A laboratory evaluation of comfrey (Symphytum officinale L.) as a forage crop for ensilage


Wilkinson, J.M.

Animal Feed Science and Technology 104(1/4): 227-233

2003


A 10 m2 area of comfrey (Symphytum officinale) was chopped and ensiled in duplicate laboratory silos, either unwilted or following a 24 h wilt, to test the hypothesis that the crop might be suitable for ensiling as animal feed. Concentrations of dry matter (DM) averaged 112 and 146 g/kg for the unwilted and wilted crops. Both crops were very difficult to chop due to the mucilaginous nature of the material. The concentrations of water-soluble carbohydrates, ash and N in the crops at ensiling averaged 107.0 and 118.0, 143.0 and 148.5, 32.2 and 34.7 g/kg DM for the unwilted and wilted materials, respectively. Buffering capacity averaged 488 and 472 mE kg/DM for the unwilted and wilted crops at harvest. Mean pH values of the silages were 5.43 and 5.16 for unwilted and wilted materials. The fermentation quality of the silages was poor, with relatively low concentrations of lactic acid and mean concentrations of n-butyric acid of 44.1 and 29.4 g/kg DM in the unwilted and wilted silages, respectively. Concentrations of soluble and ammonia N were, respectively, 708 and 470, and 238 and 179 g/kg total N in the unwilted and wilted silages, indicating extensive proteolysis, particularly in the unwilted crop. Without additive treatment and extensive wilting, the fermentation of comfrey leads to poor quality silage.

ANIMAL
FEED
SCIENCE
AND
TECHNOLOGY
Animal
Feed
Science
and
Technology
ELSEVIER
104
(2003)
227-233
www.elsevier.com/locate/anifeedsci
Short
communication
A
laboratory
evaluation
of
comfrey
(Symphytum
officinale
L.)
as
a
forage
crop
for
ensilage
J.M.
Wilkinson*
School
of
Biology,
University
of
Leeds,
The
Louis
Compton
maill
Building,
Leeds
LS2
9JT
UK
Received
16
February
2002;
received
in
revised
form
2
October
2002;
accepted
2
October
2002
Abstract
A
10
m
2
area
of
comfrey
(Symphytum
officinale)
was
chopped
and
ensiled
in
duplicate
laboratory
silos,
either
unwilted
or
following
a
24
h
wilt,
to
test
the
hypothesis
that
the
crop
might
be
suitable
for
ensiling
as
animal
feed.
Concentrations
of
dry
matter
(DM)
averaged
112
and
146
g/kg
for
the
unwilted
and
wilted
crops.
Both
crops
were
very
difficult
to
chop
due
to
the
mucilaginous
nature
of
the
material.
The
concentrations
of
water-soluble
carbohydrates,
ash
and
N
in
the
crops
at
ensiling
averaged
107.0
and
118.0,
143.0
and
148.5,
32.2
and
34.7
g/kg
DM
for
the
unwilted
and
wilted
materials,
respectively.
Buffering
capacity
averaged
488
and
472
mE
kg/DM
for
the
unwilted
and
wilted
crops
at
harvest.
Mean
pH
values
of
the
silages
were
5.43
and
5.16
for
unwilted
and
wilted
materials.
The
fermentation
quality
of
the
silages
was
poor,
with
relatively
low
concentrations
of
lactic
acid
and
mean
concentrations
of
n-butyric
acid
of
44.1
and
29.4
g/kg
DM
in
the
unwilted
and
wilted
silages,
respectively.
Concentrations
of
soluble
and
ammonia
N
were,
respectively,
708
and
470,
and
238
and
179
g/kg
total
N
in
the
unwilted
and
wilted
silages,
indicating
extensive
proteolysis,
particularly
in
the
unwilted
crop.
Without
additive
treatment
and
extensive
wilting,
the
fermentation
of
comfrey
leads
to
poor
quality
silage.
©
2002
Elsevier
Science
B.V.
All
rights
reserved.
Keywords:
Comfrey
(Symphytum
officinale);
Composition;
Silage;
Animal
feed
Abbreviations:
DM,
dry
matter;
FW,
fresh
weight;
MAFF,
Ministry
of
Agriculture,
Fisheries
and
Food
(UK);
NDF,
neutral
detergent
fibre;
N,
nitrogen;
FC,
fermentability
coefficient;
WSC,
water-soluble
carbohydrates;
BC,
buffering
capacity;
NH3-N,
ammonia
nitrogen
*
Tel.:
+44-1132332880.
