NUTRITION, FEEDING, AND CALVES
Additives Containing Bacteria and Enzymes for Alfalfa Silage'
A. C. SHEPERD,* M. MASLANKA,s D. QUINN,4 and L. KUNG, JR5
Department of Animal Science and Agricultural Biochemistry
ABSTRACT
First-cutting alfalfa was wilted, har-
vested from alternate rows, left untreated
or treated with additives containing lactic
acid bacteria and enzymes (cellulase,
amylase, and pectinase), and ensiled in
bag silos. Inoculation increased lactic
acid bacteria from 5 x IO4 to 1 x 106
cfdg of forage. Because treatments were
bagged consecutively, the DM of treated
silages was higher than that of untreated
silage. However, after 4 d of ensiling, the
pH of treated silage, about 4.3, was
lower than that of untreated silage, 4.7,
and remained lower throughout the ensil-
ing period. After 177 d of ensiling, total
lactate was about 25% higher, and am-
monia N was about 40% lower, in
treated silage. In addition, NDF and
ADF contents were lower in treated than
in untreated silage. Between 51 and 177
d of storage, glucose content increased in
treated silage, but not in untreated silage,
suggesting that some plant cell-wall
hydrolysis occurred during prolonged
storage. In vitro digestion of NDF did
not differ among treatments during early
incubation, but the extent of digestion
after 36 and 48 h was lower in treated
than in untreated silage. The microbial
and enzyme silage additives used in this
Received December 22, 1993.
Accepted November 4, 1994.
'Published as Miscellaneous Paper Number 1499 of
the Delaware Agricultural Experiment Station.
2Present address: Department of Dairy Science,
University of Wisconsin, Madison 53706.
3Present address: Department of Animal Science,
University of Minnesota, St. Paul 55108.
4Present address: 18 Wexford Road, Gibbsboro, NJ
08026.
5Corresponding author.
1995 J Dairy Sci 78565-572
Delaware Agricultural Experiment Station
College of Agricultural Sciences
University of Delaware
Newark 19717-1303
study improved fermentation characteris-
tics and reduced fiber content of silage
but decreased the in vitro digestibility of
fiber.
(Key words: silage, alfalfa, inoculant,
enzymes)
Abbreviation key: ALF = Alfazymem, LAB
= lactic acid bacteria, SFP = Silage Fresh
PlusW.
INTRODUCTION
During the anaerobic stage of ensiling,
epiphytic lactic acid bacteria (LAB) lowered
pH by producing lactic acid from soluble car-
bohydrates. The addition of exogenous LAB to
alfalfa has resulted in a more homolactic fer-
mentation (12,21). Changes in silage composi-
tion that were due to inoculation with LAB
included a more rapid decline in pH, a de-
crease in ammonia N because of inhibition of
proteolysis, decreases in acetate and butyrate,
and an increase in lactic acid content (11).
If a forage has a low to marginal content of
soluble carbohydrates, fermentable substrate
may limit the ensiling process. Addition of
enzymes that are capable of increasing the
amount of substrate available for fermentation
could alleviate this problem. Ideally, addition
of enzymes hydrolyzed polysaccharides rapidly
enough to produce glucose, which would be
converted to lactate by the LAB. Furthermore,
although hydrolysis of cellulose later in the
ensiling period had no effect on fermentation,
forage quality may be improved if enzymes
predigest parts of the rapidly degradable cell-
wall fraction. For this reason cellulase en-
zymes have been added to silages. These en-
zymes are a mixture of endoglucanase, exo-
glucanase, and cellobiase, which break the P-
(1,4)-linkages of cellulose to release polysac-
charides and, eventually, glucose monomers
(14). Addition of cellulase enzymes to silage
565
566 SHEPERD ET AL.

has been extensively studied, but the results within treatment and stored on ice prior to
have been variable. Leatherwood et al. (13) processing.
reported that addition of enzymes increased Numbers of LAB in the water-inoculant
cellulose hydrolysis and decreased the pH of mixes and on fresh forage were enumerated on
alfalfa silage. Henderson et al. (5) found an spread plates using MRS agar (DeMan,
increase in cellulose hydrolysis with addition Rogosa, and Sharpe; Difco Laboratories,
of cellulase enzymes but no changes in fer- Detroit, MI) and incubated aerobically at 30°C
mentation characteristics of grass, alfalfa, and for 48 h. Silage samples were taken 2, 4,8, 14,
clover silage. However, other researchers (2, 7, 51, and 177 d after ensiling. At each sampling,
10) reported no effects of enzyme additives on three samples were taken from the length of
silage fermentation. For sheep, DMI and timo- each bag silo. Samples were mixed well, and a
thy silage digestibility were decreased by addi- 10-g portion was homogenized in 100 ml of
tion of a cellulase-hemicellulase mixture (15). sterile Ringer's solution (Oxoid, Hampshire,
For cows, Stokes (21) reported that combined England) for 1 min, and pH was determined. A
LAB and enzyme treatment negatively affected portion of the homogenate was filtered through
DMI but increased milk production over that Whatman number 54 filter paper (Clifton, NJ),
for a mixed grass-legume silage inoculated acidified with 6N HCI, and centrifuged at
with LAB. 15,000 x g for 10 min. Samples were stored at
The objective of this study was to compare -20°C until analysis. Silage water extracts
the effects of two commercial silage additives were analyzed for acetic and lactic acids by
containing LAB and enzymes on fermentation gas chromatography (model 5890A; Hewlett-
characteristics and in vitro NDF digestibility of Packard Co., Avondale, PA) using a 10-m,
alfalfa silage. 530-pm macro bore Carbowax M column (Su-
pelco, Inc., Bellefonte, PA). Helium was used
as the carrier gas with a flow rate of 10 ml/
MATERIALS AND METHODS min. Sample (1 pl) was injected with a split
First-cutting alfalfa was at approximately ratio of 8:1. Injection port temperature was
one-tenth bloom and was cut with a mower- 200'C, and the detector temperature was
conditioner from one field. After wilting for 250°C. The oven was programmed for temper-
approximately 2 d, forage was chopped to a ature as follows: 70°C for 1 min, 5'C/min
theoretical length of .95 cm and stored in silo increase to l W C , 45Wmin increase to 170"C,
bags. Forage was wilted on clear, cool days, and a final holding time of 5 min. L-Lactic
and all forage was ensiled within a 6-h period. acid and D-lactic acid were analyzed by an
Treatments were bagged consecutively from enzymatic method (Sigma kit 826-UV; Sigma
Chemical Co., St. Louis, MO). For analysis of
alternating rows in the following order: 1) D-lactic acid, L-lactic acid dehydrogenase was
control silage (untreated), 2) silage treated with
replaced with a similar amount of D-lactic acid
Silage Fresh Plusm (SFP;Lactobacillus plan- dehydrogenase (Sigma L-9636); L-lactic acid
tarurn, Pediococcus cereviseae, cellulase, amy- (Sigma L-2250) and D-lactic acid (Sigma L-
lase, and pectinase; manufactured for Blue lO00) were used for standards for their respec-
Ribbon Products Co., Inc., Dewitt, NY by Chr. tive assays. Total lactic acid concentration was
Hansen's Biosystems, Milwaukee, WI), or 3) the sum of the concentrations of L-lactic acid
Alfazymem (ALF, L. plantarum, Pediococcus and D-lactic acid. Glucose content was deter-
acidilactici, amylase, and cellulase; Farmline mined using a Yellow Springs Instrument glu-
International, Schaumburg, IL). Additives were cose analyzer (Yellow Springs, OH), and am-
mixed in nonchlorinated water and sprayed monia N content was determined by a
onto the forage at the bagger per manufac- modified phenol-hypochlorite assay (17). Si-
turer's instructions (3 Utonne). Silage (35 to 40 lage samples were dried in a forced-air oven at
tonnes) was prepared for each treatment. Silage 55'C for 48 h and ground through a I-mm
was bagged consecutively within 6 h on a screen. The ADF content of forages was ana-
clear, cool day. During silo filling, grab sam- lyzed as outlined by AOAC (1) and NDF as
ples of forage were collected from each wa- described by Van Soest et al. (24), omitting
gon. Treated material was sampled after addi- sodium sulfite and with the addition of heat-
tion of the additive. All samples were pooled stable amylase.

Journal of Dairy Science Vol. 78. No. 3, 1995

 ADDITIVES FOR ALFALFA SILAGE 567
TABLE 1 . Lactic acid bacteria (LAB) and enzyme activities.
LAB-Water Cellulase Amylase
Treatment mixture' LAB activity2 activity3
(Cfu/d) (cWg of forage
after treatment)
Control . . 4.9 x 104 . . . . . .
S W 1.3 x 10s 1.0 x 106 5.8 7482
ALP 1.6 x 10s 2.0 x 106 222.5 532,609
'Mixture added to forage at the bagger at a rate of 3 Utonne.
2Units added per tonne of forage, based on solubilization of cellulose azure. One unit equals 1 AU (absorbance unit/
mm at 23°C (569 nm).
3Units added per tonne of forage, based on known zones of starch solubilization.
4Silage Fresh Plusm (Blue Ribbon Products Co., Inc., Dewin. NY).
*Alfazymem (Farmline International, Schaumburg, IL).

In vitro NDF digestibility was determined
on silage collected after 177 d of ensiling
using the method of Goering and Van Soest
(3). Ruminal fluid was collected from a fistu-
lated steer fed a diet of alfalfa hay. Triplicate
tubes were incubated in a 40°Cwater bath and
removed at 6, 9, 12, 16, 24, 48, and 72 h.
Immediately after removal from the water
bath, tubes were placed in a freezer at -20°C.
Representative samples of the forage and the
contents of each digestion tube were analyzed
for NDF.
Cellulase activity was determined in both
inoculates by assay of the solubilization of
cellulose azure and measurement of ethanol-
soluble fragments at 569 nm (23). Amylase
activity was determined by a modified proce-
dure of the National Feed Ingredients Associa-
tion (16).
Data were analyzed by ANOVA as a com-
pletely randomized design by day using the
procedures of SAS (20). When a significant (P
< .OS) F test was detected, means were com-
pared by Tukey's test.
RESULTS AND DISCUSSION
Samples of the inoculants taken from the
sprayer had similar numbers of viable LAB
(Table 1). Epiphytic LAB were less than 105
cfidg of forage, and inoculation increased the
numbers of LAB to about IO6 cfidg of forage.
Although numbers of LAB were similar
among the inoculant water mixes, enzyme ac-
tivity was markedly different; ALF contained
about 40 and 70 times more cellulase and
amylase activity units, respectively, per tonne
of forage than did SFP.
In general, silage fermentation occurred at a
slower rate and to a lesser extent as DM
content increased (1 1). However, silage treated
with ALF and SFP had improved fermentation
characteristics even though their DM contents
were about 6 to 7 units greater than that of the
untreated silage (Table 2). In addition, the
combinations of LAB and enzymes used in
this study did not appear to be antagonistic as
had been reported by Stokes (21). The addition
of SFP caused the most rapid decrease in
silage pH by d 2 (Figure 1). Addition of ALF
did not lower silage pH effectively as SFP on
d 2; however, between d 2 and 4,pH of silage
treated with ALF declined rapidly, resulting in
pH lower than that of the untreated silage. The
pH of both types of treated silage were lower
than that of the untreated silage from d 4 on.
The reason for lower pH in treated silages is
unclear because lactic acid was not greater in
those silages during those days (Table 2) and
cannot be explained by production of other
acids because propionic and butyric acids were
not detectable. These findings and those of
others (8,9, 11) showed that inoculation with
LAB or a combination of LAB and enzymes
could cause silage pH to drop quickly, even
when the high DM content of the forage would
normally restrict fermentation.
The glucose content of untreated silage
decreased from 2.40% at the start of ensiling to
.IO% at 177 d (Table 2). After d 2 of ensiling,
Journal of Dairy Science Vol. 78, No. 3, 1995
568 SHEPERD ET AL.

glucose content of the silage treated with SFP
was about 40% lower than that of the untreated
silage and the silage treated with ALF, sug-
gesting that the substrate was utilized at a
faster rate in silage treated with SFP than in
silage treated with ALF. From d 4 to 14,
glucose content did not differ among silage
type; however, as storage time increased, the
glucose content of silage treated with ALF
increased from -73% (after d 51) to 1.76%
(after d 177). Glucose content of silage treated
with SFP also increased during the later stages
of storage, but the increase was less than that
in silage treated with ALF. Increases in the
glucose content of treated silage suggested that
added enzymes continued to hydrolyze sub-
strate even after the fermentation was com-
plete. The ramifications of this finding are
presently unclear, but large amounts of
residual water-soluble carbohydrates could
lead to aerobic instability of silage. Further-
more, excess hydrolysis of plant cell walls in

