Medium Without Rumen Fluid for Nonselective Enumeration and Isolation of Rumen Bacteria

Transcription

1 APPLIED MICROBIOLOGY, Sept., 1966 Vol. 14, No. 5 Copyright 1966 American Society for Microbiology Printed in U.S.A. Medium Without Rumen Fluid for Nonselective Enumeration and Isolation of Rumen Bacteria DANIEL R. CALDWELL AND MARVIN P. BRYANT' Animal Husbandry Research Division, U.S. Department ofagriculture, Beltsville, Maryland Received for publication 28 April 1966 ABSTRACT CALDWELL, DANIEL R. (U.S. Department of Agriculture, Beltsville, Md.), AND MARVIN P. BRYANT. Medium without rumen fluid for nonselective enumeration and isolation of rumen bacteria. Appl. Microbiol. 14: Colony counts which approximated those in a habitat-simulating, rumen fluid-agar medium (RFM) were obtained in medium 10, a medium identical to the RFM except for the replacement of rumen fluid with 1.5 X 10-6 M hemin, 0.2% Trypticase, 0.05% yeast extract, and a 6.6 X 10-2 M volatile fatty acid mixture qualitatively and quantitatively similar to that in rumen fluid. Single deletion of Trypticase, yeast extract, or the volatile fatty acid mixture from medium 10 significantly reduced colony counts. Colony counts were also reduced when medium 10 was modified to contain higher concentrations of Trypticase or volatile fatty acids. Significant differences were found between colony counts obtained from diluted rumen contents of animals fed a cracked corn-urea diet, and the colony counts obtained from animals fed either a cracked corn-soyean oil meal or an alfalfa hay-grain diet. Qualitative differences were found between the predominant bacterial strains isolated from rumen contents of animals fed cracked corn diets and strains isolated from animals fed alfalfa haygrain. Regardless of differences in the predominant flora associated with diet, medium 10 and the RFM supported growth of similar bacterial populations. The results show that medium 10 is suitable for enumeration and isolation of many predominant rumen bacteria. Relatively nonselective culture media previously used for enumeration and isolation of rumen bacteria, have been divided into two types: (i) "habitat-simulating" media, which contain rumen fluid, several carbohydrates (e.g., glucose, cellobiose, starch), minerals, reducing agents, and a carbon dioxide-bicarbonate buffer system, but which are relatively low in organic matter (7, 16); and (ii) media without rumen fluid which contain large amounts of organic materials such as yeast extract and peptones. Habitat-simulating rumen fluid media have generally supported growth of larger numbers of organisms and a greater variety of species than have media devoid of rumen fluid, presumably because rumen fluid contains a number of growth factors (e.g., certain volatile fatty acids and heme), which are not commonly found in the extracts and hydrolysates usually added to bacteriological media (16). In addition, large amounts of certain materials containing large amounts of organic nitrogen have not been added 1 Present address: Department of Dairy Science, University of Illinois, Urbana. to most habitat-simulating rumen fluid media, because some of these materials have been shown to be inhibitory to certain rumen bacteria (18). Recent studies have indicated that many strains of the majority of species of rumen bacteria can be grown in defined media, or in media containing enzymatic casein hydrolysate as the only undefined ingredient (9), and have suggested that a medium for the enumeration and isolation of rumen bacteria could be developed in which rumen fluid is replaced by more standardized ingredients. A medium of this type would be useful, because rumen fluid varies considerably, even when collected on different days from one animal held under one set of conditions, and because rumen fluid is sometimes unavailable to workers who wish to use such a medium for study of similar bacteria in other anaerobic habitats (e.g., other animal gastrointestinal tracts or sewage sludge). Thorley, Sharpe, and Bryant (in preparation) used a medium identical to that previously used by Bryant and Robinson (7),. except that rumen fluid was replaced by a mixture of volatile fatty 794

2 VOL. 14, 1966 MEDIUM FOR RUMEN BACI ERIA 795 acids similar to that in rumen fluid, hemin, and low concentrations of Trypticase and yeast extract. This medium supported growth of numbers and species of bacteria similar to those obtained with the rumen fluid medium when rumen contents from animals fed dried-grass diets were studied in England. Since the diets fed to animals in England were substantially different from those common to the United States, the present study was initiated to evaluate the medium under American conditions, and to study the effects of variation in the concentration of certain components of the medium on colony counts. MATERIALS AND METHODS Animals. Three mature Holstein cows were fed an alfalfa hay-grain mixture (63% alfalfa hay) at a recommended maintenance level twice daily at 4:00 and 16:00. Water was available between 5:00 and 7:00, and between 13:00 and 14:00. The remaining four animals were allowed feed and water ad libitum. Two yearling Angus steers were fed a diet composed of 41.7 kg of cracked corn, 0.73 kg of urea, 2.3 kg of molasses, 0.46 kg of steamed bone meal, 0.46 kg of trace-mineralized NaCl, 200 IU of vitamin D, and 160,000 IU of vitamin A. The remaining two animals, mature Holstein cows, were fed a cracked-corn diet identical to that described above except that each 46 kg of diet contained only 38 kg of cracked corn, and soybean oil meal (11%, w/v) replaced urea as the major nitrogen source. Animals were maintained on their respective diets for at least 1 month prior to sampling of rumen contents. For studies of the effects of components of medium 10 upon colony counts, animals fed the alfalfa haygrain diet were used. For studies of components other than yeast extract, one animal fed alfalfa hay-grain was used. For studies of the effects of yeast extract, a different animal, also fed alfalfa hay-grain, was used. For studies of the predominant flora in animals fed different diets, two different animals were sampled from each of the cracked-corn diets. The results were compared with those obtained by duplicate sampling of a single animal fed alfalfa hay-grain. Roll tube and pure culture maintenance media. The compositions of the rumen fluid medium (RFM) and of the medium without rumen fluid (medium 10) used for roll tubes are given in Table 1. The RFM was similar to medium 98-5 previously used for the study of rumen bacteria (7). Medium 10 was identical to the RFM except that rumen fluid was replaced by yeast extract, Trypticase, hemin, and a mixture of volatile fatty acids (VFA) such that the concentrations of individual acids were similar to those in a 40% rumen fluid medium. The inclusion of VFA in medium 10 necessitated the adjustment of the ph of this medium to 6.5 with 2.5 N NaOH prior to autoclaving. With this exception, the anaerobic techniques, methods of media preparation and sterilization, procedures for media tubing, reduction, and inoculation, and incubation conditions were essentially those described or re- TABLE 1. Compositions of the rumen fluid and non-rumen fluid roll tube media, Component Percentage in mediumb Rumen fluid Medium 10 medium Clarified rumen fluid (v/v)' Glucose Cellobiose Soluble starch Minerals (V/V)d Na2S*9H Cysteine*HCl*H Resazurin Na2CO Trypticase Yeast extract Volatile fatty acids (v/v)e Hemin Agar CO2 gas phase The final ph of both media was 6.7 to 6.8. b Weight/volume unless otherwise indicated. c The rumen fluid was CRF-2 of Bryant and Robinson (7). d Except for K2HPO4, which was prepared separately, the minerals were prepared as a common stock solution. The final mineral concentrations in the medium were as follows: KH2PO4, 1.3 X 103 M; NaCl, 7.6 X 10-4 M; (NH)2SO4, 3.4 X 10-' M; K2HPO4, 1.7 X 10-3 M; Ca(C1)2, 4.1 X 10-4 M; MgSO4c7H20, 3.8 X 10-4 M. e The volatile fatty acid mixture used was the following: acetic, 17 ml (2.9 X 10-2 M); propionic, 6ml (8.0 X 10-8M); butyric, 4 ml (4.3 X 10-8 M); isobutyric, 1 ml (1.1 X 10-3 M); n-valeric, isovaleric, and DL-ai-methylbutyric, 1 ml each (9 X 10-4 M). The figures in parentheses indicate the final concentration of each acid when 0.31 ml of mixture was placed in a final medium volume of 100 ml. ferred to by Bryant and Robinson (7). The medium for maintenance of pure cultures picked from roll tubes was the RGCA slant medium of Bryant and Burkey (5) as modified by Bryant and Robinson (9). Sampling and inoculation procedures. Rumen contents were sampled by stomach tube between 9:00 and 13:00. Representative samples (at least 200 g) were brought immediately to the laboratory, and 10-g subsamples, containing proportions of solid and liquid material which approximated those in the original, were weighed out and were washed with 90 ml of anaerobic mineral solution (7) into a Waring Blendor which was continually gassed with a heavy stream of CO2. All subsequent operations were conducted under anaerobic conditions. The Blendor contents were agitated for 1 min. The homogenized contents were then serially diluted, in 10-fold steps, in the anaerobic mineral solution of Bryant and Burkey (5). Quantities of dilutions sufficient to produce from 30 to 120

3 796 CALDWELL AND BRYANT APPL. MICROBIOL. colonies per tube (0.2 to 1.0 ml) were inoculated via pipette into tubes of medium 10 and the rumen fluid medium which had been reduced and melted as previously described (7), and were maintained in a water bath at 46 to 48 C. Within 15 min after inoculation, roll tubes were prepared by rapidly rotating inoculated tubes under a cold-water tap. The solidified tubes were incubated in an upright position at 37 C for 3 days. Colony counts. Colony counts in roll tubes were obtained by use of an electronic colony counter equipped with an ink pen (model C 100; New Brunswick Scientific Co., New Brunswick, N.J.). Four replicate roll tubes of each medium were counted. The colonycounting criteria of Bryant and Robinson (7) were used. The counts obtained were analyzed statistically by means of analysis of variance and by Duncan's multiple-range test for separation of means (17). Isolation andpresumptive identification oforganisms. Material from well-isolated colonies was picked from roll tubes with a bent platinum-iridium wire, and was stab-inoculated into RGCA slant medium. Colonies were selected from roll tubes as randomly as possible, and as many colonies as possible were picked from the same roll tube. Inoculated slant cultures were incubated at 37 C. Growth was obtained in most of the cultures within or 2 days. Cultures in which growth was not detected within this time were incubated for week before being discarded. Organisms in the water of syneresis from cultures showing growth were examined in wet mounts by phase-contrast microscopy for morphology, motility, and homogeneity, and were also Gram stained. Cultures of organisms determined to be microscopically homogenous were subjected to the following physiological tests: (i) oxygen relation, (ii) motility (agar), (iii) appearance of growth in glucose broth, (iv) final ph in weakly buffered glucose broth, (v) growth and production of gas in glucose-agar, (vi) starch hydrolysis, (vii) cellulose digestion, and (viii) H2S production. The medium used for determination of oxygen relation, motility, and H2S production was that described by Bryant and Small (11), except that 0.1% glucose was replaced with 0.025% each of glucose, cellobiose, and soluble starch. The media used to determine the appearance and final ph in glucose broth and for determination of growth and gas production in glucose-agar were described by Bryant et TABLE 2. al. (12). The medium used for determination of cellulose digestion and starch hydrolysis was the basal medium described by Bryant and Small (11) with the addition of 30% (v/v) rumen fluid and 0.1% (w/v) each of soluble starch and ball-milled Whatman no. 1 filter paper (5). Starch hydrolysis was determined by the iodine method described by Bryant and Doetsch (6). The methods used for inoculation of test media were those described by Bryant and Burkey (5), except that the glucose broth medium was used to prepare inocula. Previously described criteria (2, 4, 10, 13) were used for presumptive identification of isolates. RESULTS Colony counts in medium 10 and the rumen fluid medium. The data in Table 2 show the colony counts obtained in preliminary experiments in which medium 10 and RFM were inoculated with rumen contents from animals fed an alfalfa hay-grain diet. Although there was some variation from experiment to experiment in the colony count, it is evident that the counts in medium 10 were often substantially lower than those in RFM. In efforts to obtain counts which more closely approximated those in RFM, the effects of Trypticase, yeast extract, and VFA upon the colony count in medium 10 were studied. The data in Table 3 show the effects of different concentrations of Trypticase upon colony counts in medium 10 with other ingredients held constant. Deletion of Trypticase reduced counts. A reduction of count also occurred in media containing Trypticase concentrations greater than 0.2% (w/v). The data in Table 4 show the effects of yeast extract concentration upon colony count in medium 10 with other ingredients held constant. Deletion of yeast extract reduced the colony count, but no significant differences were found between counts in media containing 0.05 or 0.2% (w/v) yeast extract. No significant differences were found in experiments which compared BBL yeast extract with Oxoid yeast extract. Colony countsa in roll tubes of the rumen fluid medium and medium 10 inoculated with diluted rumen contents from animals fed an alfalfa hay-grain diet Medium Expt I Expt 2 Expt 3 Expt 4 Expt 5 Meanb SDC Mean SD Mean sid Mean sd Mean SD Rumen fluid x x x x x Medium y y x x y a Colony counts in billions per gram of rumen contents. b The means are the average colony counts in four replicate tubes. Within an experiment, means followed by a given letter are statistically different (5% level of probability) from means not followed by the same letter. - Standard deviation of four replicate tubes.