E-mail
address:
(J.M.
Wilkinson).
0377-8401/02/$
see
front
matter
©
2002
Elsevier
Science
B.V.
All
rights
reserved.
PH:
S0377-8401(02)00293-6
228
J.M.
Wilkinson/Animal
Feed
Science
and
Technology
104
(2003)
227-233
1.
Introduction
Comfrey
(Symphytum
officinale),
a
member
of
the
genus
Boraginacae,
is
a
perennial
herb
reputed
to
have
beneficial
medicinal
properties.
Thus
the
leaves
have
been
used
externally
to
treat
fractures,
strains,
thrombophlebitis
and
haematoma,
and
internally
as
infusions
in
the
treatment
of
gastro-intestinal
and
respiratory
tract
diseases
(Roder,
1995).
Potentially
benefi-
cial
components
of
the
leaves
include
allantoin
and
rosmarinic
acid
(Duke,
2002).
However,
comfrey
also
contains
pyrrolizidine
alakaloids
including
acetyl-intermidine,
acetyl-
lycopsamine
and
symphytine
(Dewick,
2002;
Duke,
2002),
and
nitrates
(Cooper
and
Johnson,
1984)
which
may
act
as
hepatotoxic
factors
to
ruminant
livestock
given
silage
made
from
comfrey,
if
the
toxic
components
are
not
de-toxified
in
the
silo
or
rumen.
The
comfrey
plant
has
a
taproot
up
to
3
m
in
length,
and
can
produce
two
to
five
crops
per
year
(Bremness,
1988),
which
are
often
utilised
as
a
source
of
organic
fertiliser,
either
as
green
manure
or
as
an
ingredient
of
compost.
The
possibility
existed
that
comfrey
might
have
potential
as
a
forage
crop
for
ensiling
and
subsequent
use
as
animal
feed.
Accordingly,
a
laboratory-scale
ensiling
experiment
was
undertaken
to
test
the
hypothesis
that
comfrey
might
be
suitable
for
ensiling.
2.
Materials
and
methods
A
10
m
2
plot
of
comfrey
(Symphytum
officinale,
cf.
Bocking
14)
in
primary
growth
was
harvested
to
5
cm
stubble
height
by
hand
shears
on
12
May
1999.
The
plants
were
approximately
1
m
in
height
and
in
early
bloom
(Fig.
1).
The
entire
crop
was
weighed
and
divided
into
two
equal
parts.
One
was
chopped
to
20
mm
average
particle
length
immediately
after
being
weighed
("unwilted"),
through
an
AL-KO
Dynamic
H1600
chopper
(AL-KO
Maschinenfabrik
GmbH,
Obdach,
Austria).
The
other
was
placed
unchopped
in
a
thin
layer
on
a
clean
concrete
floor
in
an
unheated,
ventilated,
greenhouse
for
24
h
before
being
chopped
through
the
same
machine
("wilted").
Samples
of
both
unwilted
and
wilted
crops
were
frozen
immediately
after
being
chopped.
Both
lots
of
chopped
material
were
ensiled
in
laboratory
silos,
comprising
two
polyethylene
bags
enclosed
within
a
polyethylene
bucket.
The
inner
bag
of
the
pair
was
filled
by
hand
and
compressed
during
filling
by
treading
to
remove
as
much
air
as
possible.
After
being
filled,
both
the
inner
and
outer
bags
were
tied
firmly
at
the
neck
with
string
and
each
bucket
was
then
closed
with
a
sealed
lid.
The
average
quantity
ensiled
was
12.0
kg
fresh
weight
per
silo
and
the
average
packing
density
was
641.7
kg
fresh
weight/m
3
.
Due
to
lack
of
material,
it
was
only
possible
to
fill
two
silos
of
unwilted
and
two
silos
of
wilted
material.
The
silos
were
opened
after
a
period
of
202
days,
and
the
contents
were
mixed
and
sub-sampled
for
dry
matter
(DM)
and
compositional
analysis.
Dry
matter,
ash,
ammonia
N
and
pH
were
determined
using
methods
in
MAFF
(1986).
Total
N
was
determined
by
the
method
in
SI
(1991).
