BACTERIAL, BIOCHEMICAL AND ENVIRONMENTAL INTERRELATIONS IN FRESH AND ENSILED FORAGES By ALAN GEORGE KEMPTQN A THESIS Submitted to the College of Science and Arts of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1953 ProQuest Number: 10008573 All rights reserved INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted. In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest ProQuest 10008573 Published by ProQuest LLC (2016). Copyright of the Dissertation is held by the Author. All rights reserved. This work is protected against unauthorized copying under Title 17, United States Code Microform Edition © ProQuest LLC. ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106- 1346 ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to Dr. C. L. San Clemente for his advice and encouragement throughout the course of this work. Even outside the laboratory, Dr. San Clemente continuously labored to ensure that the cultural and physical advantages of Michigan State University were utilized to the fullest extent. The help of Dr. S. T. Dexter and Dr. R. N. Costilow was also frequently solicited and generously given. The author is deeply indebted to Dr. E. Benne under whose? guidance some of the chemical analyses were carried out. Thanks are also extended to Dr. J. L. Fairley and Dr. L. C. Ferguson for their continued interest in this project and their assistance in the preparation of this manuscript. VITA Alan George Kempton candidate for the degree of Doctor of Philosophy Final Examination: September 15, 1953, 10:00 A.M. Room 101, Giltner Hall Dissertation: Bacterial, Biochemical and Environmental Interrelations in Fresh and Ensiled Forages* Outline of Studies: Major subject: Microbiology Minor subject: Biochemistry Biographical Items: Born, August 21, 1932, Toronto, Canada. Undergraduate Studies: Ontario Agricultural College Guelph, Ontario, 1950-1954, B.S.A. Graduate Studies: Ontario Agricultural College Guelph, Ontario, 1954-1956, M.S.A. Michigan State University East Lansing, Michigan, 1956-1953. Experience: Graduate Teaching Assistant Ontario Agricultural College, 1954-1956. Graduate Research Assistant Michigan State University, 1956-1953. Affiliations: Society of the Sigma Xi Canadian Society of Microbiologists Society for Industrial Microbiology American Association for the Advancement of Science. BACTERIAL, BIOCHEMICAL AND ENVIRONMENTAL INTERRELATIONS IN FRESH AND ENSILED FORAGES By ALAN GEORGE KEMPTQN AN ABSTRACT Submitted to the College of Science and Arts of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Microbiology and Public Health 1958 Approved Alan George Kempton This investigation was undertaken to make simultaneous bacterial, biochemical and environmental studies on a wide range of fresh and ensiled forages, using farm-sized silos as experimental vehicles. The predominant bacteria on all fresh forages were classified as facultative-anaerobic species of Flavobacterium. on the basis of pigment production, and morphological and fermentative studies. Less than 0.1 per cent of the bacteria on the crop at the time of ensiling were capable of growing on lactobacillus selection medium. In spite of differences in plant species, weather conditions, wilting and other harvesting procedures, the number of bacteria on the fresh crops could be predicted with reasonable error solely from the harvesting date. The predominant Flavobacterium type increased from 10^ per gram of dry matter in late May, to 1 0 ^ in early August. The lactic acid bacteria showed a parallel increase from 10^ to 10^. However, it was found that the initial number of bacteria on the fresh crop bore no relationship to the final quality of the silage. A significant correlation between the moisture content and pH of fresh crops was noted, in spite of differences in crop species. For crops of 60 per cent moisture, the regression line passed through a pH of 6.2 but falls to a pH of 5*7 when Alan Georgs Kempton the moisture content rises to per cent. Since the low pH values of high moisture crops were not associated with higher numbers of acid-producing bacteria, it was concluded that plant enzymes may be more active in high moisture crops than in low moisture crops. Silage quality was determined primarily by the amount of packing the silage received. Loosely packed silages overheated, but underwent an acetic acid fermentation. Well-preserved silages were optimumly packed, and contained as much as 300 micromoles of lactic acid per gram of fresh weight. When silage was packed too tightly, vegetative cells of Clostridium tvrobutvricum could be isolated immediately after ensiling. Although tightly-packed silages contained about 100 micromoles of lactic acid after 2-3 days in the silo, all the lactic acid subsequently disappeared, to be replaced by butyric acid. During spoilage, succinic acid also disappeared and some propionic acid was produced. Crops with a low dry matter content tended to undergo butyric spoilage only because they were more apt to be Excess production of volatile base could overpacked. not be associated with any particular organism, but always occurred when the total hydrolysable carbohydrate dropped below 1 per cent of the fresh weight. High undergo a volatile base type of moisture crops tendedto spoilage because the carbohydrate content was initially diluted. Alan George Kempton Therefore, one silage was tightly packed and a butyric spoilage developed, but since the final carbohydrate content was 1.3 per cent, no excessive volatile base was produced. Although Cl. tvrobutvricum developed in a bisulfite silage because it was too tightly packed, the bisulfite inhibited the formation of butyric acid. Bisulfite also inhibited the normal utilization of carbohydrate; hence, an appreciable amount of proteolysis and deamination occurred even though the carbohydrate content was above the critical 1 per cent. The Flavobacterium group 011 the fresh forage did not multiply in the silage process, but did persist throughout the storage period in spite of the accumulation of acid. The lactic acid bacteria increased tremendously soon after ensiling, frequently approaching 10^ per gram of fresh material. Thereafter there was a relatively rapid decrease to an average of about lQ^ after three weeks in the silo. In most silages this was followed by a secondary fermentation which reached a peak of about 10? in four to five weeks, and subsequently declined. This pattern was found in both well- preserved and spoiled silages. Lactic acid bacteria did not develop at all in the overheated silages. The lactic acid bacteria of fresh and ensiled forages were very variable morphologically. On a fermentative basis, they divided equally into two main types; one which fermented Alan George Kempton all test carbohydrates and litmus milk, and one which fermented only simple sugars. There were no detectable differences in the lactic flora of fresh crops and of the primary and secondary f erment ati on s . TABLE OF CONTENTS CHAPTER Page I INTRODUCTION......... .............................. 1 II LITERATURE REVIEW................................... 4 The bacterial flora of fresh plants prior to ensiling..................... 4 The chemical composition of fresh forage crops.... 6 9 Carbohydrate metabolism in silage................ Nitrogen metabolism in silage........ 14 Silage microbiology...................... 19 III METHODS............................................. 27 ..... 27 Sampling procedure Enumeration of bacterial populations...............2$ Characterization of bacterial populations...... 30 Chemical determinations........ 33 Statistical calculations........... 35 IV THE BACTERIAL FLORA OF FRESH FORAGES................ 37 Introduction................ Results........ Discussion........ V THE CHEMISTRY OF FRESH FORAGES...................... Introduction.............. Results. ........... Discussion.. ..... VI GENERAL FERMENTATION INDICES........................ Introduction............. Results.......... Discussion............. VII 37 38 43 46 46 46 48 49 49 49 51 CARBOHYDRATE METABOLISM IN SILAGE................... 56 Introduction............... Results ..................................... Discussion............................... 56 56 60 TABLE OF CONTENTS - Continued CHAPTER VIII Page NITROGEN METABOLISM IN SILAGE....................... 63 Introduction ...... 64 Results*..*........................................ 64 Discussion. ..... 66 IX SILAGE BACTERIOLOGY.......................... Introduction........ Results........... Discussion.................... X XI 71 71 72 77 GENERAL CONCLUSIONS................................... 81 SUMMARY. ..... 86 REFERENCES............................... APPENDIX.................................................... 97 LIST OF FIGURES Page 1 2 3 4 5 6 7 8 9 10 11 12 13 14 The relationship between the number of bacteria on fresh forages at the time of ensiling and the harvesting date................ 39 Carbohydrate preference of the acid-producing strains of the chromogenic bacteria of fresh forages....... 42 The relationship between the pH and moisture content of fresh forages at the time of ensiling .......... 47 Comparison of the pH changes in a wellpreserved silage (silo 7) and a spoiled silage (silo 9l*...................... »....... 52 Changes in the dry matter content of silages during storage.. ...... 53 The production of organic acids in a wellpreserved silage (silo 7).....,................ $7 The production of organic acids in a spoiled sila ge (silo 9) ••....... 5$ The production of organic acids in an over-heated silage (silc 13)................... 59 The relationship between the volatile and lactic acids in silage and the pH...... 61 The production of amino acid and volatile base in a well-preserved silage (silo 7)....... 65 The production cf amine acid and volatile base in a spoiled silage (silo 67 The production of amino acid and volatile base in a bisulfite-treated silage (silo 1).... 68 The population of lactic-acid bacteria in well-preserved, spoiled, and over-heated silages.......... 73 Summation of the biochemical changes in a spoiled silage (silo 9) ......................... 84 LIST OF TABLES Table I II III IV Page Pigments produced by the bacteria of fresh forages isolated from TGY medium,.......• 41 General description of silos and silages ..... .......,.......... ............... 50 Fermentative ability of silage microorganisms isolated from V& and lactobaciilua-seleotion media,.,..,,,.,...,..,. 74 Comparison of the lactobaoilli of fresh and ensiled forages. ......... 76 CHAPTER I INTRODUCTION Increasing emphasis on the practice of ensiling grasses and legumes has stimulated efforts to develop a method of ensiling which would consistently produce palatable silage of high nutritive value. Silage preservation depends primarily on the attainment of a pH low enough to restrict the activities of spoilage bacteria. Providing there is sufficient fermentable carbohydrate present, the lactic acid produced by the metabolism of the indigenous microflora should lower the pH below the critical value of 4.0. Legumes are difficult to ensile because they have a low content of natural carbohydrate; hence several attempts to assure the production of good quality silage have been based on the addition of various carbohydrate sources such as molasses, beet pulp or whey. The mechanical adjustment of pH by the addition of mineral acids has been practised extensively in Europe, and recently (Knodt et al. 1952} the control of spoilage bacteria by the use of selective bacteriostatic agents has been advocated. It is also apparent that the efficacy of the fermentation is controlled by the moisture content of the crop. Within 2 limits, the moisture content can be controlled by wilting high-moisture crops before ensiling, or if necessary, water can be added during ensiling. Unfortunately, no single method of making silage has been consistently satisfactory over a wide enough range of field conditions to be universally acceptable; hence spoiled silage still accounts for a substantial loss of farm income. Successive failures have partially been due to a lack of fundamental knowledge of silage bacteriology. Although the predominant types of bacteria found at various stages of the fermentation have been briefly characterized, the precise role of each type has not been determined, nor have they been fully classified. Silage quality has often been correlated with environmental conditions or the chemical products of bacterial action, but the direct relationship between silage quality and the bacterial population requires further clarification. Furthermore, it has never been definitely proved that the fermentation occurring in large silos is identical with the fermentations in the miniature silos studied by most workers. Consequently, this present investigation was undertaken in an attempt to interrelate the Quality of forage crop silage with the bacterial and biochemical composition of the fresh and ensiled crops, and the environmental conditions at the time of ensiling. By sampling the fermentations which took place in a number of farm-sized silos located in Michigan’s 3 Baton and Ingham counties, a wide range of field conditions was included in this experiment. The results of this study are presented and discussed in two main parts. The first, part deals with the salient environmental conditions causing variation in the bacterial and chemical nature of fresh forages at the time of ensiling. The progressive changes occurring during the ensuing fermentations are presented in the second part. CHAPTER II LITERATURE REVIEW The Bacterial Flora of Fresh Plants Prior to Ensiling. The microorganisms responsible for silage fermentation are on the fresh plant material before ensiling. However, attempts to demonstrate the presence of typical silage lactic acid bacteria on fresh plants have produced conflicting conclusions. Using dilution techniques, Allen and Harrison (1936b) identified the predominant organism on grass as Lactobacillus plantarum.* the same organism as found in silage. Stone et al (1943) also concluded that fresh alfalfa contained large numbers of lactobacilli. Conversely, Stirling (1953) and Kroulik et al. (1955a) found that newly developed media which were selective for lactobacilli supported the growth of only a very small proportion of the organisms found on fresh plants The few organisms Kroulik and co-workers did isolate did not belong to the groups of active acid-producing lactobacilli typical of the silage flora. Stirling was able to demonstrate *The term Lactobacillus plantarum will be used throughout to refer to the group of organisms Bergey*s Manual (Breed et a l . 1957) includes in this single species. Individual authors may have originally used a term now recognized as a synonym. 5 the presence of a small number of lactobacilli on a variety of forage crops regardless of the season or geographical location. Kroulik et al. (1955a) characterized the predominant flora of fresh plants as chromogenic, aerobic, non-sporeforraing rods which hydrolysed starch but did not generally produce acid from carbohydrates. Thomas (1950) believed that these chromogenic organisms were facultative anaerobes, and he further concluded that typical silage bacteria were derived from organisms found on fresh plants through the action of specific lysins found in plant juice. Similar yellow colonies from cotton plants have been identified as Xanthomonas malvacearum by Clark et al. (1947). The role played by these organisms in the silage process is unknown. Kroulik suggested that they may aid in the development of the anaerobic conditions necessary for acid production. Since the type of fermentation may be a function of the number of bacteria on the fresh crop, Kroulik et al. (1955a) studied the effect of various environmental conditions on the total number of bacteria on alfalfa. They found that a substantial increase in the bacterial population accompanied an increase in the maturity of the plants, and that these aerobic, chromogenic bacteria increased in numbers if the crop was wilted. Variations in the bacterial population with the kind of plant, part of the plant, time of day and the season were less specific. As the season progressed, the population became somewhat more varied and included a number of lighter yellow, buff, pink and brown colonies. There are many coliforra-like bacteria on fresh grasses. Allen et al. (1936a) classified this group as Bacillus aerogenes gra minis because of their optimum temperature of 30°C, while Kroulik et al. (1955a) classified them as Aerobacter cloacae on the basis of their ability to liquefy gelatin. Clostridium snorogenes has been implicated as a spoilage organism in silages which evidence a marked degree of proteolysis. Small numbers of this organism have been found on fresh grass by Allen and Harrison (1937). The Chemical Composition of Fresh Forage Crops. Related to the production of silage, carbohydrates themost important constituents of fresh forages. are Since green plants contain relatively small amounts of the simple sugars, the storage carbohydrates must participate in the formation of the large amounts of lactic acid produced in the conversion of fresh crops to good quality silages. Starch has long been considered the principal reserve carbohydrate of plants. However, Laidlaw and Reid (1952) found very little true starch in grasses and clovers, but considerable amounts of a 2-6 linked levan type of fructosan. Percival (1952) concluded that this fructosan was the reserve carbohydrate utilized in the production of silage. Laidlaw and Reid (1951) 7 also chromatograph!cally identified glucose, fructose, sucrose, some oligo-saccharides, hemicellulose and cellulose in rye grass. According to Watson (1949), the total carbohydrate expressed as "starch equivalent" increases as the plant matures. On the other hand, Barnett (1954) quoted several reports which indicated that the fructosan content apparently reaches a peak at the flowering stage and subsequently decreases as the plant matures. It may be significant that the peak fructosan content corresponds to the time that the crop is most ideal for silage making. After photosynthesis has been retarded or stopped, plant enzymes continue the hydrolysis of reserve carbohydrate to simple sugars. Hence, if plants were cut and allowed to wilt in the field, especially overnight, there would be reason to expect an increase in the amount of simple sugars at the expense of the complex storage carbohydrates. The greater production of lactic acid observed in silages made from wilted crops has been ascribed to this higher level of simple sugars which are easily fermented. Brown (1893), Molisch (1921), Horn (1923), Ritschl (1929), Dickey et al. (1942), Stone (1943) and Speer (1950) all recorded sizable increases in reducing sugar content after wilting, with subsequent decreases in "starch" content. On the other hand, Wilson (194$) carried out exhaustive wilting experiments with several different crops and concluded that wilting did not 8 cause a measurable increase in fermentable sugar. During the wilting process, considerable carbohydrate may be lost due to complete respiration to carbon dioxide and water* However, a decrease in pH during wilting would indicate the formation of organic acids by respiratory enzymes, in which case the carbohydrates would not have been wasted. Although. Wilson (1%$} did observe slight decreases in pH after wilting, this decrease could be accounted for by an increase in carbon dioxide as postulated by Small (1946). Small has also listed the pH values of the juices of plants and of various parts of plants. Truog and Meacham (1919) studied the pH of plant juices which had been grown on limed and unlimed soil and harvested under various environmental conditions. The pH varied only within 0.2 of 1 pH unit regardless of the soil reaction and the amount of cloud cover when harvested. The juice was even allowed to stand for several hours without a change in pH being recorded. Watson (1952) recorded increases in the per cent fibre and dry matter as grasses mature, with a resultant decrease in the per cent protein. Fresh plant material does contain some acids. Annett and Russell (1908) have demonstrated the presence of small amounts of malic and succinic acids. Turner and Hartman (1925) added citric and malonic to this list, while Davies and Hughes (1954) were able to show the presence of acetic, lactic, 9 succinic, malic, citric and malonic acids by the use of chromatography. The malic acid content increased during the growth of the plant while citric acid decreased, but these two acids together accounted for 50 per cent of the plant acidity at every stage of growth (Davies and Hughes 1954)* In addition to the above acids, Hulme and Richardson (1954) found traces of quinic and chlorogenic acids in meadow foxtail and meadow fescue* Carbohydrate Metabolism in Silage* During the course of silage fermentation, the carbohydrates of the fresh plants are metabolised to carbon dioxide, alcohol, volatile and non-volatile organic acids. Previous to 1921, there was considerable disagreement over the relative role of plant and bacterial enzymes in carbohydrate metabolism. Babcock and Russell (1900) (1901) believed that plant enzymes were the main agents of fermentation because acids were formed even after the addition of antiseptics. Cn the other hand, 3sten and Mason (1912) believed that the rapid multiplication of bacteria and yeasts which occurred during the silage fermentation must have been responsible for the parallel increase in acidity, and they also recognized the fact that these acids were important preservative agents. Samarani (1913) and Lamb (1917) agreed that respiration of plant cells could account for the alcohol produced early in the fermentation, but that 10 lactic acid was produced from carbohydrates by microorganisms. Samarani also postulated that acetic acid arose from the respiratory oxidation of alcohol. Still other workers (Sherman and Bechdel, 191&) admitted that their results were inconclusive. In 1921, Hunter published conclusive proof that lactobacilli produced most of the acidity in a typical fermentation. He showed that the increase in acid production was always accompanied by an increase in acid-producing bacteria. No acids v;ere produced in forage treated with chloroform, but when heat-sterilized forage was inoculated with fresh silage juice containing large numbers of microorganisms, a normal fermentation was resumed. Peterson et a l . (1925) believed that alcohol in corn silage was also an end product of bacterial metabolism. As late as 1943, Pratolongo believed that "autolytic" plant enzymes produced lactic acid In silage, but the generally accepted view was recently summarized by Watson (1949), who stated that the respiration process results chiefly in the formation of large volumes of carbon dioxide. Watson (1949) quoted Virtanen as having said that respiratory losses as carbon dioxide are inhibited by the addition of mineral acids at the time of ensiling, but Watson1s own experience did not sustain this claim. Alcohol is usually present in silage between the levels of 0.2 per cent and 0.8 per cent. Determinations on corn 11 silage showed that 12, per cent of the alcohol was present as ethanol, 21 per cent as methanol and 7 per cent as propanol (Watson 1949)• Of the organic acids present in silage, acetic, lactic, propionic, but]rric and pyruvic acids are formed by the bacterial degradation of carbohydrates* Butyric acid may be formed directly from sugars or by the secondary fermentation of lactic acid* Valeric and other longer chain fatty acids are probably derived from amino acids by deamination (Barnett, 1954). Silage quality has often been directly correlated with the amount of lactic acid present, and inversely correlated with the amount of butyric and higher acids. In a summary of many years experience, Archibald (1954) has stated that good silage has a lactic acid content of 3 to 5 per cent or more and a butyric acid content of 2 per cent or less, expressed on a dry weight basis. Barnett (1954) believed that as much as £-9 per cent of the dry weight may be lactic acid, while Watson (1949) expressed the attainable limit of lactic acid as 2 per cent of the fresh weight, with 1 per cent of the fresh weight as the minimum for good quality silage. According to Barnett and Duncan (1953) even good quality silages may have a small amount of butyric acid. However, Barnett and Killer (1951) had shown that the presence of butyric acid in otherwise good silage did not affect its palatability, from which they concluded that although poor 12 silages have large amounts of butyric acid, their unpalatability stems from other side effects which accompany a butyric fermentation. Lactic acid is s much stronger acid than acetic or butyric; hence Its desirability in silage is a consequence of its ability to prc^ice a low pH. Virtanen (Barnett, 1954) had found that proteolytic spoilage did not occur when the pH was lower than 4*0 while the production of undesirable volatile acids by Clostridia and coliforms was inhibited below a pH of between 4 and 5. In general, good silage is characterised by a pH of 4.5 or less, the nearer to 4.0 the better {Archibald 1954). Barnett (1954) arrived at pH 4*2 as the critical value for silage preservation. Nevertheless, the pH is not always an indication of quality. According to Lind (1953)» lactic acid may be produced slowly in some silages, allowing a degree of butyric fermentation and proteolysis to occur before a suitably low pH has been attained. Spoilage of this type can be overcome by the addition of mineral acid at the time of ensiling, which lowers the pH below 4.0 immediately/*. McLean (1941) cautioned against the use of pH as an indicator of quality in overheated silages, which have very little nutritive value even though they may contain considerable lactic acid and have low pH values. On the other hand, legume silages rarel]/* attain the critical pH, although the quality may be good. Watson (1949) believed that high pH values of 13 legume silages are due to the presence of buffering agents, notably calcium, but he admitted that this question requires further research. The relationship between the pH and the various organic acids present can be illustrated by data presented by Watson (1949)♦ The amount of lactic acid increased from 0.1 per cent at pH 5.0 to 2.2 per cent at pH values below 4.0. Conversely, the total volatile acids decreased from 1.5 per cent to 0.75 per cent over the same pH range. Between pH 4*0 and 4*5* the volatile acids were chiefly acetic, but the increase in volatile acid between pH 4.5 and 5.0 was largely due to butyric acid. However, nolassed or acidified silage can exist at pH 5.0 with relatively less butyric acid than untreated silage. Common (1941) found that the buffering capacity of silage extracts was mainly determined by the ratio of volatile to non-volatile acids, and he has devised a formula relating the pH to the type of acids present. The relative amounts of volatile and non-volatile acids also accounted for the variations in the relationship between titratable acidity and pH observed by Pederson and 3a gg (1944). Stone (1943) has said that if the pH is not below 4.2 and there is no reserve sugar, the lactobacilli will attack lactic acid itself and the silage will not keep. This statement was made because it had been observed that poor quality silages made without preservatives or wilting had 14 little residual sugar, Archibald (1953) had also found more residual sugar in silages which had had preservatives added. In particular, the residual sugar in SO-) preserved silage was almost as high as the original sugar content. Archibald (1953) concluded that SC^ prevented the breakdown of sugars, but did not prevent the utilization of cellulose and other complex carbohydrates. Knodt et al_. (1952) claimed that the increased amount of residual sugar and the decreased production of lactic acid amounted to a saving in nutrient value in 302 treated silages. However, in 1930 Kirseh and Hildebrand (quoted by hatson 1939) had pointed out that lactic acid had almost as great a feeding value as glucose itself. Barnett (1954) listed the total hydroiysable carbohydrate content of several silages. All values ranged between 5.30 and 9.65 per cent of the dry matter, Petersen et al. (1925) detected approximately 2 per cent reducing sugar even after 146 days in the silo. They concluded that the reducing sugar was probably glucose since a glucosazone was obtained with phenylhydrazine, but it may also have been fructose. Nitrogen Metabolism in Silage. There are three principal aspects to the study of nitrogen metabolism in silage, namely: (1 ) the loss of tTcrude" protein,'(2 ) the conversion of "true,r protein, end (3) the relationship between the end products of nitrogen metabolism and silage quality. found that 14*7 per cent of the total nitrogen of the fresh plant was lost after 145 days in the silo, and their survey showed that four previous reports had listed losses of approximately the same order. Similar studies on hay crop silages were conducted by Taylor et, al. (1940) who recorded losses of total nitrogen (expressed as crude protein) in the range of 12.5 to 17*3 per cent. Seme loss is unavoidable since the bacteria which produce the acid necessary for silage preservation require nitrogen for growth. In fact, nitrogen may have to be added to crops which have a low content of natural protein. For example, Cullicon (1944) showed that a rapid fermentation of sweet sorghum could only be achieved by adding urea. However, Barnett (1954) has expressed doubt that the addition of urea to protein-rich crops would reduce nitrogen, loss unless it could be proved that silage organisms preferred urea to plant protein. Urea added to grass actually produced an unpalatable, high pH silage (Archibald 1946). Nitrogen losses vrere reduced when an earlier lactic acid fermentation was induced by crushing cr macerating the plants (Barnett 1954). Working with bottled grass sap, KeoPhersen (1952) found that 60 per cent of the nitrogen of fresh grass was contained in insoluble "true" proteins, where*-s the nitrogen cf the fermented product was only 60 per cent protein. The increase 16 in soluble nitrogen from 20 to 40 per cent was essentially all accounted for by increases in amino nitrogen and ammonia; hence, it was concluded that 25 per cent of the true protein Although Peterson et al. (1925) recorded a was hydrolysed* similar increase in soluble nitrogen in corn silage there was less proteolysis, since precipitation with tungstic acid indicated that approximately half of the soluble nitrogen was still protein in nature* IlacPherson. (1952a) noticed that the bulk of protein breakdown occurred in the short time required for the pH to fall to a value between 4.5 and 5.0. Ho, therefore, concluded that proteolysis could be reduced by rapid acidification with mineral acid. Virtanen had already proved that this could be done on a large scale using HJ1 (Barrett 1954). In Virtanen*s experience, no proteolysis occurred et pH 3 *6 . Watson snd Ferguson (1937) concurred. The concentration of HC 1 used could net itself hydrolyse various vegetable proteins (SelmaIfuse and Werner 1941)• Archibald (1946) found that losses of both true ana crude protein .‘ .ere generally less when carbohydrates vcere used as preservatives. On the other hand, IlacPherson (1952a) reported that proteolysis was unaffected by carDehydrate level. Protein losses were greater when the crop was allowed to wilt (Cwansen uni Tague 1917) (KacPherson 1952b) (autrey et al. 1947). Even if much of the protein is hydrolysed to amino 17 acid , there i s no groat l o s s i n feeding value, Kirsch and Jantzon (1933) showed that cattle responded equally well to diets of clover lay or clover silage which contained equivalent e mounts of nitrogen, even though the silage contained less protein and more non-protein nitrogen. However, breakdown of niaino acid is always indicative of spoilage. Deamination will produce ammonia and high molecular v:ai rht carhoxylic acids such as isovaleric (Lind 1953) • Decarboxylation produces other volatile basic trines. Archibald (1954) stated tliat the volatile base content expressed as ammonia •should net exceed 0.5 per cent cf the dry .ratter. The relationship between pH and velatils base content has been summarized by V.fatecn (1943). Where preservatives were used, the volatile ba coo a ocounted for less than 4 per cent of the total nitrogen at pH 4.0, increasing to 12 per cent at pH 5*0* In silages 'without preserve Lives, or in moldy silages, the volatile- base content or.: generally higher, amounting to as ranch as 25 per cent of the total nitrogen at pH 5.9. Tho relationship between silo,go quality and volatile base content h-: s been neosured in various ways. (1939) Watson estimated silage quality by the ratio of volatile base to amino acid. However, this ratio is misleading when up lied to silages prepared by the addition of minorcl acid, which have an unnc.to rally low a.,,cunt of s v;ino acid owing to the lower degree of initial proteolysis. Lind (1953) 16 proposed that the ratio of volatile base to total nitrogen be used to indicate the extent of protein lose. Barnett and Miller (19di/ demonstrated the close relationship between the pepsin digestibility of a dried silage sample and the ratio of soluble to total nitrogen. notably with overheated siia hj £ wii.■!_ In several oases, I t) L x O ^ ^ better indication of silage quality than the pH. Bender et al. (1941J discovered that ammonia production was negligible in silages which had received added carbohydrate in the form of molasses. They concluded chat fermentable earoonydru se suppressed astcnid lormation • Barnett (jl954) believed that the presence of fermentable carbohydrate directly inhibited the formation of deamineting enzymes irrespective of the pH. lie therefore reasoned that a high resiauai carbohydrate level may explain the anomalous position of those silages which have a high pH yet are of good quality. Hunter (1921) proved that oarbohydx\ate breakdown was due to bacterial action. Using the same techniques, he showed tha t plant enzymes were responsible for most of the proteolysis in normal silages. similar beliefs. itacPherson (1952s.) expressed By ensiling grass grown in a sterile chamber, habbitt (1951) was able to prove that both amino acids and volatile oases ocuii be formed by plane enzymes. On the other liana, Peterson _et al. (1925; believed that most changes in the nitrogenous constituents were due to bacterial 19 activity. Compromising, Barnett (1954) held that plant enzymes were responsible for the breakdown of protein to amino acids, but that putrefactive bacteria produced the excessive ammonia found in poor silages. Similar views had been expressed by Watson (1949)* Silage Microbiology. Almost immediately after crops are ensiled, there is an enormous increase in bacterial numbers. Working with corn, Peterson et al. (1926) reported a 100 fold increase in the first 24 hours, and a further 1000 fold increase during the second day. A maximum count of 100 billion organisms per c.c. of silage juice was recorded on the seventh day. Thereafter, there was a gradual decline in numbers although there were still 10 million active bacteria per c.c. after 132 days. Similar changes in the bacterial population of grass and legume silages have been noted by Allen et a l . (1937b), Stirling (1951) and Kroulik (1953). Normal bacterial development is modified by wilting and the use of various additives. Stirling (1951) and Kroulik et a l . (1955b) have shown that the bacterial population increases more slowly in silages made from wilted grass. However, since the wilted forage was well preserved, Kroulik concluded that the greater microbial activity in the high moisture silage was apparently not essential or even desirable. Stirling found that bacterial multiplication was stimulated 20 when the contents of the plant were made more accessible by maceration. Allen et al. (1937a) studied the bacterial changes in silages which had received additions of molasses, mineral acids, whey and bacterial inocula. They found that the numbers of microorganisms were considerably higher than those observed previously in normal grass silage. Some increase in bacterial activity was expected due to a higher moisture content, but their data also indicated that silages could support higher numbers of bacteria vdien more carbohydrate was added. Kroulik et al. (1955a) were also able to show that lactic acid bacteria were favored by an abundance of carbohydrate. It had been shown that the addition of mineral acids reduced, but did not eliminate bacterial activity (Cunningham and Smith 1939)- When Knodt et al. (1950) introduced sulfur dioxide as a silage preservative, they believed that all fermentation would be inhibited by the formation of sulfurous acid. However, Alderman et. al, (1955) concluded that the sulfite ion itself was bacteriostatic since some sulfur dioxide silages, although well preserved, had pH values above the critical limit of 4.0. Kroulik ejt aJL. (1955a) found that sulfur dioxide did not hinder the rapid development of lactobacilli in the first days of ensiling, but did inhibit subsequent bacterial activity. They concluded that sulfur dioxide selectively inhibited undesirable and unnecessary 21 bacteria. Peterson jgt al. (1925), Thomas (1950) and Burkey et al. (1953) all agreed that the typical chromogenic organisms of the fresh plant disappeared with the increase in acidity. Acetic acid was observed in silage before the main lactic fermentation had begun. Heineman and Hixson (1921) believed that this was the result of coliform activity. Allen et al. (1937b) identified the coliforras in grass silage as a type which would not grow at temperatures much above 30°C.; hence they believed that these organisms were destroyed by the normal temperature rise in the preliminary stage of the ensilage process. Burkey et al. (1953) claimed that it was the sharp decrease in pH which eliminated the coliforras as well as the chromogenic bacteria, since both groups persisted longer in silages made from wilted forages where acid development was slower. Hall et al. (1954) suggested that coliform bacteria may be susceptible to the lactic acid itself. According to Allen and Harrison (1936b), most of the lactic acid is produced by strains of Lactobacillus plantarum. These workers placed the majority of 152 strains of silage lactobacilli in this species on the basis of their homoferraentative degradation of glucose to inactive lactic acid and their comparative inactivity in litmus milk. L. plantarum has also been implicated in silages made with mineral acids (Cunningham and Smith, 1940), sugar beet pulp silage (Olsen, 1951) and sweet potato vine silage (Hall et al.. 1954). 22 The strains of L» plantarum isolated by Allen and Harrison (1936b) showed considerable differences in ability to ferment carbohydrates. Cunningham and Smith (1940) found that the L. plantarum in mineral acid silage did not exhibit the preference for sucrose and maltose over lactose which Orla-Jensen had originally ascribed to this species. Working with mixed cultures of silage organisms, Salsbury et al. (1949) showed that simple sugars were generally fermented more readily than oligo- and polysaccharides. Salsbury et al. (1949) believed that the comparatively low acid production from lactose by silage microorganisms might explain why dried whey was a less desirable preservative than molasses. However, Allen et al. (1937a) found that the lactobacilli isolated from silages made with whey differed from strains of L. plantarum previously isolated from untreated silages in that they could readily ferment lactose and rapidly produce an acid curd in litmus milk. Fred et a l . (1921) and Rogosa et al. (1953) believed that the L. plantarum species could be subdivided mainly on the basis of pentose fermentation. However, Bergey’s Manual (Breed et al. 1957) has upheld Pederson’s decision (1936) that the fermentation differences of various strains were not important enough, or constant enough, to warrant the creation of new species. Politi (1940) has proposed Lactobacillus sili as a new species for those silage lactobacilli which cannot be readily related to L. plantarum. 23 Peterson et al, (1925) found that most of the silage lactobacilli fermented xylose. These workers also reported that L. plantarum produced equimolar quantities of acetic and lactic acid from pentoses. The isotopic studies of Gest and Lampen (1952) haveshown that this is accomplished by a splitting of the bond between carbons 2 and 3 . In a later publication, Peterson et al. (1928) showed that members of the 1 . plantarum group could use protein as a source of carbon with the liberation of ammonia. Gorini (1940) believed silage was rendered more digestible by the formation of amino acids from protein by certain strains of L. plantarum which possessed "acidoproteolytic" enzymes, Peterson et al, (1925) believed that there was a second organism active in the production of lactic acid, but it was not clear whether this organism could produce a higher percentage of acid from a given quantity of sugar or whether there was a symbiotic relationship between this organism and the predominating type of lactobacilli. The possibility of a second organism being present in small numbers arose from the observation that a higher percentage of glucose was fermented by a large inoculum of silage juice than by a small inoculum which presumably contained only the preponderant type of lactobacilli routinely isolated in plate counts. Olsen (1951)-found Lactobacillus casei in sugar beet silage. Butyric and proteolytic spoilage has been associated with the presence of species of Glostridium. Beynum and 24 Pette (1935) (1936) isolated two types of butyric bacteria. Clostridium tvrobutyrieum fermented lactic acid and was inhibited at a pH of 3•5, whereas Clostridium saccharobutyricum fermented sugars but not lactic acid, and could exist below pH 3.5* However, SjostrSm (1942) demonstrated the presence of spores of C. tvrobutyrieum in silages made with mineral acids which had pH values as low as 2.7. Lactate- fermenting 3poreforming anaerobes have subsequently been isolated from silage by Martos (1949) and Rosenberger (1951)* Bryant and Burkey (1955) concluded that Clostridium tvrobutyrieum should be considered distinct from Clostridium butyricum on the basis of carbohydrates fermented. These authors also found that ATCC samples of Clostridium butyricum did ferment lactate in contradiction to the description of the species given in Bergey*s manual. Most reports of Clostridia have been based on spore counts. Rosenberger (1951) has emphasized the fact that Clostridia, form spores in an actively growing state only when under duress. He has, therefore, devised media for enumerating proteolytic and lactate-fermenting anaerobes in the vegetative state, and he has cautioned against accepting data based on spore counts, Lind (1953) was able to show good correlation between the amount of butyric acid in silage and the number of obligate anaerobes in the vegetative state. Allen et al. (1937b) and Bryant et al. (1952) have isolated proteolytic anaerobes from grass silages which were 25 similar to Clostridium sporogenes. Politi (1943) reported that some strains of proteolytic anaerobes in silage appeared to be closely related to Clostridium biferraentans. Thermophilic bacteria in overheated grass silage were identified as strains of B. subtlljs. although some of them were asporogenous (Allen et al., 1937a)* Etchells and Jones (1949) classified the predominant organisms in steamed potato silage as thermophilic, facultative anaerobes belonging to group X of the genus Bacillus. Several other bacterial species have been found in silage but their relative importance has not been determined. Allen et al. (1937a) made several silages with added whey which contained millions of Streptococcus lactis. Since very few of these organisms were isolated during the main fermentation, they were judged to be unimportant. However, Cunningham and Smith (1940) isolated S. lactis from silages made with mineral acids and in 1943 these authors recommended replenishing the microflora of steamed potatoes which were to be ensiled ivith a mixed culture of S. lactis and L. plantarum. Beynum and Pette (1939) found that acid-producing diplococci preceded the typical lactic fermentation, while Burkey et al. (1953) demonstrated the presence of pediococci, streptococci and lancet or diplo-short-rod forms in the early periods of storage. These workers agreed that these organisms died out because they could not tolerate the high acidity produced by the lactobacilli. 26 After the main fermentation, heterofermentative rods predominate since they are even more acid resistant than L. plantarum (Peterson jet al*, 1925) (Cunningham and Smith, 1940) (Olsen, 1951). Burkey et al. (1953) also found Corvnebacterium species in the later periods of storage. The presence of these different groups of organisms in predominant numbers depended on such factors as the kind of forage, its stage of growth, and other conditions of ensiling and storage. In general, the bacterial flora of silages made from wilted forage were more complex than highmoisture silages (Burkey et al. 1953)* Dobrogosz and Stone (1958) believed that pediococci were found in good silages and not in poor silages. Leuconostoc mesenteroides (Betacoccus arabinosaceas) has been isolated from sugar beet silages (Olsen 1951) nnd silages made with the addition of mineral acids (Cunningham and Smith 1940). Allen et al. (1937b) concluded that the miscellaneous flora of fresh grass, such as micrococci, yeasts and aerobic sporeformers exerted little influence 0x1 the resulting fermentation. CHAPTER III METHODS Sampling Procedure. Samples of the fresh olant material were taken at the instant the forage was being fed into the silo, when the silo was about three feet from being full. In upright silos, subsequent samoles of the fermenting silage were taken at a mean depth of three feet so that the silage samples would correspond to the samples of fresh forage. At this depth, the samples appeared to be sufficiently isolated from surface effects. Furthermore, samples from this depth could not have been unduly influenced by the percolation of juices from higher layers. In the more tightly packed bunker silos, a sampling depth of on® foot proved satisfactory. Silage samples were taken at irregular intervals, but as often as possible within 50 days after ensiling. All samples were packed in one quart polyethylene freezer bags, sealed, and immediately placed in a portable dry-ice cold chest for transport to the laboratory. At the time the samples of fresh material were obtained, the elapsed time between mowing and ensiling was estimated. The approximate composition of the crop, the quality of the stand and further details of the harvesting procedure were also 28 recorded. The temperature and relative humidity at the hour of ensiling were obtained later from the East Lansing weather bureau, but the number of days since the last appreciable rain­ fall and the amount of cloud cover had to be estimated locally. When silage samples were taken, the general appearance was recorded and the quality was categorized as good, fair, poor or overheated. Enumeration of Bacterial Populations A lOg* portion of finely cut material was placed in a Waring blendar with 90 ml. of physiological saline and agitated for 10 minutes at a slow speed obtained by setting the Powerstat variable transformer to deliver 40 v. on a 115 v. line, .Suitable dilutions were made, usually through one to one-hundred million, for the preparation of culture plates. All plates were counted after 5 days incubation at room temperature. Since the samples were not collected aseptically, counts of less than 1000 per grata were never reported. This parallels the procedure developed by Kroulik et al. (1955a). The interval between sampling and the completion of plating rarely exceeded six hours. The remainder of the sample was resealed, quick frozen, and stored in deep freeze for future chemical determinations. ^Total” aerobic count The total number of bacteria on fresh plants capable of 29 growing aerobically was determined with the tryptcne-glueoseyeast (TGY) medium which Kroulik et al. (1955a) had found to be superior for this purpose. TGY medium was also used to enumerate the aerobic bacteria in silage, since preliminary studies indicated that this medium gave consistently higher counts than the peptore-tamato agar employed by Stirling (1951). The ammonium lactate medium used by Stirling to count the gram-negative organisms on grass and in silage appeared to be less selective than claimed. The formula for TGY medium is given in the Appendix. MTotal” anaerobic count A total count of organisms capable of growing anaerobically on both the fresh and ensiled samples was made with anaerobic agar containing dextrose and Eh indicator (Baltimore Biological Laboratories) using Brewer anaerobic petri dish covers. The count for the whole plate was obtained by multiplying the number of colonies on 10 or 15 sq. cm. in the center of the plate by 6*35 or 4*23 respectively. Lactic acid bacteria count The population of lactic acid bacteria on the fresh plant material was determined with the lactobacillus selection (LBS) medium developed by Hogosa and co-workers (1951)» but at the author’s suggestion the 2 per cent glucose in the original formula was replaced by 1 per cent glucose, 0.5 per cent 30 arabinose and 0.5 per cent sucrcse. The complete formula is given in the Appendix. For enumerating the lactic acid bacteria in the silage samples, both LBS medium and the V5 agar developed by Fabian et a l . (1952) were used, since Rosen et al. (1956) demonstrated that LBS medium restricted the growth of a number of strains of lactobacilli. Since the IB medium was relatively non- selactive, its value as a counting medium was restricted to those samples where acid-forming bacteria predominated* The formula for VS agar is also listed in the Appendix. Characterization of Bacterial Populations For every sample of fresh and ensiled forage, eight isolations were made from each counting medium. To make the isolates as representative as possible, a rectangular area containing about eight colonies was marked off on a piste from a suitable dilution showing discrete colonies, and transfers were taken from all the colonies in this ores. Sack isolate from TGY, LBS and Vo media was replated once in the Isolation medium, whereas the isolates from anaerobic agar ^ere similarly purified by dispersion an tubes of the same medium which had been melted and cooled. Representative colonies -were subsequently transferred to tubes of storage medium. Isolates frcm TGY and anaerobic media were stored in the same media at 4°C while isolates from LBS and Y£> raedia were stored in micro assay culture agar 31 (Difco) at room temperature. All incubation times given in the following paragraphs refer to room temperature. Characterization of isolates from TGY medium Fermentation tests were conducted in purple base broth (Difco). The following substrates were used at levels of 1$: D-glucose, D-fructose, D-lactose, D-sucrose, D-maltose, D-mannitol, D-sorbitol, glycerol, D-xylose and inulin. The media were dispensed in 4 ml. quantities into 12 by 100 mm. test tubes containing inserts. Tubes of litmus milk (BBL), nitrate broth (Difco) and nutrient gelatin (Difco) were similarly prepared, but without inserts. All tubes were inoculated by needle from a 24 hour streak culture on TGY agar. This 24 hour growth was also used to prepare gram strains, employing the Hucker modification, and for the catalase test. The sulfanilic acid - alpha-napthylamine test for the reduction of nitrate to nitrite was performed after 3 days incubation. After 4 days incubation, observations were made for the production of acid and gas in the carbohydrate media, for changes in the indicator or any change in the physical characteristics of litmus milk, and for the liquefaction of gelatin. Pigment production on TGY agar was noted. 32 Characterization of anaerobes The anaerobes were classified as obligate or facultative by their zone of growth in tubes of anaerobic agar. Gas production and pigmentation were also noted. Gram stains were prepared from 24 hour growth in anaerobic agar. The obligate anaerobes were tested for proteolytic and lactate-fermenting ability by Rosenberger*s (1951) methods. The modified media employed are described in the Appendix. Characterization of lactic acid bacteria Fermentation tests were conducted in the modified microinoculum broth employed by Rosen et al. (1956). The following substrates were used: L-arabinose 1.0 per cent, D-glucose 1.0 per cent, dextrin 1.0 per cent, D-lactose 1.0 per cent, D-sucrose 1.0 per cent, D-mannitol 1.0 per cent, and salicin 0.5 per cent. The media were dispensed in 4 ml. quantities into 12 by 100 mm. test tubes. Arabinose was sterilized at 115 °C for 12 minutes. Tubes of litmus milk and indole-nitrite media (BBL) were similarly prepared. Inocula were prepared by transferring from the stock culture into 4 ml. of modified microinoculum broth. Yifhen abundant growth was observed (usually 24 hours) one drop of this suspension was added to each tube. This inoculum was also used for the gram stain and catalase test. 33 The test for nitrite was made after 3 days incubation at room temperature. Observations for acid production from carbohydrates and changes in litmus milk were made daily for 2 weeks. Chemical Determinations The frozen samples were thawed in a refrigerator for 2-4 hours, cut into l/4,,-l/2" lengths with hand scissors while still in a semi-frozen state, divided into 2 portions and immediately refrozen. One portion was thawed for the pH, Kjeldahl nitrogen and moisture determinations. Carbohydrate analysis was made on the dry residues from the moisture determination. The second portion was used to prepare the extract for the organic acid measurements and the extract for the total acidity, amino acid and volatile base determinations. The extractions were usually begun while the samples were still frozen. « K.ieldahl nitrogen Total nitrogen was determined on 5 g. moist samples. Digestion with sulfuric acid in the presence of and CuSO^ was continued for 90 minutes after the boiling solution had cleared. An excess of NaOH and a small measure of granulated zinc were added to the flasks before distillation. The distillate was trapped in standard acid, and the excess acid was titrated with standard base. 34 Results were expressed as crude protein, calculated as 6.25 times the nitrogen content, £H Three to 5 g. of material were added to 20 ml, of distilled water in a 50 &1. beaker. Preliminary trials indicated that the relative amounts of material and water were not critical. The mixture was briefly stirred and the pH was measured with a Beckman model G pH meter. Moisture The moisture content was calculated from the weight loss of 5 g. samples which had been heated in a dry oven at 105°C for 24 hours. The weighing dishes used had previously been tared after being dried to constant weight at 105°C and cooled in a dessicator. Carbohydrate The total hydrolysable carbohydrate was determined by the phenol-sulfuric acid method developed by Koch et a l . (1951). The experimental details outlined by Barnett (1954) were followed from the preparation of extracts to the development of the colored solutions. The color densities were determined with a Bausch and Lomb "Spectronic 20” . For the samples of fresh forages, the per cent carbohydrate was calculated by the formula given by Barnett from the optical densities of the standard and sample at 425 mp. The per cent carbohydrate in the silage 35 samples was obtained from the optical density at 490 mp. with reference to a standard curve which had a range of 10 to 60 micrograms per aliquot. Organic acids The micromoles of butyric, propionic, acetic, formic, lactic and succinic acids per gram of dry matter were determined by the Wiseman and Irvin (1957) chromatographic method. Total titratable acidity, volatile bases, and amino acids The Woodman (1925) modification of the Foreman method was used to measure the total acidity, volatile bases and amino acids. The experimental details have been outlined by Barnett (1954)* Extracts were made in 2 litre screw-cap bottles on a shaker which reciprocated through 5 inches 75 times a minute. Resu3„ts were expressed as micromoles of acid or base per 100 gm. of dry matter by an adaptation of the formula given by Watson and Ferguson (1937). Statistical Calculations The formula used for computing regression lines was: (1 ) y - My = a(x - Fix) in which My and Fix represent the true means of the independent and dependent groups of values respectively. The slope "a" of regression lines was calculated from the formula: 36 (2 ) Zxv - N *xz - M -*~~TixT2 Correlation coefficients were calculated from the formula: £xv (3) r* 553 V rx2 3RT sx^v Zy2 . *¥! Standard error was calculated from f r ^ &yj~] .Zy*- N . - a (4) TxZy ' - In formula© (2 ) (3 ) and (4)» x represents the variation from an assumed mean of a single item in the dependent variable group of values, and y the same variation in the independent variable group. N represents the total number of samples. Significance was determined by reference to the table of values for this purpose given by Snedecor (193$)* The term "highly significant" was applied to "r" values which were well beyond the point of 1 per cent significance. CHAPTER IV THE BACTERIAL FLORA OF FRESH FORAGES Introduction It has been shown that the number of bacteria on fresh forage fluctuates with the plant species, the part of the plant, the maturity of the crop, the time of day, the season, the amount of wilting and other details of the harvesting procedure (Kroulik et al* 1955a). It is also logical to expect that the bacterial population of fresh crops at the moment of ensiling may vary with the pH and carbohydrate content of the plant juice, the temperature, rainfall and other environmental factors. This present investigation was an attempt to determine the net effect of all these environmental factors on the number of bacteria on fresh forages at the time of ensiling. Ideally, there would be one or two dominant factors from which the number of bacteria could be predicted without appreciable error due to the sum of all other factors. It was hoped that the morphological and fermentative characterization of the predominant bacteria of fresh plants would lead both to their classification and to an understanding of the role these bacteria play in the ensuing fermentation. Similar studies of the lactic acid bacteria found on fresh 36 plants were made to determine whether or not they were typical silage types. In addition, the data obtained in this section will be used to investigate the relationship between silage quality and the number and type of bacteria on the fresh crop. Results A total of 33 samples of different forages were collected and analysed as outlined in Chapter III. All environmental statistics, plate counts on TGY, AA and LBS media, and other analytical data are listed in the Appendix. Bacterial counts are given on a dry matter basis to eliminate the partial effect of wilting, rainfall and humidity, which is due only to a change in dry matter content. It was found that the harvesting date was the most important factor affecting the number of bacteria on fresh plants. This relationship is illustrated in Figure I. The regression line and correlation coefficient for TGY medium was found to be: y = .033x + 7.63; r * .79 Similarly for AA medium: y * .030x + 6.63; r =* .75 .02£x + 4 *39 ; r ® .61 and for LBS medium: y * With 31 degrees of freedom, all r values are very significant at 1 per cent. JUNE JULY on fresh UJ Z < < < 00 o time < h O <0 h s i jl v w Ada do w v d 9 d 3 d i N n o o b i o v i a do 9 0 1 at the AUGUST m forages O £T The relationship between the number of bacteria of ensiling and the harvesting date. o 1. MAY 3 Figure AEROBES 39 40 Most of the 225 organisms isolated from TGY medium were chromogenic, the number of each color being given in Table I. All cultures were short, gram negative rods which often exhibited polar staining. Almost 60 per cent (132) of all strains did not produce acid in any of the carbohydrates tested. The remaining 40 per cent produced acid in various carbohydrate broths, but essentially none of the strains produced gas, even from glucose. A graphic presentation of carbohydrate preference of the 40 per cent which did produce acid indicated that the simple sugars were fermented most often (Figure 2). Both fermentative and nan-fermentative strains were found in each of the color categories of Table I. All cultures were variable with respect to nitrate reduction and gelatin liquefaction. The organisms capable of growing anaerobically were all facultative with respect to oxygen requirement. Although all cultures in AA medium were colorless, most exhibited yellow colonies when transferred to TGY medium, indicating that the bacteria isolated from both media were similar. When grown on the surface of AA medium, those isolates from both AA and TGY media classified as "dark yellow" in Table I retained their coloration. However, "light yellow" isolates from both sources were colorless even on the surface of AA medium. The bacteria isolated from LBS medium are compared to 41 TABLE I PIGMENTS PRODUCED BY THE BACTERIA OF FRESH FORAGES ISOLATED FROM TGY MEDIUM Color Number of isolates Per cent of Isolates Dark Yellow 80 35.6 Light Yellow 77 34.2 Yellow-Orange 20 a Orange 10 4.5 Pink £ 3.6 Brown 5 2.2 25 11.1 No Pigment. .9 42 40 NUMBER OF STRAI NS 80 UJ _i <0 z> Figure 2. SORBITOL 0 C a r b o h y d r a t e p r e f e r e n c e of the a c i d - p r o d u c i n g stra ins of the chro mo gen ic b a c t e r i a of fresh forages. 43 the typical silage bacteria in Chapter IX# Discussion Good correlation has been found between the bacterial numbers on fresh forages and the harvesting date, in spite of marked differences in other environmental conditions, plant species and crop composition. This indicates that the harvesting date exerts the controlling influence on the bacterial composition of the raw material from which silage is rrade. Since the regression lines for lactic acid bacteria and the predominating plant flora have similar slopes, the harvesting date apparently does not affect one group more than the other. Probably only a factor which selectively favored the lactic acid bacteria would be worth further investigation. Krouiik et al. (1955a) found that bacterial numbers increased with the maturity of the plant more than with the season. The data presented here neither deny nor confirm this, since harvesting date*1 is the sum of maturity and season, in which the contributions of the individual factors could not be separated. If a second controlling factor existed, it could be detected by a study of the deviations from the main regression line of bacterial numbers vs. harvesting date. For example, since Krouiik et al. (1955a) found that wilting greatly increased the number of bacteria on fresh alfalfa, it might be expected that wilted crops would have higher counts than the bacterial numbers vs. harvesting date relation would predict. Although this could not be demonstrated routinely, it was noted that all crops which were field chopped (essentially no wilting) had less than the minimum numbers of lactobacilli oven though one such sample was obtained as late as July 16. Exhaustive attempts to correlate these deviations with other factors were unsuccessful. The quantitative results showed that fresh plants contained almost as many bacteria capable of growing anaerobically as aerobically. Subsequent quantitative investigations inaicateu that the bacteria growing anaerobically v;ere tho same as those growing aerobically; the higher count on aerobic media may mean that some strains were strictly aerobic. These data are in accordance with the results of Thomas (1950) who found that the predominating chromogenic bacteria of fresh crops were facultative anaerobes, and at variance with the results of Krouiik et al. (1955a), who believed that they were strict aerobes. The predominating bacteria on fresh forage crops can, therefore, be described as gram-negative, rod shaped bacteria; which may show polar staining; which characteristically produce yellow, orange, red, or brown insoluble pigments, the hue often depending upon the nutrient medium; which 45 generally do riot ferment carbohydrates and which are aerobic to facultatively anaerobic. This description contains all the essential characteristics of the genus Flavobacterium (Breed et al. 1957). Since the preponderant number of bacteria on fresh crops do not produce acid, they are of little help in the preservation of silage. However, most liquefied gelatin. Where they predominate for long periods of time In silage made from wilted crops (Burkev et al. 1953), they may be responsible for some of the excessive proteolysis which occurs in this type of silage (MacFherson 3952b). Since they are facultative anaerobes, it can not, be the attainment of anaerobic conditions which causes their demise in silage. CHAPTER 7 THE CHEMISTRY OF FRESH FORAGES Introduction This study of the chemical composition of fresh forages was slanted to detect evidence of any fermentation which might have occurred up to the time the crop was placed into the silo. As in the previous chapter, an attempt was made to show what environmental factors control.led the amount of fermentation in this pre-ensiling period. Since plant enzymes are active in this period, this preliminary fermentation need not be correlated with changes in the bacterial flora to be significant in the silage process. Results It was observed that high-moisture crops hove lower pH values than low-raoisture crops at the moment of ensiling. This relationship is depicted in Figure 3, where the regression line was: y - -,22x * 7*46 The correlation coefficient was a highly significant -.69. 47 P co (U g O

Q < to H<* o in a well-preserved UJ Figure o > 52 a: 5> 6# £, 11, 12). There was no difference between the counts on LBS and V3 media. A comparative study of the fermentative capacities of 542 isolates from VC and LBS media is given in Table III. The largest single group fermented all sugars very rapidly and are very active in litmus milk. Ir: general, Group 1 and Group 2 of Table III form one division of types which can be called " M**"ru M'*vw fermentative*. In contrast,r the second division composed of Croups 3, Aj 5, and 6, are "weakly fermentative"• Ilost of the miscellaneous group differ from one of the other and spoiled, 00 o _i CO •— CO <•1 • <1 t 1 • in well-preserved, X Q. CO bacteria _J CO o > o ro o The population of lactic-acid over-heated silages. O UJ 13. CVJ o CD 39V 1IS CO HS3 d 3 3 0 WVU9 H 3 d 1 N D 0 0 ^ 318VIA 30 9 0 1 Figure X - 74 Cl rH CB 40 O H «M rH —zt O CA(H O IQ •H i>~ CM H c *\ + + s co p 03 03 0) R ■f + + O n CM i I I l i I I l + I I I I I l I I I I •f + I I + 44* t I h «S! *1 O CO M C M nO CMr^ C-'-rH ITS P'\\0 + + + + 4+ + + + + * + CO Q COS! M Jatas '-=5 O ^ • — s, + + I + g £ O C3 C3 «»-» O KJ H M m ► n CQ H >i pp C3.P ««4 ^4 ♦-3V—! W O 03 <3j PQ ir-HC3 OH o :- ^ H i—I O <*> --J- £>-rH o » a— CM -d- + + •f + + + + + 4* + + 4* + 03 CO 0) to rh !*! + CM r H o"\ + l> r H + + + iH r- + + + +■ f 4* + ■¥ + + + + + + + 4- + + + + + I cB cB <0 cfl T3 "Ci T3 TS I J Q H £3 ra < CO KJ> rH CT' rH •f + 4* + + + 4-f + + 4. ♦ 4+ > M 'S H O PS o < Qi •H &; ^ **^4 t-- •p s o a 03 •H O cfl Pi o 0) JO a (u pia, &q £si u as- * k 0 3 £3* ps o u a *H CB P* P 03 P> x s pi o 6 u o*yvO -jfr-dH rH I I I Pi rH nt-t"- © p fi C CJCw *rl *H“H CO p o c:p!c oo o *r4 *H »H *rl aJ jJ +J p y yy o cB ctl cfl ctf 03 0)03 03 Pi Pi Pi U 03 0303 03 03 l> > > > rH •rt *H -rt *H P PPPP K ose M M &3 i : r: i O o •H O o JO Prj Pi C3 p ct) rH Pt C O Pi o <0 CO •r4 rH B o c C a •H •H p Pi O •H 3 P •H a S X rH c! P r i 0 3 co C O • a CO iP •H «H*H CB *H J3 31 03 tuD Pi O O O 03 CO ix a, cu a > a itu ii n + + + + + + i* 75 groups In the fermentation of only 1 carbohydrate* 'Then these afore included with the group they most nearly resembled, the relative percentage in each group did not change significantly. iiorphological studies of the lactic acid bacteria were inconclusive. They included large diplocacci, pediococci, lancet-rods and diplo-rods. Each morphological typo was found In each fernentation group of Tabic III. All core catalase negative, and less than 1 per cent reduced nitrate. Table IV indicates that the fresh plants contained about the same number of highly fermentative types go the All attempts to show that the number or type of lactic acid bacteria varied with quality met w e f a x lure. Furthermore, the percentage of highly fermentative bacteria was the same throughout the storage period. A total of 171 obligate anaerobes was isolated. fermented lactate anc none hydrolysed gelatin. all the bacteria growing in silages with AA a high content of All Essentially aedia were ofthis type in butyric acid.These organisms were? also found in preponderant numbers in the spoiled silages before die butyric forment_tion hoc begun, ana in the bisulfite siluge where butyric acid never occurred. A few were also isolate sporadically from the well-preserved and overheated silages. The chromogonie organisms found on the fresn plants 76 TABLE IV COMPARISON OF THE LACTOBACILLI OF FRESH AND ENSILED FORAGES Fresh Crops Division Highly Fermentative Group Strains number Per cent Silages Strains number Per c* 1 5 7 91 17 2 14 19 71 13 26 Weakly Fermentative 3 10 14 20 4 4 13 16 70 13 5 1 1 34 6 6 4 6 72 13 25 175 21 M Miscellaneous 25 77 persisted throughout the storage peric-ci, hut they /.ere only For detected, when they were present in predoi ir.avit vr„tibers. example, in silo £, there was a rapid multiplication uf lactic ueid bacteria ari no ohrosogenic types w ere found on TGY medium after 24 hours. However, after 55 days in the silo, the count on LE3 median dropped below the count on TOY medium and chrorogsniu types vrere isolated a rain. In overheated silos, the number of lactic acid bacteria never exceeded the count on TGI medium and chrorcyenic organisms were isolated at every stage. Discussion Jio.it workers noted a tremendous increase in the mnuber of lactic acid bacteria immediately after ensiling, followed by a gradual decrease during the remainder of tie .storage period. Kronlik et ajL. (1955b) found that the numbers decreased acre rapidly than expected up tc about the 50th clay. Thereafter, a secondary increase in bacterial numbers occurred in most silages. Kroulik* s work with experimental silos Las been substantiated, by the results obtained from these full-sized silos. Burkey et al. (1955) and Dcbrogosz and Stone (1956), attempted to follow the changes in silage bacteria on a morphological basis. In so., a samplers, Ear key classified as many as 67 per cent of the bacteria as "variable rods” . Results of this present investigation were similarly inconclusive, a l t h o u g h nil of the v a r i ou s d e s c r i p t i o n s of bacteria given by Burkey were noted. However, when arranged in order of their abil.it}’ to produce acid fro:r. various carbohydrates and in litmus milk, they divided readily into two main divisions; those which fermented all the test carbohydrates readily, and thcnse which fermented on]}’ the simple sugars. All the lactic arid bacteria produced acid in pentose (arnblnose) as well as in glucose. This concept has been unchanged since the work of Peterson et al. (1921). The meet common lactic acid bacteria were t; or.e of Group I. which were active in oil cirbohydr; tes and litmus milk. This group is unlike the atriins of I.. G.ontarnm in normal silages described by Allen e£ al. (1937b). These same workers (1937m) found organisms with these ohara eterio t,ics on] y ir. r.i1 *go s which had received vuri ,s additives such, as -whey. Salsbury et a l . (19h°) also believed that normal oils re bacteria preferred the simp] e aag'-rs, whi *•*> 'jon Id mean the organi sms in Divi si on 2 wo.;Id he prod or,ilr.ant. Kronlik et ejl. (1955b) found that the lactic acid organisms on fresh plants were not "acid producers" like those in the initial d-v, of the errdlinu procs-s, and that the bacteria in the later stages of at:rage wore "weak -cid producers". oner-117 In this report, it was found that the fermentative cs no city of tve lactic, acid bacteria of fresh plants was as great as the silege bacteria, and that the types 79 of oact-ena. in si la re after 59 days of storage vrere n ot consistently different fro:;- those isolated earlier. There 'were no detectable differences in the number or type of lactic acid ba uteris between v/t11 -proserved silages and silages vhid eventually spoiled. Lactic acid bacteria developed normally in bisulfite silage, whereas overheated silages did riot undergo ?. lactic fomentation, and correspondinrly did r at possoc-:-. many lact5 acid T-actoi*1r , It is concluded that si la ;e <\- '.ality Is ’crelutei to the number or type of lactic acid bacteria present, evcept then overt e: ti ng cccurs. The p red osdn&nt anaerobe in butyric acid silage resembles lloatridi-m tvrohatvri c o c ii . it can a gar. since be ■■ ■ n— - tt - I sole ted --- c aunt i ■■■ K i T i r i i M u 3:1.. oe it •red oriinates. -X r a in t.h" v e g e h -ti v c> ot- However , counts the .... r of r,na erobor. 9 in an a n a e r o b i c re riot significant, an a n a e r o b i c a gar we. s o f t e n h i g h e r than on LBf* and Vf modi-i wi thout si u n i f y i n g t h e p r e s e n c e of o h l l y i t e venae rob p It s . is rirnifior- nt. Lb-1 ti n p r u d o 'dr ant c r g v i l s n acid mproduced. o r g a n i s m is fa vo re d dlostri • .• in spo il ed E v i d en tl y, t;-"v>Vtvri cam v/-: s 35.] are s b ef ore bu t y r i c the a s c e n d a n c y of th i s by the me t ho d of ensiling, rather than by f a i l u r e t o a t t i i n a Io n pH. The presence of C? outrf .■iiun t yr ohntrricur, in the Men!f: te a H a r e -.ay o v m tb?t bisulfite Inhibits its lactate, fermenting ability but not its growth. go Although the initial bacterial flora fluctuated with the harvesting date, silage quality was dependent only on what occurred after ensiling. It also appeared that some of the chromogenic bacteria of fresh plants could persist throughout the ensiling period in spite of the accumulation of acid. Whenever the count on LBS and Vg media dropped below the count on TGY medium, the chromogens reappeared. Since lactic acid bacteria also grow on TGI medium, they merely occluded the Flavobacterium species of the fresh plant. A medium which would support the growth of chromogens while restricting lactic acid bacteria might be of value in determining the exact fate of these organisms which comprised the predominant bacterial type on the fresh plant. Since quantitative and qualitative studies failed to uncover any specific organism which is exclusively present in high~ammonia silages, it is suggested that deamination is brought about by the normal silage flora in the absence of fermentable carbohydrate. CHAPTER X GENERAL CONCLUSIONS Fresh forage crops contained large numbers of chromogenic bacteria which were classified as species of Flavobacterium. and a smaller number of lactic acid bacteria which had fermentative capacities similar to the typical lactic acid bacteria of silage. As the harvesting date advanced from May to August there was an appreciable increase in the number of each bacterial type on the crops at the time of ensiling. However, examination of the resultant silages failed to show any relationship between silage quality and the number of lactic acid bacteria originally present. This is further evidence that the inoculation of silage with lactic acid bacteria would be useless if used in conjunction with unfavorable harvesting conditions, and unnecessary under optimum conditions. The predominant Flavobacterium of the fresh crop could apparently persist throughout the fermentation regardless of acid production. It might appear that they were eliminated quickly in those silages where the lactic acid bacteria multiplied rapidly, but they were actually only overwhelmed. They reappeared later in every fermentation whenever the count on LBS medium dropped below the count on TGY medium. By the time they were ensiled, high moisture crops had 32 lower pH values than low moisture crops. Since the apparent increase in acidity was not associated with higher numbers of acid-producing bacteria, it was assumed that acidproducing plant enzymes were very active in the high-moisture crops. Spoilage entailed the loss of a considerable amount of dry matter, while well-preserved silages appeared to gain in dry matter content because of moisture loss. Therefore, single dry matter determinations after the storage period would have shown that all spoiled silages had less than 20 per cent dry matter. From similar data it has often been concluded that a high moisture content causes spoilage. Since there was no correlation between the moisture content of the fresh crops and silage quality, it was concluded that high moisture silage is only the result of spoilage. However, high moisture crops are predisposed to spoilage because they can be over-packed more easily. Over-packing always led to spoilage. It was found that over-packing favored, or even caused, the development of lactate-fermenting Clostridium tvrobutvricum in the vegetative state immediately after ensiling and before butyric acid appeared. The addition of bisulfite inhibited the formation of butyric acid, but since the bisulfite silage was over-packed, Clostridium tvrobutvricum still appeared. A summation of the biochemical changes which occurred S3 in a spoiled silage is given in Figure 14. In Phase I ( 0 - 1 0 days), the initial drop in pH corresponded to the formation of lactic acid. During Phase II (10 - 25 days), the lactic and succinic acids were utilized by Clostridium tvrobutvricum. and butyric and propionic acids were produced. Since butyric acid is a weaker acid than lactic, the pH rises. Amino acids were deaminated in this interval, but this reaction does not contribute to the rise in pH because volatile base and high molecular weight acids are produced in equivalent amounts. Therefore, the pH remained low in silo 1 because no butyric acid was formed, even though there was considerable volatile base produced. Conversely, there was no appreciable deamination in silo 11, but the pH rose because butyric acid was produced. In Phase III (25 - 55 days), the pH gradually decreased. Because the pH was relatively high during this period, volatile bases were lost, leaving an acidic residue. In some cases, the level of volatile base did not drop in absolute quantity, but a loss of base could be calculated from the excess of amino acid lost to volatile base gained. When advanced spoilage occurred, the pH rose above 3, in which case essentially no acids were present, and most of the volatile bases had also been lost. As mentioned above, vegetative cells of Clostridium tyrobutvricum were found in over-packed silos from the x Q. 10 0 50 L A C T I C ACID MICROMOLES PER GRAM OF FRESH SILAGE B UT Y R I C ACID V O L A T I L E BASE 100 50 Phase I Phase II Phase III 30 40 50 DAYS Figure 14. Summation of the biochemical changes in a spoiled silage (silo 9)• 35 beginning of the fermentation, but there was no apparent factor which precipitated the formation of butyric acid. Conversely, deamination was precipitated when the carbohydrate level dropped below 1 per cent, but there was no noticeable alteration of the bacterial flora. Therefore, deamination and butyric acid formation could occur separetely; not necessarily together as in Figure 14. Butyric acid was produced in silo 11 because the silage had been too tightly packed, but deamination did not take place because the carbohydrate level remained above 1 per cent. The population of lactic acid bacteria exhibited two peaks during the storage period of both well-preserved and spoilage silages. Although this indicated a secondary fermentation, and other workers had noticed this same phenomenon, there was no corresponding changes in the biochemical composition of either type of silage. Furthermore, the lactic acid bacteria in the main and secondary fermentations were indistinguishable. CHAPTER XI SUMMARY Fresh forages have been found to contain a large number of chromogenic bacteria which were classified as species of Flavobacterium, and a small number of typical silage bacteria* The number of both types of bacteria on fresh forage at the time of ensiling can be predicted from the harvesting date. Bacterial numbers increased 100 times between late May and early August. However, the initial number of bacteria on the fresh crop bore no relationship to the final quality of the silage. The Flavobacterium group did not multiply during the silage process, but they evidently persisted in low numbers throughout the storage period in spite of the accumulation of acid. By the time they were ensiled, forages with a high moisture content had developed low pH values. Since there was no corresponding increase in acid-forming bacteria, it was concluded that plant enzymes may be very active in highmoisture crops prior to ensiling. When the silage was too tightly packed, Clostridium tvrobutvricum developed immediately after ensiling, eventually leading to the conversion of all lactic and succinic acids 37 to butyric and propionic acid. Forages with a high moisture content tended to undergo butyric spoilage because they were more apt to be packed too tightly. Excessive production of volatile base occurred when the hydrolysable carbohydrate content fell below 1 per cent of the total mass. Due to a high content of water and a relatively smaller concentration of carbohydrate, highmoisture forages were also more likely to undergo volatile base spoilage. The one example of bisulfite silage was also tightly packed; hence Clostridium tvrobutvricum developed. However, the bisulfite evidently inhibited the fermentation of lactic acid to butyric acid. Bisulfite also apparently interfered with the normal utilization of carbohydrate, since volatile base was produced even though the carbohydrate level remained above the critical level of 1 per cent of the wet weight. Quantitatively, the lactic acid bacteria in both wellpreserved and spoiled silages revealed two peaks; one in the first few days of ensiling and the second after 3 to 5 weeks. The significance of this secondary fermentation was not determined. Qualitatively, the lactic acid bacteria divided into a group which rapidly fermented all carbohydrates and was active in litmus milk, and a second main group which preferred simple sugars. Bacteria of each type v/ere isolated at every stage of the fermentation from every quality of silage. Overheated silages •were intermediate in quality snd pH. Acetic acid was the only acid produced in appreciable quantity; hence they contained only a few lactic acid bacteria. REFERENCES Alderman, G,, Cowan, E. L., Bratzler, J. W., and Swift, R* W, 1955 Some chemical characteristics of grass and legume silage made with sodium metabisulfite. J. Dairy Sci., 305-310. Allen, L. A., and Harrison, J. 1936a The character of some coliform bacteria isolated from grass and grass silage. Ann. Appl. Biol., 22, 533-545. Allen, L. A., and Harrison, J, 1936b A comparative study of lactobaciili from grass silage and other sources. Ann. Appl. Biol., 22, 546-557. Allen, L. A., and Harrison, J. 1937 Anaerobic spore-formers in grass and grass silage. Ann. Appl. Biol., 24* 143-153. Allen, L, A., Watson, S. J., and Ferguson, W. S. 1937a The effect of the addition of various materials and bacterial cultures to grass silage at the time of making on the subsequent bacterial and chemical changes. J. Agr. Sci., 22, 294-303. Allen, L. A., Harrison, J., Watson, S. J., and Ferguson, W, S. 1937b A study of the chemical and bacteriological changes occurring in grass silage. J. Agr. Sci., 27. 271-293. Annett, H. E., and Russell, E. J. 1908 The composition of green maiae and of the silage produced therefrom. J. Agr. Sci., 2, 382-391. Archibald, J. G. 1946 Studies in the chemistry of grass silage. J. Agr. Res., 72. 277—267. Archibald, J. G. 1953 Sugar and acids in grass silage. J. Dairy Sci., 26, 385-390. Archibald, J. G,, Blaisdell, M. L., Gersten, B., and Kinsman, D. l«i. 1954 Grass Silagej a reappraissx. mass* Aur • Exp. Sta. Hill., 477. Autrey, J. M., Knodt, G. B., and Williams, P. S. 1947 Grass and legume silage studies using 2-quart glass jars as miniature silos. J. Dairy Sci., 22> 775-785. Babcock, S. M., and Russell, H, L. 1900 Causes operative in the production of silage. Wis. Agr. Exp. Sta. 17th Ann. Rpt•, 123-141. Babcock, S. M., and Russell, II, L. 1901 Causes operative in the formation of silage. Wis. Agr. Exp. Sta. 18th Ann. Rpt., 177-184. Barnett, A, J. G. 1954 Press Inc. New York. Silage fermentation. Academic Barnett, A. J. G., and Duncan, R. E. B. 1953 The volatile fatty acids in fresh and infermented kale. Biochem. J., 53. xxx. Barnett, A. J. G., and Miller, T. B. 1950 Studies on the soluble nitrogen content of dried samples of grass silage, J. Agr, Sci,, ^0, 50-54. Barnett, A. J. G., and Miller, T. B.1951 The laboratory determination of the protein digestibility of silage. J, Agr. Sci., itl, 317-321. Bender, C. B., Bosshardt, D. K., and Garrett, 0 . F. 1941 Some chemical changes occurring in molasses grass silage. J. Dairy Sci., 2 1 4 7 - 1 5 1 . Beynum, V. J. van^, and Pette, J. W. 1935 Zuckervergarende und Laktatvergarende Butteraaurebakterien, Zentr. Bakteriol. Parasitenk,, Abt. II, £2* 198-214. Beynum, V. J. van, and Pette, J. W. 1936 Buttersaure Garung und MilchsauregarungSilofutter. Zentr. Bakteriol. Parasitenk., Abt. II, 94. 413-433. Beynum, V. J. van, and Pette, J. W. 1939 Die^Stabilitat der G^rungssilage. Zentr. Bakteriol. Parasitenk., Abt. II, 99, 353-374. Breed, R. S., Murray, E. G. D., and Hitchens, A. P. 1957 Bergey’s manual of determinative bacteriology, 7th ed. The Williams and Wilkins Co., Baltimore, Md. Brown, H. T., and Morris, G. H. 1893 A contribution to the chemistry and physiology of foliage leaves. J. Chera. Soc. (London), 62, 604-677. Bryant, M. P., and Burkey L. A. 1955 The characteristics of lactate-fermenting sporeforming anaerobes from silage. J. Bact., 21» 43-46. 91 Bryant, M. P., Kroulik, J, T., Burkey, L. A., and Wiseman, H. G, 1952 A study of anaerobic bacteria present in grass silage. Bacteriol. Proc. 1952. 21. Burkey, L« A,, Kroulik, J. T., Bryant, M. P., and Wiseman, H. G. 1953 Bacterial activity in forage-crop silage as indicated by the predominant groups or species of bacteria at different periods of storage. USDA-BDIInf-154. Clark, F. E., Hervey, R. J., and Blank, L. k, 1947 Occurrence of cotton fiber contaminated by Aerobacter cloacae. USDA Tech. Bull., 9351. Common, R. H. 1941 The buffer curves of silage extracts. Analyst., 66, 407-412. Cullison, A. E. 1944 The from sweet sorghum. J. use ofurea in making silage Animal Sci., 2> 59-62. Cunningham, A., and Smith, A. M. 1939 The microbiology of silage made by the addition of mineral acids to crops rich in protein. I.Quantitative chemical and bacteriological data. Zentr. Bakteriol. Parasitenk., Abt* 1Q2» 394-406'. Cunningham, A., and Smith, A. M. 1940 The microbiology of silage made by the addition of mineral acids to crops rich in pi’otein. II.The microflora* J. Dairy Res. 2, 243-265. Cunningham, A., and Smith, A. M. 1943 Observations on the microbiology of potato silage. Proc. Soc. Agric. Bact. (Gr. Brit.), 1943. 16-18. Davies, C. W., and Hughes, R. B. 1954 The organic acids grass extracts. J. Sci* Food Agr., 200-205* of Dickey, 44. Mabbitt, L. A. 1951 The role of plant cells in the ensilage process: an approach to the problem. Proc. Soc. Appl. Bact., 14. 147-150. MacPherson, E. T. 1952a Changes in nitrogen distribution in crop conservation. I. The rate and extent of protein breakdown in ensilage. J. Sci. Food Agr., 2* 362-365* MacPherson, H. T. 1952b Changes in nitrogen distribution in crop conservation. II. Protein breakdown during wilting. J. Sci. Food Agr., 2* 365-367* I4artos, V. 1941 The lactate-fermenting anaerobic and facultative organisms in silage. J. Bact., £2, 140. 94 McLean, W. 1941 Note on the significance of the pH determination in the evaluation of quality in silages. J. Agr. Sci., 21, 518-521. Molisch, H. 1921 tfber, den Einfluss der Transpiration auf der Yerschwinden der Starke in den Slattern* Ber. Bot. Ges., 29, 339-344. Glsen, E. J. 1951 The fermentation of sugar-beet pu IP silage. J. Gen. Microbiol., 817-829. Pederson, G. S. 1936 A study of the species Lactobacillus plantarum (Orla-Jensen). Eergey et al. J. Bact., 31, 217-224* Pederson, C. S., and Bagg, J. V. 1944 The cause of variation in the relationship between titratable acidity and hydrogen ion concentration among lactic acid bacteria. J. Bact., 48. 559-566. Percival, S. G. Y. 1952 The carbohydrate constituents of herbage. Brit. J, Nutr., 6 , 104-109. Peterson, W. H., Hastings, E. G., and Fred, E. B. 1925 A study of the principal changes which take place in the making of silage. Wis. Agr. Exp. Sta. Res. Bull. 6l. Peterson, W. H., Pruess, L. M., and Fred, E. B. 1928 The proteolytic action of certain lactic acid bacteria. J. Bact., 1£, 165-171. Politi, I. 1940 Researches on lactic-acid bacteria. I. Study and systematic position of acid bacteria of ensiled fodders. Ann. Microbiol., 1, 65-84. Politi, I. 1943 Researches on anaerobic sporogenous putrefying bacteria. Ann. Microbiol., 104-109. Pratolongo, tJ. 1943 Biochemical researches on ensilage. Sci., 2Z» 493-502. R. 1st. Lombardo. Sci. S. Lett. Ritschl, A. 1929 Weiterer untersuchungen liber das gegenseitige raengenverhaltniss der Kohlenhydrafce irn Laubblatt unter verscheiden Aussenbehengungen. Bot. Arch,, 26, 349-384. Rogosa, M., Mitchell, J. A*, and Miseman, R. F. 1951 A selective medium for the iviolation and enumeration of or.._l lucloQucill^« 1. Dental, noa•, 3o« 082—6b")• 95 Rogosa, M., Wiseman, R. F r, Mitchell, J. A., Disraely, M. N., and Beaman, A. J. 1953 Species differentiation of oral lactobacilli from man including descriptions of Lactobacillus salivarius nov. spec., and Lactobacillus cellobiosus nov. spec. J. feet., 6j>, 681-699. Rosen, S., Ragheb, H. S., Hunt, H. R«, and Hoppert, G. A. 1956 A comparative study of oral lactobacilli isolated on two different media from the Hunt-Hoppert rats. J. Dental Res., 291-293* Rosenberger, R. F. 1951 The development of methods for the study of obligate anaerobes in silage. Proc. Soc. Appl. Bact., 1/fc, 161-164. Salsbury, R. L., Mather, R. E., and Bender, C. B. 1949 Various carbohydrates as energy sources for some mixed cultures of silage organisms. J. Dairy Sci., 32. 901- 906. Samarani, F. 1913 Studi intorao alia conservazione dei foraggi alio stato verde. Bol. Min. Agr. Indus, e Com. (Rome) ser. C, anno 12. 87-103* Schmalfuss, H., and Werner, H. 1941 Sauerbindung und Eiweissabbau in Garfutter. Zeit. Tierernahr* U. Futtermittelk., 6, 96-96. Sherman, J. M., and Bechdel, S. I. 1913 silage. J. Agr. Res., 12, 539-600. Corn-stover SjostrBm, G. 1942 The presence of lactate-fermenting and lactose-fermenting butyric acid bacteria in varieties of A.I.V. silage. K. Lantbruks-Akad. Handl. Och. Tidskr, (Stockholm), 81, 451-453. Small, J. 1946 New York. pH and plants. D. van Vostrand Go. Inc., Snedecor, G. W. 1938 Statistical methods. Iowa State College Press, Ames. 1st ed. Stone, R. W., Bechdel, S. I., McAuliffe, H. D., Murdock, F. R., and Malzahn, R. G. 1943 The fermentation of alfalfa silage. Pa. Agr. Expt. Sta. Bull. 444. Speer, W. 1950 Ensiling withered green fodder. Forschung., 2, 111-114. Stirling, A. G. 1951 laboratory silage. Landw. Bacterial changes in experimental Proc. Soc. Appl. Bact., 14, 151-156. 96 Stirling, A. C, 1953 Lactobacilli and silage making, Proc. Soc. Appl. Bact., 16, 27-29. Swanson, G. 0., and Tague, E. L. 1917 Chemical studies in making alfalfa silage. J. Agr. Res. 10, 275-292. Taylor, M. W., Bender, C. B., and Russell, W. C. 1940 Effect of ensiling upon the composition of forage crops. M. J. Agr. Exp. Sta. Bull., 633. Thomas, R. C. 1950 Microbiology of silage. Expt. Sta. Res. Girc. 2. Ohio Agr. Truog, E., and Meacham, M. R. 1919 Soil acidity: II. Its relation to the acidity of the plant juice. Soil Sci., 469-474. Turner, W. A., and Hartman, A. M. 1925 The non-volatile organic acids of alfalfa. J. Am. Chem. Soc., 47. 2044-2047. Watson, S. J. 1939 The science and practice of conservation: grass and forage crops. The Fertilizer and Feeding Stuffs Journal. London. Watson, S. J. 1949 The chemistry of ensilage. Industry., 1949. 699-703* Chem. and Watson, S. J., and Ferguson, W. S. 1937 The chemical composition of grass silage. J. Agr. Sci., 27. 1-42. Wilson, J. K. 194$ The pH and sugar content of forage crops before andafter wilting as related to the making of silage. J. Am. Soc. Agron., 40. 541-552. Wiseman, H. G., and Irvin, H. M. 1957 Determination of organic acids in silage. J. Agr. Food Chem., j>, 213-215. Woodman, H. E. 1925 Critical examination of the methods employed in silage analysis, with observations on some special chemical characteristics of "sour” silage. J. Agr. Sci., !£, 343-357. APPENDIX 95 LACTOBACILLUS SELECTION MEDIUM Rogosa et al., (1951) Trypticase (BBL) 10 g * Yeast extract (Difco) 5 g. k h 2p o 4 6 g- Ammonium Citrate 2 g- Glucose 10 g* Sucrose 5 g* Arabinose 5 g- Tween 50 1 g* Sodium acetate {MaC2H-502 . 3H20) 25 g* 1.32 ml Glacial Acetic Acid (99.5$) Agar 15 g* 5 ml *Salt Solution to 1 liter Distilled Water adjusted to 5.40 pH Do Not Autoclave *Salt Solution MgS04 • 7H2O 11.5 g- MnS04 * 2H20 2.4 g* FeS04 . 7H20 0.6 8 g- Distilled Water to 100 ml. 99 V8 MEDIUM Fabian et al. (1953) Tryptose 10 g. Glucose 5 g. Agar IS g. Filtered V# Juice 125 ml. to 1 litre Distilled Water 0.1 g. Brom Gresol Green adjusted to 5.6 pH TGT MEDIUM Kroulik et al. (1955) Tryptone 5 g. Glucose 1 g- least Extract 3 g« Agar Distilled Water pH 15 g. to 1 litre adjusted to 7.0 100 MEDIUM FOR THE ANAEROBIC FERMENTATION OF LACTATE Rosenberger, (1951) Sodium lactate syrup (70%) Sodium acetate Yeast extract (Difco) (NH4 )2S04 p-amino-benzoic acid 14.3 ml. 8.0 g. 20.0 g. 0.5 g. 100 ug Cysteine hydochloride 0.5 g. fhioglycollic acid 0.5 ml. Resazurin 5*0 mg. KnCl2*4H20 0.15 g* NaCl 2.5 g. K2HP04 1-0 g. (NH4 )2Mo04 0.005 g. FeS04 .7H20 0.04 g. MgS04.7H20 0.80 g. Agar Tap water pH 2,Q 61 liter adjusted to 6.0 MEDIUM FOR THE STUDY OF ANAEROBIC PROTEOLYSIS Rosenberger, (1951) Trypticase (B3L) 10 g. Phytone (BBL) 5 g* Gelatin 120 g. Yeast extract (Difco) 10 g. Cysteine hydrochloride 0,5 g* Resazurin 5*0 mg. Tap water to 1 liter pH adjusted to 7*0 WATER AGAR Rosenberger, (1951) Thioglycollic acid 0.5 ml* Cysteine hydrochloride 0.5 g* Agar Resazurin Tap water pH 10 S* 5*0 mg. 1 liter adjusted to 7*0 102 o rH O O O un H to CO o o AT THE 4 f—I O iP on UN -4 o Q On o O C- CM -CO ON ON CM C*> CM ITS O O o rH O rH O 0 O Q O K-) CO o 4 0 P» ON O O O NO UN ON nQ UN -4 CM UN O ■ • ON UN • UN O v ON 4 -4 -4 -4 NO • UN CO t> ON ON -4 ON • NO • UN -4 » UN to O* -4 ON • O to xO xO nO © £3 UN 3 t> NO r~H E"- On -co O 4 O UN C3 f-H o\ CO O V O o ON O ON o R- un C*- CM O UN ►4 O Q O NO o ON to o O OF THE ON no ON CM ON nO O Q ON on o ON CM on rH NO on O O ON r4 O Q ON ON on t'» C3 NO on O ON >4 O Q CM to V >1 3 ON V * CM O O vO rH to o• un nO R- NO O UN O Q CM O to xQ nO CM • o to * C'- # ON V NO ON NO • UN 4 ON to to o to CM • ON CM rH • ON o ON • o ■00 CM ON ON CM CM • CM to ON • O On « NO UN CM 1—1 4 NO CM NO • to • to • UN r-" • -4 • o• • ON O CM • ON 4 CM -4 UN O to CM • O* On C^ On O- • • ON NO ON 4 CM ON • UN • O C^ NO On rH to • UN « r"l to ON UN ON UN CM O ON • NO Cn•H • UN CM ON • CM CM • O • • ■4 NO UN • 4 ON On • 4 co * © UN O O ON on rH C"- OON ON o on UN NO ON V O CM NO • ON ON • UN to i—f fr- 4 n O UN to O on o ON ON nO NO ON • -4 nO 4 O ON FRESH CROPS rH to nO CM ANALYSES -4- O o o TIME OF ENSILING o o nO NO ON On • * p w P O i—I &4 o * S-. © «H O © rH P S © © P s CO © Q © a a © >> P E-« P © © o P o E 3 X o W TJ 3 o r*H O © > u P •rH © © >> © Q P © x •H P Cw O © O T) O o •H P, a o o X P © 4^ 4 © P 3 O X T3 © P P P p 3 O O 3 O o 3 rH © rH c u © P © u © •H © © o P o © © © P •H P 3 © P p p o © QQ o o 00 0Q © RR a © X a p a •H © P o P ru T3 X o X p © o 103 O O' i—1 nO TIME in iH IA CM O - Isv£> CM M3 CM nO O rH o rH rH O O CO CO o o • ON E>nO • C"- to On • to CM in • C^- nO On m ON CM • 4 CM on \/ in ON • to CO HI CM H CM CM O O 4 rH 'O H CnO r-> nO H On CM nO CM vO O 4 o «H O H rH rH O 4 3t H H H © CO o CM CM • nO CO co m Cn• NO NO 4 cn NO • 4 NO * in NO 4 • On nO i—I NO • o- 4 to • m un m • C"NO i—1 O' • 4 «H O MD • O' nO NO ♦ o- O to • in to • o to On CM • cn NO un * H rH -4 in • in cn -4 # O' CM rH • to O rH • NO ON 4 • in On Cn• 4 [>to • c*- o o on On O • m CM TO • in NO ■to O • in CM r- NO CM • m • r- o • NO O to • m i—1 to • 4 NO in in • in to O o C'H to ON in O H m O »-3 • 03 -C p o H Cxj o © Q • Cm O s © £h P •H G •H 3 p (X O o *d G o H O CO * * p © © g •H > Ct) as g © o •H CO CQ >9 0) Q «G O *H P •H © O ct) w a ex o o - —» U) u p o «H o 33 —• TJ o X) ® p p © rH S3 o ■—' P P G O o G O O H © •H G © i—1 © •H G © P P G O « CO G o © m • in 4 • CM 9 • nO • un <—» < s CL, O p 23 P O © O G G © 104 O' CV CO CV o rH 0 0 rH rH vO rH O CV CA rH VD >H UA rA 4 rH O O O CA rH 0 rH 3c rH CO 0 UA ITS 4 O CO CO CV CV >> rH £ *~D O CA O rH o CV O O UN rH 0 cv O O 4 rH vO r- OA c- l> j> - rA C~ rH _4 4 vO 0 0 UA 4 rH 4 \ rH 4) I* C!5 4* CO CO rH CV CD £ *"3 O CA VO rH t>- c- nO vO O rH rH O O • NO 4 CV • cn~ nO CA vO ON • 0 V to rA O • to NO 4 • NO CA O • C"UA NO CV 0 UA O • UN ON CA CV • -00 O On • UA • ON \D 4 CA • 4 CV ON • 4 vO r•0 ■ 0 rH to • UA CV rA • ON nO to C'• 4 O CV • 4 • 0 0 4 C"• CA ON 4 • UA £>ON • to sO CA O • UA CV r• 4 rcv • rH CH O' CA • 4 £> CA • 4 O C”• ON nO 4 UA • 4 4 CV • UA O 4 On » CA to C"• UA to 9 On 0 UA CA O • ON O ON • to 4 ON * 0 UA nO • UA to On • UA NO 0 • to ON CA • 0 rH CV • 4 [> O • \0 CV NO • to UA to • O- nO 0 ON ON • 4 nO nO 4 O * 4 to • 0 UA 4 CA • ON 0 O UA • UA NO CV • ON • "co CQ ■4—i' J* -—. ■—' ■ —* P P # * p to X P O rH for O £Jj Q «h O © Q & co CO © 43 CO 0 U ■ PL s © £-1 © e •H * •H c >* -P •H T3 «H s S3 SC © > O O TJ S3 O rH O 0 0 £ •H cQ 04 © O £ DO CO © Q * £ O •H P •H CO O a s 0 0 CO © > !H © X Cm O xs 0 4: p © SI O UA • to NO UA • to to ON • M- to • UA vO nO CV CA O • UA 0 CV 4 • O • SO O to • CA O C• UA tc O CV C^• 4 0 rH » to to m 0 to to UA • ON CV THE FRESH CROPS AT THE t> CV CA rH • UA -4 O • UA OF ENSILING o 4 UA * nO nO r-» CO P £ r-' O £ S3 O O £ S« O O rH © ■rH O as ■H rH © •H T3 © U U P P © CQ aS CQ p rH •H 3* ffl 13 UA to 9 OO- nO • CA C3 S3 O se — - 1 —1 O C'NO • UA •to O • 4 © O •—" © P © U £ P © P O © CQ X cx, © •H O Si —■' £ •H © P O • M Qu © U xs !>» r£ O JO £ © O C5 0 OF TIME THE AT CROP © © 0 0 s CD o U H so © © a > t o 3 Wt) a fcScd © O P o © Si © rH C» o ^0300 p U8 Ua a S CO £ •H a o £3 O *d *H © P a 05 o Cu s o o * FRESH 0 •*2*— * -__ «■ S !U © rJ X G NO rA rA O rH CV o rA to CO • CO x * p co © p .— v Cxi ,—, 0 ._« © • Q >» CL p O O G •H © © Ci G •H © © O Sa iH © O a TJ S3 * g 0 *H P *H • UA • HO CA to CO X p? M O 6* • +3 a CO 3 © U o X P o O 1—I cm X O nl > u © TJ TJ O O © CL .£ P s P© rH •H 0 0 2! 35 P c <-« O CV to NO — • 0 r- ($) iH rH C^Nj 3 ’s o (AA) CV 5*37 O NO On ON . 4*79 CV •CO nO • UA 4.40 o UA !>- o o cv rH to CO Carbohydrate o NO * A CA 5.38 NO NO• rH 5.47 CV ca o W) o P «c UA rA rA r- tc 4.52 r •H •H ■H Si 5-4 U 0) © © P P p O O o X CO OS CO CO CS CL o o 0) s »<1 p co •H o £ Protein THE Hr ca ANALYSES TJ t J>H iH H i <3 P 0 a© Cm XI »-l O a p © © p o p 01 iH H 0 H rO o NO •H G CO bfl O rH CO cO CO TJ £3 C D O * * * CO V to i—t to • • l> UA nO 1 CO ON • UA 03 £a • ON 0- CO O co • tr\ UA CO • 03 UA nO rH m 0rH o UA o 1 to UA 1 1 1 CP „_„ 03 *33 .— -CO fc> -- o e-i < «t * -P G TJ CD rH •H CO G

■ CO a P o o rH CO •H G CD P o ca cq P a p o o 1— 1 ct) •H G CD P o CO 03 p a p o o rH Ct) •H G CD P O Ct) 03 a> G O, X p C G O o rH Ct •H G CD P O ct 03 ^“A w a* G p p 00 a CO •rH o S3 CD 3 -— * P ct ~o G O TJ >> q> g •H J2 rH O 0 p Ct P P G O O G ct CU O •—A •j—1 r"A •=--H ^GL, -r-*. SC 3 s CD G, CO V—* TJ •H a) O P TJ •H a) O Q) t> ct rH •rH •H O G P rA Cfl G rH •H P O a P r» -=c cq (0 1 —' •"A s •A '"A tSTl TJ •rl O Ct O •H G O •H a o G a, ■'-A Gk "S 3 -' 3 TJ •rH O Ct TJ •rl O CO o •H P CD O < o a G O CG T) •H O CD O *H P O ct G3 TJ •H O CO o •H G •H O V P 00 G CP H § cO •H G CP +3 CJ cS us 4 ■ M3 Os -4 • M> O CM • 4 rH CM • Os sO M3 CM • u\ CM • CM vO O 4 4 CM US Os o CM o CM • UA o r• -4 O CH • M3 to Os *■ 4 »H CM • 4 4 CH • 4 4 -4 CM • 4 rH • CM o 4 CM OS 4 CM pH to O• us M3 Os • v3 vO 4 • )—1 rH C'* M3 Os CH • 4 to M3 • to o 4 os • OH 4 rH rH US pH CM SO ■CO o o us 4 • sQ oM3 • MO u\ C"• CM • H 4•oo rH O CM r4 o o rH H O vO • to 4 C'• to CM • os pH rrH sO rH «-i us o o CH £4 4 Os • OH 4 O rH to 1 t 1 O to > o »«—• < «=S -p ti 3 /-\ -p c 3 p TJ © rH -rl a> CJ 0) CQ > oi G> o rH CO *H CD p O cfl (32 o o I—i cO *H C-i 0) p o cO cq S3 d c o 1—1 r—( cO ■H JH © P O © 02 cO p O cd 02 <+\ 1 —— S3 CD 'SR © u d *H to •H O p rH Al P- c O a CD P u (X. P CO P TJ !>i P o p u CO C3 3 T3 *H O CO rH 03 P O 6i ANALYSES 300 0 o —~ Ar*i* --S b? P S3 d o o *H u CD o to 3 © CO co p CD rH *H __ i *£i 3 . —a« ■ n *rt O TJ H ct O o CO P o <0 rH 0 > •H E «C a H rS P d CQ 3 TJ *H O 03 o •H P © O • IN­ (V • nO rH O CH vO Cn• UN UN rH 4 4 On NO rH • 4 vO • CV 4 CV 4 UN CH On O rH 4 CH 4 Oi—1 • 4 ON -4 • O 14 On On • CH CH • CH O ■co CH UN CH NO ON O UN • CV c- rH ON • CH NO CV rH to CV O CH CV On 4 i— 1 • CV 14 rH CV CV • P s H O o P G G o o iH © •H rH © •H © P © p u O © CO CO - P £ G O o iH Cl) •H £ © p O V O CH UN to rH £> NO CV sO CV • 4 • • -— • rH © •H © G 3 u © P o © m ■ —- X Cu P 93 •H O S3 p •H © P O G rn CH -CO un C"" rH CV CH CH 4 4 o O o to O to to o o o © CL. O © E-* 4 CH O CH NO O ON NO UN CV ■*—IS -— Si Gs —- f1—u 3s —^ 3 s; 3 tj -o •H O rH O •H P o s <£ £ 03 u Cu rH P O i—1 UN o rH U © G O On O- o 4 JO On CV •H P g i— 1 4 to ON rH © JO o CV CH ■^ rt T5 V X >o UN 4 CV •ri o © o •H © £ *—. <>* 3 © m ® 'sR. rH On i— I UN ■•a; P G p> O t) o ON 4 rH o .— . © CO CH O • o o 4 CH V to to • UN CO On CH CH 4 o rH CH V nO Q O« 4 CH On • to NO C"- o © CH • CH NO • CH ■ CO • 4 o • T J *H O £3 •H - O 45 •H O •H © © £ O ■H a. o *H P © O < 45 •H O © o •H s £ O Ce» 4) •H O © CJ T) *H O © o •H £ U ■H O o 4 CO H P © £ 109 <—1 i —I to to £>•o NO CM • • • cn CO CN tO V O •to UN wn CM ANALYSES FOR SILO UN o o• On C*• NO CN • ON to NO UN CM CM O o •60 H CN rH -4 rH CM o to o rH CM CM Cn- CN t—1 CM o o CM ON iH ON to ON iH ■00 CM O • to o rH CN O• n • ON • • CM « -4 rH NO 1 to UN -4 O' •4 rH CM CN ■oo o o CN NO N! -4 O Crt rH t 1 1 1 to CM <-H ON o o iH UN • un -4 [ > • -4 UN • m • • • CN NO• UN • CN to t> sO c- -4 O- -4 CN CN CN rH to UN «H rH r- CN CN • C nQ CO v£> O O n • • » • • NO M• -4 • ON O CM CV C"- C" ■60 to -4 O- -4 H CM ■to to • o «H r~- o 73 © rH •rt to c CD CO © 03 CN CM CM ■4 rH o -Jt CO ■ un to CQ •_-jJ» s +5 c! 3 O o rH © •H in © p o 0035 o c^ o -4 CN O CN H CM CM NO UN nO -4 -4 NO CM CN 1 -co .—» to '—p c? o o rH £0 •H U CD P O C0O5 >-! CD E-r -«_»» P £ r3 t O o rH cti •H u © p o © £23 s «c "4 ■—P a 3 o o rH © © ■rt u fl u os © p » p o p S-. © o SrIt -s 73 •rt o © N a o •rH P 5-. s 03 Q a. *=s —• •—» * 1 Prrt SI 3 3 T5 ■a •rt •rt O o © ca o o •rt •rt P S © Or O o 73 •rt O © o •rt P O © P '—• SI •—' X5 •rt O © o ■rt c •rt a o 3 co -4 cn cn V o o• 110 CV TO • cn rt- cn in • cn to • rH NO -4 cn O cn cv • rte- to o• cn to • rH cn rH CV rtHO • -4 O cn • -4 to cn V NO CO cn V cn V rH rH • rt- cn V NO rH • -4 to rH • ON ON cn • ON CV -4 • to -4 nO * -4 TO • rtc-- rtto • CV O cn cn -4 • On • CV vO cn • O rH rH • un a -4 • rt- -oo On • in to rt• o trt- -4 rt• cn CN -4 * m cn -4 « cn c^- cn -4 * -4 cv • rH nO rH TO • rH rt- cv rt• -4 o i —1 TO NO © a rt- cv TO rH „_„ --* iH a < iH ■— *■ o co CQ nQ UN 53 3 © •p CO rt ® rt rt as (X p to ■H O S3 e'­ en CV o <—i CO *H rt ffl +3 TO cn rH rH ON • -4 rH CO •H rt © -P O CO CQ 1 o -4 A rH CO •H rt © +1 O cO ra 1 o cn V rH CO ♦H rt © P O CO pq 1 O o rH cn V V 1 ON in O rH • rH rt- p rt rt o o o o rH cn -4 O • rt- +3 rt rt o o -4 -4 -4 • rH cn cn cn • m o m rt- rH rtr• -4 "rt © rH vH 01 £3 0) O c- to cv 1— 1 o rtto • in P rt rt -4 -4 cv cn ON in • \0 P « rt o o 1— 1 «H ON o rH i —1 nO CV o TO > w -4 NO rH rt- ON o O ON • cn CO cq vH o rH rH rH cn V • -4 o o -4 cn rtm o> cn cn V O -4 • -4 O NO NO in rH 1 On cn • -4 CV o cv CV cn T0 1 O cn • rt- O rH rH rH • m o> rH O o to m o rH o f to TO ■o C >» £1 © O x> •H P O rt co rt CO O £3 TJ *H O c0 <—1 ctl P o © w © P © rH •H P © rH O fc» 1 3 TJ *H O © o rt •H s UN • no UN vO v CM to NO UA NO ON nO • CA • ON 9 O' 4 CN- t4 -Q CV 4• r-i • nO UA 'V UA to 0 nO O O• n C V • • -4 N UA •to 4 IN cv UN to 4 • • CV CV 0 cv r-i r- o- o fA O n N ON - UA CA O • C"• CA _4 CV O CV r-i 4 u\ O 4 cn ON 111 CN- r-i r-i Cn- c\i cv 4 CA r-i O' rH rn CA -4 O' ca ca CA -4* O CA -4 CA CA CA • UA CA nO • CN• • 4• O « -4 NO• UA • -i CS’ ' NO !>- £>- UA to CA O rCA r-i O UA UA • • O' c- to CV sO CV nO • • to 1 N UA 14 rH 9 nO O £4 UA UA to -4 • 1—i rH to O ON • • • • • On UN UA UN O 4 J4 •co O CV r-i rH -4 CA • f—i • NO 4 4 CV CA ON r-i ON ON NO rH to r-i P p fl P S3 a rt 0 d 0 > rt a 0 rt 3 rt d p rH rt -H -P CO rt CO 0 „_, *»• rt rt •H rH 0 rt Pt 0 rH ct) rt p 0 rt rt rt O rt rt rH to o 0 0 rr* ci*» 3 3 ■co UA rH CV CV r-i UA •A 4 rH NO 4 cv -4 UA CA On CA __* t~-i a 0 0 rH *ri CO ON £n- rH CA CA H, O «£ &-• ’ — to o r- O- UA to -4 O CA UA CA NO to UA * NO to ON CA to • • • 0 CA rH CV 4 NO UA UN to to o iH P O rt rt t-S. rt rt Ch P 0 rt rt rt Ct a •H CD ■P O rt Oh u •H __ (D Q rt rt O rt (— 1 CD 1— 1 •H P rt 0 O rt rH O T5 >A rt O cc rt P Eh ____ > 3 d •H 0 rt O rt •H £ < -3 H d •H O rt 0 •H rt i=A P rt CQ t) rt 0 c 0 ■H a 0 rt a, 3 d •H O rt TJ •H ■ TJ •H d •H rt rt 0 O 0 0 •H P a) -H s •H 0 rt rH 0 •aj rt 0 p 0 O rt 0 •H •r-i O O rt cn 112 on UN ON •4 -j- Ch to mj « V to i—I v IN4 \*/ ON 9 SILO FOR ANALYSES ON to O • vO 4 o cv ft 60 CV 4 O • On O o • oO ON ft in- un rH • On o C cv tCOO OonN O V v N 0O 00 ON • O to .♦ * 4 4 un\ oo u co ON •• UN "\ vO vO to • \ vO • jj c ,»- 0 C'o-• • • UN • UN • • O >A o u\ o • * vO rH •• 4 *• {t>>- t>on O vO v0 660 0 *• I C--S m • 4 • • 4 •• UN on un vO sO UN un on to • O 4 * vD vO o vQ • o- o 4 OJ ON on 4 4 O ' vO O u \ on (rN • v vO Q to O- oN O H * -•i v0 vO 4 ON • to to o to to • 4 0N■ 4 • to oC"• to £> o o 1 CO j_q 03 1 — > to 1 * vO •• HH uunn CV CV 0n ON oO r— H1 TO —1 Hi CV CV OO H rH tto c O UN O' • ON 4 • H ON CV vO to vO 60 4 ON ft to C- vO n 2 O o +3 c [H O o c: 0 O o ON • 4 CV • 4 O' ON 4 rH .___ k +3 s 1 » ,—» — I— I I— I I— I I— 1 fij CO 03 © © •H Sh CD p *H S* *H Sh *H Pt U 3 O CO a) CD +3 O CO tc PI 0) ffl +3 © p P CQ P O O *H O cdCrt raffl K as: H 4 C V o o O 4 H vO vO O O cv CV H Q\ cv un vO O o n O H o 60 o O H 4 on VO kO CV o H to O' H O' ON O o ON ON r -i NO vO cv « 1 1 4 s: rH 4 H CV O' o to ON ON o cv ___ c 3 O o un N r -i 4 S>H O 4 £-» < p V vO f-r a. CD *5? —4 -— . 3 p 03 TJ u T) P*» SI o rO U •H CO o O CO i—1 co P o (0 cO r-H jo* TJ •H © H -H P CO rH O > t **<■ -1 3 - C ON CV CV ■£> • ON to • O o H to • to ft H• VfN ■cO cv H O ON • ON O' • » 1A to ON • ON to ON ■60 N O cf- o c >H s 4 X3 vH O CO TJ «.N O cO o •H C) r^, i‘ |H P tJ 03 ,C o •H CL o f-r a. 3- 3 T3 O <0 TJ TJ •H o P © O 4 rri O cO o £ O a TS •H O CO O ct! o •H t) ■H P o CO •H O o s C/3 113 ON 4 P V V • f- iH to • -4* sO ON on (V CM o on O • on to • to E>• CM to • s£> • •co P CO CO « p < O en P • -4 CO 03 TJ Q) i —I »rl 05 a 0) CO > CO m • t P to > P © P © p o o co •H U © p o ($ ■ -4 O o E'­ o -4 • vO CO OS • -4 Os sO O O UN sO Pon V -4 o p M -4 • -4 ■ ON p o o UN sO O -4 O CM • Os Os • C"- to -4 OO• ON to • sO vO UN • rH P• ON CM CM « H to -4 to sO O * P- sO CM • Os O- • ON • Os • sO P- rH O H CM ON • sO rH CM to • Os • UN ON • -4 P © o O o a P vO Os Os to CM -4 O rH to Os • o CM rH ON rH -4 P o o H P- ON ON 1 t CM CM sO ON • ON ON rH rH CM • -4 Os rH UN * CM to Os o ■— » © © © 0 P (0 © © P p 05 eg a *H O 15-0 a •H 0 P O © Q- O vO P vO O' rH CM ON O CM rH vO ON i 1 1 US 3 a H O sO <■=* 3 CM o ON >■* o •< £-« > P o rO © CO o o £0 •H CO P O £-4 © r~i •H P © r—1 o > TJ •H O © O © •H a rH • Os UH to p- -DO * o• pOs • SO SO sO • sO O o • V rH V sO CM Os rH o O • to to V to -400 P*■ P- CM €■ OH CH • rH Pto rH CM u* to us CM -4 CM CH I I O O t— I r—I CM OH CM US us to O O H -4- sO CTS P- CM rH P• 4 JH ttJ O TC CM • sO CM P- Hi -4 • rH sO CM SO rH sO US CM o rH P- us vO « rn sQ 4• O P- -4 UH • 4 UH UH CM CM o -co CM -4 O' CM rH ' O • i> on UH O -4 CH P~ iH O vO P- OH o C'N rH o o - o oh un sO rn rn 4 t I I CO sa co PI O p- sO UH • P- « o cn -co m O * us CM P- • 4 rH < CO ■T| ----, , to tr* CD t-< •4 >— ■* TJ CD rH ■r4 CO c CD co > cr Q -P C 3 o o rH ce *H u CD P O CD £D P G T 5 o P G 3 O o P G 3 O cC rH cC 1 — i C tJ •H 3 0 P o cn •H 3 CD P O cc CQ s=U •~m0* *H 3 0 P O C D cq — - — *. o (X> —— , —^ 0 3 3 P C O •ri S3 a, O '■S' a •r-l 0 o 3 » ® P Cfl 3 Vfn -3 P O 32 3 CL, TJ 3 •X, S3 3^ Gs CD (0 CTS TJ *H y Ctf rH CD P Q Eh -O CD rH •iH P cfl rH o >• TJ •H O GJ O q *ri TJ *H O CO O •n 3 P 3, TJ *H a co o *r‘ C G P G 3 a. TJ •H a co y •H P o y TJ •H O CO o ■H a 3 o Jx, TJ •H a cfi O •H P O cfl PI TJ •H O co o •H a ■H O O 3 CO 115 *• o> to r• rH * CV IH o CV CN •to « On £.rH nO • cN rH • • O O• rH• CN to •to CN rH CN t—1 rH 0 rH ♦ CN -CO • 0 O UN • to un vO CV UN UN • CN O CN • • O' to to CN rH •co u n CN CN un CV F- vO CN CV CN V-' UN O ^n SILO 12 cv \0 CV O UN ■co rv• CN UN c- UN to CN O -oo to -4 CN o ANALYSES FOE V -4 CV IrH -4 rH • NO -4 O • NO rCV in- UN to o -4 CO X to '— - -— p p X G G 3 O O cd GJ TO G G P G 3 O O -4 to CN rH • c*~ UN • 00 -4 CV rH • "CO •CO • >-» 0 eu -— *■ <4 P O G X 3s p u •H 3 -4 CD CO 3 0) p o rH CV NO CV • NO cd X O' • O' UN • O to CD P O CtJ *H G 0 UN rH • O • • -4 n£> to CN C'- UN NCN rH G > AT O JO G cd 0 G _5n CD rH CtJ P O H X •rH O •ri ct 02 3 cd P 3 O •H C d •H y ct —• -r (4* rH O cd P 0 ® •rH 0 •rl G rH •H G JN P O > a «< 3 a cB "TG '“‘V 3 s i2, 3 xf ■d G •H •H O Ct O ct O •H y •H 0 Or O G P 3) a •H G O a. G CC X G CO O •H •H y 0 r—^ rt <4 0 B O (-1 r\ un • un un • to o> to rn on V V un .A. -4“ • O O' CV • 4 un e'­ cn ON m • 4 •CC' to CV ISCV O NO o CV un rH rH cn cn o 4 ts cn IS un 4 O rH CS • cn un cn CV 4 un CN • CV IS un rH IS • un sun rH • 4 -to o o rH to o NO _-• 1 On CV o C^ 4 CV 4 4 rH CV UN CV i 1 ON CV O 4 4 CV t 1 rH O ir\ rH r-l rH O o o NO CV cs CV C on cn NO o o 4 4 rH cn CV O rH NO CV 1 1 to o to • NO O H CD '- IH _^ +3 G G O o i —l © •H © p O CO CO P G G a o i—t © © P O CO CO --—■* © G p CO DC a, •H o J3 rr—^ ■ — G © P O u Or ’3'- ^—» 3 o P CO TJ •H u O d Js CO ,g i— I o X3 CO P p o CO o E~« 3 © © © © !—1 P © rH o > —* 3 TJ •H O CO o »-) *H £ -—* 3 <23 3 TJ •H O TJ © O o © *H o G *H O ■H U >» Cu o P u g C L , CQ •"0 ■ —rd •H a © o •H P a> o ■u; at. 3 "•~ 7 -L 3 3 TJ o © o ♦H TJ O © o •ri B G O Cu •H TJ O © o *H •H P a © iH •H G *H O o d cn