from glucose (Table 3). The predominant bacteria (90%) isolated were gram positive and consisted offacultatively anaerobic streptococci, Eubacterium sp., Clostridium sp., and Propionibacterium acnes. Several minor groups made up about 8% of the total number isolated and were designated (on the basis of features given in Table 4) as species similar to Treponema, Selenomonas, Lactobacillus, Veillonella, Bacteroides, and E. coli. The anaerobic bacterial groups in Table 4 were further characterized by additional fermentation, physiological, and biochemical tests. (i) Facultatively anaerobic bacteria. Fecal streptococci (group I), comprising 44% of the isolated microflora, were the predominant group of bacteria recovered from all samples.

4 82 SALANITRO, BLAKE, AND MUIRHEAD TABLE 3. Presumptive identification features ofpredominant groups offecal bacteria isolated from six adult swinea Percent of Morphological group Gram re- Oxygen re- Glucose ph Glucose fermen- isolated Tentative identification action lationsb tation productsc microflorad I. Streptococci + F L(f) 44.2 Streptococcus (nongroup D, viridans) II. Nonfermentative rods + A lf(a, b) 26.6 Eubacterium III. Fermentative rods + A ALf(e) 9.5 Eubacterium IV. Sporeforming rods + A ABLE (f); H2 6.8 Clostridium V. Pleomorphic rods + A 4.5 PAs(1) 3.5 Propionibacterium acnes VI. Minor groups similar to +/- A/F 7.8 Treponema, Selenomonas, Lactobacillus, Veillonella, Bacteroides, and E. coli VII. Miscellaneous groups 1.6 a Data from 381 roll tube strains isolated. Abbreviations: F, Facultatively Anaerobic; A, anaerobic. c Abbreviations: L, 1, Lactic acid; f, formic acid; A, a, acetic acid; B, b, butyric acid; E, e, ethanol or mixture of shortchain alcohols; P, propionic acid; S, succinic acid. Capital letters refer to acids produced in amounts of 10 Lmol/ml of medium or greater; lowercase letters refer to amounts less than 10,umol/ml. Products in parentheses are formed by few strains in a group. H2 was produced by group IV strains in amounts of 41 to 46% (vol/vol). d Number of strains in a group/total number isolated. e Also based on tests of Table 4. TABLE 4. Distribution of bacterial groups isolated from swine fecesa Percent of isolated strains in fecal sample: Bacterial groupb A B C D E F I. Streptococcus II. Eubacterium C 2.9 III. Eubacterium IV. Clostridium V. Propionibacterium acnes VI. Minor groups Total strains isolated and cultured a The percentage of the microflora (culture counts/microscopic bacterial counts) cultured per gram (wet weight) was 24% (mean) for samples A-D and 5% for samples E and F. Morphological groups given in Table 3. '-, No strains isolated. Thirty strains were presumptively identified according to the criteria (bacitracin sensitivity, hippurate and bile-esculin hydrolysis, and NaCl tolerance) of Facklam et al. (5) as belonging to the viridans non-group D streptococci. They were all alpha-hemolytic and readily fermented several sugars (fructose, glucose, maltose, mannose, and sucrose). Other facultatively anaerobic bacteria were identified (Improved Enterotube, Roche Diagnostics) as E. coli. The predominant organisms isolated from aerobic plate media (BHIA or EA) accounted for 2 to 3% of the direct microscopic cell count of fecal material and were also similar to the fecal streptococci and E. coli isolated from roll tubes. (ii) Nonfermentative (group II) and fermentative (group III) gram-positive rods. Anaerobic bacteria in group II made up a large portion APPL. ENVIRON. MICROBIOL. of the isolated strains (26.6%) and were identified as Eubacterium sp. (11). Group II organisms were relatively nonreactive in sugar fermentations or biochemical reactions and produced small amounts (2 to 5,umol/ml of culture) of lactate and formate from glucose. Strains in group III (9.5% of the recovered microflora) were also identified as Eubacterium sp. (11). They formed large amounts of acetic and lactic acids (35 and 14 g.mol/ml, respectively) and smaller quantities of formic acid (8 to 9,umol/ ml) from glucose. A few strains also produced small amounts of ethanol (3.2,umol/ml). Crude nutritional experiments indicated that growth of these strains was stimulated by rumen fluid (or a VFA mixture) and yeast extract. (iii) Sporeforming anaerobic rods (group IV). These strains comprised 6.8% of the total