Neutral
detergent
fibre
(NDF)
was
determined
using
the
method
in
MAFF
(1994),
with
the
addition
of
sodium
sulphite
and
alpha
amylase,
and
expressed
with
residual
ash.
Buffering
capacity,
soluble
N,
and
amino
acid
N
were
assessed
by
electrometric
titration
of
juice
(Offer
et
al.,
1993)
and
water-soluble
carbohydrates
were
determined
by
the
method
of
Thomas
(1977).
Lactic acid
in
the
silages
Wilkinson/Animal
Feed
Science
and
Technology
104
(2003)
227-233
229
t
.
"
e
.
Fig.
1.
Comfrey
(Symphytum
officinale)
at
the
early
bloom
stage
of
growth.
was
determined
using
the
method
of
Bergmeyer
et
al.
(1984).
Ethanol,
acetic,
propionic
and
n-butyric
acids
were
determined
in
the
silages
by
gas
liquid
chromatography
(Jones
and
Kay,
1976).
Data
for
both
fresh
and
ensiled
materials
were
subjected
to
analysis
of
variance
using
a
general
linear
model
in
Minitab
12.1,
with
"crop"
(crop
before
ensiling
versus
silage,
1
d.f.)
and
"wilt"
(unwilted
silage
versus
wilted
silage,
1
d.f.)
as
main
effects.
Data
derived
solely
from
the
silages
were
analysed
by
ANOVA
with
"wilt"
(1
d.f.)
as
the
sole
main
effect.
3.
Results
and
discussion
The
yield
of
the
crop
was
61.3
t
fresh
weight
(FW)
per
hectare
(6.9
t
DM/ha).
On
chopping,
the
unwilted
crop
formed
a
sticky
mass,
which
did
not
pass
through
the
chopper
without
considerable
physical
assistance
from
the
operator.
After
a
24
h
wilt,
the
crop
still
required
physical
assistance
to
aid
its
passage
through
the
chopper.
The
composition
of
the
crops
before
and
after
ensiling
are
in
Table
1.
The
crops
were
relatively
wet
at
the
time
of
ensiling,
averaging
112
and
146
g
DM/kg
FW
for
the
unwilted
and
wilted
materials,
respectively.
The
laboratory
silos
did
not
allow
liquid
effluent
to
escape,
and
the
DM
concentrations
were
slightly,
though
not
significantly,
lower
in
the
silages
than
in
the
fresh
crop
before
ensiling
(Table
1).
The
presence
of
liquid
within
the
silos,
coupled
with
the
relatively
low
DM
concentrations
may
have
influenced
fermentation.
Table
1
Composition
of
comfrey
(Symphytum
officinale)
at
harvest
and
after
ensiling
Crop
S.E.D.
Significance
of
effectsa
Before
ensiling
Silage
Crop
Wilt
Crop
x
wilt
Unwilted
Wilted
Unwilted
Wilted
Dry
matter
(g/kg
FW)
112.0 146.0
99.0
134.5
6.15
NS
**
NS
Buffering
capacity
(mE/kg
DM)
488
472
3669
3216
974.1
NS
NS
Water-soluble
carbohydrates
g/kg
FW
12.0
17.2
1.25
2.70
0.255
*** ***
**
g/kg
DM
107.0 118.0
12.9
20.2
3.47
***
NS
NS
Ash
(g/kg
DM)
143.0
148.5 125.5 144.5
10.63
NS NS
NS
Total
N
(g/kg
DM)
32.2
34.7
28.4
30.5
1.53
NS NS
NS
Neutral
detergent
fibre
(g/kg
DM)
333.5
350.0
192.0
252.0
9.08
***
**
NS
pH
5.43
5.16
0.710
NS
Lactic
acid
(g/kg
DM)
18.9
28.7
24.25
NS
Acetic
acid
(g/kg
DM)
3.30
13.3
7.95
NS
Propionic
acid
(g/kg
DM)
20.5
10.4
8.22
NS
n-Butyric
acid
(g/kg
DM)
44.1
29.4
32.45
NS
Total
fermentation
acids
(g/kg
DM)
87.5
83.2
45.41
NS
Ethanol
(g/kg
DM)
10.8
15.0
3.78
NS
Soluble
N
(g/kg
total
N)
708
470
172.6
NS
NH3-N
(g/kg
total
N)
238
179
114.9
NS
Amino
acid
N
(g/kg
soluble
N)
483
639
114.5
NS
a
NS:
not
significant.
*
P
<
0.05.
**
P
<
0.01.
***
P
<
0.001.
Wilk
i
nson/A
ni
mal
F
eed
S
ci
ence
and
T
ech
nol
o
gy
1
0
4
(2
003
)
22
7
-2
33
Wilkinson/Animal
Feed
Science
and
Technology
104
(2003)
227-233
231
The
water-soluble
carbohydrate
(WSC)
concentrations
of
12.0
and
17.2
g/kg
FW
for
the
unwilted
and
wilted
materials,
respectively,
were
lower
(P
<
0.001)
than
the
critical
value
of
20
g/kg
FW
essential
for
satisfactory
fermentation
in
the
silo
(Wilkinson
et
al.,
1981).
In
these
respects,
comfrey
resembled
sunflower
(Helianthus
annus)
grown
in
northern
European
conditions
and
harvested
at
the
onset
of
flowering
at
maximum
plant
height
(Edwards
et
al.,
1978).
The
mean
buffering
capacities
of
the
fresh
crops,
which
averaged
488
and
472
mE/kg
DM for
unwilted
and
wilted
materials,
respectively,
were
higher
than
the
typical
value
of
380
mE
kg/DM
for
perennial
ryegrass
(Lolium
perenne)
and
close
to
those
of
449
and
472
mE/kg
DM
for
sunflower
and
lucerne
(Medicago
sativa),
respectively
(McDonald
et
al.,
1991).
The
relatively
high
buffering
capacity
of
the
comfrey
might
have
been
associated
with
relatively
high
concentrations
of
organic
acids,
and
anions
such
as
salts
of
organic
acids,
orthophosphates,
chlorides
and
nitrates
(Playne
and
McDonald,
1966),
although
these
constituents
were
not
determined.
However,
the
concentrations
of
ash
in
the
crops
at
ensiling
and
in
the
resultant
silages
were
substantially
higher
than
mean
values
for
grass
(93
g/kg
DM),
lucerne
(105
g/kg
DM)
and
clover
(Trifolium
spp.,
116
g/kg
DM)
silages,
and
similar
to
the
average
value
of
143
g/kg
DM
for
whole
pea
(Pisum
sativum)
silage
(MAFF,
1992).
The
comfrey
was
relatively
leafy
(Fig.
1),
which
may
have
been
a
contributory
factor
to
the
relatively
high
buffering
capacity,
since
Muck
(1985)
recorded
substantially
higher
buffering
capacities
for
lucerne
leaf
than
stem
(723
and
365
mE/kg
DM
for
leaf
and
stem
material,
respectively).
The
contents
of
total
N
in
the
crops
at
harvest
(32.2
and
34.7
g/kg
DM
for
unwilted
and
wilted
materials,
respectively),
and
in
the
silages
(28.4
and
30.5
g/kg
DM for
unwilted
and
wilted
materials,
respectively),
were
relatively
high
compared
to
grass
silages,
and
similar
to
concentrations
for
lucerne
silage
(31.0
g/kg
DM),
and
whole
crop
pea
silage
(28.6
g/kg
DM)
made
in
the
UK
(MAFF,
1992).
The
relatively
high
concentration
of
nitrogenous
components
in
the
comfrey
at
the
time
of
ensiling
may
have
been
a
further
contributing
factor
to
its
high
buffering
capacity,
although
the
relative
importance
of
plant
proteins
to
total
buffering
capacity
has
been
found
to
be
relatively
small
(Playne
and
McDonald,
1966).
The
concentrations
of
NDF
were
low
compared
to
other
forage
crops,
with
the
exception
of
that
of
a
sample
of
whole
pea
silage,
which
contained
280
g
NDF/kg
DM
(MAFF,
1992).
There
was
a
decrease
(P
<
0.001)
in
NDF
during
ensiling,
suggesting
degradation
of
plant
cell
wall
material,
most
likely
hemicelluloses,
either
by
plant
enzymes,
cellulolytic
clostridia
or
by
acid
hydrolysis
(McDonald
et
al.,
1991).
Working
with
grass
crops,
McDonald
et
al.
(1960)
found
that
up
to
half
the
original
hemicellulose
was
degraded
during
ensiling.
The
concentration
of
NDF
was
higher
in
the
wilted
silage
than
in
the
unwilted
silage
(P
<
0.01).