TABLE 2. Fermentation characteristics of forage and silage during 177 d of ensiling

Total
Treatment DM Glucose D-Lactate L-Lactate lactate Acetate Ammonia N
(%) (% of DM)
d O
Control 38.3 2.40 .06
SFP' 44.0 2.91 . . . ... . . . ... .05
ALP 45.7 2.25 . . . ... . . . . . . .04
SE ... , . . . . . ... . . . . . . . . .
d 2
Control 3 1.2b ,798 1.w 1.3 3.2 1 .Sa .23a
SFP 42.58 .49b 1.2ab 2.5 3.7 .5b .I Ib
ALF 46.38 ,898 .8b 1.7 2.5 .6b .lob
SE 1.9 .07 .2 .5 .5 .1 .01
d 4
Control 33.98 .74 1.9 2.3 4.2 1.9 .2g8
SFP 43.2b .68 1.5 2.8 4.3 .8 .14b
ALF 44.Ob .7 1 1.7 2.7 4.4 .6 .14b
SE 1.4 .06 .1 .I .2 .2 .01
d 8
Control 33Xa .90 2.2 2.8 5.0 2.3a .40a
SFP 45.ob .70 2.8 3.0 5.8 .Hb .17b
ALF 46.e .80 2.5 3.2 5.7 .7b .17b
SE 2.3 .I4 .2 .3 .5 .2 .02
d 14
Control 33.e .78 2.7 2.7 5.4 2.78 ,428
SFP 38.gb .64 3.0 3.2 6.2 1.3b .22b
ALF 48Sa .66 2.7 3.1 5.8 .7b .
SE 1.1 .07 .2 .2 .4 .3 .03
d 51
Control 31.7b ,878 1.8 2.4 4.1 4.78 .53a
SFP 45.w .48b 2.5 2.8 5.2 1.3b .17b
ALF 41.41 .73ab 3.1 3.0 6.1 1.3b .20b
SE 1.4 .09 .4 .2 .5 .3 .03
d 177
Control 33.9b .1W 2.7b 3.3b 6.W 3.58 ,658
SFP 43.21 .7ob 3.wb 4.41 7.4b 1.2b .42b
ALF 47.1' 1.76a 3.38 4.58 7.W 1.4b .37c
SE 1.1 .04 .1 .1 .l .I .01
a*b*cMeans within a day within a column with unlike superscripts differ (P e .05).
'Silage Fresh PlusTM(Blue Ribbon Products Co., Inc., Dewitt. NY).
*AlfazymeTM(Farmline International, Schaumburg, IL).

Journal of Dairy Science Vol. 78, No. 3, 1995

 ADDITIVES FOR ALFALFA SILAGE 569
? content of silage treated with SFP was inter-
I mediate and did not differ either from the
7 Silage Fnsh Plus untreated or ALF silage. Because L-lactate is
6 Alfazyrna metabolized more readily by the cow, more L-
lactate than D-lactate in silage might be desira-
I, ble (14).
0)
0) Acetate content in the untreated silage was
E 5 higher than that of the treated silage through-
out ensiling (Table 2). The acetate content in
the untreated silage increased from 1.8 to 4.7%
4 through d 51 and then decreased to 3.5% at d
177. In contrast, the acetate content of the
d 2 i a 1'4 51 177 treated silage increased from only about .5 to
Days of Ensiling 1.41% through 177 d. No difference was de-
tected in acetate content between the two
treated types of silage. Lower concentrations
Figure 1. The pH of forages and silage during ensiling. of acetate and higher concentrations of lactate
Treatments within a day with unlike letters (&b) differ (P<: acid support the suggestion of a more
.Os). Figures without standard deviation bars had standard
deviations 4 4 .
homolactic fermentation in inoculated silage,
as found previously (11). When mixed grass-
legume forage was ensiled with a cellulase-
enzyme complex, Stokes (21) found that inocu-
high moisture silage could increase effluent lation alone resulted in a more homolactic
loss. fermentation; however, inoculation in combi-
There were no differences in L-lactate or nation with enzyme treatment resulted in a
total lactate concentration among treatments heterolactic fermentation.
until d 177 [Table 2). Treated silage tended to As expected, ammonia N increased during
have less D-lactate at d 2, but only silage ensiling and was lower in treated than in un-
treated with ALF differed from the untreated treated silage (Table 2). The CP on d 0 was
silage. The ratio of L-1actate:D-lactate isomers 19.7, 21.5, and 20.7% and on d 177 was 20.2,
on d 2 in the untreated silage was about .76; 19.7, and 20.0% for untreated silage, silage
the ratio of treated silage was nearly 2. Al- treated with SFP, and silage treated with ALF,
though many of the homofermentative strains respectively. The reduction in ammonia N
of silage LAB produce L- and D-isomers of likely was a result of the restriction of proteol-
lactate, the high ratio of L-1actate:D-lactate in ysis because of DM content in the silage.
treated silage suggested that the fermentation However, microbial inoculation also reduced
was dominated by species that produce L- ammonia N content in silages and could be
lactate. Although L-lactate is formed initially, partially responsible for these findings (4, 10,
lactate racemase in some of the Lactobacillus 19).
species and, perhaps, in other microorganisms The fiber contents of forage and silage are
may convert the L-isomer to the D-isomer (14). shown in Table 3. Because fiber content of
This theory explains why all ratios were ap- silage was numerically different at the start of
proximately 1.1 by d 8 of ensiling. No differ- the study, 0 values were used as covariants in
ences in D-lactate among treatments were the statistical analysis. The NDF was approxi-
noted between d 4 and 51. However, at d 177, mately 35% at ensiling and did not differ
total lactate content of silage treated with ALF between treated and untreated silage after 14 d
was higher than that of silage treated with of ensiling. Differences were found in the NDF
SLP, and the lactate content of both treated content of silage after d 51 of ensiling. Neither
silages was about 27% greater than that of the type of treated silage differed from the un-
untreated silage. Treated silage had about 35% treated silage, but they did differ from each
more L-lactate than did untreated silage, and other. Silage treated with ALF contained
ALF silage had approximately 20% more D- 36.5% NDF, and silage treated with SFP con-
lactate than the untreated silage. The D-lactate tained 41.1% NDF. By d 177, no difference

Journal of Dairy Science Vol. 78, No. 3. 1995

570 SHEPERD ET AL.

TABLE 3. Fiber content of forage and silages.

d 01 d 14 d 51 d 177
Treatment NDF . ADF HC? NDF ADF HC NDF ADF HC NDF ADF HC

Control 37.4 28.4 9.0 38.8 33.9 4.8 40.2ab 34.2 5.9 43% 38.9a 4.9a
sFp3 33.5 28.4 5.1 38.3 32.4 5.9 41.la 31.5 9.5 40.6b 37.3b 3.3b
ALF4 34.2 26.2 8.0 40.3 35.2 5.1 3 6 . 9 32.7 3.8 40.0b 35.6C 4.3a
SE' . . . . .. . .. 1.3 1.3 1.1 .9 1.0 1.8 .3 .8 .3
a.b.cMeans in a column with unlike superscripts differ (P e .05).
lDay 0 values were determined from a pooled sample and were used as covariants for the remaining days.
2Hemicellulose.
3Silage Fresh PlusTM(Blue Ribbon Products Co., Inc., Dewitt, NY).
4AlfazymeTM(Farmline International, Schaumburg, IL).
5n = 3.

existed between types of silage treated with
ALF and SFP, but treated silages were about
8% lower in NDF than untreated silage. The
ADF content was similar among treatments
until d 177 of ensiling, at which time the
untreated silage had 38.9% ADF, and silage
treated with SFP and ALF had about 4 and 8%
less ADF, respectively. Hemicellulose content
tended to be low in all forages and tended to
decrease with storage. After 177 d, silage
treated with SFP was lower in hemicellulose
content than were other treatments. Because
microbial inoculation usually does not de-
crease the fiber content of silage (ll), these
findings suggested that the enzyme additives
effectively hydrolyzed the plant cell walls. At
d 177, the ADF content was lowest, and the
glucose content highest, in the silage treated
with ALF, which probably reflected the greater
amount of enzymes applied to this silage. Plant
cell-wall hydrolysis in silage by cellulase en-
zymes has been variable (7,22). Our data agree
with results of Henderson et al. (5) and other
researchers (6, 18), who found changes in cell-
wall components with enzyme treatment, but
contradict other findings (10, 12). However, as
previously noted (lo), results of studies with
enzyme-treated forages are difficult to compare
because of the variability between enzyme
complexes, different amounts of activity, and
interactions with moisture content, plant matu-
rity, and plant species.
Traditionally, the energy content of a forage
and silage is calculated from an equation based
Journal of Dairy Science Vol. 78, No. 3, 1995
on its ADF content; therefore, a lower ADF
content in treated silage would increase calcu-
lated energy content. Thus, when feeding a
silage treated with enzyme, the proportion of
dietary concentrate theoretically could be
decreased while maintaining a similar energy
content with a less costly formulation. To our
knowledge, this concept has not been ade-
quately tested and it is questionable whether a
silage that has been treated with enzymes with
reduced ADF content can be subject to energy
calculations using standard procedures.
Treatments had no effect on in vitro NDF
digestion among silage treatments during the
early stages of digestion (Figure 2). During this
time, digestibility increased to about 40% at 20
h and then remained relatively constant. How-
ever, NDF digestion was lower in treated than
in untreated silage after 36 and 48 h of incuba-
tion. Others (2, 12, 15) have also reported
decreases in fiber digestion when forages were
treated with enzymes. Fiber digestibility
decreased when alfalfa silage that had been
treated with cellulase was fed to sheep (2). In
addition, Narasimhalu et al. (15) reported a
decrease in digestibility of enzyme-treated
timothy silage. These findings indicate that the
cellulase enzymes used in this study acted on
the more readily degradable portion of the
fiber.
CONCLUSIONS
The addition of ALF to silage resulted in a
substantially higher application of fiber-
 ADDITIVES FOR ALFALFA SILAGE
50-
a 4Gordon. H. K. 1989. An evaluation through lactating
'
c
40-
b
C cows of a bacterial inoculant as an additive for grass
5 5Henderson, A. R., P. McDonald, and D. Anderson.
% 30-
6 from Trichoderma viride on the chemical changes
L
D during the ensilage of grass, lucerne and clover. J. Sci.
z 20-
1 D-
0 8 ,
0 10 20 30 40 50 60
Fermentation, h
Figure 2. In vitro digestion of NDF of silages. Treat-
ments within an hour with unlike leners (a,b,c) differ (P<
.05). Treatments: control @), Silage Fresh PlusN (0;Blue
Ribbon Products Co., Inc., Dewitt, NY). and Alfazymem
(A; Farmline International. Schaumburg, IL).
digesting enzymes than did addition of SFP,
but both additives contained similar numbers
of LAB. Both additives improved silage fer-
mentation by decreasing the pH more rapidly
in silage with higher DM content than in the
untreated silage. The addition of SFP caused
the most rapid decrease in silage pH, but the
addition of ALF more effectively decreased the
final ADF content of silage. In vitro NDF
digestion was decreased by both treatments
relative to that of the untreated silage, probably
because the added enzymes had already hydro-
lyzed the more readily digestible portion of the
forage in the silo.
ACKNOWLEDGMENTS
The authors thank A. 0. Hession, P. Garcia-
Lopez, and E. M. Kreck for technical as-
sistance. This project was funded partially by
Chr. Hansen's Biosystems, Inc., Milwaukee,
WI.
REFERENCES
1 Association of Official Analyucal Chemists. 1990.
Official Methods of Analysis. 15th ed. AOAC,
Washington, DC.
2 Bolsen, K. K., A. Laytimi, and J. White. 1989. Effects
of enzyme and inoculant additives on preservation and
nutritive value of alfalfa silage. J. Dairy Sci. 72(Suppl.
1):297.(Abstr.)
57 1
3 Goering, H. K., and P. J. Van Soest. 1970. Forage
Fiber Analyses (Apparatus, Reagents. Procedures. and
Some Applications). Agric. Handbook No. 379. ARS-
USDA, Washington, DC.
silage. Grass Forage Sci. 44:169.
1982. The effect of a cellulase preparation derived
Food Agric. 33:16.
6 Jaakkola, S. 1990. The effect of cell wall degrading
enzymes on the preservation of grass and on the silage
intake and digestibility in sheep. J. Agric. Sci. Finl.
6251.
7 Jaster, E. H., and K. J. Moore. 1991. Quality and
fermentation of enzyme treated alfalfa silages at three
moisture concentrations. Anim. Feed Sci. Technol. 31:
261.
8 Jones, B. A., R. D. Hatfield, and R. E. Muck. 1992.
Effect of fermentation and bacterial inoculation on
lucerne cell walls. J. Sci. Food Agric. 60:147.
9 Jones, B. A.. L. D. Saner, and R. E. Muck. 1992.
Influence of bacterial inoculant and substrate addition
to lucerne ensiled at different dry matter contents.
Grass Forage Sci. 47:19.
10Kung. L., Jr., B. R. Carmean. and R. S. Tung. 1990.
Microbial inoculation or cellulase enzyme treatment
of barley and vetch harvested at three maturities. J.
Dairy Sci. 73:1304.
11 Kung, L., Jr., D. B. Grieve, J. W. Thomas, and J. T.
Huber. 1984. Added ammonia or microbial inocula for
fermentation and nitrogenous compounds of alfalfa
ensiled at various percents of dry matter. J. Dairy Sci.
67:299.
12 Kung. L. Jr., R. S. Tung, K. G. Maciorowski, K.
Buffum, K. Knutsen, and W. R. Aimutis. 1991. Ef-
fects of plant cell-wall-degrading enzymes and lactic
acid bacteria on silage fermentation and composition.
J. Dairy Sci. 74:4284.
13Leathenvood, J. M., R. D. Mochrie, and W.E. Tho-
mas. 1959. Chemical changes produced by a cellulo-
lytic preparation added to silages. J. Anim. Sci. 18:
1539.
14McDonald, P., A. R. Henderson, and S.J.E. Heron.
1991. The Biochemistry of Silage. 2nd ed. Chalcombe
Publ., Marlow, Bucks, England.
15 Narasimhalu,P., L. J. Halliday, J . B. Sanderson, H. T.
Kunelius, and K. A. Winter. 1992. The composition,
intake, and digestibility of timothy silage preserved
untreated or treated with formic acid or a cellulase-
hemicellulase preparation. Can.J. Anim. Sci. 72:431.
16 National Feed Ingredients Association. 1991. NFIA
Lab Methods Compendium. Vol. 2. Natl. Feed Ingred.
Assoc., Des Moines, IA.
17 Okuda, H., S. Fuji, and Y.Kawashima. 1965. A direct
colorimetric method for blood ammonia. Tokushima J.
Exp. Med. 12:11.
18 Raununaa, A., J. Setala, T. Moisio, T. HeikkIia, and
M. Lampila. 1987. The effect of inoculants and cellu-
lase on the fermentation and microbiological composi-
tion of grass silage. I. Biochemical changes in the
silages. J. Agric. Sci. Finl. 59:361.
Journal of Dairy Science Vol. 78. No. 3, 1995
572 SHEPERD ET AL
19Rooke. J. A., F. M. Maya. J. A. h o l d , and D. G.
Armstrong. 1988. The chemical composition and
nutritive value of grass silages prepared with no addi-
tive or with the application of additives containing
either Lambacillus plantarum or formic acid. Grass
Forage Sci. 43:87.
20 SAS@ User’s Guide: Statistics. Version 5 Edition.
1985. SAS Inst., Inc., C q , NC.
21 Stokes, M. R. 1992. Effects of an enzyme mixture, an
inoculant. and their interaction on silage fermentation
and dairy production. 1992. J. Dairy Sci. 75:764.
22Stokes, M. R.. and L E. Tomy. 1990. Preservation
Journal of Dairy Science Vol. 78. No. 3. 1995
and nutritive value of hay crop forage ensiled with
two enzyme additives or Protectein RTU. J. Dairy Sci.
73LSuppl. 1):225.(Abslr.)
23Turner. P. M.. and R. J. Johnson. 1991. A spec-
trophotometric assay for determining cellulase activity
that is suitable for enzyme kinetics. Page 53 in Proc.
91st Mtg. Am. Soc.Microbiol.. Dallas. TX. Am. Soc.
Microbiol., Washington, DC,
24Van Soest, P. J., J. B. Robertson, and B. A. Lewis.
1991. Methods for dietary fiber, neutral detergent
fiber. and non-starch polysaccharides In relation to
animal nutrition. J. Dairy Sci. 74:3583.