5 VOL. 33, 1977 isolated microflora and were identified as Clostridium sp. on the basis of morphology and fermentation features (24). Large amounts of acetic, butyric, and lactic acids (22, 16, and 13 gmol/ml, respectively) and hydrogen were produced from glucose fermentation. Strains of this Clostridium species all produced ethanol or a mixture of short-chain alcohols (11,umol/ ml) from glucose. These bacteria were stimulated by addition of rumen fluid (or VFA) and Trypticase to growth media. (iv) Pleomorphic gram-positive anaerobic rods (group V). This group of bacteria comprised only 3.5% of the total isolated and was morphologically and biochemically similar to P. acnes (18), producing mainly propionic and acetic acids (50 and 28 umol/ml, respectively) from glucose metabolism. Rumen fluid (or VFA) and Trypticase stimulated growth of these bacteria. Variation in fecal microflora. Table 4 shows the distribution of the various groups and indicates that there was considerable variation in the bacterial types isolated from individual samples. In samples A through D, from which 19 to 27% of the microflora was isolated, the predominant bacteria identified from all samples were fecal streptococci and Eubacterium (group II); the remaining major groups were not consistently recovered from all four samples. In samples E and F, from which only 5% of the total microscopic bacterial count was isolated, the predominant organisms isolated were the facultatively anaerobic streptococci. It was noted, however, that fecal smears (wet mounts) from these latter two samples showed large numbers of spores and small motile spirochetelike bacteria. The unusually low recovery of isolates in some samples suggests that these bacteria may have stringent nutrient requirements for growth that were not satisfied by the roll tube media used. If a prior heat shock treatment had been given to fecal samples in which large numbers of spores were observed, increased recovery of bacteria presumably would have been obtained. In addition, we have only been able to recover the spirochete organisms as mixed-colony isolates in association with other fecal bacteria identified in this study. DISCUSSION These results clearly indicate that a rumen fluid roll tube medium such as M98-5 or a similar medium with swine cecal extract is better than other media tested for isolating and culturing the predominant bacteria in pig fecal material. With these media, 23 to 34% of the FECAL BACTERIA OF SWINE 83 organisms present in microscopic smears could be recovered. Our previous work in chickens (22) also demonstrated the superiority of a rumen fluid medium for isolation of the majority of the anaerobic bacteria from the chicken cecum. Other workers (4, 19) have also used rumen fluid roll tube media for isolating human fecal bacteria. It was unexpected that media containing swine fecal extract would not be better than rumen fluid media for isolating fecal bacteria. There may exist some natural inhibitors in feces that interfere with the growth of many fecal organisms. Our studies also illustrate that nonselective, highly nutritious, and complex primary isolation media such as RCM and BHIA are less suitable for isolating swine fecal anaerobes. RCM was been used (roll tubes or plates) by several workers (7, 14, 20, 27) in attempts to isolate the predominant intestinal bacteria from pigs. The results in Table 1 indicate that 11 to 17% lower percentages of recovery of direct counts were obtained as compared with rumen fluid media (M98-5). Addition of blood to these media does not improve isolation. Previous experience (22) with blood agar media also indicated that substantially fewer numbers of bacteria were recovered from chicken cecal digesta. In this paper we have determined that by using strict anaerobic methods, the predominant fecal microflora isolated from adult swine consisted of several bacterial groups; namely, fecal streptococci, Eubacterium sp., Clostridium sp., and P. acnes (Table 3). No previous study in pigs has demonstrated the isolation of these bacterial types or the distribution of anaerobic species in swine feces. It should be mentioned, however, that the types of intestinal bacteria recovered will be influenced by several factors, such as (i) the isolation media and methods used, (ii) the sampling site (upper or lower gastrointestinal contents and digesta), (iii) dietary composition and nutrition of animals, (iv) environmental conditions (housing and sanitation), and (v) sex and age of the animals. There were marked variations in the proportions of bacterial types isolated from the feces of swine (Table 4). Although not considered in this work, the microfloral variation between animals is yet another aspect to consider when assessing the importance of intestinal microflora populations. LITERATURE CITED 1. Aalbaek, B Gram-negative anaerobes in the intestinal flora of pigs. Acta Vet. Scand. 13: Bryant, M. P., and L. A. Burkey Cultural methods and some characteristics of some of the more numerous groups of bacteria in the bovine rumen. J. Dairy Sci. 36:

