Bacterial Developments on Silage Production and Aerobic Stability

The silage fermentation is predominantly a lactic one brought about through the activities of lactic acid bacteria found naturally on the crop ensiled. These bacteria ferment sugars into varying amounts of lactic acid, acetic acid, ethanol, the sugar alcohol mannitol, and carbon dioxide. Ideally there should be a dominance of lactic acid as it is by far the strongest acid produced in silage (has the lowest pKa value at 3.08), and is extremely palatable to the ruminant consumer. Acetic acid (pKa 4.64) makes less contribution to acidification, and is not palatable to the ruminant. Least palatable of all is mannitol. Carbon dioxide is a very weak acid unlikely to make any significant contribution to acidification.

If there is insufficient sugar, or the dry matter (DM) content of the crop is low, acidification may not be of sufficient magnitude to prevent an unwanted fermentation in which another group of bacteria, known as clostridia, become significant. Clostridia ferment lactic acid and sugars to butyric acid and later ferment proteins to a whole host of end products, including ammonia, which reverses the good work on the part of the lactic acid bacteria. Hopefully, a pH in silage of sufficiently low level will be reached to prevent the clostridia from growing.

Fortunately nature has decreed that lactic acid bacteria are more tolerant of acid than are the clostridia, but the latter are more tolerant of low pH if the DM is low. The lower the DM the lower the pH required to prevent clostridial growth. Conversely, stability (i.e., prevention of clostridial fermentations) is achieved in higher DM silage at a higher pH as clostridia are less tolerant of low water availability than the lactic acid bacteria. When crop DM is in excess of 30% clostridial fermentation is suppressed through a lack of moisture availability. Up to this level a combination of acid, pH and moisture availability will inhibit clostridia. Generally, at a DM of 20% stability is achieved at a pH of about 4.0.

To place the bacteria in context, the lactic acid bacteria have been referred to as man’s best friend. They are responsible for the formation of cheese, yogurt and sauerkraut to name a few instances where they are exploited.




Figure 1.A sequence of bacterial types involved in the silage fermentation.



However, some of them are know as the dentists’ best friend – as they are implicated in dental caries! Some clostridia are responsible for the retting of flax (loosening the fiber from the woody tissue), gas gangrene and botulism in man and animals. None of these have been found in silage, so it is safe to continue eating this form of conserved forage – though it might rot your teeth! The microbial changes which occur (or can occur) are depicted in Figure 1. The succession of lactic acid bacteria from streptococci (syn. enterococci), lactobacilli and finally pediococci, is dictated by the tolerance to acid. Streptococci are the least tolerant, pediococci are the most tolerant.

The silage fermentation, its microbiology, biochemistry and nutritional considerations have been covered in great detail elsewhere (Woolford, 1984: McDonald et al., 1991; Woolford and Pahlow, 1998) and many aspects are out of the scope of this work.


The efficiency of the silage fermentation

The word ‘efficiency’ in the present context refers to efficiency of acidification by the microorganisms concerned, and not efficiency of conservation of nutrients, but the two ‘efficiencies’ are interlinked. The efficiency with which nutrients in the crop are conserved is of cardinal importance since it matters not how well the silage is preserved if nutrient losses are high and the ruminant will not consume with relish and perform well on it.

However, efficiency here refers to efficiency with which sugars are fermented to acid and how one can realize optimum value from the sugar in terms of acid production to prevent the spoilage activities of clostridia. It is essential that the decline in pH is as rapid as possible and such a rapid reduction is critical in the first two or three days of ensilage to stem proteolysis. In uncontrolled silage fermentations 45% of the protein will be broken down (Carpintero et al., 1979) and the figure can be somewhat higher (Whittenbury, 1968).

Indeed, it is this very subject of retention of intact or bound protein that has been linked some thirty years later with the improvements in animal performance which accompany the use of an inoculant on silage (Jones et al., 1996). Jones will cover this subject in detail in another chapter in this volume.

The principal sugars in silage crops are hexoses such as fructose and glucose whilst some pentoses also occur. Fructose is the dominant hexose, being in a ratio of 1.1 to 3.9 (MacKenzie andWylam, 1957; McDonald et al., 1960).

The lactic microflora of silage is invariably dominated by heterofermentative species which are less efficient in the production of lactic acid (Stirling and Whittenbury, 1963; Lindgren et al., 1985). The homolactic fermentation of hexoses yields lactic acid and lactic acid only. Since the heterolactic fermentation of fructose is less efficient in the production of lactic acid than is fermentation of glucose (see Table 1), fructose is the dominant hexose, and the fermentation of pentoses always yields equimolar amounts of lactic and acetic acids (Table 1) irrespective of whether a given lactic acid bacterium is heteroor homofermentative with respect to hexoses.


Table 1.Fermentation changes which occur in silage.



*The most desirable of all the fermentations in silage.



In terms of strong acid (i.e., lactic), the silage fermentation is not all that efficient. To compound this, heterofermentative lactic acid bacteria dominate the microbial scene at the outset (Stirling and Whittenbury, 1963). There is some evidence to suggest homofermentative lactobacilli dominate the standing crop in corn; but streptococci are the principal ones on ryegrass and alfalfa (Andrieu and Gouet, 1991). It is of little consequence how low the final pH may be; but it matters how quickly this final pH is achieved.

This raises the question of how to exercise control over the silage fermentation and what means are available to realize this goal. Historically, an empirical approach was made by the use of extraneous inorganic and organic acids, salts and a whole host of chemicals with the object of obviating the clostridial fermentations. The products from these fermentations, like acetic acid from the heterolactic fermentation of hexoses, depreciated silage quality since most yardsticks of silage quality were based on the ratio of lactic acid to the products of less efficient/unwanted fermentations (acetic and butyric acids plus ammonia). The higher the ratio the greater the quality.

However, the guidelines to the production of good quality silage are now laid down in tablets of stone and abided by most silage makers and so the incidence of poor quality high butyric evil-smelling silage is largely history.

However, there are still a few silage makers who are adept at converting a good silage crop into manure without first passing it through the alimentary tract of the ruminant animal!

During the initial stages of ensilage, when oxygen is becoming depleted through the respiratory activities of plant enzymes and the aerobic components of the epiphytic microflora, there is a shift from a microflora of aerobic bacteria, yeasts and other fungi to one dominated by facultative and obligate anaerobic bacteria, although in some silages (e.g., corn or whole-crop cereals) yeasts, some of which are facultative anaerobes, will also be present. This shift is the Achilles’ heel of making silage.

The removal of oxygen and maintenance of oxygen exclusion is critical as the influences of oxygen (air) can cause the lion’s share of the potential nutrient losses (Zimmer, 1980; Table 2). But the magnitude and significance of protein losses due to delays in acidification can outstrip losses in other nutrients within the context of animal production (Jones, 1998).

In contrast, with hay making it is essential to remove water (3.5 tonnes of water are removed to make 1 tonne of hay), often with extended field curing leading to added nutrient losses derived from respiratory sources (greater than indicated for silage) and mechanical handling. Furthermore, in some climes uncertain weather makes the whole hay making operation more risky than with silage making.

The shift to a microflora dominated by bacteria capable of anaerobic growth reduces intermicrobial competition and, in theory at least, makes the job of fermentation control somewhat easier.


Table 2.Summary of the sources, extent and causative factors of loss during ensilage and after removal of the feed from the silo.






Control of the silage fermentation

EPIPHYTIC PLANT POPULATIONS

In stark contrast to other industrial processes which depend for their existence on fermentation such as brewing or antibiotic production, silage making for the most part is uncontrolled when left to Mother Nature. A modicum of control can be imposed by harvesting the crop at a stage of growth when sugar content is at its height, by raising the DM of the silage crop through wilting and by the observance of the established principles of silo management.

In the other industrial examples cited, the medium is freed of indigenous microorganisms, either by boiling or sterilization, and inoculated (or pitched to use the brewers’ jargon) with specific microorganisms. In contrast, in silage the process depends on the lactic acid bacteria gaining dominance over the remainder of the microflora, aided and abetted by the anaerobic conditions imposed. It would be highly impractical to sterilize 1,000 or 3,000 tonnes of silage on the farm – not to mention the dire effects this might have on heat labile nutrients, e.g., protein!

Evidence available points in favor of there being an adequate supply of lactic acid bacteria by the time the harvested forage reaches the silo. This can be of the order of 100,000 colony-forming units (CFU)/g of crop (Stirling and Whittenbury, 1963). This is a monumental explosion in numbers compared to those found on the standing crop, which often contains bacterial numbers in the region of 100 CFU/g.

The harvesting equipment has been suggested to act as a mass-inoculator, and the ease of access to substrate (sugar) as a result of the release of juice has been implicated in this apparent wholesale microbial growth. However, antagonistic or antibiotic-like substances on the plant material have been suggested as the reason for the presence of such a small number at the outset (Nilsson and Nilsson, 1956). Pahlow and Muller (1990) and Muller and Seyfarth (1996) proposed the ‘somnicell’ concept whereby the ability of lactic acid bacteria to form CFU is impaired by environmental constraints, i.e., ‘viable but not cultivable,’ on the plant surface.

These constraints are considered to be eliminated when a ‘recovery’ period has elapsed. This recovery period, in association with access to nutrients, ‘wakes up and invigorates’ the cell thus making it capable of growing unaided to form CFU. This is a very plausible principle.

Numerical adequacy of lactic acid bacteria by the time the crop reaches the silo naturally leads to the topic of acid-producing efficiency discussed earlier. There is a dominance of heterofermentative types, which is not entirely desirable, but there is a school of thought which considers presence of these bacterial types a means of enhancing the bunk life of silage.


THE CHEMICAL APPROACH

The notion of using chemicals, and extraneous acids in particular, to augment acid produced by fermentation stems back to the end of the last century. Their popularity has been confined to Europe, Japan and Australia. Acids were never accepted to any great degree in North America.


The popularity of chemical

additives for silage has waned considerably in the last decade in the wake of the wider acceptance of biological treatments. Chemicals commonly used are listed in Table 3.

Formaldehyde, which is normally available as formalin (a stabilized solution of formaldehyde in water and methanol), was used to prevent fermentation.

Subsequently it enjoyed a transient popularity owing to the discovery that protein in silage was protected from the degradation activities of rumen microbes thus making protein available for digestion in the small intestine (Ferguson et al., 1967). Formaldehyde application rate was critical. Underapplication resulted in encouragement of a clostridial or butyric fermentation, whilst over-application gave over-protected protein, i.e., it was totally unavailable for digestion (Wilkinson et al., 1974; Wilkinson et al., 1976; Ferguson et al., 1967). Other obnoxious substances such as sodium metabisulfite (a source of sulfur dioxide) and hexamine (a source of formaldehyde) have been used.

Specific antimicrobial agents directed at clostridia have been tried but have not stood the test of time. Sodium nitrite still enjoys some popularity in Europe in spite of disquiet relating to its reactions with secondary amines to form highly carcinogenic nitrosamines. Antibiotics (of biological origin?) have been largely proscribed by legislation and so little chemical choice remains available.

Undoubtedly inorganic acids such as sulfuric (introduced by Virtanen in 1933) and organic acids, principally formic, have dominated the chemical scene in silage production. Though formic acid use was adopted in the early 1960s in Europe it was studied much earlier in Germany (Dirks, 1923, cited by Watson and Nash, 1960).

The downside of using acids, over and above their corrosive nature to skin, clothing and machinery, has become apparent over the years. People did not read their ‘history books’ before the revival of interest in sulfuric acid in the 1980s owing to its low cost. Sulfuric acid use in silage led to the locking-up of trace elements, especially copper, and fertility problems in animals fed silage treated with this acid (Suttle, 1978).

Indeed, because sulfuric acid is much stronger than formic (pKa 1.92), its ability to acidify silage immediately is greater. Moreover, the fact that formic is an organic acid explains why there is delay in acidification upon application to a crop. Before gaining access to plant juice the acid must first render soluble the cuticular waxes. In addition, there is a buffering effect of the plant juice acids (Woolford, unpublished results). Further, it seems less mature crops and legumes would need high application rates to ensure a satisfactory low pH was achieved (Henderson and McDonald, 1976).


Table 3. The classification of silage additives.





Negative effects of both sulfuric and formic acids on animal performance are summarized in Table 4. This survey, done by Jones (1994), revealed both sulfuric and a formic acid-based additive increased DM intake (as a percentage of the control) but the liveweight gains experienced were negative for the former and mediocre for the latter. In contrast, inoculants, which will be discussed later, both improved DM intake and performance under both difficult and easy ensiling conditions. Similar trends were found in the works of Keady and Steen (1995) and Kennedy (1990).


Table 4. The effects of acids on silage quality and performance.*


*After Jones, 1994.



THE BIOLOGICAL APPROACH: LACTIC ACID BACTERIA, ENZYMES AND NUTRIENTS

The quest for animal, consumer and environmentally-friendly materials in our everyday lives has meant silage is no exception. Not only has this led to a revival of interest in sugar sources such as molasses (Table 3), but also interest in the most logical approach to control of the silage fermentation, i.e., to supply efficient homofermentative lactic acid bacteria which have a far stronger ability to yield lactic acid than the related native bacteria.

There had been serious setbacks to the notion of using biological additives to control the silage fermentation, especially in Europe. In the early 1970s US-derived inoculants intended for use on corn were used for low DM European grass and resulted in bad silages.

Another setback was the employment of enzymes which had pH optima lower than the level at which most silages stabilize resulting in many silages being euphemistically described by UK farmers as ‘meltdown’! This has endowed biological additives with the label of ‘fair-weather additives,’meaning they only work when good weather ensures good ensiling conditions. This is far from true; and in a recent publication from the Milk Development Council in the UK proven biological additives are recommended for a wide range of conditions (Anonymous, 1997).

There is a growing volume of evidence to show that biological additives, particularly those containing lactic acid bacteria, alone or especially if supplemented with saccharolytic enzymes, bring about a more rapid fall in pH than attained through the use of no additive. This leads to a rapid decline in coliform bacteria and a more rapid rate of fall in pH (rather than the final pH achieved) thus stemming proteolysis. The advantage is improved animal performance (Gordon, 1992; Patterson et al., 1997).


Molasses

The use of molasses to augment naturally-occurring sugars in silage could loosely be described as a biological treatment since sucrose is commonly found in most silage crops and it is derived from plant sources. It is a by-product of sugar refining, and it is worth noting that if more sugar could economically be extracted, molasses would not be on offer for use in silage production.

References to molasses use in silage are legion dating back before other biological approaches evolved. The subject has been reviewed in depth by Woolford (1984) and McDonald et al. (1991). There is little evidence that molasses applied at rates which are ergonomically as well as economically efficient (i.e., up to 10 kg/tonne ensiled) has any significant positive effect in controlling the silage fermentation. Psychologically, the user of molasses believes the nutritional value of silage is going to be enhanced by its application.

However, the question of ‘how much of the nutrients provided by the molasses are going to finally reach the animal’ needs to be asked. The answer is very little.

The consensus is that to provide a significant quantity of sugar, and therefore, have some positive influence over the fermentation, would require in the region of 20 to 40 kg/tonne ensiled (Woolford, 1984; Keady, 1997). However, in order to meter this viscous material on to the crop special equipment is required.

Alternatively, it has to be diluted with an equal weight of water, a marked disadvantage since molasses tends to promote the discharge of effluent.

Additionally, owing to its fructose content molasses encourages growth and activity of the less efficient acid-producing heterofermentative lactic acid bacteria (Whittenbury, 1968). Moreover, it must be appreciated that molasses feeds the less desirable as well as the desirable bacteria in silage.

Molasses is a perfectly suitable supplement to any ruminant ration if applied at the time of feeding. It is extremely palatable and its full energy value can, no doubt, be realized by this approach.


Enzymes

Forage plants have a large reserve of carbohydrate which, unfortunately insofar as silage is concerned, in its natural state occurs as polysaccharides. For instance, grasses and legumes have much cellulose, and corn has much cellulose and starch. It is logical to assume that one would, by the addition of sources or preparations of cellulolytic and amylolytic enzymes to forage crops, increase the fermentable capacity of silage by virtue of the release of fermentable substrate by the enzymes. Moreover, cellulolytic enzymes would promote cellular breakdown and thus render cell contents more accessible to the silage microflora.

Early work concentrated on the use of malt as a source of enzymes in cereal.

Malt treatments of a variety of silage crops brought improvements in fermentation quality (Rydin, 1961; Nilsson and Rydin, 1960). It is for this reason malt is used in the production of whiskey!

Enzyme usage was considered in depth byWoolford (1984) and Muck and Kung (1997). Generally, the potential for enzymes in silage is great with silage made from crops with a DM of up to 40%; but beyond that limitations to moisture availability restrict enzyme effectiveness. In very low DM conditions loss of effluent and indeed promotion of effluent, may result from enzyme application. However, selection of appropriate enzymes is critical. Enzyme pH optima must not be less than 4.0.


Bacteria inoculants

Lactic acid bacteria (homofermentative)

Of the various types of fermentative bacteria available, homofermentative lactic acid bacteria are the first choice; and the vast majority of commercially available inoculants contain these bacteria.

Before a culture can be considered suitable for inclusion in an inoculant, it must satisfy a number of criteria (Whittenbury, 1961):

1. It must have a high growth rate and be able to compete with and dominate other organisms likely to occur in silage.

2. It must be homofermentative.

3. It must be acid tolerant and produce a final pH of 4.0 quickly.

4. It must be able to ferment glucose, fructose, sucrose, and preferable fructosans and pentosans.

5. It must not produce dextran from sucrose or mannitol from fructose.

6. It should have no action on organic acids (Author’s note: action on acids such as citric, malic and succinic would enhance buffering and thereby, hinder acidification).

7. It should have a good growth temperature range, preferably up to 50wC, in order to survive any rise in temperature during the early stages of ensilage.

Essentially similar criteria were employed by Wieringa and Beck (1964) in the selection of bacteria suitable for inoculation trials. In addition, they listed a lack of proteolytic activity as an essential factor. A culture containing such activity is not much use because the intent of the silage maker, like the cheese maker, should be to preserve protein!

Ignoring the consideration of growth rate, the chance of a successful outcome from the use of an inoculant is greatly improved if, directly upon application, it provides a population which out-numbers the indigenous population of organisms of the same type. Such a notion places less onus on criterion 1 in the list of Whittenbury (1961). No doubt this idea was borne in mind by several investigators. Generally, additions of the order of 10⁶ to 10⁷ organisms per g of fresh crop have produced well-preserved silages (Gross, 1969; Ohyama et al., 1973; Moon, 1981; Woolford and Pahlow, 1998).

The number applied might be one issue, but the viability (the capacity to grow and produce acid) is of paramount importance. These factors will vary according to the stage in the life cycle of the bacterial culture. This topic is over and above the concept of ‘somnicell’ referred to earlier. Typically, all bacterial cultures can pass through up to four phases in their life cycle, provided there is no environmental intervention which might cause a sudden change in viability such as rapid change in temperature, physical factors such as pH and pressurized applicator systems in the case of silage inoculants. These four phases are depicted in Figure 2.

The first phase is called the ‘lag phase.’ In this phase the microbes are in a non-active, non-growing state (e.g. such as when bacteria are being revived from a freeze-dried state or they are getting acclimatized to a new environment).

The second is the ‘logarithmic phase’ when there is a rapid growth in numbers. The third phase is the ‘stationary phase.’ In this phase numbers are no longer increasing and a fairly steady level is maintained, with death being replaced by life. The fourth is the ‘death phase’ where there is a decline in numbers, aided and abetted by an exhaustion of substrate, accumulation of substances toxic to bacterial growth such as acid, as is the case in silage, or antagonists.

The life cycle goes to completion unless, some environmental factor intervenes. Such environmental factors might be a drop in temperature as would occur when a culture is applied to a forage crop (likened to rolling in snow directly from exiting from a sauna!), not to mention the adverse environmental changes being imposed by passage through an applicator nozzle at great velocity following the encounter of the various mechanical components on route through the forage harvester! This could induce the ‘somnicell’ state.

Furthermore, if nature is allowed to take its course, a culture could be stale or past its expiration date by the time it is finally applied. This can occur if a delay in silage making is brought about by inclement weather and by the fact that fermentation cannot get underway until oxygen trapped in the silo has become completely exhausted, which in farm-scale silage may be after several days, especially if there has been a delay in sealing (Langston et al., 1958).




Figure 2. Typical growth curve of a bacterial culture (after Schlegel, 1988).



This conflicts with the principle of bacteria being applied to silage in the form of a pre-culture (variously described as ‘fully active’ or ‘activated’).

This may be far from the case and inappropriate terminology to use. One of the disadvantages of culture prior to application (pre-culture) to the silage crop, in addition to microbial contamination, is that by the time it is used it might be in the third or fourth stage of its life cycle described above. Inoculants derived from freeze-dried sources will have had the opportunity to become rejuvenated and to grow in aerobic/microaerophilic conditions and possibly through several life cycles by the time anaerobic conditions are established in the silo. This microbiological fact casts doubt on the efficiency and concept of ‘pre-cultured’ inoculants.


Lactic acid bacteria (heterofermentative)

Acetic acid, like propionic acid, if well mixed with silage will enhance the aerobic shelf life of silage. Indeed, any silage with more than 40 g/kg of acetic acid in the DM is very stable (Woolford, 1978). This acid is a product of the fermentation of both hexoses and pentoses as referred to in Table 1; but it must be re-iterated that it is not at all palatable to the ruminant. Nevertheless, there has been some work whereby heterofermentative lactic acid bacteria have been incorporated into a commercially-available inoculant to promote the formation of acetic acid, a choice of bacterium contrary to the criteria of Whittenbury (1961) and Wieringa and Beck (1964).

Driehuis et al. (1996) investigated the role of the heterofermentative bacterium, Lactobacillus buchneri (strain ID-DLO Lsp, isolated from corn silage) alone and in combination with an inoculant comprised of three bacteria (L. plantarum, Pediococcus pentosaceus and Propionibacterium jensenii) in the enhancement of the aerobic stability of laboratory-scale silage made from corn, high DM grass and whole-crop wheat. With the L. buchneri alone inoculated into corn at ten-fold intervals in the range 10²–10⁶ CFU/g fresh weight, acetic and propionic acid contents were enhanced in direct proportion to the level of inoculum. Better sealing was also associated with higher levels of these acids; aerobic stability was accordingly enhanced.

In-silo losses of DM increased with inoculum amount. However, in the trials with high DM grass and whole-crop wheat silage, only L. buchneri alone significantly improved aerobic stability. Likewise, the commercial inoculant alone improved the stability of the grass silage (Table 5)! Clearly, with the high in-silo losses undoubtedly due to the evolution of carbon dioxide encountered in the treatments containing L. buchneri, its choice alone could not be entertained. Losses encountered in the combination treatment with the commercial inoculant (and the latter alone, although higher) do cast doubt on the wisdom of using heterofermentative lactic acid bacteria in silage. Shelf life is enhanced, yes, but in-silo losses may be unacceptably high.


Table 5.Chemical composition and aerobic stability of high dry matter grass and whole-crop wheat silages.*


*From Driehuis et al., 1996.
†Commercial inoculant containing two species of lactic acid bacteria.
a,b,c Means differ, P<0.05.



Propionic acid bacteria

Propionic acid is better able to enhance the shelf life of silage; and if extraneous acid is well mixed with unloaded silage it will improve aerobic stability (Woolford, 1978). The notion of using propionic acid bacteria, which elaborate the acid from lactic acid and sugars, to improve the aerobic stability of silage has been considered. This activity is responsible for the characteristic flavor and odor of Jarlsberg and Emmental cheeses; and propionic acid is in part responsible for body odor! Propionic acid bacteria have on occasion been isolated from silage and are restricted to a few species, namely P. zeae, P. freundenreichii and Veillonella (syn. Micrococcus) gazogenes (Woolford, 1975).

By using a combination of lactic acid bacteria and propionic acid bacteria in an inoculant one would derive two major benefits in theory. First, lactic fermentation would be enhanced. Second, aerobic shelf life of silage would be improved through production of propionic acid. Any rise in pH on account of propionic acid having a higher pKa than lactic acid (4.87 versus 3.08) would be minimal and the benefits obtained by better aerobic stability would offset this slightly negative event.

The work conducted by Woolford (1975), Wyss et al. (1991) and Driehuis et al. (1996) revealed that under laboratory conditions propionic acid bacteria could elaborate the said acid, but what about on a farm-scale? There is one small problem with the cultivation of the propionic acid bacteria: they are strict anaerobes with no known resting state (e.g,. spore). Oxygen is very toxic to these bacteria.

The mere act of spraying through a commercial applicator would introduce air into the culture and delays in sealing and final exhaustion of oxygen from the clamp would definitely have a detrimental effect on viability. Perhaps encapsulation would be a way of circumventing this problem; but the capsule would need to survive for as long as oxygen persisted in the silo.


The potential role of fructans

Fructans have a minor role in the silage fermentation, but are the major storage carbohydrates in grasses indigenous to northern Europe (de Cugnac, 1931).

However, starch assumes this function in tropical and sub-tropical grasses, legumes, corn and grass seeds (Hunter et al., 1970;Wilkinson, 1978). Fructans and sucrose are hydrolyzed to their component monomers to some extent within four hours of ensilage (Wylam, 1953). Since microorganisms of relevance to the silage fermentation would have little opportunity to develop in such a short space of time, it is likely that enzymes of plant origin are involved. Fructan hydrolases have been shown to be present in temperate grasses (Schlubach and Grehm, 1968) and amylases have been shown present in the leaves of many grain crops including oats, rye and corn (Gates and Simpson, 1968).

However, what about bacteria being able to hydrolyze fructans in the later stages of ensilage? Kuhbauch and Kleeberger (1975) and Kuhbauch et al. (1975) have shown lactobacilli are able to utilize fructans. Merry et al. (1995) found both plant and microbial enzymes play a role in breaking down fructan during ensilage and the rate of breakdown was enhanced by the use of an inoculum of a fructan-hydrolyzing strain of lactic acid bacteria. But of the isolates of lactic acid bacteria obtained from grass, 90% could ferment glucose while only 5% could ferment fructans (Muller et al., 1992). Indeed, the ability to ferment fructans is not common with the lactic acid bacteria (Woolford, 1984).

Recent work by Merry et al. (1996) has shown that the use of fructanhydrolyzing strains of lactic acid bacteria may enhance the availability of fermentable sugar and could be of use in inoculant fermentations, especially with forage crops low in sugar. The work showed that a fructan positive strain of lactobacilli became dominant as fermentation proceeded, and that the strain of L. paracasei employed had the best potential (Tables 6 and 7).


Antagonist substances

Antimicrobial and antifungal agents

Lactic acid bacteria, like other groups of microorganisms, are able to produce a variety of antimicrobial compounds (Korzybski et al., 1967). Amongst these are nisin, produced by a strain of Streptococcus lactis which has strong anticlostridial properties. Nisin is of no clinical value and is therefore used to prevent clostridial activity in dairy products, especially processed cheese and canned vegetables, where it supplements heat treatment (Tramer, 1966).


Table 6. Changes in pH and lactic acid concentration during ensilage of grass with different inoculant lactic acid bacteria.*


*After Merry et al., 1996.
†Able to degrade fructan.
‡Unable to degrade fructan.



Table 7.Utilization of grass fructan by strains of lactic acid bacteria isolated from silage.*


*After Merry et al., 1996.
†Able to degrade fructan.
‡Unable to degrade fructan.



On the assumption that the strain mentioned might elaborate nisin in a silage environment (as a more economic alternative to using the antibiotic itself) Flam (1967) embarked on a series of trials with silage but found no nisin or anticlostridial effect.

The strong antimycotic properties of pimaricin, used as a mold control agent in Dutch cheese, led to it being examined as a potential agent for preventing the growth of yeasts and other fungi, the causal agents of aerobic deterioration in silage. Preliminary studies byWoolford andWilkins (1975) andWoolford and Cook (1980) on a laboratory scale gave reason for hope that treatment of silage with pimaricin might improve the aerobic stability of silage. However, in the extensive trials carried out on a farm-scale by Woolford et al. (1980) with pure pimaricin, the outcome was inconclusive.

Attention in the silage context has of late focused on antagonistic substances produced by bacteria other than the lactic acid types. Gram-positive bacteria known as Bacillus are related to clostridia, but unlike the strict anaerobic nature of clostridia can grow in both anaerobic/aerobic/microaerophilic conditions.

They are known to produce potent antimicrobial/antagonistic substances, particularly ones comprised of polypeptide, such as one strain of B. licheniformis (Woolford, 1972).

More recent laboratory studies have revealed that specific strains of B. subtilis and B. licheniformis are able to elaborate an antimycotic substance referred to as ‘zymocin,’ a polypeptide substance able to inhibit a range of yeasts implicated in the aerobic deterioration of silage (Goodman et al., 1995). Farmscale work is on-going.


Clostridial bacteriophages

Phages, more correctly called bacteriophages, are very host-specific viruses which can attack and colonize a bacterium.Agiven phage infects and multiplies in only a single bacterial strain or a limited number of closely-related strains, species or genera. Phage can cause havoc in cheese or yogurt-making plants, which rely on very specific, often very secretly maintained, starter strains of lactic acid bacteria. Such infestations lead to total loss of production and the particular starter strain is never used again in a plant so affected. Phage can be developed or enriched for most bacteria.

The concept of using phages which affect clostridia (clostridial bacteriophage) as a treatment for preventing the activities of clostridia in silage has been used in certain formulations of silage inoculants in Europe. The fact that host specificity can be confined to genera, and to species and strains within species, and the fact that resistance can be acquired by the host, limits the effectiveness of clostridial bacteriophage. If such phages had a wide spectrum of activity there would be no clostridia on this planet, no botulism, no gas gangrene and no retting of flax, all of which are brought about by clostridia.

The studies with low DM ryegrass silage, where clostridia can normally exert the greatest influence (Done, 1990), revealed that fermentation products symptomatic of clostridia such as butyric acid and ammonia were not suppressed by the inclusion of phages in a commercial inoculant (Table 8).

Moreover, the number of clostridia was found to be higher in silages treated with an inoculant containing clostridial bacteriophage than in untreated or formic acid-treated silages. The studies of Wight (1990) confirmed hostspecificity and found that the clostridia in a commercially-available inoculant targeted by the phages specific for C. tyrobutyricum and C. sporogenes, were unlikely to be controlled. These studies did not address the subject of other clostridia of significance in silage.

Developments with bacterial inoculants for improved fermentation and bunk life

Short of genetic modification of lactic acid bacteria to improve their ability to generate more lactic acid and(or) to induce in such modified organisms the ability to produce propionic acid, developments will focus on enhancement of the bunk life, or aerobic shelf life, of silage during the inevitable period when it has to be exposed to air for feeding. Genetic modification nowadays is a ‘dirty word’ evoking much controversy and we are left with traditional selection techniques. This leaves little choice when it comes to developments to improve silage quality/palatability.

The work of Merry et al. (1995) in accessing sugars from fructans will lead to improved ensiling potential, especially those crops naturally low in sugars that are difficult to ensile such as legumes.


Table 8. Chemical composition of silages (means of six samples).*


*Afer Done, 1990.
†Predicted D value from OMD (organic matter digestibility) determined by near infra-red spectroscopy.



To date, the best option is to use inoculant:enzyme formulations to improve silage quality. The work of Buchannan-Smith (1988) stands testimony to the value of the synergism between the inoculant’s ability to improve quality and the enzymatic release of sugars to sustain the fermentation, especially if sugars are low in the crop at the outset.

It is fact that the use of inoculants enhance the level of lactic acid over and above other fermentation products, i.e., to increase the quality of silage.

This very factor potentially reduces the stability of silage during feed-out as lactic acid, along with sugars, is the first potential substrate for the yeasts, other fungi, and bacteria involved (Woolford, 1978). Hence, the mere use of an inoculant in silage would appear, at first sight, to encourage instability during feed-out. It must be borne in mind that the opposite might be the case, i.e., the more lactic acid there is the more must be utilized by the fungi before a rise in temperature or pH (the most noticeable parameters which accompany aerobic deterioration in silage) occurs (Table 9).


Table 9. The effect of inoculum, enzymes and a combination of these on fermentation in alfalfa silage.*


*Buchanan-Smith, 1988.
†Sil-All, Alltech Inc.



Losses from the various sources outlined in Table 2 are potentially high and to induce a heterolactic fermentation and exacerbate those losses, not to mention depreciation in palatability, seems folly. Likewise, given the low incidence nowadays of clostridial/butyric silages, use of bacteriophages and attempts to employ nisin-elaborating cultures seem unnecessary.

The option of using antagonistic substances likely to be elaborated by other microorganisms in silage warrants further examination, especially bacilli and the potential improvements they might endow to the aerobic stability of silage.


References

Andrieu, J.P. and J. Gouet. 1991. Forage Conservation Towards 2000 Proceedings. (G. Pahlow and H. Honig, eds), Landbauforschung Volkenrode 123:287–288.

Anon. 1997. Silage additives – Making the best choice, London: Milk Development Council. p. 5.

Buchannan-Smith, J.G. 1988. Inoculants and enzymes as silage additives. Alltech Research & Technical Bulletin. p. 4.

Carpintero, C.M., A.R. Henderson and P. McDonald. 1979. The effect of some pre-treatments on proteolysis during the ensiling of herbage. J. Br. Grassland. Soc. 34: 311.

de Cugnac, A. 1931. Research on sugars in grass. Ann. Sci. Naturelis. 13:1.

Done, D.L. 1990. The effect of an inoculant additive containing clostridiaphages on grass of low ensiling potential. Proc. 9th International Silage Conference, University of Newcastle-upon-Tyne. p. 98.

Driehuis, F., S.F. Spoeistra, S.C.J. Cole and R. Morgan. 1996. Improving aerobic stability by inoculation with Lactobacillus buchneri. Proc. 11th International Silage Conference, University of Wales, Aberystwyth. p. 106.

Ferguson, K.A., J.A. Hemsley, P.J. Reis. 1967. The effect of protecting dietary protein from microbial degradation in the rumen. Aus. J. Sci. 30:215.

Flam, F. 1967. Studies on the possibility of the initialization of nisin and nisinwhey in silage making. Ziv. Vyroba. Praha. 12:693.

Gates, J.W. and G.M. Simpson. 1968. The presence of starch and alphaamylase in the leaves of plants. Can. J. Bot. 46:1459.

Goodman, S.A., P.J.Warner and C. Orr. 1995. A novel bacterial inoculant to reduce aerobic deterioration of silages. In: Biotechnology in the Feed Industry, Proceedings of the 11th Annual Symposium. Nottingham University Press, Loughborough, Leics, UK. p. 451.

Gordon, F.J. 1992. Improving the feeding value of silage through biological control. Proc. 6th Alltech Eur. Lecture Tour. p. 18.

Gross, F. 1969. Directing the silage process with additives. Proc. 3rd General Meeting of Europe Grassland Fed., Braunschweig. p. 139.

Henderson, A.R. and P. McDonald. 1976. The effect of formic acid on the fermentation of ryegrass ensiled at different stages of growth and dry matter levels, J. Br. Grassland. Soc. 31:47.

Hunter, R.A., B.L McIntyre and R.J. McIlroy. 1970.Water-soluble carbohydrates of tropical pasture grasses and legumes. J. Food Sci. Agric. 21:400.

Jones, R. 1994. The importance of quality fermentation in grass silage. Proc. Alltech 8th European Lecture Tour. p. 33.

Jones, R. 1998. Bridging the protein gap – potential of forage crops for UK livestock production. In: Biotechnology in the Feed Industry, Proceedings of the 14th Annual Symposium. Nottingham University Press, Nottingham, UK.

Jones, R., A.Winters and J. Cockburn. 1996. Changes in amino acid content of inoculated grass silage and its effect on animal production. In: Biotechnology in the Feed Industry, Proceedings of the 12th Annual Symposium. Nottingham University Press, Nottingham, UK. p. 249.

Keady, T.W.J. 1997. Making more profit from grass silage. Alltech UK Fermentation Forum, 13–15 October. p. 16.

Keady, T.W.J. and R.W.J Steen. 1995. The effect of treating low dry matter, low digestibility grass with a bacterial inoculation on the intake and performance of beef cattle, and studies on its mode of action. Grass and Forage Sci. 50:217.

Kennedy, S.J. 1990. Comparison of the fermentation quality and nutritive value of sulfuric and formic acid-treated silages fed to beef cattle. Grass and Forage Sci. 45: 17.

Korzybski, T., Z. Kowszyk-Gindifer and W. Kurylowicz. 1967. Antibiotics: Origin, Nature and Properties, in 2 Vols. Oxford, Pergamon. p. 1651.

Kuhbauch, W. and A. Kleeberger. 1975. Bacterial decomposition of grass fructosan of different degrees of polymerization. J. Br. Grassland. Soc. 30:223.

Kuhbauch,W., G. Voigtlander and A. Leeberger. 1975. Decomposition of grass polysaccharides of different degrees of polymerization by lactobacilli. Proc. 6th Mtng. Eur. Grassland Fed., Madrid. p. 311.

Langston, C.W., H. Irvin, C.H. Gordon, C. Bouma, H.G.Wiseman, C.G. Melin, L.A. Moore and J.R. McCalmont. 1958. Microbiology and Chemistry of Grass Silage. USDA Tech. Bul. No. 1187. p. 73.

Lindgren, S., K.L. Petterson, A. Johnson, P. Lingvail and A. Kasoersson. 1985. Silage inoculation: selected strains, temperature, wilting and practical application. Swed. J. Agric. Res. 15:9.

Mackenzie, D.J. and C.B.Wylam. 1957. Analytical studies on the carbohydrates of grasses and clovers. 8. Changes in carbohydrate composition during the growth of perennial ryegrass. J. Sci. Food. Agric. 8:38.

McDonald, P., A.C. Stirling, A.R. Henderson,W.A. Dewar, G.H. Stark,W.G. Davie, H.T. MacPherson, A.M. Reid and J. Slater. 1960. Studies on ensilage. Edin. Sch. Agric. Tech. Bul. No. 24. p. 83.

McDonald, P., A.R. Henderson and S.J.E. Heron. 1991. The Biochemistry of Silage. Marlow Bottom: Chalcombe. p. 340.

Merry, R.J., A.L.Winters, P.I. Thomas, M. Muller and T. Muller. 1995. Degradation of fructans by epiphytic and inoculated lactic acid bacteria by plant enzymes during ensilage of normal and sterile hybrid ryegrass. J. Appl. Bact. 78:583.

Merry, R.J., A.L.Winters, D.R. Davies, M. Muller, T. Muller and G. Pahlow. 1996. Degradation of fructans during the ensilage of grass with inoculants. Proc. 11th International Silage Conf., University of Wales, Aberystwyth. p. 110.

Moon, N.J. 1981. Effect of inoculation of vegetable processing wastes with Lactobacillus plantarum on silage fermentation. J. Sci. Food. Agri. 32:675.

Muller, T. andW. Seyfarth. 1996. Viable but non-culturable state in epiphytic lactic acid bacteria and resuscitation of the cells. Proc. Internat. Silage Conf., University of Wales, Aberystwyth. p. 240.

Muller, T., M. Muller and W. Seyfarth. 1992. Fermentation of fructans by epiphytic lactic acid bacteria. Proc. 2nd International Symp. Fructan. IGER, Aberystwyth, p. 24.

Muck, R.E. and L. Kung. 1997. Effects of silage additives on ensiling. Proc. Silage: Field to Feedbunk, Hershey, PA. p. 187.

Nilsson, G. and P.E. Nilsson. 1956. The microflora on the surface of some fodder plants at different stages of maturity. Archiv. Mikrobiol. 24:412.

Nilsson, R. and C. Rydin. 1960. The effect of malt enzymes on the biochemical changes occurring during ensilage. Proc. 8th International Grassland Congr. p. 493.

Ohyama,Y., S. Masaki and T. Morichi. 1973. Effects of temperature and glucose addition on the process of silage fermentation. Jap. J. Zootech. Sci. 44:59.

Pahlow, G. and T. Muller. 1990. Determination of epiphytic microorganisms on grass as influenced by harvesting and sample preparation. Proc. 9th Internat. Silage Cong. University of Newcastle-upon-Tyne. p. 23.

Patterson, D.C., C.S. Mayne, F.J. Gordon and D.J. Kilpatrick. 1997.An evaluation of an inoculant/enzyme preparation as an additive for grass silage for dairy cattle. Grass and Forage Sci. 52:325.

Rydin, C. 1961. Malt as a supplement in biological ensiling. Archiv. Mikrobiol. 38:156.

Schlegel, H.G. (1988). General Microbiology. Cambridge: University Press. p. 587.

Schlubach, H.H. and M. Grehm. 1968. Studies on the polyfructosans. 68. The relationship between beta-fructofuranosidase (invertase) and the polyfructosanases. Hoppe-Seyler’s Z. Physiol. Chem. 349:1141.

Stirling, A.C. and R. Whittenbury. 1963. Sources of lactic acid bacteria occurring in silage. J. Appl. Bact. 26:86.

Suttle, N.F. 1978. Sulphur in forages. Proc. Symp., An Foras Taluntais, Dublin. p. 197.

Tramer, J. 1966. Nisin in food preservation. Chemy. Ind, No. 11:446. Virtanen, A.I. 1933. TheAIV method of processing fresh fodder. Emp. J. Exper, Agric., 1:143.

Watson, S.J. and M.J. Nash. 1960. The conservation of grass and forage crops. Edinburgh, Oliver & Boyd. p. 758.

Whittenbury, R. 1961. An investigation of the lactic acid bacteria. Ph.D. Thesis, University of Edinburgh.

Whittenbury, R. 1968. Microbiology of grass silage. Proc. Biochem., Feb. p. 27.

Wieringa, G.W. and T. Beck. 1964. Investigations on the use of cultures of lactic acid bacteria in the preparation of silage in small containers. 1. Obtaining active Lactobacillus cultures for inoculation trials. Wirtschaftseigene Futter 10:34.

Wilkinson, J.M. 1978. The ensiling of forage maize: effects on composition and nutritive value. Forage Maize. London: ARC. (E.S. Bunting, B.E. Pain, R.H. Phipps, J.M. Wilkinson and R.E. Gunn, eds). p. 346.

Wilkinson, J.M., R.F.Wilkinson and T.N. Barry. 1976. Factors affecting the nutritive value of silage. Outlook on Agric. 9:3.

Wilkinson, R.F., J.M. Wilkinson and M.K. Woolford. 1974. The effects of formaldehyde on the silage fermentation. Vaxrodling 29. Proc. 5th Gen. Mtng. Eur. Grassland Fed. p. 197.

Woolford, M.K. 1972. The semi large-scale production, extraction, purification and properties of an antibiotic production by Bacillus licheniformis Strain 2725. J. Appl. Bact. 35:227.

Woolford, M.K. 1975. The significance of Propionibacterium species and Micrococcus lactlyticus to the ensiling process. J. Appl. Bact. 39:301. Woolford, M.K. 1978. The aerobic deterioration of silage. ARC Res. Rec. 4:8.

Woolford, M.K. 1984. The Silage Fermentation. New York: Marcel Dekker. p. 350.

Woolford, M.K. and J.E. Cook. 1980. A note on the effect on the aerobic deterioration of maize silage of the manipulation of the microflora by means of antibiotics. Anim. Food Sci. Technol. 3:89.

Woolford, M.K. and G. Pahlow. 1988. The Silage Fermentation. In: Microbiology of Fermented Foods, 2nd Edition. 2Volumes. (B.J.B.Wood, ed.) London, Chapman & Hall. pp. 72-102.

Woolford, M.K. and R.J. Wilkins. 1975. The evaluation of formaldehyde, bronopol, tylosin and pimaricin as additives in simulated silage. J. Sci. Food. Agric. 26:869.

Woolford, M.K., J.E. Cook, D.M. Hall and A. Bonis. 1980. The use of pimarcin as an additive to improve the aerobic stability of silage. J. Sci. Food. Agric. 31:558.

Wight, P.N. 1990. An investigation of clostridiaphage additives as an aid to silage production. Proc. 9th Silage Conference, University of Newcastleupon- Tyne. p. 106.

Wylam, C.B. 1953. Analytical studies on the carbohydrates of grasses and clovers. 3. Carbohydrate breakdown during wilting and ensilage. J. Sci. Food. Agric. 4:527.

Wyss, U., H. Honig and G. Pahlow. 1991. The influence of air stress and the effect of specific additives on the aerobic stability of wilted grass silage. Wirtschftseigene Futter 37:129.

Zimmer, F. 1980. Efficient silage systems. Occ. Symp. Br. Grassland. Soc. No. 11:186.