Characteristics of Selenomonas ruminantium var. bryanti var. n. from the Rumen of Sheep

JOURNAL OF BACTERIOLOGY, Mar. 1971, p. 820-825 Copyright 1971 American Society for Microbiology Vol. 105. No. 3 Printed in U.S.A. Isolation, Culture, and Fermentation Characteristics of Selenomonas ruminantium var. bryanti var. n. from the Rumen of Sheep R. A. PRINS' Department of Bacteriology, University of California, Davis, California 95616 Received for publication 23 November 1970 Large forms of Selenomonas sp. were isolated from the sheep rumen on a rumen fluid-glucose-agar medium by using a differential centrifugation technique to purify the inoculum. The cells from the six isolated strains were curved, gram-negative, strictly anaerobic crescents, and rapidly motile by flagella attached to the concave side of the cell. One or more of the volatile fatty acids were essential for growth. None of the strains produced indole or reduced nitrate. All strains grew on fructose, glucose, mannose, cellobiose, maltose, sucrose, and salicin. Fermentation end products from glucose were mainly lactate, acetate, propionate, and formate. Small amounts of succinate were formed. The final ph in a glucose medium ranged between 4.3 and 4.5. On the basis of the sugar fermentation characteristics and the capacity to form hydrogen sulfide from cysteine, it is suggested that one of the strains is a large form of Selenomonas ruminantium. The other five strains are designated S. ruminantium var. bryanti, var. n. Although many rumen bacteria have been isolated and identified (8), the large species in the rumen of sheep (13) have not yet been cultured and virtually nothing is known of their nutrition and physiology. Woodcock and Lapage (18) described Selenomastix ruminantium from the rumen of goats as including crescentic and oval forms, both motile. Subsequently the designation Selenomonas ruminantium (Certes) Wenyon (12) has been accepted for the large crescentic forms in the rumen (14). The oval form has not been named. Both large forms also occur in sheep. Crescentic bacteria cultured from the bovine and ovine rumen and classified as Selenomonas ruminantium (1, 5) are usually smaller than the large selenomonads in sheep. They are serologically related to some of the large selenomonads of sheep but not to others (6). Attempts to obtain pure cultures of large selenomonads by picking single cells have been unsuccessful (16). In the present work, pure cultures of large selenomonads were obtained by classical dilution methods after eliminating most of the small cells by centrifugation. MATERIALS AND METHODS Rumen fluid was removed from one of two tistulated sheep fed a pelleted alfalfa hay ration at maintenance level. Equal portions of the daily ration were fed at 2-hr intervals by means of a mechanical feeder. The rumen fluid (5 ml) was placed in a small culture tube, flushed out with CO2, and stoppered. The tube was centrifuged for 2 min at 150 x g to sediment feed particles and protozoa. The supernatant fluid, drawn off with a pipette connected to a mouth tube, was placed in a fresh tube under CO2. This was stoppered and centrifuged at 500 x g for 5 min. The creamy-white sediment was taken up in a l-ml syringe and injected through the recessed stopper (9) into a tube containing 6.5 ml of anaerobic sterile wash medium (Table 1). This tube was centrifuged for 5 min at 500 x g, the supernatant fluid was discarded, and the sediment was taken up and again diluted into a fresh tube of wash medium. A suspension containing 90% large cells was obtained by repeating the washing procedure 8 to 10 times. Reducing the centrifugal speed or the centrifugation time slightly improved the quality of the suspension, but it was never possible to remove all small cells. During all steps in the preparation of the suspension of large cells, extreme anaerobic precautions were taken. The sediment was finally suspended in 6.5 ml of fresh wash medium, and the number of cells per milliliter was counted with a Petroff-Hauser counting chamber under a phase-contrast microscope at a magnification of l,000x. Based upon the number of large cells in the I Present address: Laboratory of Veterinary Biochemistry, suspension, the inoculum was suitably diluted through a Utrecht, The Netherlands. series of culture tubes (12 x 70 mm) containing 3.4 ml 820