6 84 SALANITRO, BLAKE, AND MUIRHEAD 3. Carle, G. C Gas chromatographic determination of hydrogen, nitrogen, oxygen, methane, krypton, and carbon dioxide at room temperature. J. Chromatogr. Sci. 8: Eller, C., M. R. Crabill, and M. P. Bryant Anaerobic roll tube media for non-selective enumeration and isolation of bacteria in human feces. Appl. Microbiol. 22: Facklam, R. R., J. F. Padula, L. G. Thackers, E. C. Wortham, and B. J. Sconyers Presumptive identification of group A, B, and D streptococci. Appl. Microbiol. 27: Fewins, G., L. G. M. Newland, and C. A. E. Briggs The normal intestinal flora of the pig. III. Qualitative studies of lactobacilli and streptococci. J. Appl. Bacteriol. 20: Fuller, R., and M. Lev Quantitative studies on some of the gram-negative anaerobic bacteria in the pig alimentary tract. J. Appl. Bacteriol. 27: Gilliland, S. E., M. L. Speck, and C. G. Morgan Detection of Lactobacillus acidophilus in feces of humans, pigs, and chickens. Appl. Microbiol. 30: Graber, C. D., R. W. Moore, M. Suzuki, W. E. Redmond, R. M. O'Neal, and B. M. Lockhart Autochthonous intestinal bacterial flora and cholesterol levels in specific pathogen-free swine fed highlipid and high-sucrose diets. J. Bacteriol. 92: Holdeman, L. V., and W. E. C. Moore (ed.) Anaerobe laboratory manual. Virginia Polytechnic Institute and State University Anaerobe Laboratory, Blacksburg. 11. Holdeman, L. V., and W. E. C. Moore Genus Eubacterium, p In R. E. Buchanan and N. E. Gibbons (ed.), Bergey's manual of determinative bacteriology, 8th ed. The William and Wilkins Co., Baltimore. 12. Hungate, R. E A roll tube method for cultivation of strict anaerobes, p In J. R. Norris and D. W. Ribbons (ed.), Methods in microbiology, vol. 3B. Academic Press Inc., New York. 13. Kenworthy, R., and W. E. Crabb The intestinal flora of young pigs with reference to early weaning, Escherichia coli, and scours. J. Comp. Pathol. 73: Kolacz, J. R., R. B. Westcott, and A. R. Dommert Influence of age and rations on fecal microflora of hormel miniature swine. Am. J. Vet. Res. 32: APPL. ENVIRON. MICROBIOL. 15. Miller, T. L., and M. J. Wolin A serum bottle modification ofthe Hungate technique for cultivating obligate anaerobes. Appl. Microbiol. 27: Mitsuoka, T Vergleichende Untersuchungen uber die Laktobazillen aus den Faeces von Menschen, Schweinen und Huhnern. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 1 Orig. 210: Mitsuoka, T., A. Terada, K. Watanabe, and K. Uchida Bacteroides multiacidus, a new species from the feces of humans and pigs. Int. J. Syst. Bacteriol. 24: Moore, W. E. C., and L. V. Holdeman Genus Propionibacterium, p In R. E. Buchanan and N. E. Gibbons (ed.), Bergey's manual of determinative bacteriology, 8th ed. The William and Wilkins Co., Baltimore. 19. Moore, W. E. C., and L. V. Holdeman Human fecal flora: the normal flora of 20 Japanese-Hawaiians. Appl. Microbiol. 27: Rall, G. D., A. J. Wood, R. B. Westcott, and A. R. Dommert Distribution of bacteria in feces of swine. Appl. Microbiol. 20: Salanitro, J. P., I. G. Blake, and P. A. Muirhead Studies on the cecal microflora of commercial broiler chickens. Appl. Microbiol. 28: Salanitro, J. P., I. G. Fairchilds, and Y. D. Zgornicki Isolation, culture characteristics and identification of anaerobic bacteria from the chicken cecum. Appl. Microbiol. 27: Salanitro, J. P., and P. A. Muirhead Quantitative method for the gas chromatographic analysis of short-chain monocarboxylic and dicarboxylic acids in fermentation media. Appl. Microbiol. 29: Smith, L. D. S., and G. Hobbs Genus Clostridium, p In R. E. Buchanan and N. E. Gibbons (ed.), Bergey's manual of determinative bacteriology, 8th ed. The William and Wilkins Co., Baltimore. 25. Snedecor, G. W Statistical methods, 5th ed. The Iowa State University Press, Ames. 26. Terada, A., K. Uchida, and T. Mitsuoka Bacteroidaceae flora in faeces of pigs. Zentralbl. Bakteriol. Parasitenkd. Abt : Vervaeke, I. J., and C. J. van Nevel Comparison of three techniques for the total count of anaerobes from intestinal contents of pigs. Appl. Microbiol. 24: Zani, G., B. Biavati, F. Crociani, and D. Matteuzzi Bifidobacteria from faeces of piglets. J. Appl. Bacteriol. 37: