A Study of the Microorganisms from Grass Silage I. The Cocci

C. WV. LANGSTON AND CECELIA BOUMA Dairy Cattle Research Branch, Animal Husbandry Research Division, Agriculture Research Service, 1griculture Research Center, Beltsville, Marlyland Received for ptiblication November 16, 1959

In a natural silage , the mass is acidified their taxonomical relationship. The data presented are by lactic and acetic acid forming that ferment the results of detailed colonial, morphological, and sugars in the plant material. When forage is ensiled the physiological studies on the cocci important in the plant cells continue to respire for a time, using up the silage fermentation. oxygen and giving off CO2 and heat. As conditions be- come favorable, acid producing bacteria increase MIATERI.ALS ANDV IETHODS rapidly and, at the end of 3 or 4 days, each gram of Preparationi of silages and techniques used in iso- silage will contain several hundred million bacteria. lating and grouping the bacteria from the These organisms produce acid until the sugar is ex- silages have been outlined in an earlier publication hausted or until the pH becomes unfavorable for further (Langston et al., 1958). growth. This phase of w-ork includes detailed studies on repre- A recent study (Langston et al., 1958) on the micro- sentative strains of obtained from organisms in orchard grass and alfalfa silages showed 30 silages. The strains were picked from highest dilution that the total numbers of acid producing bacteria had roll tubes (Trypticase)l and plates (Rogosa et al., 1951). little bearing on the final quality. The increase in num- Preliminary grouping was based on the following bers of these organisms showed similar trends in all tests: Gram stains, production, growth at silages studied. The counts reached about the same 45 C, growth in 6.5 per cent of sodium chloride, hy- maxima in both good and poor quality silages, and the drolysis of arginine, reduction of nitrate, production of few silages that had relatively high initial counts of gas from glucose, and reaction in litmus . acid producing bacteria were no better in quality than About 440 strains from the 3142 isolates grouped as those with few or none. The difference in quality of the indicated above were chosen for further study. The silages was correlated directly with the appearance and strains wNere replated at least twice to insure purity and increase of sporeforming anaerobes. The poor quality all tests were stanidardized as far as possible. Unless silages had high numbers of these organisms. otherwise stated, the media were inoculated with one It is not clearly understood why some silages show a drop of culture (18 to 20 hr incubation) that had been rapid drop in pH and others yield only a limited amount transferred 2 or 3 times in tomato juice glucose broth. of acids. The carbohydrate source, however, is not Except for temperature studies all tests were incubated always the limiting factor. Various reasons have been at 30 C. put forth to account for the variability in acid produc- All of the strains studied were gram-positive and tion. They are: (1) variability in sequence changes of nonsporeforming. During the preliminary study, several microorganisms, (2) antagonism among certain groups strains were observed to produce catalase and reduce of bacteria early in the fermentation process, (3) defi- nitrate. Some of these strains were chosen for detailed cient nutrients in the plant material for bacterial study. This prosved to be a judicious choice because, as growth, and (4) occurrence of weakened strains of will be shown later, most of the strains that had these bacteria. variable characteristics w-ere similar in most respects Although the preservation of forage by natural fer- to the true lactic acid bacteria. Characteristics studied: mentation is an old and useful method of preserving 1. Colony formation. Colony formation (arrange- fodder, little work has been done to characterize the ments, size, and color) was observed with the aid of a types of organisms responsible for the acid productioni. wide-field microscope. Many of the early workers investigating the micro- 2. Gram stain (Burke modification), form, size, and organisms in silage spoke only generally of "lactic arrangement of cells. Smears were made from 18 to 24-hr ferments" or "acidifying bacteria." cultures grown in tomato juice glucose broth and semi- The object of this study was to learn more about the ' Baltimore Biological Laboratory, Inc., Baltimore, Mary- types and occurrence of microorganisms in silage and land. 212 19'60] 190]IICROORGANISM\S FRO'M GRASS SILAGE. I 21:3 solid deeps. Tomato juice glucose broth had the follow- was used to determine growth at 45 C (1 week of incu- ing composition and will be designated as medium A: bation) and growth at 15 C (2 weeks' of incubation) Trypticase, 10 g; phytone, .5 g; yeast extract, 5 g; glu- was determined in a Cenco3 refrigerating incubator. cose, 5 g; sodium chloride, a g; potassium phosphate 7. Growth in 6.5 per cent of sodium chloride. The or- (dibasic), 1.5 g; "Tween 80" (sorbitan monooleate),2 ganisms were tested for their ability to grow in sodium 0.5 ml; tomato juice, 200 ml; bromeresol purple, 0.016 g; chloride in medium A. The salt concentration was in- and distilled water to make 1 L. The medium was creased from 5 g to 65 g per L. The medium was dis- adjusted to pH 7 and dispensed inito test tubes in 9-ml pensed in 9-ml quantities into test tubes and incubated quantities. It was autoclaved at 15 lb pressure for 15 for 2 weeks. min. The semisolid medium was similar to medium A 8. Growth at pH 9.6. After the medium was auto- with the addition of 5 g agar and deletion of the brom- claved and dispensed in 9-ml quantities into test tubes, cresol purple. sterile sodium hydroxide wras added aseptically to pro- 3. Motility. Hanging drop slides and in some cases duce the desired pH. The tubes were incubated for 2 semisolid medium were used to detect motility. F'lagella weeks. stains were prepared in the followiing manner: Cells 9. Reduction of 0.1 per cent methylene blue. Skimmed were washed from the surface of slants with saline solu- milk' coiitaining 0.1 per cent methylene blue was inocu- tion (0.85 per cent) centrifuged twice, and made to a lated with test cultures and observed for changes from slightly cloudy suspension with distilled water. The light blue to colorless. The iuicubation period was for slides were prepared and stained by the method de- 1 week. scribed by Leifson (1951). Best results were obtained 10. Reaction in litmus milk. Litmus milk' was inocu- with cells 8 to 12 hr old. The slant medium was similar lated with test cultures anid incubated 1 month. Tubes to medium A with the deletion of bromeresol purple wvere examined and medium changes recorded at 24 hr, and the addition of 15 g of agar. 48 hr, 1 week, and 1 mouth. 4. Catalase production. Catalase production was de- 11. Final acidity itn skimmed milk. Dehydrated termined on broth cultures and on agar streak plates skimmed milk' was reconstituted, dispensed in 9-ml containinig low carbohydrate (0.05 per cent). The broth quanitities into test tubes, sterilized, and inoculated with and plating media were similar to medium A. To deter- test cultures. After 1 month of incubation, the contents mine catalase from cultures growing in broth, a few of the tubes were washed into small beakers. Clots wi-hen drops of the broth were transferred to a spot plate and formed were broken up by magnets kept in motion by 3 per cent hydrogen peroxide was added. The evolution a 1Mag-mix.4 The pH was recorded and titration values of gas constituted a positive test. Precautions should obtained by titrating electrometrically to the phenol- be taken in reading tests, especially when the cultures phthalein end point with N/10 NaOH. Acid values from are weakly catalase positive. Flaming of the pipette the controls were substracted from the cultures to ob- before transferring the culture to the spot plate should tain correct acidity alues (acidity values are expressed be avoided and the test should be observed for at least as per cent lactic acid). A Beckman' pH meter was used. 5 min before discarding. Streak plates were flooded with 12. Final acidity in glucose broth. Procedures for ob- 3 per cent hydrogen peroxide and observed for gas taining final acidity values in glucose broth were essen- evolution. Strains that produced catalase always tially the same as those described for skimmed milk. showed greater activity on plates than in broth. The medium had the following composition: Trypti- 5. Production of gas. To detect gas, medium A was case, 10 g; yeast extract, a g; NaCl, 5 g; glucose, 20 g; used with the following modifications: Bromeresol and distilled water to make 1 L. The medium was ad- purple was deleted, glucose increased to 20 g, and 20 g justed to pH 7 and dispensed into test tubes (18 by of agar added. The medium was dispensed into test 150 mm) in 15-ml quantities. The tubes were inocu- tubes in 7-ml quantities. An agar layer (2 per cent) and lated with test cultures aiid incubated for 2 weeks. oil seal were used to prevent loss of gas. Heavy inocula- 13. Production of a.mmonia from arginine. The tions were made and the tubes wvere incubated for 2 method and medium described by Niven et al. (1944) weeks. It was not uncommon for some cultures to push was used to determine ammonia production from ar- the agar and oil layers to the top of the tubes. ginine. After 3 days of incubation, ammonia was de- 6. Growth at 15 C and 45 C. Medium A was used with tected by placing 3 or 4 drops of culture into a spot plate the following modifications: Five grams of agar were and adding 1 drop of Nessler's reagent. added; the tomato juice was filtered, and bromeresol 14. Production of acetylmethylcarbinol. Test cultures purple deleted. The medium was dispensed in 9-ml were inoculated into the following medium: polypep- quantities into test tubes. After sterilization and inocu- tone,' 7 g; glucose, 5 g; potassium phosphate (dibasic), lation (inoculating loop), the tubes were sealed with 3 Central Scientific Company, Chicago, Illinois. rubber stoppers to avoid evaporation. A wvater bath 4Precision Scientific Company, Chicago, Illinois. 2 Atlas Powder Company, Wrilminigton, Delaware. 5 Beckman Instruments, Inc., Ftillerton, California. 214 C. W. LANGSTON AND C. BOUMNIA [VOL. 8

5 g; and distilled water to make 1 L. The pH of the to pH 1.0 with concentrated sulfuric acid and extracted medium was adjusted to 7.0. After inoculation the tubes with anihydrous ether for 36 hr in a water bath at 60 C were incubated 48 hr. To detect acetylmethylcarbinol, (Langston, 1955). Aliquots of the extract were titrated 1 ml of culture was placed in a test tube and 0.6 ml of a to determine total acids produced and a fraction was 5 per cent a-naphthol solution in 100 ml of alcohol and removed for chromatographic analysis of the individual 0.2 ml of 40 per cent potassium hydroxide wtere added. acids. Butyric, propionic, acetic, formic, succinic, and The development of a red color in the mixture from 2 lactic acids were determined by a method similar to the to 4 hr constituted a positive test. No tubes were read one described by Neish (1950). later than 4 hr after the addition of the reagents. The remaining acid solutioni was brought to about 15. Gelatin liquefaction. Cultures were inoculated into half volume on the steam bath, boiled for a few minutes, tubes of gelatin medium that had the following composi- and neutralized with zinc carbonate (Brin et al., 1952). tion: Beef extract, 3 g; peptone, a g; gelatin, 120 g; and The zinc lactate crystals were usually collected in three distilled water, 1 L. The pH of the medium was adjusted fractions and characterized by the water of crystalliza- to 7.0. After 48 hr incubation the tubes were chilled to tion and optical rotationi. determine if the gelatin had been liquefied. 16. Production of mnucoid colonies on 10 per cent RESULTS sucrose plates. The medium described by _McCleskey et The genera Streptococcuts and Letconostoc are of prime al. (1947) was used to determine dextran production. importance in many phases of agriculture and industry 17. Nitrate reduction. N\itrate reduction Awas deter- from a standpoint of desirable or undesirable fermenta- mined after incubation for 3 days in indole-nitrate tions. The genus Streptococcuts includes the homofer- medium.' Tests were made for nitrite and residual mentative cocci which produce primarily dextro-lactic nitrate. acid as a by-product of glucose fermentation. Sherman 18. Indole test. The test for indole was carried out in (1937) separated the streptococci into four groups on indole-niitrate medium' after 24 and 48 hr incubation. the basis of certain physiological characteristics. In Kovacs reagent, was used to determine indole produc- contrast, the genus includes the hetero- tion. fermentative cocci which usually produce a limited 19. Carbohydrate . Fermentation reac- amount of levo-lactic acid, C02, acetic acid, and alcohol tions were carried out in the following basal medium: from glucose. Three species have been recognized in the Trypticase, 10 g; yeast extract, 5 g; NaCl, 5 g; and dis- genus Leuconostoc (Hucker and Pederson, 1930). tilled water to make 1 L. Substrates wiere added at the The lactic acid producing cocci from the forages were 1 per cent level. Rhamnose, glucose, lactose, sucrose, identified and divided inlto five more or less distinct trehalose, raffinose, melezitose, starch, dextrin, inulin, groups. Included were strains of (a) Streptococcus fae- salicin, esculin, sodium hippurate, glycerol, mannitol, calis, (b) Streptococcus liquefaciens (Streptococcusfaecalis inositol, sorbitol, and sodium lactate were sterilized with var. liquefaciens), (c) Leuconostoc mesenteroides, (d) the base. Arabinose, xylose, ribose, fructose, galactose, variable Leuconostoc (Leuconostoc, type I), and (e) mannose, sorbose, maltose, cellobiose, and melibiose . All cultures had the following charac- were sterilized separately (Seitz filtration) and added to teristics in common: colonies were usually small anid the autoclaved base. The material was then tubed subsurface and the commonly observed forms were aseptically into test tubes in 7-ml quantities. The tubes lenticular or circular. The margin of the colonies were were incubated for 2 weeks. Fermentation reactions usually entire although some were slightly wavy or un- were detected through the use of a Beckman pH meter. dulate. Surface colonies were uniform, smooth, and Hydrolysis of sodium hippurate was detected by round. Mlost of the colonies observed had a whitish to adding 0.5 ml of 50 per cent sulfuric acid to 2 ml of grayish appearance. All strains were facultative anaer- culture. Appearance of benzoic acid crystals constituted obes, gram-positive, and nonsporeforming. Growth on a positive test. Hydrolysis of esculin was determined by the surface of agar slants was thin to moderate. None adding 0.5 ml of ferric citrate to 7 ml of culture. The of the strains reduced nitrate to nitrite or produced typical blackening reaction of the medium constituted indole. (a) S. faecalis and (b) S. liquefaciens were usually a positive test. oval shaped and occurred singly, in pairs, and some- 20. Fermentation products. Strains of lactic acid bac- times short chains (4 to 6 cells). They averaged about teria were grown in medium similar to that described 0.7 ,1 by 0.8 to 1.1 ,. None of the strains produced gas by Harrison and Hansen (1950b). However, the glucose from glucose or dextran on 10 per cent sucrose agar and was sterilized separately and added to the autoclaved no catalase was observed. All of the cultures produced base and tomato juice was deleted except in cases where ammonia from arginine, grew in 6.5 per cent of sodium acid production was very low. Initial and residual glu- chloride and at pH 9.6, and reduced 0.1 per cent methyl- cose was determined by the anthrone method described ene blue in milk. by Morris (1948). Aliquots of the medium were acidified (a) Streptococcus faecalis. The strains of S. faecalis 19601 MICROORGANISMS FROM GRASS SILAGE. I 215

were further characterized by their ability to completely fied, curdled, and peptonized litmus milk. The litmus reduce litmus milk prior to acidification and curdling, was completely reduced prior to curdling. All strains and most of the strains produced acetylmethylcarbinol. liquefied gelatin, produced acetylmethylcarbinol, and They grew at 15 C and eight of the 13 strains showed grew at 15 C and 45 C. growth or slight growth at 45 C. None liquefied gelatin. Final pH values in glucose broth ranged from 4.2 to The final pH in glucose broth ranged from 3.8 to 4.3 4.3 and titratable acidity values from 0.47 to 0.61 (avg 4.2), and in skimmed milk, 4.4 to 5.8 (avg 4.8). (avg 0.55). Carbohydrate fermentations (table 1) were Titratable acidity values in glucose broth and skimmed similar to those of S. faecalis with the exception that S. milk ranged from 0.40 to 0.86 (avg 0.55) and 0.13 to liquefaciens failed to ferment arabinose but did ferment 0.66 (avg 0.46), respectively. melezitose and sodium hippurate. All of the strains (table 1) fermented ribose, glucose, Streptococcus liquefaciens gave a carbon distribution fructose, mannose, galactose, maltose, lactose, treha- (table 2) similar to that of S. faecalis, however, two of lose, cellobiose, dextrin, salicin, esculin, glycerol, and the strains studied produced inactive lactic acid. Al- mannitol; most fermented arabinose, xylose, rhamnose, though organisms in this group usually produce dextro- sucrose, starch, and sorbitol. They varied with respect lactic acid, it is not unusual for strains to produce a to melibiose and raffinose and none fermented sorbose, mixture of the dextro- and levo- types (Long and melezitose, inulin, sodium hippurate, inositol, or Hammer, 1936). The two strains that gave inactive sodium lactate. lactic acid were replated and picked to insure purity Dextro lactic acid (table 2) was the main by-product and the by-products redetermined. This procedure was of glucose fermentation produced by three strains of repeated twice and in each trial inactive lactic acid was S. faecalis. Recoveries showed that 91.2 to 96.0 per cent obtained. of the glucose fermented was converted to lactic acid. The strains of S. faecalis and S. liquefaciens described Traces of other products were also formed. These in- here are generally in close agreement with earlier de- cluded propionic, acetic, formic, and succinic acids. scriptions presented by Sherman et al. (1937), Smith and (b) Streptococcus liquefaciens. S. liquefaciens acidi- Sherman (1942), Harrison and Hansen (1950a) and as

TABLE 1 Final pH of broth cultures after 14 days of incubation at 30 C Streptococcus faccalis Streptococcus liquefaciens

Substrate pH pH Basal Medium Culture reaction Culture reaction Range Avg Range Avg

Arabinose ...... 1-; 12+ 4.3-5.8 4.6 7- 6.4 Xylose ...... 3-; 10+ 4.3-5.1 4.7 6-; 1+ 4.3 6.4 Ribose ...... 13+ 4.3-4.5 4.4 7+ 4.3-4.5 4.4 5.4 Rhamnose ...... 2-; 11+ 4.6-5.5 5.1 1-; 6+ 5.1-5.6 5.4 6.7 Glucose ...... 13+ 4.0-4.3 4.1 7+ 4.0-4.2 4.1 6.7 Fructose ...... 13+ 4.0-4.4 4.2 7+ 4.0-4.2 4.1 6.6 Mannose ...... 13+ 3.9-4.5 4.2 7+ 4.0-4.1 4.1 6.8 Galactose ...... 13+ 3.8-4.6 4.4 7+ 4.1-4.7 4.5 6.7 Maltose ...... 13+ 4.1-4.6 4.4 7+ 4.2-4.3 4.3 6.8 Lactose ...... 13+ 4.4-5.0 4.6 7+ 4.8-4.9 4.9 6.9 Sucrose ...... 3-; 11+ 4.1-4.5 4.4 7+ 4.4-4.7 4.6 7.0 Trehalose ...... 13+ 3.9-5.0 4.5 7+ 4.3-4.9 4.8 7.0 Cellobiose ...... 13+ 4.1-4.5 4.3 7+ 4.3-4.5 4.4 6.9 Melibiose ...... 5-; 8+ 4.4-5.2 4.0 6-;1+ 4.3 6.9 Raffinose ...... 6-; 7+ 4.3-5.5 5.2 6-; 1+ 4.6 7.0 Melezitose ...... 13- 7+ 4.2-4.8 4.6 7.0 Starch ...... 1-; 12+ 4.6-5.6 5.0 7+ 5.0-5.7 5.4 7.1 Dextrin ...... 13+ 4.3-5.6 5.0 7+ 5.0-5.2 5.1 7.0 Salicin ...... 13+ 4.2-4.6 4.5 7+ 4.3-4.5 4.4 7.0 Esculin ...... 13+ 7+ Sodium hippurate ...... 13- 7+ Glycerol ...... 13+ 5.4-5.8 5.6 7+ 4.8-4.9 4.8 7.0 Mannitol ...... 13+ 4.4-5.0 4.7 7+ 4.6-4.8 4.7 7.0 Inositol .... 13- 3-; 4+ 4.3-5.7 5.0 7.0 Sorbitol ..... 4-; 9+ 5.0-5.2 5.1 7+ 4.5-5.0 4.6 7.0 None of the strains fermented sorbose, inulin, or sodium lactate. 216; C. W. LANGSTON AND C. BOUA[IA [VOL. 8 given in Bergey's Manual of Determinative Bacteriology 6.5 per cent of sodium chloride. Only slight acid reac- (Breed, Murray, and Smith, 1957). tions were observed in litmus milk. Final pH and titrat- The heterofermentative cocei studied were divided able acidity values in glucose broth ranged from 4.1 to into 2 groups mainly on the basis of dextran production. 4.4 (avg 0.43) and 0.41 to 0.63 (avg 0.53), respectively. Strains of L. mesenteroides produced this material None of the cultures produced ammonia from arginine, readily. A closely related group (Leuconostoc, type I) acetylmethylcarbinol, liquefied gelatin, or grew at 45 C. produced dextran only after serial transfers in enriched The unusual feature of this group was the ability of media. some strains to produce catalase. Five of the 10 strains (c) Leuconostoc mesenteroides. Cells of L. mesente- studied exhibited this property. Weak catalase reactions roides were usually oval to slightly elongated and oc- were first observed in tomato juice glucose broth. When curred singly, in pairs, and short chains. They averaged the strains were streaked on low carbohydrate medium about 0.7 to 0.8,u by 1.2 A. It was not uncommon to find (0.05 per cent glucose) they gave a relatively strong elongated cells up to 3 or 4 in length. The strains pro- reaction. Two strains that failed to produce catalase in duced gas from glucose, grew at 15 C and 45 C and in the tomato juice glucose broth gave a positive reaction on the low carbohydrate medium. Some of the cultures TABLE 2 exhibited a "delayed reaction" in production of 02 Per cent acids and optical type of lactic acid produiced following the addition of hydrogen peroxide. The test by Streptococcuis material should be observed for several minutes before Per Per Cent Fermented Glucose Converted to recording results. Visual observations indicated that the Optical Type Culture GlucoseCent of Lactic amount of gas evolution by the more active strains of No. Fr Buty- Pro: ctcFri Suc-Lci Acid i cinic Lactc was to mentedFer-uric psonic cd acid acid Produced Leuconostoc similar that produced by cultures of acdai acidc Pediococcus. Catalase production by organisms in this Streptococcus faecalis genus comes as no surprise since various workers have T-10 51.8 0 0.5 1.1 1 0.6 1.5 91.2 Dextro shown that other facultative lactic acid producing bac- T-16 47.8 0 0 2.0 0.4 0.5 96.0 Dextro T-843 45.5 0 0 0 0.3 1.5 91.8 Dextro teria (pediococci, lactobacilli) exhibit this property. Streptococcuts liquefaciens Tw^o colonial types were observed when strains of L. T-848 49.3 0 0 0.4 0.4 0 92.1 Dextro mesenteroides were streaked on 10 per cent sucrose agar. T-865 51.2 0 0 0 0.02 0. 5 92.2 Inactive The colonies resembled type B and D described by T-873 41.6 0 0.2 0.8 0.03 0 95.5 Inactive XvIcCleskey et al. (1947). Although, they noted varia-

TABLE 3 Final pH of broth culttures after 14 days of incuibation at 30 C Leuconostoc niesenteroides Leuconostoc Type I Basal Medium Substrate pH pH pH Culture reaction - Culture reaction Range Avg Range Avg

Arabinose ...... 3-; 7+ 4.3-4.9 4.5 10-; 33+ 4.0-4.6 4.4 6.4 Xylose ...... 10+ 4.4-5.1 4.8 21-;22+ 4.1-4.9 4.3 6.4

Ribose ...... 5-; 5+ 4.2-5.5 4.6 4-; 39+ 4.3-4.9 4.7 5.4 Glucose ...... l0+ 4.0-4.4 4.3 43+ 4.0-4.7 4.3 6.7 Fructose ...... 10+ 4.2-4.5 4.4 1-; 42+ 4.2-5.6 4.4 6.6 Mannose ...... 10+ 3.7-4.5 4.3 43+ 4.1-4.8 4.4 6.8 Galactose ...... 10+ 4.2-4.7 4.4 43+ 4.1-5.4 4.4 6.7 Maltose ...... 10+ 3.6-4.4 4.3 4-; 39+ 4.0-4.5 4.2 6.8 Lactose ...... 2-; 8+ 4.3-5.1 4.6 8-;35+ 4.1-5.5 4.6 6.9 Sucrose ...... 10+ 4.2-4.5 4.4 4-;39+ 4.1-4.7 4.4 7.0

Trehalose ...... 10+ 4.0-4.4 4.3 43+ 4.0-4.7 4.3 7.0 Cellobiose ...... 2-; 8+ 4.4-4.9 4.6 13-; 30+ 4.0-4.8 4.4 6.9 Melibiose 3-; 7+ 4.3-4.7 4.5 4-; 39+ 4.1-4.6 4.3 6.9 Raffinose 3-; 7+ 4.3-4.5 4.4 11-; 32+ 4.1-5.6 4.5 7.0 Starch .10- 37-; 6+ 5.2-5.9 5.4 7.1 Dextrin .10- 32-; 11+ 5.2-5.6 5.4 7.0 Salicin .10+ 4.3-4.7 4.5 20-;23+ 4.3-5.4 4.9 7.0 Esculin 0...... lO+ 19-; 24+ Mannitol ...... 3-; 7+ 4.8-5.2 5.0 14-; 29+ 4.6-5.2 5.0 7.0 None of the strains fermented rhamnose, sorbose, melezitose, inulin, sodium hippturate, glycerol, iI1ositol, sorbitol, or sodium lactate. 1 960] 9MICROORGANISNIS FROM GRASS SILAGE. I 217 tions in fermentation reactions that we did not observe. later, the cultures produced dextran after serial trans- Carbohydrate fermentations (table 3) were almost fers in enriched media. The cells occurred singly, in identical with those described in Bergey's Manual pairs, and occasionally short chains and clumps. Some (Breed et al., 1957) for L. mesenteroides. All strains cells tapered and longer ones had a tendency to bend. fermented xylose, glucose, fructose, mannose, galactose, Physiological reactions were similar to those of L. maltose, sucrose, trehalose, salicin, and esculin. They mesenteroides with the following exceptions: 12 of the 43 varied but usually fermented arabinose, ribose, lactose, strains showed slight growth at 45 C; some gave a cellobiose, melibiose, raffinose, and mannitol. None of reduced, acid, and curdled reaction in litmus milk; and the strains fermented rhamnose, sorbose, melezitose, a slightly wider range was noted for titratable acidity starch, dextrin, esculin, sodium hippurate, glycerol, values in glucose broth (0.23 to 0.77). inositol, sorbitol, or sodium lactate. Carbohydrate fermentation reactions (table 3) were Strains of L. mesenteroides converted less than 43 per usually similar to those produced by L. mesenteroides cent of the glucose fermented to lactic acid (table 4). although some of the strains failed to ferment xylose, Acetic and traces of other acids were also found. All of maltose, sucrose, and esculin and a few showed slight the cultures tested produced the levo- form of lactic growth in starch and dextrin. acid. The cultures fermented less glucose than L. mesente- (d) Leuconostoc, type I. Leuconostoc, type I, showed roides, although the carbon distribution was similar and greater morphological variation than was found with levo-lactic acid was produced (table 4). L. mesenteroides. The cells varied from about 0.4 to 0.7 The close physiological similarities between L. ,u by 0.8 to 4 ,. Under some conditions of growth it was mesenteroides and Leuconostoc, type I, indicate that they difficult to determine if the organisms were actually are closely related. Pederson and Albury (1955), in a cocci or rods. Cultures grow-n in stabs were usually more study on atypical nondextran producing Leuconostoc, coccoid in shape than those grown in broth. This rela- show-ed that by serial transfers in tomato juice, orange tionship did not always hold true, however, and it was juice, and sucrose broths many of their strains recov- not unusual to find within a smear small oval and ered the ability to produce dextran and some were able elongated rod forms from solid or liquid media. This to utilize sugars not fermented previously. In experi- morphological variation presented some problems in the ments similar to those described by the above authors, early phase of work. It was difficult to determine if they silage strains of atypical Leuconostoc and heterofer- were more closely related to the heterofermentative mentative lactobacilli w-ere subjected to serial transfers. cocci or rods. The problem w-as further complicated in Sixteen of the 34 atypical Leuconostoc, type I, strains that although they usually fermented sucrose they were able to form varying amounts of dextran on 10 failed to produce dextran on plates containing 10 per per cent sucrose agar after 10 serial transfers in tomato cent sucrose. Later results show-ed that they w-ere closely juice glucose broth and most of the strains show-ed related to the genus Leuconostoc and, as will be show-n dextran after 32 transfers in tomato juice glucose broth follow-ed by 29 transfers in orange juice glucose broth. TABLE 4 A small number of cultures produced dextran only after Per cent acids and optical type of lactic acid produiced 91 serial transfers in tomato and orange juice glucose by Leuiconostoc broth. The dextrani produced varied from visible to Per Per Cent Fermented Glucose Converted to Optical moderate amounts. Few- of the cultures produced dex- Cent Type of Culture No. Glucose Lactic tran as profusely as typical strains of L. mesenteroides. Fer- Buty- Pro: Acetic Formic Suc- Lactic Acid mented ric pionic Iacid acid cinic acid Produced It w-as interesting to follow- the population change of a acid acid acid culture from nondextran to dextran production. As the Leuiconostoc mtesenteroides cultures w-ere serially transferred, isolated colonies and T-13 95.6 0 0.2 2.6 0.04 0.4 40.1 Levo the periphery of streaked areas first showed the smooth, T-60 77.0 0 0.3 3.0 0.02 0.3 24.0 Levo shiny, transparent appearance characteristic of dextran T-79 78.0 0 0.3 2.9 0.07 0.6 42.7 Levo T-328 86.0 0 0.4 4.0 0.2 1.0 38.6 Levo colonies. After further serial transfers, small, raised, Leuconostoc type I shiny areas w-ithin a streak appeared and continued to T-17 44.5 0 0.08 5.3 0.08 0.2 37.5 Levo increase until the entire population converted. T-40 68.0 0 0.01 4.2 0.2 0.6 43.7 Levo (e) Pediococcus. The high lactic acid producing cocci T-57 43.0 0 0 5.2 0.1 0.7 37.7 Levo that form tetrads were first described as spoilage agents 0 0.1 3.1 0.2 0.7 40.0 T-312 76.0 Levo the the of these T-323 51.0 0 0.04 5.7 0.1 0.1 26.9 Levo in beer and through years relationship T-357 62.0 0 0.07 4.4 0 0.3 35.8 Levo organisms to other types of homofermentative lactic acid T-1600 60.0 0 0 2.8 0 0.9 39.5 Levo bacteria has been confused. In the 6th edition of R-610 58.0 0 0.06 4.3 0.08 0.5 37.6 Levo Bergey's Manual of Determinatite Bacteriology (Breed, R-943 59.0 0 0.1 4.0 0.02 0.6 42.5 Levo .Murray, and Hitchens, 1948), the genus Pediococcus w-as 0 0.2 3.2 0.1 0.4 38.7 R-962 75.0 Levo included in the appendix in the family Iicrococcaceae. 218 C. W. LANGSTON AND C. BOUMAv [VOL. 8.

A critical review and detailed study of the group by fermented was converted to lactic acid along with l'ederson (1949) suggested that they should be placed acetic and traces of other acids. Pederson's (1949) study in the family of lactic acid producing cocci and rods. The on 121 strains of the genus Pediococcus from fermenting 7th edition of Bergey's MIanual (Breed et al., 1957) now vegetables showed some differences among strains, but includes the genus Pediococcus in the tribe Strepto- on the basis of carbohydrate fermentations no signifi- cocceae of the family along with the cant trends were demonstrated. He concluded that all genera Diplococcus, Streptococcus, and Leutconostoc. of his strains should be considered as belonging to one The pediococci examined in this study were rather species. His strains produced inactive lactic acid, a trace uniform in their morphology. The cells occurred singly, of volatile acid, and a small amount of CO2. They in pairs, as tetrads (especially in acid medium), and fermented glucose, fructose, mannose, galactose, occasionally short chains (4 to 6 cells). They averaged maltose, and usually arabinose, sucrose, lactose, raffi- 0.6 to 0.7 ,u by 0.8 to 1.1 u. All of the cultures grew in 6.5 nose, salicin, and amygdalin. Most strains failed to per cent of sodium chloride and at 15 C. None liquefied ferment mannitol, a-methylglucoside, inulin, dextrin, gelatin, gave gas from glucose, or produced acetyl- and starch. Pederson brought out that slight differences methylcarbinol. All of the cultures examined produced in ability to utilize certain sugars and related com- inactive lactic acid (table 5) and most of the glucose pounds should not be considered sufficient grounds for excluding a strain from the species or establishing a new TABLE 5 species. Per cent acids anid optical type of lactic acid produiced The silage isolates had many characteristics similar to by Pediococcus those outlined above. Some additional characteristics not noted by Pederson were observed in the present Per Per Cent Fermented Glucose Converted to Cent Optical Group and C Type of work. For comparative purposes and to show certain Culture No. Glucose Bu- Pro- A Formi Suc- Lci Lactic Acid Fer- ormc variable characteristics, the 79 strains in this study were mented tyric pionic aciacd acid cnccinic acid Produced n acid acid acid divided into five groups. In table 6 it may be noted that the majority of the Grouip 1 T-33 56.4 0 0.1 0.4 0 0 94.4 Inactive strains produced ammonia from arginine. However, all T-1599 54.8 0 0.8 0.06 0 0 97.2 Inactive strains in group 4 failed to hydrolyze this compound. T-1746 58.6 0 0 2.3 0.3 0.7 91.2 Inactive Some of the strains in each group produced catalase Grou2p 2 with the exception of those in group 5. Organisms in T-1 52.2 0 1.8 0.2 0.3 0.3 95.2 Inactive group a also grew at 48 C and the other groups failed T-3 46.6 0 0.2 2.2 0.1 0.9 94.9 Inactive T-351 55.8 0 0.1 1.0 0.08 0.4 94.4 Inactive to grow at this temperature. Strains varied in litmus T-2002 52.4 0 0.2 1.8 0.02 0.5 91.2 Inactive milk, some produced only a slight acid and others Grou2p 3 caused an acid curdled reaction. Many of the cultures T-330 56.4 0 0 0.9 0.06 0.3 95.8 Inactive reduced the litmus in the bottom of the tubes and a few T-1206 57.6 0 0 0.2 0 0.8 93.0 Iniactive were capable of almost complete reduction of the litmus. R-800 72.6 0 0.1 1.0 0.07 0.3 93.7 Inactive Grouip 4 The curdling reaction in litmus milk always occurred T-2058 53.3 0 0.2 0.06 0.1 0.5 996.8 Inactive after 1 week of inicubation. Final pH and titratable 1t-247 47.8 0 0 0.7 0 0.4 92.0 Inactive acidity -alues ranged from 3.6 to 4.1 and 0.78 to 1.4 Grou2p 5 per cent, respectively. Strains in group 4 gave the lowest TI"-953 50.4 0 1.2 1.2 0.04 I 97.4 Inactive pH and the highest titratable acidity values. T-1 367 40.0 0 0.9 2.0 0.01 94.7 Inactive All of the strains (table 7) fermented glucose, fruc-

TABLE 6 Physiological characteristics of the yen us Pediococcsls Group No.

1 2 3 4 5

Ammoniia from arginiine...... 11+ 21+ 20+ 14- 13+ Catalase prodtuction...... 8-; 3+ 1(- ; 11+ 15-; 5+ 12-; 2+ 13- Growth at 15 C...... 11+ 21+ 20+ 14+ 13+ Growth at 45 C...... 11+ 21+ 20+ 7-; 7+ sl 13+ Growth at 48 C...... 11- 21- 20- 14- 13+ Litmtus milk (1 month)...... A, C 9A, C IIA, C 7A, C sl A 12 sI A 9 sl A 7 sl A Final pH gltucose broth...... 3.8-4.0 3.9-4.0 3.9-4.1 3.6-3.8 3.8-4.0 Titratable acidity glucose broth...... 0.8-1.09 0.78-1.0 0.79-1.0 1.1-1.4 0.95-1.1 19601 MICROORGANISMS FROM GRASS SILAGE. I 219

Cd .)§ *I1 -4tt - k- = I>- 0 0 0 0 0 = VD C CD c9 C Ci =0 Co' t - C C6 r t- 0 t o q oo 0 i cteNno~~~~~~~~~~~~~NSb N LOc

. b5J 0 4M S A A 4 4 . > 0 o v t t en I I O s s ~~~~~~++ + + + + + + + o+ e+ + + +

0 cs Co C-c cs

~~10 -~ -

< I e dq t t U: < U: t c< d 4J2ou co

ce

0.- -l -i -i -~ - -- O~~~~~~~~~~." IR1 cy cc du C6 IIII ~~~~~~~~~~~~~"tcc I", IC CXO X0 00 0 let, c0 N O~~~~~~~~~ I-- .t+t+!,,t+ + +--.+ +t +-1+- -!+ +

______

cC I*~ -CI ii o II bED

on "0 ¢ L CC II4cq C-1 -1)b._

IC 10C O0'-' t1C1OX CO 00

m O -tCQ " COD CD Oe t It"Vc t" 0~ ~ ~ ~ C ___ _

0.W° ._ .) ......

°O1 Iq Cq C4 O _q _q _N _q N N CSCDN ._ ce0 ;> I IXII II

Cd tt ^ d d t ^t Q tN,C"

0C 0 C)~~~~10°0 ff-~~cCOO-N C C '-I>OC- r- -l LOQ . 0q ccCOS_ 0cc- SC

- O.>C : O- O-O C _ bOb 0.0C ...... ;-

1 0o CO t COt OCOCOCOCOCOCOt't + + + + + + + + + + + + + + + + + 0 :.W°~ _ - ~4--~ - - ~ - - ._ ._

- CS - - -

CQ 1. d .4. .0 /I)

C-c ~ 220 C. W. LANGSTON AND C. BOUMA [VOL. 8 tose, mannose, galactose, cellobiose, salicin, and esculin. Recent work has been submitted which indicates that They usually failed to ferment sorbose, melezitose, Streptococcus faecium (Orla-Jensen, 1919) should be starch, dextrin, inulin, sodium hippurate, mannitol, given species rank along with S. faecalis. For many inositol, sorbitol, or sodium lactate. Variations were years these organisms w-ere regarded as being synony- observed in the fermentation of arabinose, xylose, mous. The following characteristics have been used to ribose, rhamnose, maltose, lactose, sucrose, trehalose, differentiate between the organisms: S. faecalis grows melibiose, and raffinose. The greatest variation occurred in the presence of /1500 potassium tellurite, usually in groups 3, 4, and 5. Pederson (1949) reported that the shows a vigorous fermentation of mannitol and sorbitol, majority of his strains fermented sucrose, lactose, and fails to ferment arabinose, and gives a strong reduction raffinose. Many of the strains in the above mentioned of litmus milk prior to acid and curdling. Streptococcus groups failed to ferment these compounds. Further- faecium is inhibited by 12500 potassium tellurite, al- more, Pederson showed that all of his strains fermented ways ferments arabinose, varies on mannitol, usually maltose. None of the silage strains in group 5 fermented does not ferment sorbitol, an-d may show only acidity this sugar. Some of the strains in group 2 and most of in litmus milk with little or no reduction. Although four the strains in groups 3, 4, and 5 were not able to fer- of the silage strains resembled S. faecium in that they ment melibiose. The greatest v-ariation in the ability to fermented arabinose and failed to ferment sorbitol, ferment carbohydrates occurred in group 4. None of they must be regarded as S. faecalis because all of the these strains fermented arabinose, xylose, ribose, sucrose, strains studied were able to grow in the presence of melibiose, raffinose, or glycerol and it is interesting to 21500 potassium tellurite and gave a strong reduction note that they also failed to hydrolyze arginine. This of litmus milk prior to acid and curdling. The cultures group deserves further attention with regard to final also exhibited a-hemolysis on blood agar. The diversity pH values in carbohydrates. It may be seen in table 6 of fermentative reactions within this species was pointed that organisms in this group produced lower pH and out by Harrison and Hainsen (1950a). In fact, they higher titratable acidity values than were found in the showed greater variation in carbohydrate fermentation other groups. However, in table 7 it may be seen that among strains of S. faecalis isolated from the intestinal the same organisms gavre the highest final pH in glucose flora of healthy turkeys than was reported here or by and other carbohydrates. This seems to illustrate Sherman et al. (1937). variability among strains in response to conditions for The 6th edition of Bergey's Mlanual (Breed et al., growth. 1948) regards S. faecalis and S. liquefaciens as separate Whether the group differences indicated are impor- species. The 7th edition of Bergey's Manual lists S. tant enough to warrant their division inlto new species liquefaciens as a vrariety of S. faccalis. Apparently the or subspecies remains to be seen. Certainly, too few editors felt that differences wvere not great eniough to strains have been studied to draw any definite conclu- warrant species distinctioni. Although it is indicated sions. At this time, however, the authors agree with the that the primary differences between the two organisms suggestion of Pederson that the strains studied should are that of gelatin li(uefaction and peptonization of be considered as belonging to one species (Pediococcus milk, it is interesting to note that strains of S. liqute- cerev,isiae Balcke). When a greater number of strains are faciens isolated from forages varied from S. faecalis in studied from different environments, it is possible that that they consistently fermeinted melezitose and sodium the group differences indicated above will become im- hippurate. Harrison and Hansen (1950a) also showed portant. this distinction with regard to the fermentation of Discussio-N melezitose. They did not, however, test the ability of The results showed that the cocci isolated from their strains to hydrolyze sodium hippurate. forages fell into two primary groups. Those which pro- A comparison of the data obtained in this study wvith (luce mostly lactic acid from glucose and those which the detailed work of Hucker anid Pederson (1930) on the produce limited amounts of lactic acid along with C02, genus Leuconostoc showed that strains of L. mesentcroides acetic, and traces of other acids. from forage are identical to theirs in most respects. The 13 strains of S. faecalis examined compared The forage strainis differed mainly in their ability to favorably with the published data of other workers. The produce catalase. MIembers of the genera , only variation of significance as compared to the wvork Pedicoccuzs, and Leutconostoc have been shown to possess of Sherman et al. (1937) and described in Bergey's the enzyme catalase (Harrisoni and Hansen, 1950a; Mlanuial (Breed et al., 1957) is that 12 of the 13 silage Felton et al., 1953; Dacre and Sharpe, 1956; Vankova, strains hydrolyzed starch and nonie of their strains were 1957). Since this property has become so common able to hydrolyze this carbohydrate. The hydrolysis of among the facultatively anaerobic lactic acid producing starch by this group would suggest some similarity to bacteria, it is doubtful that it should be used as one of Streptococcus bovis but, other basic characteristics de- the main lines of identification anid classification. Harri- scribed earlier discount this relationship. son and Hansen (1954) in to the genus Lacto- 19601 1IICROORGANISMIS FROM GRASS SILAGE. I 221 bacillus stated that the presence of catalase excludes strains from a variety of sources. They have been found any member from this group. By the same token this in fermenting forages, fermenting vegetables (Pederson, should also apply to the genus Leuconostoc. If taxono- 1949), the rumen (Bauman and Foster, 1956), summer mists agree with this opinion then it will be necessary sausages (Deibel and Niven, 1957), the cecal feces of to remove the catalase producing strains to new and turkeys (Harrison and Hansen, 1950a), and in cheese different genera. If, however, the decision is made to (Dacre, 1958). Pederson indicated that these organisms maintain the organisms in their present taxonomic posi- may be important in vegetative fermentations when tions, it will be necessary to change the genera defini- more is learned about their true role in nature. In view tions to include strains which are able to produce of the part played by this group in the silage fermenta- catalase. tion, his prediction is true. They occur in fermenting The evidence is clear that Leuconostoc, type I, should forage in high numbers and because of their ability to be regarded as a variable of Leuconostoc mesenteroides. produce and withstand high acid conditions they are This is obvious not only from the standpoint of glucose undoubtedly important in the preservation process. fermentation and by-product distribution but also from In Pederson's (1949) work on the genus Pedio- reactions in different carbohydrates. The sucrose fer- , he stated they were catalase negative. Later menting, nondextran producing strains of Leuconostoc work by Felton et al. (1953) showed that catalase was and Lactobacillus studied by Pederson and Albury produced by these organisms when they were streaked (1955) gave submaximal amounts of acid in glucose and grown on low carbohydrate medium. Weak cata- broth (average of about 0.24 per cent). Some of the lase activity in the group had been shown earlier, how- forage strains also produced low acid in glucose but the ever, by Harrison and Hansen (1950a). Work by Jensen range was from 0.23 to 0.77 per cent. It was interesting and Seeley (1954) further defined the group. They de- to note that when their strains recovered the ability to termined nutritional requirements of 34 strains and produce dextran that they were also able to utilize several physiological characteristics not determined by certain sugars that they had not previously fermented. Pederson. Later work was presented by Dacre (1958), They suggested that the changes observed in variants on a strain of Pediococcus isolated from cheddar cheese. may involve dormant or latent enzyme formation, It was concluded that the strain studied was a variant spontaneous mutation or a combination of the two of the type species, Pediococcus cerevisiae Baleke. It was systems. catalase negative, failed to grow in 6.5 per cent of Other investigators have also observed variations in sodium chloride, and showed variation on certain- dextran production among the Leuconostoc and Lacto- carbohydrates. This strain showed some characteristics bacillus. Niven et al., (1949) studied the lactic acid that resembled the forage strains described in groups bacteria which caused surface discoloration on sausages 4 and 5 (table 7). and found that 14 of 20 cultures of the genus Lacto- With the exception of the pediococci, the cocci de- bacillus studied produced large mucoid colonies on 5 per scribed in this paper occurred early in the forage fer- cent sucrose gelatin medium. Ten strains of Leuconostoc, mentations, then were lost when the higher acid produc- on the other hand, failed to produce mucoid colonies on ing lactobacilli appeared. It was not unusual to find this same medium. Carr (1957) studied strains of lactic pediococci both early and late in the fermentation acid producing cocci and rods from apple juice, ciders, process. The effect the cocci have upon the fermentation and perries and reported that they fell into twro groups of forage will be discussed in a subsequent paper. which he considered belonging to the species L. mesente- roides. The groups showed variation in fermentation ACKNOWLEDGMENTS patterns and none of the strains produced dextran in the The authors wish to express their appreciationi to Dr. presence of sucrose. One strain, a rod that he designated P'. Arne Hansen of the University of Maryland for his as Lactobacillus pastorianus, was able to produce slime helpful suggestions and critical comments oni the manu- from sucrose. script, and to Barbara Stafford and Patricia Augustine Other workers (Ward, 1892; Perquin, 1940; Koba- whose technical assistance aided materially in this work. yasha, 1944; Deibel and Niven, 1959) have reported incidence of lactobacilli that produced mucoid colonies SUMMARY from sucrose. Continued reports of lactobacilli that Cultural and physiological data have been presented synthesize dextran from sucrose have narrowed the on the homo- and heterofermentative cocci isolated taxonomic differences between the genera Lactobacillus from forages. anid Leuconostoc and more work is needed to clearly Results obtained with cultures of Streptococcus define the relationships and boundaries of these groups. faecalis and Streptococcus liquefaciens compared favor- Following the review and description of the genus ably with the published reports of other workers. Pediococcuts by Pederson (1949), various investigators Leuconostoc mesenteroides and variant strains of this have examined his strains aDd have isolated and studied species were described. The strains of L. mesenteroides 222 C. W. LANGSTON AND C. BOUMA [VOL. 8 examined showed only minor variations from their usual physiology of the genus Pediococcus. J. Bacteriol., 67, 484-488. description. The most striking feature exhibited was the KOBAYASHA, T. 1944 Mucous fermentation of sucrose by ability of some strains to produce catalase. Data were Lactobacillus mucicus isolated from tobacco leaves. J. presented to show that atypical, sucrose fermenting, Agr. Chem. Soc., Japan, 20, 60-64 (Chem. Abst. 1948, 42, nondextran producing cocci isolated from forage were 3810a). similar to L. mesenteroides. The nondextran producing LANGSTON, C. W. 1955 Microbiology and chemistry of grass cocci recovered their ability to produce dextran after silage. Ph.D. thesis, University of Wisconsin, Madison, Wisconsin. serial transfers in enriched media. LANGSTON, C. W., IRVIN, H., GORDON, C. H., BOUMA, C., Strains of the genus Pediococcus studied were divided WISEMAN, H. G., MOORE, L. A., AND MCCALMONT, J. R. into five groups on the basis of physiological charac- 1958 Microbiology and chemistry of grass silage. U.S. teristics. Although group variations were observed, it Dept. Agr., Tech. Bull., 1187, 73 pp. E. 1951 Staining, shape, and arrangement of was concluded that a greater number of these organisms LEIFSON, bacterial flagella. J. Bacteriol., 62, 377-389. should be studied before specific names are applied to LONG, H. F. AND HAMMER, B. W. 1936 Classification of the them. organisms important in dairy products. I. Streptococcus liquefaciens. Iowa State Coll. Agr. Expt. Sta. Bull., 206, REFERENCES 41 pp. BAUMAN, H. E. AND FOSTER, E. M. 1956 Characteristics of MCCLESKEY, C. S., FAVILLE, L. W., AND BARNETT, R. 0. 1947 organisms isolated from the rumen of cows fed high and low Characteristics of Leuconostoc mesenteroides from cane roughage rations. J. Bacteriol., 71, 333-338. juice. J. Bacteriol., 54, 697-708. BREED, R. S., MURRAY, E. G. D., AND HITCHENS, A. P. 1948 MORRIS, D. L. 1948 Quantitative determination of carbo- Bergey's manual of determinative bacteriology, 6th ed. The hydrates with Dreywood's anthrone reagent. Science, Williams & Wilkins Co., Baltimore, Maryland. 107, 254-255. BREED, R. S., MURRAY, E. G. D., AND SMITH, N. R. 1957 NEISH, A. C. 1950 Analytical methods for bacterial fermen- Bergey's manual of determinative bacteriology, 7th ed. tations. Report No. 46-8-3 (revised). National Research The Williams & Wilkins Co., Baltimore, Maryland. Council of Canada, Saskatoon, Canada. BRIN, M., OLSON, R. E., AND STARE, F. J. 1952 Microbiologi- NIVEN, C. F., JR., SMILEY, K. L., AND SHERMAN, J. M. 1944 cal synthesis of C'4-uniformly labeled L(+)- and D(-)- The hydrolysis of arginine by streptococci. J. Bacteriol., lactate. Arch. Biochem. Biophys., 39, 214-218. 43, 651-660. CARR, J. G. 1957 Occurrence and activity of some lactic NIVEN, C. F., JR., CASTELLANI, A. G., AND ALLANSON, V. acid bacteria from apple juices, cider and perries. J. 1949 A study of the lactic acid bacteria that cause sur- Inst. Brewing, 63, 436-440. face discolorations of sausage. J. Bacteriol., 58, 633-641. DACRE, J. C. 1958 Characteristics of presumptive Pedio- ORLA-JENSEN, S. 1919 The lactic acid bacteria. Mem. coccus occurring in New Zealand cheddar cheese. J. acad. roy. sci. Danemark sect. sci., 8 ser., 5, 81-197. Dairy Research, 25, 409-413. PEDERSON, C. S. 1949 The genus Pediococcus. Bacteriol. DACRE, J. C. AND SHARPE, N. E. 1956 Catalase production Revs., 13, 225-232. by lactobacilli. Nature, 178, 700. PEDERSON, C. S. AND ALBURY, M. N. 1955 Variation among DEIBEL, R. H. AND NIVEN, C. F., JR. 1957 Pediococcus the heterofermentative lactic acid bacteria. J. Bacteriol., cerevisiae: a starter culture for summer sausage. Bac- 70, 702-708. teriol. Proc., 1957, 14-15. PERQUIN, L. H. C. 1940 On the incidental occurrence of rod- DEIBEL, R. H. AND NIVEN, C. F., JR. 1959 Microbiology shaped, dextran producing bacteria in a beet-sugar factory. of meat curing. II. Characteristics of Lactobacillus oc- Antonie van Leeuwenhoek. J. Microbiol. Serol., 6, 227- curring in ham curing brines which synthesizes a polv- 249. saccharide from sucrose. Appl. Microbiol., 7, 138-141. ROGOSA, M., MITCHELL, J. A., WISEMAN, R. F. 1951 A selec- FELTON, E. A., EVANS, J. B., AND NIVEN, C. F., JR. 1953 tive medium for the isolation and enumeration of oral Production of catalase by the pediococci. J. Bacteriol., lactobacilli. J. Bacteriol., 62, 132-133. 65, 481-482. SHERMAN, J. M. 1937 The streptococci. Bacteriol. Revs., HARRISON, A. P., JR. AND HANSEN, P. A. 1950a The bac- 1, 3-97. terial flora of the cecal feces of healthy turkeys. J. SHERMAN, J. M. 1938 The enterococci and related strepto- Bacteriol., 59, 197-210. cocci. J. Bacteriol., 35, 81-93. HARRISON, A. P., JR. AND HANSEN, P. A. 1950b Lacto- SHERMAN, J. M., MAUER, J. C., AND STORK, P. 1937 Strepto- from turkeys. J. Bacteriol., 60, 543-555. coccus faecalis. J. Bacteriol., 33, 275-282. HARRISON, A. P., JR. AND HANSEN, P. A. 1954 Characteriza- SMITH, P. A. AND SHERMAN, J. M. 1942 The lactic acid fer- tion of the Lactobacteriaceae encountered as secondary mentation of streptococci. J. Bacteriol., 43, 725-731. invaders in blackhead liver lesions of turkevs. Zentr. VANKOVA, J. 1957 Motile catalase-producing strains of Bakteriol. Parasitenk., 161, 125-137. Lactobacillus delbruickii. Nature, 179, 204. HUCKER, G. J. AND PEDERSON, C. S. 1930 Studies on the WARD, H. M. 1892 The ginger-beer plant, and the organisms coccaceae. XVI. The genus Leuconostoc. N. Y. State composing it: A contribution to the study of fermenta- Agr. Expt. Sta., Tech. Bull. 167, 80 pp. tion yeasts and bacteria. Phil. Trans. Roy. Soc. London, JENSEN, E. M. AND SEELEY, H. W. 1954 The nutrition and Ser. B., 183, 125-197.