4 VOL. 14, 1966 MEDIUM FOR RUMEN BACIERIA 797 TABLE 3. Effect of Trypticase concentration on colony counts in medium 10 Trypticase in medium Colony count X 10-9 per gram Meana Statistical b significance x y x z Mean of data from two experiments. b Means followed by a given letter are statistically different at the 5% level of probability from means not followed by that letter. TABLE 4. Effect ofyeast extract concentration on colony count in medium 10 Yeast extract Colony count X 10-9 per gram Mean' Statistical significance b x y y a Mean of data from two experiments. b Means followed by a given letter are statistically different at the 5% level of probability from means not followed by that letter. Data in Table 5 show the effects of varying the concentration of the VFA mixture on colony counts in medium 10. Deletion of the mixture severely reduced the counts, and the optimal concentration was about 6.6 X 10-2 M. That the stimulation in count was due to the VFA, rather than ion concentration, is indicated by the data showing a similar growth response when the media were held at a constant salt concentration by appropriate additions of NaCl. That the depression in count in media containing additions of the VFA mixture above 6.6 x 10-2 M was due to the ionic concentration, rather than to the volatile fatty acids as such, is suggested by the data in Table 6, which show that additions of either NaCl or NaHCO3 to the medium containing 6.6 X 10-2 M VFA mixture reduced colony counts. Evaluation of modified medium 10 for study of rumen bacteria. To compare medium 10, modified to contain 6.6 X 10-2 M total VFA, with the RFM, colony counts and presumptive identifications of large numbers of strains were carried out with cattle fed three diets. Although large and significant differences in colony counts were found between different animals fed different diets (Table 7), e.g., the corn-urea compared with either the alfalfa hay-grain or the corn-soybean diet, and in one case between different animals on the TABLE 5. Effect of additions of various molar concentrations of sodium salts of volatile fatty acids (VFA) and chloride (Cl) on colony counts in medium 10 minus VFA Expt Molarity of added Colony count 10-9 salts X 10-2 per gram Statistical VFA Cl Meana significanceb x yz yz z yz xy x xy y y Means of media in experiment 1 are the average of three replicate experiments. Means of media in experiment 2 are the average of two replicate experiments. bmeans of media containing unequal salt cannot be directly compared to those from media containing equal salt, but, within these two groups, footnote b of Table 3 applies to this table. TABLE 6. Effect of addition of sodium bicarbonate (HCOa) or sodium chloride (Cl) to medium 10 modified to contain 6.6 X 10- M volatile fatty acids Molarity of added salts X 102 Colony count X 10-9 per gram HCO, HM ci Cl Meana ~~~~~~~Statisticalb significance x Xy yz z z x x x y y Average data from two experiments for each salt. bwithin media containing a given salt, footnote b of Table 3 applies to this table but media containing corresponding concentrations of each salt cannot be directly compared.

5 798 CALDWELL AND BRYANT APPL. MICROBIOL. TABLE 7. Colony counts in medium 10 modified to contain 6.6 X 10-2 M volatile fatty acids and in the rutmen fluid medium (RFM) inoculated with rumen contents from animals fed different diets Animal diet Medium Colony count X 10-3 per gram Sample 1 Sample 2 Meanb Alfalfa hay-grain a x 2.48 x 2.84 x i 0.51 RFM 2.93 x 2.43 x 2.68 x i 0.27 Corn-soybean x 3.88 x 3.12 x i 0.91 RFM 4.80 y 4.65 x 4.73 x ± 0.60 Corn-urea x 8.12 y y RFM y 9.80 y y a Means of four observations. Sample means within a ration followed by a particular letter are statistically different (5% level) from sample means not followed by that letter. bmedium means followed by a particular letter are significantly different (1% level) from medium means not followed by that letter. TABLE 8. Distribution of strains of groups of anaerobic bacteria isolated from modified medium 10 and the rumen fluid medium (RFM) inoculated with rumen contents from cattle fed three diets Percentage of total strainsa Presumptive identification Alfalfa hay-grain Corn-urea Corn-soybean 10 RFM 10 RFM 10 RFM Butyrivibrio sp Butyrivibrio (atypical) B-385-like Succinivibrio sp Selenomonas sp Lachnospira sp Succinimonas sp Bacteroides ruminticola B. succinogenes B. amylophillus Borrelia sp Eubacterium rumilnanitium Peptostreptococcus elsdenii Ruminococcus sp Aniaerobic lactobacillus Unidentified groups Nonmotile rods (a) Gram-negative (b) Gram-positive or, variable Cocci Total strains a The data for each medium within a ration is the average of two replicate experiments. same diet, i.e., the corn-urea diet, there were no soybean and corn-urea diets. In one corn-urea significant differences (5% level of probability) sample, counts in the RFM were significantly between the colony counts obtained from the two higher (5% level of probability) than those obmedia, except in one sample each from the corn- tained with modified medium 10. In one corn-

6 VOL. 14, 1966 MEDIUM FOR RUMEN BACTERIA 799 soybean sample, the counts in medium 10 were significantly higher (5 % level of probability) than those in the RFM. The distribution of strains among the presumptively identified groups of bacteria also varied considerably with the diet and animal, but considerably less variation was found between the two media (Table 8), and most of the groups were isolated with both media. Presumptively identified groups isolated only from medium 10 included Lachnospira, Borrelia, and Ruminococcus species, but only one strain of each of these was found. Extensive previous work indicates that these groups grow well on the RFM. The B-385 group was the only group isolated from the RFM which was not isolated from medium 10, but, again, only two strains were isolated. The frequency of occurrence of bacteria listed among the unidentified groups (Table 8) was similar on the two media. The anaerobic, gramnegative, nonmotile rods belong in the genus Bacteroides, but differ from the presumptively identified Bacteriodes species in one or more of their physiological characteristics. Previous detailed studies of similar organisms (1, 13) suggest that most of these strains were actually variants of, or closely related to, B. ruminicola. Many of these strains from animals fed the corn diets were identical to B. ruminicola, except for a more regular rod shape and a final ph of 4.4 or lower in glucose medium. With the exception of B. amylophilus, which does not ferment glucose, most strains of other previously described rumen bacteroides species produce a final ph in glucose medium of 5 to 6. It is possible that the Bacteroides strains which produced a ph of 4.4 or lower are actually acidoduric members of the species B. ruminicola which were selected by the low ph (5.8 or less) of the rumen contents of animals fed the corn diets. DIscussIoN Evaluation of the medium. The present study confirms earlier work of Thorley, Sharpe, and Bryant (in preparation) showing that a medium in which rumen fluid is replaced by hemin, VFA, and low levels of Trypticase and yeast extract supports growth of bacteria similar to that obtained by use of the RFM previously developed as probably the best nonselective medium presently available for isolation and enumeration of carbohydrate-fermenting rumen bacteria (7). The two media are quite similar as judged by similar viable-colony counts and by isolation of similar species of bacteria in similar proportions from rumen contents of cattle fed diets showing large differences in numbers and species of bacteria. Both media support growth of bacterial populations which appear to reflect the microscopically observed predominant flora. Medium 10 therefore appears to be useful for enumeration and isolation of many species of rumen bacteria, and is superior to the RFM in some respects, i.e., that it contains well-standardized ingredients which are readily available to workers who may not have access to rumen fluid. This is most likely to be the case with workers interested in studying similar anaerobic bacteria present in other habitats. We still believe, however, that rumen fluid media should be used where possible for obtaining new isolates, because certain strains of some species of rumen bacteria require as yet unknown growth factors (9, 16), which might not be present in medium 10. Some species, e.g., Methanobacterium ruminantium, definitely require unknown factors not present in medium 10 (3). It should also be noted that some species will not grow in either of these media because of lack of appropriate energy sources. This includes bacteria such as the lipolytic vibrio of Hobson and Mann (15), veillonellae, and M. ruminantium (19), which require glycerol, lactate, and hydrogen gas or formate, respectively, as energy sources. In the study of Thorley et al. (in preparation), no attempt was made to determine the optimal concentrations of the various ingredients added in place of rumen fluid, because, under the conditions studied, colony counts in medium 10 compared favorably with those in the RFM. In the present study, counts in medium 10 were often somewhat lower than those of the RFM (Table 2), suggesting a study of the efficacy of the concentrations of VFA, Trypticase, and yeast extract added. The present results showing that deletion of the VFA mixture drastically lowers colony counts agrees with previous results of Wegner and Foster (Bacteriol. Proc., p. 24, 1957) showing that a similar Trypticase-Phytone-glucose-cellobiose medium gave higher colony counts when mixtures of VFA were added. Also, many strains of many species of rumen bacteria require or are highly stimulated by one or more of the acids, acetate, isobutyrate, and the three valerate isomers (9, 16). Propionate and n-butyrate are not known to be required for growth by any rumen bacteria, but were included in the medium becaise they are normal constituents of rumen fluid. Tryticase and yeast extract were added to the medium as sources of B vitamins, amino acids, peptides, and possibly other organic and inorganic growth factors. Most species of rumen bacteria require one or more of the B vitamins, and some species require or are stimulated by amino

7 800 CALDWELL AND BRYANT APPL. MICROBIOL. acids or peptides (16). It is of interest that the combination of yeast extract and Trypticase increased colony counts above those obtained with either ingredient alone, and that relatively low levels of each substance resulted in the highest colony counts. The reduced colony counts in media containing levels of Trypticase above 0.2% (w/v) agrees with previous studies showing that high concentrations of various organic nitrogenous materials are inhibitory to certain rumen bacteria (18). The addition of hemin is necessary, because most strains of B. ruminicola, usually one of the most numerous rumen anaerobes, require it or related tetrapyrroles which are present in rumen fluid (14). Flora and fauna differences between diets and animals. The present experiments were not well designed to study differences in the flora and fauna due to diet, since it was not possible to study the same animals being fed different diets. Therefore, some of the differences in colony counts and distribution of strains among the presumptively identified groups from animals fed different diets could be due to animal differences rather than to dietary differences. However, some dietary effects are evident. As with previous results (5, 7, 10), the colony counts from rumen contents of the present cattle fed high grain diets were usually higher than those from cattle fed alfalfa hay-grain. This was especially evident in the counts from animals fed the corn-urea diet, which were statistically significantly different ( <1 %level of probability) from the counts in samples from either alfalfa hay-grain or corn-soybean fed animals. The colony counts in samples from the corn-soybean fed animals were usually also higher than those in samples from the animals fed alfalfa hay-grain, but the differences could not be shown to be significant at the 5% level of probability. Large and significant differences often occur between samples from the same or different animals fed the same high grain diet, and this was the present case in the samples from the two animals fed the corn-urea diet. The fact that few or no ciliate or flagellate protozoa were seen in samples from the cattle fed cracked corn correlates with the low ph (5.8 or less) obtained for these samples, and, in the case of the cracked corn-urea diet, with higher bacterial colony counts (10). The ciliate protozoa were always present in large numbers in the animals fed alfalfa hay-grain. The distribution of presumptively identified strains of bacteria (Table 8) in the animal fed alfalfa hay-grain was similar to that previously found (1, 13) in that most of the strains were gram-negative, many different species were isolated, and butyrivibrios and B. ruminicola were quite numerous. The only somewhat unusual feature was the fail;re to isolate any ruminococci. The distribui oun of presumptively identified strains of bacteria (Table 8) from the animals fed the corn-urea diet was drastically different from resuits previously obtained with cattle fed high grain diets (5). In previous studies, in which yellow corn meal or barley was the main source of carbohydrate, and soy bean oil meal was the main source of nitrogen in the diet, butyrivibrios and succinivibrios were usually the most numerous organisms isolated. B. ruminicola and various other species were present in smaller numbers. In some cases, the B 385-like group of spirilla largely replaced the vibrios. The present study contrasts with previous results. Selenomonads were the dominant organisms (Table 8). Phase-contrast microscopic observation of wet amounts of samples suggest that these large bacteria accounted for well over 50% of the total microbial protoplasm. Gram-positive bacteria, including anaerobic latobacilli, were quite numerous, as was also the lactate fermenter, Peptostreptococcus elsdenii. Large numbers of unidentified gram-negative, nonmotile rods similar to B. ruminicola, except for a more regular rod shape and final ph of 4.4 or lower in glucose broth, were also obtained. The distribution of presumptively identified strains in the animals fed corn-soybean oil meal (Table 8) differed somewhat from results with the animals fed corn-urea, because, although selenomonads, anaerobic nonmotile gram-positive rods including lactobacilli, and anaerobic nonmotile gram-negative rods were still quite numerous, more strains of other groups such as B. ruminicola, B. amylophilus, and succinivibrios were obtained. Whether the difference in the floras from animals on the two corn diets was due to the main nitrogen source (urea as opposed to soybean oil meal), or to animal differences, cannot be determined from the present results. The present study appears to be one of very few reports of a condition in the ramen of mature cattle where one species of bacterium is probably responsible for much of the metabolic activity, and appears to be the only report of such a high incidence of S. ruminantium except for the case of samples taken shortly after first exposure to penicillin of cattle fed a high grain diet (10). The results suggest that cracked corn may be a controlling factor in the development of this unusual microbial balance. However, the urea of the diet could also be involved in interactions causing the selection of selenomonads.