However,
the
mean
concentration
of
lactic
acid
tended
(P
=
0.73)
to
be
higher
and
the
mean
pH
value
tended
(P
=
0.74)
to
be
lower
in
the
wilted
than
in
the
unwilted
silage,
suggesting
that
acid
hydrolysis
of
hemicelluloses
was
unlikely
to
have
been
the
cause
of
the
difference
in
NDF
between
the
unwilted
and
wilted
silages.
Possibly,
degradation
of
hemicelluloses
by
plant
enzymes
continued
for
a
longer
period
of
time
in
the
unwilted
than
in
the
wilted
material.
The
mean
concentrations
of
soluble
and
ammonia
N
were,
respectively,
708
and
470
in
the
unwilted
silages,
and
238
and
179
g/kg
total
N
in
the
wilted
silages,
indicating
proteolysis
232
J.M.
Wilkinson/Animal
Feed
Science
and
Technology
104
(2003)
227-233
and
deamination
of
amino
acids,
either
by
plant
enzymes
or
proteolytic
clostridia
(McDonald
et
al.,
1991).
The
epiphytic
microflora
may
have
been
inappropriate
for
ensilage,
in
which
case
treat-
ment
of
the
crop
with
an
effective
additive
might
have
produced
beneficial
results.
Never-
theless,
the
mean
fermentability
coefficients
(FC)
of
the
fresh
crops
(FC
=
DM
(%)
+
8
WSC/BC),
of
12.9
and
16.6
for
the
unwilted
and
wilted
crops,
respectively,
were
con-
siderably
lower
than
the
critical
value
of
45
considered
to
be
essential
for
production
of
anaerobically
stable
silage
free
ofbutyric
acid
(Weissbach,
1996).
Thus,
it
was
not
surprising
that
the
fermentation
quality
of
the
silages
was
generally
poor,
as
indicated
by
the
relatively
high
pH
values,
low
concentrations
of
lactic
acid,
and
measurable
levels
of
n-butyric
acid
in
both
unwilted
and wilted
silages.
The
very
low
ratio
of
WSC
to
BC
(0.22
for
the
unwilted
crop)
indicated
that
comfrey
required
extensive
field
wilting,
to
a
minimum
of
432
g
DM/kg
FW,
in
order
to
achieve
well-fermented
silage
(Weissbach
et
al.,
1974).
The
short
period
of
wilting,
although
insufficient
to
raise
the
concentration
of
DM
to
the
required
level
for
anaerobically
stable
silage,
was
reflected
in
some
improvement
in
fermentation
quality,
as
evidenced
by
the
trend
towards
higher
concentrations
of
lactic
acid
and
amino-N,
and
lower
concentrations
of
propionic
and
n-butyric
acids,
soluble
N
and
NH3-N
in
the
wilted
than
in
the
unwilted
silages
(Table
1).
4.
Conclusions
The
mucilaginous
nature
of
comfrey
made
it
difficult
to
chop
and
ensile.
The
crop
con-
tained
too
low
a
concentration
of
fermentable
carbohydrate
and
too
high
a
buffering
capacity
for
satisfactory
fermentation
in
the
silo
without
extensive
field
wilting
and/or
treatment
with
an
effective
additive
at
harvest.
References
Bergmeyer,
H.U.,
Bergmeyer,
J.,
Grass!,
M.,
1984.
Methods
of
Enzymatic
Analysis,
Third
ed.
Verlag,
Weinheim,
Germany.
Bremness,
L.,
1988.
The
Complete
Book
of
Herbs.
Dorling
Kindersley,
London,
UK.
Cooper,
M.R.,
Johnson,
A.W.,
1984.
Poisonous
Plants
in
Britain
and
Their
Effects
on
Animals
and
Man,
Reference
Book
161.
Ministry
of
Agriculture,
Fisheries
and
Food,
Her
Majesty's
Stationery
Office,
London,
UK.
Dewick,
P.M.,
2002.
Medicinal
Natural
Products:
A
Biosynthetic
Approach,
Second
ed.
Wiley,
Chichester,
UK.
Duke,
J.A.,
2002.
Chemicals
and
their
biological
activities.
In:
Symphytum
officinale
L.
(Boraginaceae)—Comfrey.
Phytochemical
Database,
USDA—ARS-NGRL,
Beltsville
Agricultural
Research
Center,
Beltsville,
MD,
USA.
http://www.ars-grin.gov/cgi-bin/duke/farmacy2.pl.
Edwards,
R.A.,
Harper,
F.,
Henderson,
A.R.,
Donaldson,
E.,
1978.
Sunflower
silage.
J.
Sci.
Food
Agric.
29,
332-338.
Jones,
D.W.,
Kay,
J.J.,
1976.
Determination
of
volatile
fatty
acids
C1—C6,
and
lactic
acid
in
silage
juice.
J.
Sci.
Food
Agric.
27,1005-1014.
McDonald,
P.,
Stirling,
A.C.,
Henderson,
A.R.,
Dewar,
W.A.,
Stark,
G.H.,
Davie,
W.g.,
Macpherson,
H.T.,
Reid,
A.M.,
Slater,
J.,
1960.
Studies
on
Ensilage,
Technical
Bulletin
24.
Edinburgh
School
of
Agriculture.
McDonald,
P.,
Henderson,
A.R.,
Heron,
S.J.E.,
1991.
The
Biochemistry
of
Silage,
Second
ed.
Chalcombe
Publications,
Marlow,
UK.
Ministry
of
Agriculture,
Fisheries
and
Food
(MAFF),
1986.
The
Analysis
of
Agricultural
Materials,
Reference
Book
427.
Her
Majesty's
Stationery
Office,
London,
UK.
Wilkinson/Animal
Feed
Science
and
Technology
104
(2003)
227-233
233
Ministry
of
Agriculture,
Fisheries
and
Food
(MAFF),
1992.
Feed
Composition,
Second
ed.
Chalcombe
Publications,
Canterbury,
UK.
Ministry
of
Agriculture,
Fisheries
and
Food
(MAFF),
1994.
Prediction
of
the
Energy
Value
of
Compound
Feeding
Stuffs
for
Fann
Animals.
Her
Majesty's
Stationery
Office,
London,
UK.
Muck,
R.E.,
1985.
Variations
in
alfalfa
buffering
capacity.
In:
Proceedings
of
the
Winter
Meeting
of
American
Society
Agricultural
Engineering,
Chicago.
Paper
85-1535,17-20
December
1985.
Offer,
N.W.,
Rooke,
J.A.,
Thomas,
C.,
1993.
Rapid
assessment
of
silage
fermentation
characteristics
by
electrometric
titration.
Anim.
Sci.
56,423.
Playne,
M.J.,
McDonald,
P.,
1966.
The
buffering
constituents
of
herbage
and
of
silage.
J.
Sci.
Food
Agric.
17,
264-268.
Roder,
E.,
1995.
Medicinal
Plants
in
Europe
Containing
Pyrrolizidine
Alkaloids.
http://www.ibiblio.org/herbmed/
PAs/PAs-plants.html.
SI,
1991.
The
Feeding
Stuffs
Regulations.
Statutory
InstrumentNo.
2840.
Her
Majesty's
Stationery
Office,
London,
UK.
Thomas,
T.A.,
1977.
An
automated
procedure
for
the
determination
of
soluble
carbohydrates
in
herbage.
J.
Sci.
Food
Agric.
28,639-642.
Weissbach,
F.,
1996.
New
developments
in
crop
conservation.
In:
Jones,
D.I.H.,
Jones,
R.,
Dewhurst,
R.,
Merry,
R.,
Haigh,
P.M.
(Eds.),
Proceedings
of
the
Kith
International
Silage
Conference.
Aberystwyth,
UK,
pp.
11-25.
Weissbach,
F.,
Schmidt,
L.,
Hein,
E.,
1974.
Method
of
Anticipation
of
the
Run
of
Fermentation
in
Silage
Making,
Based
on
the
Chemical
Composition
of
the
Green
Fodder.
Proceedings
of
the
XlIth
International
Grassland
Congress,
Moscow,
Soviet
Union.
vol.
III,
Part
II,
pp.
663-673.
Wilkinson,
J.M.,
Chapman,
RF.,
Wilkins,
R.J.,
Wilson,
R.F.,
1981.
Inter-relationships
between
pattern
of
fermentation
during
ensilage
and
initial
crop
composition.
In:
Proceedings
of
the
14th
International
Grassland
Congress,
Lexington,
Kentucky,
USA,
pp.
631-634.