VOL. 105, 1971 SELENOMONADS FROM SHEEP RUMEN 821 TABLE 1. Composition of media used in the isolation and cultivation of large selenomonads from the sheep rumen of rumen fluid-glucose agar (Table 1). The final agar (Oxoid lonagar no. 2) concentration was 0.53%. Agar shake cultures (not roll tubes) were prepared, and the tubes were incubated in an upright position at 39 C. Growth was followed by examining the tubes daily under a dissecting microscope for colonies. All medium ingredients were heat-sterilized except sodium bicarbonate, glucose, and rumen fluid, which were sterilized by filtration. To obtain filtered rumen fluid, fresh rumen fluid was bubbled with oxygen-free H2 for about 15 min, and the fluid was dispensed anaerobically into 50-ml cellulose acetate centrifuge tubes which were capped with tight-fitting caps. The tubes were centrifuged for 30 min at 20,000 x g in a cold room. The top half of the supernatant fraction was carefully removed with a pipette connected to mouth tubing and transferred to an Erlenmeyer flask under H2. The flask was incubated at 45 C with occasional shaking for about I hr, during which time a precipitate formed. The precipitate was removed by centrifugation under anaerobic conditions for 30 min at 20,000 x g, and the supernatant fluid was drawn off and saturated with oxygen-free CO2. This liquid was then passed through a filter (Gelman GA-8, 0.2 tim), by using suction. The sterile filtration funnel and the suction flask were both completely filled with oxygen-free CO2 before the fluid was added, and air was excluded from the funnel with a gentle stream of CO2. The sterile rumen fluid, dispensed anaerobically in sterile culture tubes (16 by 150 mm) fitted with butyl rubber stoppers, was stored in the refrigerator until used. Stringent anaerobic techniques (9) were used throughout. The stock agar medium containing one part of solution A, one part of solution B (15), and two parts of deionized water, was autoclaved separately in small anaerobic culture tubes; other ingredients were added later by injection through the rubber stopper. Culture purity of the isolated strains was assured by subculturing through agar dilution series at least three successive times. At each transfer colonies were picked from the highest dilution showing growth. If all the colonies and the bacteria were identical in morphology, the culture was considered pure. Glucose was used as the energy source in all tests other than the tests on the fermentation of various carbon sources. The composition of the liquid medium used to determine fermentation products was the same as that of the agar medium (Table 1), except that agar was omitted and glucose was supplied at a final concentration of 0.5% (w/v). Tubes containing 10 ml of this liquid medium were inoculated with one drop of a vigorously growing culture in the same medium. Fermentation products from glucose were determined in the spent liquid cultures. Acid fermentation products were first analyzed by partition chromatography on cellulose columns using mixtures of acetone and n-hexane as eluents (10). By this method n-butyrate was not separated from its branched isomer or from higher volatile fatty acids. Samples were analyzed also for volatile fatty acids with a flame ionization gas chromatograph equipped with a column of Chromosorb W-DCMS (80 to 100 mesh) coated with neopentylglycolsuccinate and orthophosphoric acid in a ratio of 10: 1:0.1. Spent cultures were also analyzed for neutral fermentation products by thin- Agar Wash- Ingredients medium medium (ml) (ml) Rumen fluid, FSa... 1.00 2.00 NaHCOs, 10% (w/v), FS... 0.18 0.36 Cysteine * HCI, 3% (w/v)... 0.035 0.070 Na2S *9H2O, 3% (w/v)... 0.035 Casamino Acids, 2% (w/v)... 0.050 Yeast extract, 2% (w/v)... 0.050 Glucose, 10% (w/v), FS... 0.030 Carbon dioxide atmosphere... 100% 100% Mineral solution A (15)... 0.50 1.00 Mineral solution B... 0.50 1.00 Deionized water... 1.00 2.00 Agar (mg)... 18 afs, sterilized by filtration through a membrane filter (Millipore; porosity, 0.20 Arm); other ingredients were autoclaved. layer chromatography on Silica Gel G (3). Sum values of free carbon dioxide and residual bicarbonate in spent cultures and controls were determined by the syringe technique (Hungate, personal communication). One milliliter of 2.0 N H2SO4 was injected into a tube, and the syringe was quickly withdrawn. An empty 20-ml glass syringe was then rapidly inserted through the stopper of this tube, and tube and syringe were shaken vigorously several times to equilibrate gas and liquid phases. The glass syringe was lubricated with sterile distilled water. The amount of excess gas was recorded and compared with the control values. Values obtained were corrected to standard temperature and pressure. Fermentation gases were separated on a Perkin Elmer Vapor Fractometer with a silica gel column and nitrogen as carrier gas. Glucose in the spent cultures was determined with the glucose oxidase method (Biochimica Test Combination; Boehringer, C.F. & Soehne, Germany). Growth was measured in liquid cultures as optical density at 600 nm on a Spectronic-20 colorimeter (Bausch & Lomb, Inc., Rochester, N.Y.). Growth at different temperatures was assessed by measuring optical density of liquid cultures that were placed for 40 hr in different incubators with temperatures ranging from 24 to 45 C with 3 C intervals. The medium used to detect H2S production was the agar medium (Table 1) plus 0.05% (w/v) ferric ammonium citrate and 1.5% Trypticase (w/v). This medium was inoculated and incubated for 6 days. A black color indicated that H2S was produced. Nitrate reduction was tested for in the liquid medium with 0.5% (w/v) glucose and 0.2% (w/v) KNO, added. Indole and Voges-Proskauer tests were made in the liquid medium with 1 % (w/v) Casitone added, and rumen fluid was decreased to 10% (v/v). RESULTS In successful experiments, the number of colonies of large selenomonads appearing in the

:'; *F; F...: $< XAv Ne {.:... : t. j 822 PRI NS J. BACTERIOL. higher dilutions agreed with the number expected a maximum diameter of 6 mm. Growth in agar on the basis of the dilution and the cell count in or gelatin stabs was filiform. Strain L22 differed the washed inoculum. Direct counts ranged between 1.1 and 3.6 x 107 of large selenomonads/ media, colonies usually not appearing before the from the other strains by slower growth in agar ml and culture counts averaged not less than 80% third day of incubation. of the direct counts. Strains A22, E22 and L22 The cells measured 2 to 3,um in width and 5 to were isolated from sheep no. 1, and strains NSI 10 um in length. In liquid culture, cells were and NS2 were isolated from sheep no. 2. Strain slightly thicker than in agar, were more uniform SS was isolated from the rumen of a female blacktailed deer (Odocoileus hemionus columbianus) Cells were curved, crescentic, gram-negative, in size, and helicoidal filaments were never seen. shot at the Hopland Field Station, Hopland, Calif. rapidly motile with bluntly tapered or rounded Morphology. The deep colonies in agar were ends, occurring mostly as singles (Fig. 1). The creamy white, lenticular, and sometimes umbonate at both sides. They were visible after 1 to to the concave side of the cells in one or what organisms moved by means of flagella, attached 2 days of incubation as glassy, round, translucent seemed sometimes two tufts (dividing organisms). colonies with the appearance of tiny droplets of In some electron micrographs they were distributed over the concave side of the cell. water. After 2 to 3 days, colonies changed from translucent to opaque and grew in 5 or 6 days to Physiological characteristics. All six strains L. -wg 4L E._ *...:..... _r.d_ 4& V. ;." NI.. WE. o.:....-... * *b-. 4 ^: j,, FIG. 1. Electron micrograph of the large selenomonad A 22 isolated from the sheep rumen (x 10,000). Fixed with osmic acid. 4_

VOL. l05, 1971 SELENOMONADS FROM SHEEP RUMEN were strict anaerobes; no growth occurred in media in which the resazurin was oxidized. Optimal growth was obtained at 39 C. No growth occurred at or below 33 C, at 42 C or at 45 C in 40 hr. Growth in a liquid medium fortified with fermentable sugars was heavily turbid with slight sediment formation. As judged from optical density, only slight growth occurred in the liquid medium with bicarbonate, cysteine, sodium sulfide, Casamino Acids, and yeast extract omitted. Bicarbonate stimulated growth when added to such a medium. This effect would be mainly a ph effect. Cysteine and sodium sulfide were inhibitory when added to the complete medium in concentrations higher than the 0.03% normally employed. The medium was well reduced even without their addition. Yeast extract stimulated growth appreciably but hydrolyzed casein or other organic nitrogen sources had no effect on growth when added to the liquid medium containing all other ingredients (Table 1). Ovine rumen fluid could be replaced by bovine rumen fluid in the medium, and the cells retained their large size even after repeated transfers with the bovine medium. Filtered rumen fluid could be replaced by autoclaved rumen fluid or by the steam distillate of the same amount of acidified rumen fluid. Acidified rumen fluid was steam distilled after removal of cells by using the procedure of Fenner and Elliot (4). The steam distillate was neutralized with 0.1 N NaOH and concentrated with a rotatory evaporator. The concentrated solution thus obtained was filter sterilized and added to the liquid medium without rumen fluid to replace the rumen fluid. No growth occurred in the liquid medium containing all other ingredients (Table 1) but no rumen fluid or volatile fatty acids. The required acids were not identified but a considerable uptake of n-valeric acid from the rumen fluid-glucose-broth was evident from a comparison of gas chromatographic analyses of the spent cultures and the inoculated but unincubated medium. Isobutyrate and iso-valerate concentrations remained constant during incubation. None of the strains produced indole or reduced nitrate. Only strain SS produced H2S from cysteine, the other strains did not. All strains fermented fructose, glucose, mannose, cellobiose, maltose, sucrose, and salicin. None of the strains grew on arabinose, xylose, fucose, galactose, rhamnose, melibiose, lactose, melezitose, trehalose, glycerol, arabitol, dulcitol, sorbitol, inositol, or pinitol (5-O-methyl-D-inositol). Pectin and cellulose were not hydrolyzed by any of the strains, and starch was only hydrolyzed by strain SS. Lactate was not fermented, when it was the only carbon source nor with small amounts of glucose. In their sugar fermentation characteristics, strains A22, E22, L22, NSI and NS2 were alike but differed from strain SS in some respects (Table 2). It took 40 to 50 hr to reach maximal optical density values on 0.5% concentrations of the sugars that were fermented. Strain L22 grew more slowly on agar, but was not noticeably slower in liquid cultures. Fermentation products. The fermentation products of the strains grown in a 0.5% glucose-30% rumen fluid broth are shown in Table 3. Control values from inoculated tubes with the same medium minus glucose were subtracted. The pattern of products is much the same for the different strains. All of the glucose was fermented; none could be detected in the spent cultures. Final ph values ranged between 4.3 and 4.5. The principal acids formed were lactate, acetate, propionate, and formate, with small amounts of succinate and butyrate produced in most cultures. Strain SS did not form butyrate. Insignificant amounts of carbon dioxide (0.01 mmole/mmole of glucose fermented) were fixed. Carbon recoveries from glucose corrected for control values and the oxidation-reduction index values, indicating the sum of oxidized products divided by the sum of reduced products, are also given in Table 3. The Voges-Proskauer test was negative on all strains and neutral fermentation products were not formed. Products formed by strain L22 on different energy sources at the 0.5% level are shown in Table 4. With mannitol and salicin production of propionate and acetate was favored at the expense of lactate. To test the strains for iodophilly, cells were grown on 0.2% glucose for 24 hr in the liquid TABLE 2. Sugar fermentation characteristics oflarge selenomonadsa Strain Substrate A22, E22, L22, NS I, SS and NS2 Raffinose... _ + Dextrin... _ + Starch... _ + Inulin... _ + Xylan... _ + Mannitol... + 823 a All strains grew on fructose, glucose, mannose, sucrose, maltose, cellobiose, and salicin. None of the strains grew on arabinose, xylose, galactose, rhamnose, melibiose, lactose, trehalose, melezitose, glycerol, inositol, pinitol, arabitol, dulcitol, sorbitol, fucose, sorbose, or lactate. Pectin and cellulose were not hydrolyzed.

824 PRI NS J. BACTERIOL. TABLE 3. Fermentation products ofstrains of large selenomonads isolated from the sheep rumena Products Strain A22 E22 4122 NSI SS Formic acid... 0.57 0.62 0.49 0.42 0.50 Acetic acid... 1.56 1.12 1.20 2.00 1.41 Propionic acid... 0.51 0.35 0.70 0.40 0.75 Butyric acid... 0.05 0.09 0.05 0.10 0 Lactic acid... 2.86 4.01 2.60 2.50 3.01 Succinic acid... 0.11 0.02 0.30 0.21 0.20 Carbon recovery (%) 86.7 98.3 88.8 86.2 92.4 Oxidation-reduction index. 1.11 1.21 0.98 1.05 0.93 a Values expressed as millimoles per 100 ml of rumen fluid-glucose medium. medium after which they were harvested by centrifugation and transferred to fresh medium without an energy source. After incubation in this medium for another 5 hr to deplete the cells, fermentable sugars were added at the 0.1 % level, and tests were done after 10 min of incubation. All strains were strongly iodophilic, and the addition of fermentable sugars resulted in a rapid deposition of reserve polysaccharide. This storage polysaccharide gave a violet-brown color when stained with Lugol's solution. Selenomonas ruminantium var. bryanti var. n. The name Selenomonas ruminantium var. bryanti is proposed for the strains A22, E22, L22, NSI and NS2. It is a gram-negative, motile, crescentic rod with rounded ends and is 2 to 3 um wide by 5 to 10 gm long. In liquid culture cells are slightly thicker than on agar and more uniform in size. The organism occurs as singles and is motile by means of flagella attached to the concave side of the cell. The cells contain reserve polysaccharides that stain violet-brown with Lugol's iodine solution. Deep colonies are lenticular, creamy white, and reach a diameter of 6 mm. Growth in agar stabs TABLE 4. is filiform. Growth in liquid medium is heavily turbid with slight sediment formation. Final ph in glucose medium is 4.3. Good growth occurs at 36 or 39 C but no growth is obtained at 33, 42, or 45 C after 40 hr of incubation. The organism is a strict anaerobe and will not grow in media in which resazurin is oxidized. No growth occurs in a medium without rumen fluid. Indole or H2S are not produced. Nitrate is not reduced. Starch is not hydrolyzed. The Voges-Proskauer test is negative. The organism ferments fructose, glucose, mannose, cellobiose, maltose, sucrose, salicin, and mannitol. Arabinose, xylose, fucose, galactose, rhamnose, melibiose, trehalose, raffinose, dextrin, starch, inulin, xylan, glycerol, arabitol, dulcitol, sorbitol, inositol, pinitol, lactate, pectin, and cellulose are not fermented. It does not form gas splits in glucose agar shake cultures. Fermentation products from glucose include lactic, acetic, propionic, formic, butyric and succinic acids. Carbon dioxide or hydrogen are not produced. Gelatin is not liquefied. The source is the reticulo-rumen of sheep. DISCUSSION The cultured organisms were morphologically similar to the large crescentic cells seen during microscopic examination of rumen contents of sheep and deer. No diminuition in size nor change in morphology was observed when the cells were subcultured for over 7 months. This is in contrast to the results of Purdom (16), who stated that all large curved gram-negative rods which grew in pure culture after isolation with a micromanipulator were smaller than the original selenomonads. Strains of S. ruminantium cul- Fermentation products by a large strain (L22) ofselenomonas isolated from the sheep rumena Product Glucose Sucrose Cellobiose Maltose Mannitol Salicin Formic acid... 0.49 0.65 0.51 0.19 0.30 0.72 Acetic acid... 1.20 2.43 1.96 1.89 3.12 2.28 Propionic acid... 0.70 0.82 0.69 0.86 1.54 1.05 Butyric acid... 0.05 0.05 0.14 0.09 0.23 0.06 Lactic acid... 2.60 2.87 1.92 2.17 0.54 0.22 Succinic acid... 0.30 0.22 0.40 0.87 0.44 0.30 Carbon recovery (%)...... 88.8 100.7 82.2 96.3 93.8 100.3 Oxidation-reduction index...... 0.98 0.94 0.94 1.02 0.87 a Values expressed as millimoles per 100 ml of rumen fluid-sugar media.

VOL. 105, 1971 SELENOMONADS FR( :)M SHEEP RUMEN 825 tured earlier (1, 5, 7) were all smaller in size than the present isolates. Growth was relatively slow. When the inoculum was not sufficiently washed, the large forms could not be isolated because they were rapidly overgrown by small cells, including small forms that resembled S. ruminantium. Some of these small selenomonads were isolated and grown in the same medium as was used for the large isolates. The composition of the medium had no effect on size or shape of the small cells, and they did not become large under these culture conditions. Strains of S. ruminantium have been shown to be stimulated or to require one or more of the following acids: isobutyric, n-valeric, isovaleric, and DL-2-methyl-butyric (2). Small forms of lactate-fermenting S. ruminantium var. lactilytica from the sheep rumen required straight-chain saturated fatty acids with a C5 to C10 carbon skeleton, n-valerate being most effective in low concentrations (1 1). A small bovine strain also grew with n-valeric acid as the only volatile acid added (17). Thus, it is of interest that the large forms required volatile acid and that n-valerate was decreased in the medium during growth. The abnormal morphology observed with n- valerate-deficient media (l1) could explain our observation of long helicoidal filaments of cells in older agar cultures, since it may be assumed that n-valerate was rapidly exhausted at the site of the colony. Hobson et al. (6) studied the relationship between rumen selenomonads in vitro and in vivo with the fluorescent-antibody technique. Some of the large forms were serologically similar to the small form but others were not. The results obtained in the present study suggest that there are at least two types of large selenomonads. One type (strain SS) shows the characteristics of S. ruminantium and differs from the described strains only by its larger size. Another type (strains A22, E22, L22, NSl, and NS2) differs from S. ruminantium by not producing H2S from cysteine and in fermentation of arabinose, xylose, galactose, lactose, and dulcitol. It differs from S. sputigena by fermenting cellobiose and salicin and by its size. None of the large isolates grew on glycerol or lactate so they cannot be classified as S. ruminantium var. lactilytica. They are considered to be a hitherto undescribed variety for which the name Selenomonas ruminantium var. bryanti is proposed. ACKNOWLEDGMENTS I am indebted to R. E. Hungate for his guidance and for his help in preparing the manuscript. Leave of absence from the University of Utrecht and the award of Public Health Service postdoctoral Fellowship I F05 TW 1444 from the Fogarty International Center are gratefully acknowledged. LITERATURE CITED 1. Bryant, M. P. 1956. The characteristics of strains of Selenomonas isolated from bovine rumen contents. J. Bacteriol. 72:162-167. 2. Bryant, M. P., and 1. M. Robinson. 1962. Some nutritional characteristics of predominant culturable ruminal bacteria. J. Bacteriol. 84:605-614. 3. Drucker, D. B., and T. H. Melville. 1966. A high resolution system for chromatographic analysis of end products of bacterial fermentations. J. AppI. Bacteriol. 29:536-539. 4. Fenner, H., and J. M. Elliot. 1963. Quantitative method for determining the steam volatile fatty acids in rumen fluid by gas chromatography. J. Animal Sci. 22:624-627. 5. Hobson, P. N., and S. 0. Mann. 1961. The isolation of glycerol-fermenting and lipolytic bacteria from the rumen of the sheep. J. Gen. Microbiol. 25:227-240. 6. Hobson, P. N., S. 0. Mann, and W. Smith. 1962. Serological tests of a relationship between rumen selenomonads in vitro and in vivo. J. Gen. Microbiol. 29:265-270. 7. Hobson, P. N., S. 0. Mann, and W. Smith. 1963. Growth factors for Selenomonas ruminantium. Nature (London) 198:213. 8. Hungate, R. E. 1966. The rumen and its microbes. Academic Press Inc., New York. 9. Hungate, R. E. 1969. A roll tube method for cultivation of strict anaerobes, p. 117-132. In J. R. Norris and D. W. Ribbons, (ed.) Methods in microbiology, vol. 3B. Academic Press Inc., New York. 10. Hungate, R. E., W. Smith, T. Bauchop, 1. Yu, and J. C. Rabinowitz. 1970. Formate as an intermediate in the bovine rumen fermentation. J. Bacteriol. 102:389-398. 11. Kanegasaki, S., and H. Takahashi. 1967. Function of growth factors for rumen microorganisms. 1. Nutritional characteristics of Selenomonas ruminantium. J. Bacteriol. 93:456-463. 12. Lessel, E. F., Jr., and R. S. Breed. 1954. Selenomonas Boskamp, 1922-a genus that includes species showing an unusual type of flagellation. Bacteriol. Rev. 18:165-167. 13. Moir, R. J., and M. J. Masson. 1952. An illustrated scheme for the microscopic identification of the rumen micro-organisms of sheep. J. Pathol. Bacteriol. 64:343-350. 14. Oxford, A. E. 1955. Some observations upon the status of the generic name Selenomonas Prowazek. Int. Bull. Bacteriol. Nomencl. Taxon. 5:131-132. 15. Paynter, M. J. B., and R. E. Hungate. 1968. Characteristics of Methanobacterium mobilis, sp. n., isolated from the bovine rumen. J. Bacteriol. 95:1943-1951. 16. Purdom, M. R. 1963. Micromanipulation in the examination of rumen bacteria. Nature (London) 198:307-308. 17. Tiwari, A. D., M. P. Bryant, and R. S. Wolfe. 1969. Simple method for isolation of Selenomonas ruminantium and some nutritional characteristics of the species. J. Dairy Sci. 52:2054-2056. 18. Woodcock, H. M., and G. Lapage. 1914. On a remarkable type of protistan parasite. Quart. J. Microscop. Sci. 59: 43 1-458.