8 VOL. 14, 1966 MEDIUM FOR RUMEN BACTERIA 801 ACKNOWLEDGMENTS We thank Chester H. Gordon for useful criticisms and suggestions concerning the statistical analysis. LITERATURE CITED 1. BLADEN, H. A., M. P. BRYANT, AND R. N. DOETSCH A study of bacterial species from the rumen which produce ammonia from protein hydrolysate. Appl. Microbiol. 9: BRYANT, M. P Symposium on microbial digestion in ruminants: identification of groups of anaerobic bacteria active in the rumen. J. Animal Sci. 22: BRYANT, M. P Rumen methanogenic bacteria, p In R. W. Dougherty, R. S. Allen, W. Burroughs, N. L. Jacobson, and A. D. McGilliard [ed.], Physiology of digestion in the ruminant. The Williams & Wilkins Co., Baltimore. 4. BRYANT, M. P., B. F. BARRENTINE, J. F. SYKES, I. M. ROBINSON, C. V. SHAWYER, AND L. W. WILLIAMS Predominant bacteria in the rumen of cattle on bloat-provoking ladino clover pasture. J. Dairy Sci. 43: 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: BRYANT, M. P., AND R. N. DOETSCH A study of the activity cellulolytic rod-shaped bacteria of the bovine rumen. J. Dairy Sci. 37: BRYANT, M. P., AND I. M. ROBINSON An improved nonselective culture medium for ruminal bacteria and its use in determining diurnal variation in numbers of bacteria in the rumen. J. Dairy Sci. 44: BRYANT, M. P., AND I. M. ROBINSON Studies on the nitrogen requirements of some ruminal cellulolytic bacteria. Appl. Microbiol. 9: BRYANT, M. P., AND I. M. ROBINSON Some nutritional characteristics of predominant cultural ruminal bacteria. J. Bacteriol. 84: BRYANT, M. P., I. M. ROBINSON, AND I. L. LIN- DAHL A note on the flora and fauna in the rumen of steers fed a feedlot bloat-provoking ration and the effect of penicillin. Appl. Microbiol. 9: BRYANT, M. P., AND N. SMALL The anaerobic monotrichous butyric acid producing curved rod shaped bacteria of the rumen. J. Bacteriol. 72: BRYANT, M. P., N. SMALL, C. BOUMA, AND H. CHu Bacteroides ruminicola n. sp. and Succinimonas amylolytica the new genus and species. J. Bacteriol. 76: BRYANT, M. P., N. SMALL, C. BOUMA, AND I. ROBINSON Studies on the composition of the ruminal flora and fauna of young calves. J. Dairy Sci. 41: CALDWELL, D. R., D. C. WHITE, M. P BRYANT, AND R. N. DOETSCH Specificity of the heme requirement for growth of Bacteroides ruminicola. J. Bacteriol. 90: HOBSON, P. N., AND S. 0. MANN The isolation of glycerol-fermenting and lipolytic bacteria from the rumen of the sheep. J. Gen. Microbiol. 25: HUNGATE, R. E., M. P. BRYANT, AND R. A. MAH The rumen bacteria and protozoa. Ann. Rev. Microbiol. 18: LECLERG, E. L Mean separation by the functional analysis of variance and multiple comparisons. U.S. Dept. Agr. ARS McNEILL, J. J., R. N. DOETSCH, AND J. C. SHAW Some nutritional requirements of bovine rumen bacteria. J. Dairy Sci. 37: SMITH, P. H., AND R. E. HUNGATE Isolation and characterization of Methanobacterium ruminatium, n.sp. J. Bacteriol. 75: