The Fundamentals of Making Good Quality
T.A. McAllister and A.N. Hristov
Agriculture and Agri-Food Canada Research Centre, Lethbridge, AB, Canada T1J 4B1
Take Home Messages
8 Harvest the forage at optimal maturity
8 Harvest or wilt to optimal dry matter content
8 Once optimal dry matter is reached, ensile as soon as possible
8 Ensure proper packing and oxygen exclusion.
8 Consider the use of a silage inoculant.
8 Minimize exposure of silage to oxygen both at the silo face and in the feed
The ensiling process can be divided into four principal phases: i) the pre-seal
phase, ii) the active fermentation phase, iii) the stable phase, and iv) the feed-
out phase. In the pre-seal phase, as plant cells and aerobic epiphytic
microorganisms continue to respire, oxygen, water soluble carbohydrates
(WSC) and protein are converted to water, CO2, heat and free ammonia. This
phase continues until either all of the oxygen is utilized or excluded or supplies
of WSC are exhausted. As oxygen levels decline, the active fermentation
phase begins. Anaerobic acetate-producing bacteria (enterobacteria) and other
heterofermentative, lactic acid-producing bacteria (LAB) proliferate. These
bacteria produce ethanol, acetic acid, lactic acid and CO2, utilizing
carbohydrates such as glucose, fructose, xylose and ribose as substrates.
Production of these acids reduces pH, and as the pH declines below 5.0,
heterofermentative bacteria decline and homofermentative bacteria (i.e., those
that produce only lactic acid) begin to dominate the fermentation process. In
good quality cereal silage, growth of LAB and consequent production of lactic
acid continues until the pH of the silage is reduced to between 4.5 and 3.8. At
Advances in Dairy Technology (2000) Volume 12, page 381
382 McAllister and Hristov
these low pH values, providing oxygen exclusion is maintained, growth of all
microorganisms (including LAB) is inhibited, and the silage enters the stable
phase. In this stage, nutrient quality in the silage can be maintained almost
indefinitely. However, upon feed-out, silage is once again exposed to oxygen.
Although the numbers of spoilage microorganisms in the silage are reduced by
fermentation, they are not all killed. Thus, upon exposure to oxygen, these
aerobic microorganisms (e.g., bacteria, yeast) metabolize lactic acid, and the
associated increase in silage pH allows additional spoilage microorganisms
(e.g., molds) to become established. Consequently, the largest losses in silage
dry matter (DM) can occur during the feed-out phase (Rotz et al. 1992). Silage
fermentation is a dynamic process involving interactions among the forage,
microbial populations and the ensiling environment. The objective of the
present paper is to give readers an Consideration of how these various factors
influence the fermentation process and ultimately the nutrient composition and
quality of ensiled feed.
Consideration of the Plant
Next to the rate and extent of oxygen exclusion, forage characteristics at the
time of ensiling is likely the predominant factor which dictates the final quality of
silage. Factors such as type of forage to be ensiled, and maturity, DM content
and WSC content of that forage, all influence the ease of ensiling and ultimately
the quality of silage that is produced.
Type of Forage
In general, cereals are easier to ensile than legumes or grasses, because their
buffering capacity is lower, and their WSC content is higher (Table 1). The
WSC in cereals provide a readily available source of energy that is rapidly
fermented to lactic acid by anaerobic microorganisms. Consequently, pH
decline is usually more rapid, and the final pH is usually lower in cereal silage
than in grass or legume silage (Table 2). Because the buffering capacity of
grasses and legumes is higher, it takes more acid production to achieve the
same reduction in pH decline as obtained in cereal silages. Thus, even when
lactic acid production is similar, the pH of alfalfa silage remains considerably
higher (Table 2) than that of cereal silage. Because fermentation is inhibited at
these low pH values, cereal silages usually contain more residual WSC than do
legume silages. This conservation of WSC in cereals can be advantageous
because the WSC is also a readily available source of energy for the animal.
However, upon exposure to air, these WSC can also be readily utilized by
spoilage microorganisms. As a result, cereal silage is often more prone to
aerobic deterioration than legume silage.
The Fundamentals of Making Good Quality Silage 383
Table 1. Composition and characteristics of barley, corn, and alfalfa prior
Barley Corn Alfalfa
pH 6.64 5.60 6.56
Dry matter (%) 25.7 37.3 30.1
Buffering capacityz 411 91 940
--------------Analysis (g/kg DM)--------
Organic matter 916 962 889
Crude protein 103 97 167
Neutral detergent fibre 454 463 496
Acid detergent fibre 281 224 374
Starch 223 262 NM
Water soluble carbohydrate 109 78.8 36
Ammonia (% of N) 6.3 0.02 23
Expressed as meq/kg DM, calculated from titration from pH 6.6 to 4.0 with
0.1 N HCl.
Adapted from McAllister et al. 1995; 1998 and McAllister and Hristov, unpublished data.
Table 2. Composition of barley, corn and alfalfa silage.
Barley Corn Alfalfa
silage silage silage
pH 3.70 4.20 4.52
Dry matter (%) 25.5 36.4 35.3
-----------Analysis (g/kg DM)-------------
Organic matter 903 946 897
Crude protein 109 106 179
Neutral detergent fibre 432 496 509
Acid detergent fibre 278 240 400
Acid detergent insoluble nitrogen 1.8 1.6 2.7
Water soluble carbohydrate 36 10.9 27.7
Ammonia N (% of N) 7.1 3.3 6.7
Starch 172 283 NMz
Lactic acid 84.6 18.79 86.9
NM: Not measured.
Adapted from: McAllister et al. 1995, 1998; McAllister and Hristov, unpublished data.
As forage matures from the vegetative stage into a reproductive stage (i.e.,
flowering, for alfalfa; heading, for cereals) stems and leaves become more
lignified, and the digestibility of these plant components declines. Concurrently,
WSC levels decline as more carbohydrate is deposited in the plants as
384 McAllister and Hristov
cellulose, hemicellulose and starch. The rate of maturation of a crop is
influenced by several factors, including variety, moisture level, temperature,
nutrient stress and time of season. For example, rate of plant development and
maturation increases with increasing temperature (Fick et al. 1988). This
means that the timing of harvest should be based on physiological
characteristics of the plant, rather than on calender date or time after seeding.
Optimal timing of harvest usually encompasses a compromise between DM
yield, nutrient yield and, with perennial forages, persistence of the stand.
For alfalfa, this optimal harvest time is typically defined as first flower.
Harvesting at earlier stages can increase nutrient yield, but at a cost of
reducing total digestible yield throughout the season as well as stand
persistence. With annual forages, persistence is not a concern, so maximum
nutrient yield becomes the primary consideration. In vitro studies suggested
that DM digestibility differs little among barley silages harvested in the soft-
dough to the late-dough stages (Baron et al. 1992). Other researchers have
proposed that the digestibility of cereals remains constant with advancing
maturity, due to the deposition of highly digestible cereal starch which offsets
the reduction in digestibility associated with lignification of the vegetative
components of the plant (Kilcher and Troelsen 1973; Cherney and Marten
1982). As barley matures, DM increases and WSC levels in the plant decrease
(Table 3). Higher DM content renders barley of later maturity more difficult to
ensile. Because it is more difficult to pack, exclusion of oxygen is less
complete. Furthermore, the lower moisture and WSC concentrations may
impede fermentation, and as a result, the final pH of ensiled barley harvested in
the late-dough stage is usually higher than that of silage from barley harvested
The Fundamentals of Making Good Quality Silage 385
Table 3. Composition of early- and late-dough barley forage before and
Early Late Early Late
dough dough dough dough
pH 6.38 7.62 3.87 4.77
Dry matter (%) 30.1 47.5 35.0 53.5
Buffering capacityz 220 104 NM NM
--------------------Analysis (g/kg DM)--------------------
Organic matter 925.0 928 922 930
Crude protein 132.4 126.3 135 120
Neutral detergent fibre 530.0 528 464 513
Acid detergent fibre 225.7 275.3 245 279
Starch 198.0 322.8 185 225
Water soluble carbohydrate 48.3 12.2 32.1 19.6
Ammonia (% of N) 0.92 0.63 6.6 6.3
Lactic acid (g/kg DM) NM NM 97 21
DM intake (kg/d) NM NM 5.75 5.54
Gain (kg/d) NM NM 0.989 0.772
Feed efficiency (feed gain-1) NM NM 6.0 7.3
Expressed as meq/kg DM, calculated from titration from pH 6.6 to 4.0 with 0.1 N HCl.
(McAllister and Hristov, unpublished data).
Barley harvested at the late-dough stage contains more starch than when
harvested at early-dough, but our findings suggest that this higher starch
content does not offset the decline in the digestibility of the vegetative
components of the plant. In fact, we observed that intake, average daily gain
and feed efficiency were reduced by 3.7, 22 and 22%, respectively, when
feedlot cattle were fed barley ensiled at the late-dough stage as compared to
the early-dough stage. This apparent discrepancy between laboratory
predictions and feedlot trials may be attributable to methodology. Typically,
relative digestibilities are determined among groups of forage samples (e.g., at
varying stages of maturity) by measuring the plant material remaining after the
samples are incubated in ruminal fluid for 48 h (Cherney and Marten 1982;
Baron et al. 1992). Plant material not fermented within this period of time
represents the indigestible fraction of the silage. For these in vitro studies,
silage samples are usually ground through a 1-mm screen. Thus, the starch in
these samples is readily accessible to the ruminal microorganisms that carry
out the fermentation. In contrast, we observed that a large portion of the barley
kernels in silage harvested in the late-dough stage in our study remained intact
during ensiling and eating, and were visible in the feces. It seems possible that
a significant proportion of the whole kernels observed in the feces of cattle fed
diets of barley grain and late maturity barley silage may actually have originated
386 McAllister and Hristov
in the silage component, as opposed to the grain portion, of the diet. This
hypothesis is supported by the fact that the average apparent digestibility of
barley silage DM at the late-dough stage (59%) was lower than estimates (61 to
68%) given for barley ensiled in the early- to mid-dough stage (Christensen et
al. 1977; Hingston and Christensen 1982, Acosta et al. 1991; McCartney and
Vaage 1994; McAllister et al. 1995). Measured in vivo, therefore, higher starch
content in late-dough barley silage did not compensate for the maturity-
associated reduction in digestibility of the fiber components of barley silage.
Including kernel processors on the silage choppers used to prepare corn silage
for dairy cattle reduced the amount of whole corn kernels appearing in the
feces, and increased the cows' milk production by 9 to 10% (Harrison et al.
1997). A kernel processor may result in similar improvements in milk
production in cattle fed barley silage harvested at the late-dough stage.
Forage Dry Matter Content
In some cases, it is desirable to increase the DM content of forage. High
moisture silages (20 to 27% DM) promote a very active fermentation and they
are often associated with increased in seepage losses from the silo.
Furthermore, intake of high moisture silages also tends to be reduced relative
to intake of forage ensiled at optimal (27 to 38%) DM (Demarquilly et al. 1977;
Thomas et al. 1961). As discussed above, DM content of forage tends to
increase with advancing maturity of the crop, but silage DM can also be
increased by wilting a less mature forage in the field prior to ensiling. In this
case, the higher DM is achieved without the increased lignification associated
with more mature plant cell walls. Thus, wilting can be used as an effective tool
to elevate forage DM into an acceptable range for ensiling. Over wilting the
forage, however, can reduce silage quality. Loss of DM from forage during
wilting can be as high as 4% per day (McDonald et al. 1991) and prolonged
plant respiration can further reduce WSC levels. Reducing available WSC
reduces the amount of lactic acid produced. Consequently, the pH of wilted
silage is often higher than silage that is chopped directly (Table 4). Lower WSC
levels may also account for the lower rate of digestion and effective
degradability observed with wilted silage as compared to direct-cut silage.
Wilting, therefore, can also affect ruminal digestive characteristics as well as
nutrient value of silages, although not to as great an extent as forage maturity.
The Fundamentals of Making Good Quality Silage 387
Table 4. Effect of wilting on the characteristics of barley silage in
Ensiled Ensiled Treatment
directly after wilting effect
pH 4.7 5.1 ***
Dry matter (%) 30.7 37.8 ***
Water soluble carbohydrate (g/kg) 52.0 38.0 ***
Lactic acid (g/kg) 37.7 23.9 ***
Soluble protein (% of N) 49.6 45.9 ***
Rate of degradation (% per h) 3.8 1.9 *
Soluble fraction (%) 32.4 36.3 NS
Degradable fraction (%) 44.4 54.7 NS
Effective degradability (%) 51.4 46.3 **
Significance: * = (P < 0.05), ** = (P < 0.01) and *** = (P < 0.001).
NS = not significant (P > 0.05).
(Hristov and McAllister, unpublished data).
Although there is no doubt that forage variety can impact in vitro digestibility of
silages, this varietal effect has seldom resulted in consistent changes in animal
performance (i.e., milk production, growth rate). Neutral detergent fibre
(NDF)digestibilities differing by 7 to 8 percentage points have been measured
among corn hybrids (Vattikonda and Hunter 1983), although these differences
were less dramatic when multiple-year studies were conducted in numerous
locations (Allen et al. 1993). Barley silages differed in their chemical
composition, but not in their in vitro digestibility (Edwards et al. 1968), nor did
barley variety affect milk production in dairy cows (Kennelly and Khorasani
1999). Assessing the effect of forage variety on animal performance is
technically difficult, as it is almost impossible to ensure identical harvest
conditions among the varieties being studied. In general, varietal differences
are overshadowed by differences in growing conditions, harvest time and
ensiling practices. A variety selected for optimal silage production will still
result in poor quality silage if it is harvested at improper maturity or it is poorly
ensiled. As one cannot control the weather, selection for maximum yield may
still be the best bet when trying to select a variety for silage.
Consideration of the Microbial Populations
Epiphytic Microbial Populations
Epiphytic microbial populations refers to those microorganisms associated with
the forage prior to ensiling. The survival and activity of these populations is
388 McAllister and Hristov
also among the factors influenced by the characteristics of the crop at the time
of ensiling. For example, epiphytic microbial populations tend to decline with
increasing forage DM and with increased exposure to solar radiation. This may
also partially explain why ensiling high (> 40%) DM forages is difficult. It has
been proposed that the forage should contain at least 1,000,000 colony forming
units (CFU) per gram for successful fermentation (Spoelstra 1991). In our
experience, however, the microbial populations associated with barley are
seldom that high, yet the silage produced is still of acceptable quality
(McAllister et al. 1995; McAllister, unpublished data). The types of lactic acid
bacteria present in the epiphytic populations, as opposed to the total numbers
of bacteria, may be more important in determining the efficiency of the
fermentation process. Homofermentative LAB produce only lactic acid as an
end product, whereas heterofermentative LAB produce additional end products
such as acetic acid. Enhanced production of lactic acid can cause a more rapid
decline in silage pH, thereby conserving the nutrients in the silage. This
relationship is the foundation for selection of bacterial strains as silage
Forage destined for the silo may also harbor undesirable epiphytic
microorganisms that can contribute to spoilage and could be a potential health
risk. Application of manure onto forage prior to ensiling can increase numbers
of epiphytic enterobacteria such as Bacillus and Clostridium spp. (Rammer et
al. 1994). Contact of the forage with soil can also increase yeast and mold
counts in the silage. Although these microorganisms are usually inactivated
during ensiling, they can become active and contribute to accelerated spoilage
when the silage is exposed to air upon feeding.
Selection of microbes for inclusion in a silage inoculant is the principal factor
that will influence the impact of the product on silage fermentation and
subsequently, animal performance. Lactobacillus plantarum, Lactobacillus
acidophilus, Enterococcus faecium, Pediococcus acidilactici and Pediococcus
pentosaceus are the most frequently used species, but other species such as
Streptococcus bovis (Jones et al. 1991), Streptococcus thermophilus, Serratia
rubidaea (Phillip and Fellner 1992) and Lactobacillus buchneri (Weinberg and
Muck 1996) have been considered. Many recently developed inoculants
contain multiple species, building on evidence that growth of one bacterial
species may facilitate growth of another (Fitzsimons et al. 1992). For example,
Pedicococcus spp. and Enterococcus spp. grow more rapidly and are more
tolerant of high DM conditions than are Lactobacillus spp. Rapid growth of
Pedicococcus and Enterococcus increases the rate of acid production in
recently ensiled forage, and the rapid decline in pH facilitates the establishment
of lactobacilli as predominant microorganisms in the fermentation process.
This synergism has only been demonstrated experimentally with combinations
The Fundamentals of Making Good Quality Silage 389
of two or three species, however. Other than as a marketing feature, there is
likely little benefit to including numerous species in a product.
Selecting particular strains of a single species for inclusion in an inoculant is
likely just as important as deciding what combination of species to include.
Several studies have shown that fermentation responses differ widely among
strains of the same species (Woolford and Sawczyc 1984; Hill 1989; Fitzsimons
et al. 1992). Thus, two products containing identical bacterial species, but
different strains, could exert different effects on silage fermentation and,
consequently, animal performance. This makes it extremely difficult to assess
the relative value of inoculants simply by reading the product label. Reputable
inoculant producers will employ a screening process to select component
species based on the properties described above. Some organizations have
also developed more advanced selection criteria that include targeting
inoculants for specific crops, improved aerobic stability, pathogen control and
improved animal performance. Improved animal performance is of particular
importance because it is often the principal economic justification for using
inoculants. Obviously, organizations that use these more stringent selection
criteria in developing inoculants are more likely to develop products that will
positively affect fermentation and animal performance.
The mechanisms by which inoculants may alter silage fermentation and
potentially improve animal performance are numerous. Additional benefits may
arise from nutrient conservation due to improved DM recovery (Henderson
1993) and aerobic stability (McAllister et al. 1995). Presently, it is difficult to
predict which of these responses is most likely to elicit a positive effect on
animal performance. In fact, improvements in growth performance have been
reported with no measurable changes in fermentation (McAllister et al. 1998).
Perhaps the most frequently observed effects of inoculants are accelerated
post-ensiling decline in pH and increased numbers of lactobacilli (Figure 1,
Hristov and McAllister, unpublished). Water soluble carbohydrates in the silage
are conserved by the increased rate of fermentation (Gordon 1989), and
ammonia levels are reduced as a result of suppressed deamination and
proteolysis (Heron et al. 1989). Higher lactic acid levels can also raise
propionate levels in the rumen. This may give rise to the energetic advantages
associated with increased propionate absorption (Van Vuuren et al. 1995).
Conservation of protein and WSC is likely more critical with legume and grass
silages than with cereal silages, as the former contain higher levels of protein
and lower levels of WSC. Inoculants commonly reduce the pH of grass and
legume silages, but seldom alter pH in cereal silages, in which the final pH
often falls below 4 even without additives.
390 McAllister and Hristov
Figure 1. Effect of inoculants on (a) the rate of pH decline and (b)
numbers of lactic acid-producing bacteria (LAB) in barley silage (Hristov
and McAllister, unpublished
To our knowledge, there are no published reports on the effect of inoculants on
the performance of cattle fed barley silage. In a series of 19 studies with corn
silage, inoculants improved DM recovery by 1.3% and feed efficiency by 1.8%
(Bolsen et al. 1992), but improved animal performance was observed in only
three of the 14 peer-reviewed studies published in North America (Kung 1998).
Of course, comparisons such as these must be taken in the context that rate of
inoculant application, species of bacteria, moisture levels, and diet composition
differed markedly across the studies.
The Fundamentals of Making Good Quality Silage 391
In a series of 14 lactation studies, the inoculant L. plantarum MTD1 improved
DM intake by 4.8% and milk production by 4.6% when it was applied to grass,
corn or alfalfa (Moran and Owen 1994). A single inoculant from a US
manufacturer was used in five studies with lactating dairy cows and in four
studies with beef cattle. Milk production increased by an average of 0.816 kg/d
and average daily gain by 11.9% (Kung and Muck 1997). Assuming the
inoculants improved DM recovery by 1.25 to 2.5% and milk production by 0.1 L
per cow per day, the net returns offered by these additives were estimated at
US $5.76 and $14.40 per tonne of corn and alfalfa silage, respectively (Table 5;
Bolsen et al. 1999).
Table 5. Effect of inoculantz on efficiency of silage preservation and feed
utilization by dairy cows.
Corn silage Alfalfa silage
CONT TRT CONT TRT
DM recovery (kg/tonne) 900 912.5 875 900
Gain in feed DM (kg/tonne) 12.5 25
DM intake by cows (kg/d) 16.0 16.0 13.0 13.0
Cow days fed/tonne DM 56.2 57.0 67.3 69.2
Gain in cow days (per tonne DM) 0.78 1.92
Gain in milk (L/tonne DM)y - 28.5 - 72.2
Value of milk gain ($/tonne)x - 7.84 - 19.86
Return from increase in milk ($)w - 1.56 - 1.90
Net return ($/tonne)v - 5.76 - 14.40
CONT = uninoculated silage; TRT = inoculated silage.
Calculated using milk yield of 38 L/cow≅day.
Calculated using return of 0.275 $/L milk.
Assuming increase in milk production of 0.1 L/cow≅day due to inoculation, irrespective of increase
in DM yield during ensiling.
Factoring in increased DM yield, and cost of inoculation.
(from Bolsen et al. 1999).
Positive animal responses appear to occur more frequently with some
inoculants than with others. Unfortunately, the majority of inoculants marketed
have never been evaluated in animal studies. Further, the practice of
extrapolating results from one product to those of another product is
scientifically invalid and likely no better than a random guess at the possible
response that may be achieved. For example, we found that average daily gain
and feed efficiency of feedlot cattle were entirely different when the alfalfa
silage they were fed had been inoculated with L. plantarum alone, as compared
with the same strain of L. plantarum in combination with E. faecium (McAllister
392 McAllister and Hristov
et al. 1998). In general, a product for which research exists to support animal
response claims is likely more valuable than one without such research.
Consideration of the Ensiling Environment
Selection of a Silo
Several factors, including capital investment, amount of forage to be ensiled
and associated machinery costs, must be considered when selecting a silo
system. In general, more expensive silo systems such as concrete or oxygen
limiting towers are associated with lower DM losses, but increased capital
expenditure per tonne of forage ensiled (Table 6). Bunker silos are relatively
inexpensive to construct, but suffer from the highest DM losses and are the
least forgiving of errors in silage management. For example, attempts to ensile
forage at 50% DM in a bunker silo could be a recipe for disaster, whereas
storage of this same material in a tower or plastic tube silo may produce silage
of acceptable quality. This difference reflects the fact that compared to bunker
silos, tower and plastic bag silos offer greater control over oxygen exclusion
and packing density, and are less affected by weather. Dry matter losses from
bagged silage are low, and bagging is an economical method of storing
relatively small amounts of silage for short periods of time. A silage bag with a
9-ft diameter will store about 1 ton of silage per lineal foot. However, because
deer, coyotes and rodents have been known to puncture the bags to access
silage, the bags should be inspected regularly for holes. As the amount of
forage to be ensiled increases, the attractiveness of tower and plastic tube silos
declines due to their limited storage capacity. Low construction cost makes the
bunker silo the most desirable method of storage for amounts of silage over
The Fundamentals of Making Good Quality Silage 393
Table 6. Comparison of estimated feed losses and silage production costs
using various silage systems under good conditions
Horizontal Concrete Oxygen Silage Round
pit tower limiting bags bale
tower (tubes) silage
Moisture content (%) 65 65 45 65 65
Loss (%) due to:
Respiration and 4 4 6 4 4
Harvesting loss 2 2 3 2 4
Storage loss 15 9 5 7 18
Feeding loss 4 2 2 4 4
Total loss 25 17 16 17 30
Cost comparisons ($/ton DM)
(preparing 500 tons)
Total cost 31.32 NC NC 50.27 34.80
Fixed cost 19.45 NC NC 28.67 13.13
(preparing 1500 tons)
Total cost 23.95 NC NC 34.53 24.84
Fixed cost 12.20 NC NC 12.85 6.47
Adapted from the Alberta Agriculture Silage Manual and the Manitoba Agriculture Information
Exclusion of Oxygen in Bunker Silos
Although plant cells and aerobic microorganisms utilize oxygen after the forage
is ensiled, steps must be taken to minimize the penetration of oxygen into the
forage from the surrounding atmosphere. In bunker silos, exclusion of oxygen
is obtained by packing the silage to an optimal density and sealing the packed
silage with an atmospheric barrier such as polyethylene sheets. Packing
increases the density of the forage in the silo and consequently reduces
porosity and the penetration of air into the silage mass (Ruppel 1993). As air is
excluded, aerobic respiration is precluded and losses in DM are minimized
(Table 7). Also, storage of forage at a higher density increases the capacity of
the silo and as a result reduces the capital investment in the storage structure
per tonne of ensiled forage.
394 McAllister and Hristov
Table 7. Effect of silage density on dry matter loss from silage
Silage density DM loss at 180 d
(kg DM/m3) (% of DM ensiled)
Adapted from Rupple (1992).
The density of the silage will depend on the type of implement used during
packing as well as the total time spent packing per tonne of forage. Packing
density is maximized by using implements with tires that apply the greatest
weight per unit of surface area. Several factors including forage delivery rate,
number and weight of packing tractors, and total time spent packing can
influence the packing time per tonne of forage. Optimally, forage should be
packed at a rate of 1 to 4 minutes per tonne of silage. Packing rates lower than
1 min per tonne of forage may indicate that the forage is being delivered to the
bunker too quickly, that there are insufficient packing tractors, or that the
packing layer thickness is too large. Maintaining a shallow packing layer of 15
cm or less can also be used as management strategy to increase packing
density. Surprisingly, chopping silage to a finer particle size was shown not to
be related to packing density (Ruppel 1993). It is recommended that silage be
packed to a minimal density of 225 kg of DM per m3. The improved DM
recovery associated with packing to higher densities (e.g., 240 to 260 kg of DM
per m3 ) has to be weighed against the increased labor and machinery costs
associated with longer packing times. Extrapolation from work with corn silage
in the United States (Bolsen, unpublished data) suggests that the improved DM
recovery associated with similar increases in barley silage packing density
would translate to savings of $0.50 to 1.00 per tonne.
Because bunker silos have substantial surface area exposed to the
atmosphere, penetration of both air and moisture can affect the ensiling
process and consequently, the nutritional value of the silage (Bolsen 1997).
Polyethylene sheeting weighted with tires is the most common and effective
method of excluding oxygen and moisture. Other surface coverings such as
limestone, molasses and sawdust are of limited value. The economic savings
of covering bunker silos has been well documented, to the point that it is the
“no-brainer” of silage management. Losses of DM in an uncovered silo are
extensive. Up to 30% of the DM may be lost from the uppermost 1 m of silage
which, in a typical bunker silo, can represent 25% of the total forage in the silo.
Economic losses through spoilage and DM losses are over 20 times the cost of
covering a bunker silo with plastic (Table 8).
The Fundamentals of Making Good Quality Silage 395
Table 8. Comparison of costs (US $) of covering a 9' × 40' × 100' bunker
silo with plastic or leaving silage uncovered
Cost per Cost per ton Cost for
square foot as-fed silo
Cover material ($) ($) ($)
Plastic - black, 6-mil 0.023 0.11 92
Plastic - black and white, 6-mil 0.034 0.16 136
Silagez (valued at $30/ton) 0.653 2.99 2,612
Silage (valued at $100/ton) 2.175 9.97 8,700
Calculation based on loss of 30% more silage DM in the top 3' when bunker silos are not covered,
as compared to bunker silos covered with plastic (Bolsen et al. 1993).
Black or white plastics of varying thicknesses are available for covering silos.
White plastic may radiate more heat, and be more resistant to deterioration via
UV rays, but is more expensive than black plastic. Once adequate plastic is
chosen, research shows that increasing the number of tires used to hold down
the plastic may be an effective method of further reducing surface spoilage
Consideration of the Feed-Out Phase
Given that the majority of silage DM loss takes place after the forage is ensiled
(Table 6), procedures adopted during the feed-out phase represent an
important component of any total silage management program. Some silages
begin to heat within hours of aerobic exposure, whereas others remain stable
for days or even weeks (McDonald et al. 1991). Legume silages with low DM
and WSC contents are generally more resistant to aerobic deterioration than
are cereal silages, in which WSC concentrations are sufficiently high to
promote the growth of yeasts and molds (Weinberg et al. 1993). Furthermore,
legume silages may also contain other unidentified microbial inhibitors that
prevent the growth of spoilage microorganisms (Jones et al. 1995). We have
found that some inoculants can improve the aerobic stability of barley silage by
inhibiting the growth of both yeast and molds (Figure 2, McAllister et al. 1995).
To minimize aerobic deterioration, silos should be designed to match the feed
demand rate such that at least 4 to 6 inches of silage are removed from the
face on a daily basis. This technique prevents any one location in the silo from
being exposed to oxygen for a period of time sufficient to foster proliferation of
the aerobic microorganisms responsible for spoilage. Accumulation of loose
silage at the bottom of face should also be avoided as this unpacked material is
especially vulnerable to rapid aerobic decomposition.
396 McAllister and Hristov
Figure 2. Changes in the populations of yeasts (a) and molds (b) in
uninoculated (control) barley silage and in barley silage inoculated with
product 1174 or 2637 during 13 d of aerobic exposure (McAllister et al.
To make good quality silage, one must have an appreciation of the plant,
microbial and environmental factors that influence the fermentation process and
ultimately, the nutrient value of the silage. These factors must be considered
as an integrated package, as neglect of any one component can lead to a
breakdown in this forage preservation process. Silage inoculants can facilitate
the ensiling process, but they are not a replacement for paying attention to the
fundamental factors (plant maturity, dry matter content, oxygen exclusion) that
are the keys to making good quality silage.
The Fundamentals of Making Good Quality Silage 397
Acosta, Y.M., Stallings, C.C., Polan, C.E. and Miller, C.N. 1991. Evaluation of
barley silage harvested at boot and soft dough stages. J. Dairy Sci.
Allen, M.S., Oba, M. and Choi, B.R. 1993. Nutritionist’s perspective on corn
hybrids for silage. In: Silage: Field to Feedbunk, Proceedings from the
Silage: Field to Feedbunk North American Conference, Hershey,
Pennsylvania, pp. 25-36.
Baron, V.S., Dick, A.C. and Wolynetz, M.S. 1992. Characterization of barley
silage - maturity relationships for central Alberta. Can. J. Plant Sci.
Bolsen, K.K., Sonon, R.N., Dalke, B., Pope, R., Riley, J.G. and Laytimi, A.
1992. Evaluation of inoculant and NPN silage additives. A summary of
26 trials and 65 farm-scale silages. Report of Progress, Kansas State
Bolsen, K.K. 1997. Issues of top spoilage losses in horizontal silos. In: Silage:
Field to Feedbunk, Proceedings from the Silage: Field to Feedbunk
North American Conference, Hershey, PA, pp. 137-150.
Bolsen, K.K., Huck, G.L., Siefers, M.K., Schmidt, T.E., Pope, R.V. and Uriarte,
M.E. 1999. Silage management: five key factors. Kansas State
University, Manhattan, KS.
Christensen, D.A., Owen, B.D., Steacy, G., Crowle, W.L. and Mtimuni, J.P.
1977. Nutritive value of whole crop silage made from seven cereal
cultivars. Can. J. Anim. Sci. 57:537-542.
Cherney, J.H. and Marten, G.C. 1982. Small grain crop potential: II.
Interrelationship among biological, chemical, morphological and
anatomical determinants of quality. Crop Sci. 22:240-245.
Demarquilly, C. and Dulphy, J.P. 1977. Effect of ensiling on feed intake and
animal performance. In: Proceedings of International Meeting on Animal
Production from Temperate Grasslands, Dublin, pp. 53-61.
Edwards, R.A., Donaldson, E. and MacGregor, A.W. 1968. Ensilage of whole-
crop barley. I. Effects of variety and stage of growth. J. Sci. Food Agric.
Fick, G.W., Holt, D.A. and Lugg, D.G. 1988. Environmental physiology and
crop growth. In: Alfalfa and Alfalfa Improvement, Eds. A.A. Hanson et al.
Agron. Mongr. 29. ASA, CSSA and SSSA, Madison, WI, pp. 163-194.
Fitzsimons, A., Duffner, F., Curtin, D., Brophy, G., O’Kiely, P. and O’Connell, M.
1992. Assessment of Pediococcus acidilactici as a potential silage
inoculant. Appl. Environ. Microbiol. 58:3047-3052.
Gordon, F.J. 1989. An evaluation through lactating cattle of bacterial inoculant
as an additive for grass silage. Grass Forage Sci. 44:169-179.
Harrison, J.H., Lohnson, L., Hunt, C. and Siciliano-Jones, J. 1997. Use of
kernel-processed silage in dairy rations. In: Silage: Field to Feedbunk,
398 McAllister and Hristov
Proceedings from the Silage: Field to Feedbunk North American
Conference, Hershey, PA, pp.95-110.
Henderson, N. 1993. Silage additives. Anim. Feed Sci. Technol. 45:35-56.
Heron, S.J.E., Edwards, R.A. and Phillips, P. 1989. The effect of pH on the
activity of ryegrass proteases. J. Sci. Food Agric. 46:267-277.
Hill, H. A. 1989. Microbial ecology of lactobacilli in silage. In: Food for
Thought - Proceedings of the 2nd Forage Symposium, Pioneer Hi-Bred
International, Johnston, IA, pp. 47-64.
Hingston, A.R. and Christensen, D.A. 1982. The effect of type of silo and
formic acid preservation on the nutritive value of barley, wheat and oat
silages for growing Hereford steers. Can. J. Anim. Sci. 62:155-162.
Jones, B.A., Muck, R.E. and Ricke, S.C. 1991. Selection and application of
Streptococcus bovis as a silage inoculant. Appl. Environ. Microbiol.
Jones, B.A., Muck, R.E. and Hatfield, R.D. 1995. Red clover extracts inhibit
legume proteolysis. J. Sci. Food Agric. 67:329-333.
Kilcher, M.R. and Troelsen, J.E. 1973. Contribution and nutritive value of major
plant components of oats through progressive stages of development.
Can. J. Plant Sci. 53: 251-256.
Kennelly, J. and Khorasani, R. 1999. Nutritive values and fermentation
characteristics of barley silage varieties. In: Silage Production in Relation
to Animal Performance, Animal Health, Meat and Milk Quantity. XII
International Silage Conference, Uppsala, Sweden, pp. 104-105.
Kung, L. Jr. 1998. Direct-fed microbial and enzyme feed additives. In: Direct-
fed Microbial, Enzyme and Forage Additive Compendium. Vol. 4. The
Miller Publishing Company, Minnetonka, MN, pp. 15-19.
Kung, L. Jr. and Muck, R.E. 1997. Animal responses to silage additives. In:
Silage: Field to Feedbunk, Proceedings from the Silage: Field to
Feedbunk North American Conference, Hershey, PA, pp. 200-210
McAllister, T.A., Selinger, L.B., McMahon, L.R., Bae, H.D., Lysyk, T.J., Oosting,
S.J. and Cheng, K.-J. 1995. Intake, digestibility and aerobic stability of
barley silage inoculated with mixtures of Lactobacillus plantarum and
Enterococcus faecium. Can. J. Anim. Sci. 53:171-181.
McAllister, T.A., Feniuk, R., Mir, Z., Mir, P., Selinger, L.B. and Cheng, K.-J.
1998. Inoculants for alfalfa silage: Effects on aerobic stability,
digestibility and growth performance of feedlot steers. Livest. Prod. Sci.
McCartney, D.H. and Vaage, A.S. 1994. Comparative yield and feeding value
of barley, oat and triticale silage. Can. J. Anim. Sci. 74:91-96.
McDonald, P., Henderson, N. and Heron, S. 1991. The Biochemistry of Silage.
Chalcombe Publications, Marlow, UK.
Moran, J.P. and Owen, T.R. 1994. The effects of Ecosyl-treated silage on milk
production by lactating dairy cows. Proceedings of the National
Conference on Forage Quality, Evaluation and Utilization. University of
Nebraska, Lincoln, NE.
The Fundamentals of Making Good Quality Silage 399
Phillip, L.E. and Fellner, V. 1992. Effects of bacterial inoculation of high-
moisture ear corn on its aerobic stability, digestion and utilization for
growth of beef steers. J. Anim. Sci. 70:3178-3187.
Rammer, C., Ostling, C., Lingvall, P. and Lindgren, S. 1994. Ensiling of
manured crops - effects on fermentation. Grass Forage Sci. 49:343-351.
Rotz, C.A., Pitt, R.E., Muck, R.E., Allen, M.S. and Buckmaster, D.R. 1992.
Direct-cut harvest and storage of alfalfa on the dairy farm. In: American
Society Agriculture Engineers, ASAE Paper No. 921001, St. Joseph, MI.
Ruppel, K.A. 1993. Bunker silo management and its effects on haycrop quality.
In: Silage Production from Seed to Animal, Proceedings from the
National Silage Production Conference, Syracruse, New York pp. 67-84.
Spoelstra, S.F. 1991. Chemical and biological additives in forage conservation.
In: Proceedings on Forage Conservation - Towards 2000. Eds. G.
Pahlow and H. Honig. Braunschweig, Germany, pp. 48-70.
Thomas, J.W., Moore, L.A., Okamoto, M. and Sykes, J.F. 1961. A study of the
factors affecting rate of intake of heifers fed silage. J. Dairy Sci. 44:1471-
Van Vuuren, A.M., Huhtanen, P. and Dulphy, J.P. 1995. Improving the feeding
and health value of ensiled forages. In: Recent Developments in the
Nutrition of Herbivores. Eds. M. Journet, E. Grenet, M.-H. Farce, M.
Theriez, C.M. Demarquilly. Proceedings of the IV International
Symposium on the Nutrition of Herbivores. INRA Editions, Paris, France,
Vattikonda, M.R. and Hunter, R.B. 1983. Comparison of grain yield and whole-
plant silage production of recommended corn hybrids. Can. J. Plant Sci.
Weinberg, Z.G. and Muck, R.E. 1996. New trends and opportunities in the
development and use of inoculants for silage. FEMS Microbiol. Rev.
Weinberg, Z.G., Ashbell, G., Hen, Y. and Azrieli, A. 1993. The effect of
applying lactic acid bacteria at ensiling on the aerobic stability of silages.
J. Appl. Bacteriol. 75:512-518.
Woolford, M.K. and Sawczyc, M.K. 1984. An investigation into the effect of
cultures of lactic acid bacteria on fermentation in silage. 1. Strain
selection. Grass Forage Sci. 39:139-148.