Directory UMM :Data Elmu:jurnal:A:Animal Feed Science and Technology:Vol82.Issue3-4.Dec1999:

Animal Feed Science and Technology
82 (1999) 243±259

The use of tannins as silage additives: effects on
silage composition and mobile bag disappearance
of dry matter and protein
M.B. Salawua,b,c,*, T. Acamovica,1, C.S. Stewartc,
T. Hvelplundd, M.R. Weisbjergd
a

Scotland Agricultural College, Aberdeen, AB24 5UD, Scotland, UK
University of Aberdeen, 581 King Street, Aberdeen AB24 5UD, Scotland, UK.
c
Rowett Research Institute, Bucksburn, Aberdeen AB21 9SB, Scotland, UK
d
Danish Institute of Agricultural Sciences, Research Centre Foulum, P.O. Box 50,
DK-8830, Tjele, Denmark
b

Received 19 January 1999; received in revised form 30 July 1999; accepted 26 August 1999

Abstract
The effects on silage composition of ensiling perennial ryegrass (PRG) with three commercial
tannins (mimosa, myrabolam and quebracho tannins) were assessed (Experiment 1). The effects on
silage composition and mobile bag disappearance of DM, nitrogen and true protein of the addition
of these tannins or a combination of tannin plus formic acid, formaldehyde alone, formic acid alone,
or a combination of formaldehyde and formic acid (Experiment 2) were also studied. In Experiment
1, the PRG was a third cut with a mean oven-dry DM content of 200 g/kg. All silages were prepared
on a small laboratory scale (500 g), with the additives added in 20 ml aliquots/kg herbage fresh
weight. The tannins were added at the rate of 5 or 50 g/kg herbage DM and in Experiment 1,
samples were examined after 7 or 32 days of ensiling. In Experiment 2, a 1st cut PRG with a mean
oven-dry DM content of 188 g/kg was used and the samples were taken on days 7, 14 and 49. In
Experiment 1, treatment with tannins reduced the soluble nitrogen (SN) and ammonia content of the
silages. In Experiment 2, the tannins used also reduced silage SN during ensiling, and were able to
reduce degradation of silage nitrogen and true proteins in the rumen. Of the tannins used, quebracho
tannin was also able to reduce SN and rumen degradation better than mimosa tannin. However, the

*

Corresponding author. Present address: Welsh Institute of Rural Studies, University of Wales, Aberystwyth,
SY23 3AL, Wales, UK. Tel.: ‡44-1970-621678; fax: ‡44-1970-611264

E-mail address: mbs@aber.ac.uk (M.B. Salawu)
1
Department of Biochemistry and Nutrition, SAC Auchincruive, Ayr, KA6 5HW, Scotland. UK.
0377-8401/99/$ ± see front matter # 1999 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 7 - 8 4 0 1 ( 9 9 ) 0 0 1 0 5 - 4

244

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

tannins were not as good as formaldehyde at protecting silage proteins both during ensiling and
in the rumen, neither were they better than formic acid in enhancing silage quality. For both
the tannins and formaldehyde, formic acid addition further reduced the SN content as a result
of the combined effect of rapid acidification and protein binding. # 1999 Elsevier Science B.V. All
rights reserved.
Keywords: Mimosa tannins; Mobile bags; Myrabolam tannins; Proteolysis; Quebracho tannins; Silage

1. Introduction
Throughout the world, grass is often ensiled and the resultant silage used as feed when
fresh grass is unavailable. During ensilage, some grass proteins are degraded to nonprotein nitrogen (NPN). Furthermore, residual silage protein is readily degraded in the

rumen, with the liberation of ammonia. There is considerable evidence to suggest that
ruminal undegradable protein is a desirable component of some feeds (ARC, 1984), as
proteins that by-pass the rumen digestive process are a potential source of amino acids for
absorption. Treatments of silage with chemicals like formaldehyde (Barry et al., 1978a, b;
McDonald et al., 1991; Henderson, 1993) have been used to protect proteins from
microbial and plant enzyme hydrolysis. High concentrations of formaldehyde may,
however, irreversibly bind the proteins and make them unavailable to the animal (Ashes
et al., 1984). Formic acid is also used to improve silage fermentation by lowering the pH
(Henderson, 1993), and thus reducing the activities of plant proteases. Reducing the pH
also prevents the growth of undesirable microbes (listeria, clostridia, enterobacteriaceae,
moulds etc. McDonald et al., 1991). Both formaldehyde and formic acid are toxic
chemicals which may cause severe skin, eye and respiratory irritation, and which may
release toxic fumes. Formaldehyde is also potentially carcinogenic (Sigma-Aldrich,
1988). Although, improvement in ensiling technologies has made safe handling of
chemical additives possible, bulk handling of these chemicals is hazardous and less toxic
alternatives would be valuable. Tannins have been identified as showing potential value
for the protection of proteins from degradation in the rumen (Waghorn et al., 1987;
McNabb et al., 1993; Salawu, 1997). However, questions remain regarding the release of
protein by the dissociation of tannin-protein complexes post-ruminally. Salawu et al.
(1999), has for example reported that with the tannin-rich browse legume Calliandra

calothyrsus, the utilization of by-pass protein in the intestine is poor. Separately, other
tannin-protein complexes have been reported to disappear almost totally in the rumen
(Muhammed, 1997). Notwithstanding the above, tannins appear to have potential as
protein protection agents during the ensilage of grass and its subsequent utilization by
ruminants. These studies were aimed at evaluating the potential of selected tannins as
silage additives. Experiments were conducted to study the effect of treating silages with
commercially available tannins on silage composition. The effect on silage composition
and total tract disappearance of tannins alone or in combination with formic acid, were
also compared with those of formic acid alone or formaldehyde alone or a combination of
formic acid and formaldehyde.

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

245

2. Materials and methods
2.1. Silage preparation, sampling and treatments
2.1.1. Experiment 1
The silages were made from third cut perennial ryegrass (PRG; Lolium perenne)
harvested in August 1995 and stored for 2 weeks at ÿ208C. One day prior to silage

making, the grasses were removed from the freezer and thawed overnight. The tannins
(mimosa tannin, myrabolam tannin and quebracho tannin; Hodgson Chemicals, London,
UK) were dissolved in water and dispensed in aliquots of 20 ml/kg fresh PRG. Two levels
of tannins (5 and 50 g tannin/kg DM) were used. For the 500 g model ensiling, and an
estimated DM of 200 g/kg, 5 or 0.5 g tannins were dissolved in 10 ml distilled water to
achieve the desired level of 50 or 5 g/kg DM, respectively. An equal volume of distilled
water was also added to the controls. Hand mixing of tannins with the grasses was carried
out in bulk at the same time for each treatment, after which, the grass was stored in 500 g
amounts in polythene bags simulating conditions in a silo. Four replicate treatments were
set up and the grass was ensiled for 7 and 32 days. Opened silages were divided into two
parts, of which one was used for the extraction of silage juice using a silage press. pH was
determined in the silage juice. The other part was freeze-dried and milled to pass through
a 1 mm aperture before chemical analyses. The treatments were thus:
1.
2.
3.
4.
5.
6.
7.

Control
Mimosa tannin (5 g/kg DM)
Mimosa tannin (50 g/kg DM)
Myrabolam tannin (5 g/kg DM)
Myrabolam tannin (50 g/kg DM)
Quebracho tannin (5 g/kg DM)
Quebracho tannin (50 g/kg DM).

2.1.2. Experiment 2
Perennial ryegrass (1st cut) was mown with a precision chopper on 12th June, 1996.
Grasses were chopped to about 5 cm in length using a silage chopper and ensiled on the
same day. Silages were prepared as described in Experiment 1 above. The tannins
(mimosa and quebracho) were added at a rate of 10 g/kg FW or 50 g/kg DM, and the
formaldehyde and formic acid were added at a rate of 2.5 g/kg FW or 12.5 g/kg DM
(w/w). The additions were based on an estimated DM content of 200 g/kg.
Three bags were opened from each treatment 7, 14 and 49 days after ensiling. The tie
end of each bag was discarded, and samples taken from the mid-sections. Samples were
divided into three portions, one of which was used to determine the DM and nitrogen.
Silage juice was extracted from the second portion for the determination of VFAs, lactate,

ethanol, ammonia, SN and WSC. The third portion was stored at ÿ208C, and later freezedried. Freeze-dried samples were ground to pass through 3.2 mm or 1 mm apertures (for
chemical analysis). Samples milled to pass through 3.2 mm screens were used for the
mobile bag degradability study in the gastrointestinal tract (GIT) of dairy cows. The
treatments were thus:

246

1.
2.
3.
4.
5.
6.
7.
8.

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

Control
Control

Control
Control
Control
Control
Control
Control

‡12.5 g formic acid/kg DM (FA)
‡12.5 g formaldehyde/kg DM (F)
‡12.5 g formaldehyde/kg DM ‡12.5 g formic acid/kg DM (F ‡ FA)
‡50 g mimosa tannins/kg DM (MT)
‡50 g mimosa tannins/kg DM ‡12.5 g formic acid/kg DM (MT ‡ FA)
‡50 g quebracho tannins/kg DM (QT)
‡50 g quebracho tannins/kg DM ‡12.5 g formic acid/kg DM (QT ‡ FA)

2.2. Chemical analysis
Nitrogen (Kjeldahl nitrogen) was determined in wet silage. Lactate, ethanol, volatile
fatty acids (VFAs), ammonia, soluble nitrogen (Kjeldahl nitrogen) and water-soluble
carbohydrates were determined in the silage juice. The silage oven DM and nitrogen
were determined following the procedure of the official method of analysis (AOAC,

1990). The DM content of the silages used for the mobile bag experiment was
determined before incubation, and on residual materials after incubation. The residual
materials were first freeze-dried and the freeze-dried weights taken. Six bags from
the freeze-dried residual material were randomly taken and further oven-dried at 508C
for 48 h. The calculated oven-dried DM was then used to correct the DM content
of the freeze-dried residual materials. For the analysis of DM, N and amino acid
disappearance in the mobile bags, dried residues were pooled into groups of two
according to the cows in which they were incubated in the rumen. This gave a final
total of three replicates per sample.
The pH of the silage was determined directly from the silage juice using a portable
Corning 103 pH meter (Corning Labware and Equipment, Corning, NY). The ammonia
contents of the silages were determined using the automated phenol-hypochlorite
procedure (Technicon method 321-74A). The fermentation product concentration of
silage juice was determined by high performance liquid chromatography (HPLC) using a
modification of the method described by Salawu et al. (1997b). Silage juice (0.1 ml) was
prepared for analysis by adding 0.1 ml of 50 mM H2SO4 plus 0.1 ml of internal standard
(1 ml of 2-ethyl-n-butyric acid made to volume with deionised distilled water in 100 ml
volumetric flask) plus 0.7 ml of deionised distilled water. The mixture was centrifuged at
9853  g for 15 min., and the aliquot analysed immediately. The stock standard
contained 5, 2.5, 2.5, 5, 2.5 and 2.5 mg/ml of acetate, propionate, butyrate, lactate, formic

acid, and ethanol, respectively. The fermentation products were quantified by reference to
an internal standard (2-ethyl-n-butyric acid), and standard chromatogram obtained from
the stock standards.
Water soluble carbohydrate (WSC) concentrations were determined using the
anthrone reaction rate assay (Koehler, 1952). Extractable proanthocyanidin (PA) was
determined from extracts of freeze-dried silages using the butanol-HCl method (Porter
et al., 1986). The nitrogen content of the residues were determined according to
the official method of analysis (AOAC, 1990), and amino acids were determined
following the Altech modified procedure of AOAC (1984) as described elsewhere
(Salawu et al., 1997b).

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

247

2.3. Total tract disappearance in mobile bags
The total tract disappearance of DM, nitrogen and true protein (sum of amino acids) of
the silages were studied in Friesian dairy cows using the mobile bag technique described
by Salawu et al. (1999). Mobile bags (18 in total i.e. six replicates each for rumen
degradability, pepsin-HCl (200 mg pepsin in 2 l 0.004 M HCl pH 2.4) digestion and

intestinal disappearance) were used for each silage sample. Silage samples were ground
to pass through a 3.2 mm sieve and 1 (0.2) g samples were weighed into each bag. The
bags were made of 100% mononylon with pore size of 11 mm (Sefar AG,
Mesh ‡ Technology, 8803 RuÈschlikon, Switzerland). The external dimensions of the
bags were 6  6 cm and were closed by heat sealing (Elwis pack AS, Denmark).
Rumen degradability was measured in three fistulated non-lactating Friesian cows. The
cows were about 3 years old and were maintained on a standard diet consisting of 675 g/
kg grass hay and 325 g/kg concentrate. The concentrate was composed of 30 g/kg soya
bean meal, 440 g/kg DM barley grain, 440 g/kg DM oat grain, 30 g/kg DM rape seed,
30 g/kg DM low ash fish meal and 30 g/kg DM sugar-beet molasses. The cows also
received 200 g/day mineral mix and 21 g/day vitamin mix. Six cows fitted with T-shaped
cannulae at the duodenum and the terminal ileum were used for the intestinal
disappearance study. These cows were fed a complete diet made up of 550 g/kg grass
silage, 245 g/kg grass hay, 133 g/kg concentrate, 68 g/kg barley, 2.7 g/kg mineral mix and
1 g/kg limestone. All the cows were fed twice daily at 0800 h and 1600 h in two equal
portions. A maximum of 12 mobile bags was introduced per cow per day into the
duodenum through T-shaped cannulae at the proximal duodenum. Bags were recovered
from the faeces after 24 h by washing the faeces through a sieved bucket before washing
in a front loading washing machine with cold water for 15 min without spinning, and then
frozen.
2.4. Statistical analysis
All the data collected were analysed using the Minitab statistical package (Minitab
Incorporation, 1995) as one way classified data. The means were separated using the least
significant difference method of Steele and Torrie (1980).

3. Results
3.1. Silage pH
The pH of the silages used in experiments 1 and 2 are presented in Tables 1 and 2,
respectively. In Experiment 1 the control silage had the lowest pH compared to those that
were treated with the tannins. Between the tannin treated silages, those made with the
lower level of the tannins (5 g/kg DM) had a higher (p < 0.05) pH after Day 32 than those
made with higher tannins. In Experiment 2, the PRG had an average pH of 6.1 before
ensiling. The addition of formic acid however, reduced (p < 0.05) the starting pH to 4.5
(Table 2). The pH of the silages dropped to between 4.0 and 4.5 after 7 days of ensiling,

248

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

Table 1
pH values of silage (Experiment 1)
Silage

Day 7

Day 32

Control
Mimosa (5 g/kg)
Mimosa (50 g/kg)
Myrabolam (5 g/kg)
Myrabolam(50 g/kg)
Quebracho (5 g/kg)
Quebracho(50 g/kg)
SEMa
Prob.b

4.6
4.9
5.3
5.0
5.2
5.2
5.3
0.11
***c

4.9
5.6
5.4
5.5
5.3
5.6
5.4
0.09
***

a

SEM : standard error of means.
Prob.: probability.
c
***p < 0.001.
b

Table 2
pH of silage (Experiment 2)a
Silage

Control
FA
F
F ‡ FA
MT
MT ‡ FA
QT
QT ‡ FA
SEM
Prob.

Day
0

7

14

49

6.1
4.6
6.2
4.5
6.1
4.6
6.2
4.5
0.07
***

4.0
4.1
4.4
4.3
4.0
4.5
4.0
4.3
0.12
***

4.0
4.4
4.1
4.5
3.9
4.0
4.0
3.8
0.08
***

3.9
4.4
3.9
4.7
3.9
3.8
3.8
3.8
0.11
***

a
FA: Perennial ryegrass (PRG) ‡ formic acid (2.5 g/kg FW); F: PRG ‡ formaldehyde (2.5 g/kg FW);
F ‡ FA: PRG ‡ F ‡ FA; MT: PRG ‡ mimosa tannin (50 g/kg DM); MT ‡ FA: PRG ‡ MT ‡ FA; QT:
PRG ‡ quebracho tannin (50 g/kg DM); QT ‡ FA: PRG ‡ QT ‡ FA; Prob.: probability; ***p < 0.001; SEM:
standard error of means.

with all the silages without added formic acid or formaldehyde having a lower pH. The
silages FA and F ‡ FA had a significantly higher pH than other silages after 49 days of
ensiling.
3.2. Dry matter, WSC and PA content of the silages
The oven-dry DM, WSC and PA contents of the silages in Experiment 1 are presented
in Table 3. Ensiling for 32 days generally resulted in a decrease in DM content of all the
silages. After correcting for the added tannins, there was no significant difference in the
DM of the silages. Silages treated with tannins generally had lower (p < 0.05) water

249

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

Table 3
Dry matter (DM; g/kg), water soluble carbohydrate (WSC; g/kg DM) and proanthocyanidin (PA; A550 nm/g)
content of silage samples opened after 32 days of ensilage in Experiment 1a
Silage

DM

WSC

PA

Control
Mimosa (5 g/kg)
Mimosa (50 g/kg)
Myrabolam (5 g/kg)
Myrabolam (50 g/kg)
Quebracho (5 g/kg)
Quebracho (50 g/kg)
SEM

183
182
186
185
193
179
185
3.26

20.3a
7.2be
8.2be
6.4b
14.2c
11.3d
9.3de
0.80

ND
30.0a
126.3b
30.3a
45.3c
38.7ac
172.3d
2.19

a
Means with common letters in columns are not significantly different (p > 0.05); SEM: standard error of
means; ND: not determined.

soluble carbohydrate (WSC) content than the control silage. With the exception of
myrabolam tannin-treated silages, treatment with high level of quebracho or mimosa
tannins did not significantly affect the WSC content. The relationship between the WSC
and PA contents of the silages was negative and strong (r ˆ ÿ0.84).
Table 4 shows the oven-dry DM (corrected for added tannins) and WSC content of the
silages in Experiment 2. The bags were filled immediately after harvesting, so the DM
content of the silages was very low (188 g/kg), the control silage having a significantly
lower (p < 0.05) DM content than the others. The DM (g/kg) content ranged between 151
for the control to 166 for the silage MT ‡ FA after 49 days. All of the silages containing
formic acid had significantly (p < 0.05) higher WSC concentrations than the
corresponding silages without formic acid after 7 and 14 days of ensilage. After 49
days of ensilage, concentration of WSC was numerically higher (p > 0.05) in tannin alone
silages in comparison with those containing a mixture of tannins and FA. Except for
silage F ‡ FA, the WSC content of the silages decreased with length of ensilage. The rate
Table 4
Dry matter (DM; g/kg) and water soluble carbohydrate (WSC; g/kg DM) content of silage (Experiment 2)a
Silage/days of ensilage

Control
FA
F
F ‡ FA
MT
MT ‡ FA
QT
QT ‡ FA
SEM
a

DM

WSC

7

14

49

7

14

49

166a
181b
172c
180bd
178de
180bd
170c
176e
0.9

168a
177b
165a
175bc
173c
173c
176bc
173c
1.2

151a
162bcd
160bc
159b
165cd
166d
164bcd
161bcd
1.7

42.9a
69.4b
63.2c
70.5b
39.2d
63.1c
39.6d
65.0c
1.1

40.0a
65.3bd
33.6c
66.5b
41.0a
61.6d
39.4a
52.8e
1.5

35.3a
62.4b
18.6c
73.2d
31.9ae
28.9e
34.0a
31.6ae
1.3

Means with common letters in columns are not significantly different (p > 0.05); SEM: standard error of
means.

250

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

Table 5
Fermentation products (g/kg DM) of silages opened after 32 of ensiling in Experiment 1a
Silage

Lactate

Acetate

Ethanol

Butyrate

Control
Mimosa (5 g/kg)
Mimosa (50 g/kg)
Myrabolam (5 g/kg)
Myrabolam (50 g/kg)
Quebracho (5 g/kg)
Quebracho(50 g/kg)
SEM

27.5a
2.1b
17.3c
8.2d
28.0a
1.5b
5.9d
0.5

11.2a
43.8b
17.9c
33.4d
16.9c
46.9e
30.6f
0.6

26.0a
40.9b
45.8c
33.4d
23.8a
31.9de
26.7ae
1.2

34.9a
18.5b
1.3c
20.5d
9.0e
17.2b
4.2f
0.4

a

Means with common letters in columns are not significantly different (p > 0.05); SEM: standard error of
means.

of decrease in WSC was, however, higher in silages made with a combination of tannins/
formic acid or formaldehyde alone, in comparison with silages made with tannins alone,
formic acid or the control silage.
3.3. Fermentation products
The fermentation products of the silages opened after 32 days of ensiling in
Experiment 1 are presented in Table 5. The lactic acid concentration was generally lower
(p < 0.05) in silages treated with lower level of tannins in comparison with the controls or
that of silages treated with higher level of tannins. All silages treated with tannins had
higher (p < 0.05) acetate concentration than the control. Treatment of the silages with
higher levels of tannins, however, significantly reduced their acetate concentration when
compared with silages made with the lower levels of tannins. The ethanol concentration
was higher (p < 0.05) in silages treated with mimosa tannins than the control or other
tannin treated silage. Ethanol concentration was comparable (p > 0.05) between the
control silage and silages made with higher level of myrabolam or quebracho tannins.
Butyrate concentration was lowest in silages treated with the higher level of the tannin,
and was generally higher (p < 0.05) in the control than in the tannin treated silage after 32
days of ensiling. The PA content of the silages was negatively related with the lactate
(r ˆ ÿ0.21) and butyrate (r ˆ ÿ0.84), positively but weakly related to the ethanol
(r ˆ 0.16) and showed no relationship with the acetate concentrations.
The fermentation products of the silages opened after 49 days of ensiling in
Experiment 2 are presented in Table 6. The predominant products of fermentation in all
the silages were lactate, acetate and ethanol. The lactate concentration was higher
(p < 0.05) in the control silage than all the other silages. Whereas a combination of
formaldehyde and formic acid significantly reduced the lactate content than when
formaldehyde was used alone, the effect of formic acid on lactate was not significant
when used with tannins. Acetate concentrations was generally lower (p < 0.05) in all the
silages which received formic acid compared to the silages without formic acid. With the
exception of silages treated with a combination of tannins and formic acid, the ethanol
concentration of all the other silages was significantly lower than that of the control.

251

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259
Table 6
Fermentation products (g/kg DM) of silages opened after 49 days of ensiling in Experiment 2a
Silage

Lactate

Acetate

Propionate

Ethanol

Butyrate

Control
FA
F
F ‡ FA
MT
MT ‡ FA
QT
QT ‡ FA
SEM

116a
10.9b
74.6c
2.9d
96.1eg
91.3e
104f
98.8fg
2.0

21.7a
6.7bd
20.7a
6.8bd
18.4c
6.4b
16.6c
8.5d
0.6

0.1a
2.2b
1.0c
8.7d
0.0a
0.2a
0.0a
0.0a
0.2

43.5a
24.5b
32.2c
18.3d
26.5b
46.0e
32.4c
54.7f
0.8

0.0a
20.5b
2.9c
16.5d
0.0a
2.7ce
0.1a
1.6e
0.4

a
Means with common letters in columns are not significantly different (p > 0.05); SEM: standard error of
means.

A combination of formic acid and tannins appeared to increase (p < 0.05) the ethanol
concentration than when the tannins were used alone. Propionate was not detected in
silages MT, QT and QT ‡ FA, and only traces (p > 0.05) was detected in the control
silage and MT ‡ FA. Propionate concentration of silage F ‡ FA was significantly higher
than that of other silages. The butyrate concentration was highest (p < 0.05) in silage FA,
and was significantly higher in all silages containing formic acid in comparison with
silages that did not receive formic acid.
3.4. Total nitrogen (TN), ammonia nitrogen (NH3±N) and soluble nitrogen (SN) content
of silage
The nitrogen contents of the silages are presented in Tables 7 and 8, respectively, for
experiments 1 and 2. In Experiment 1, significant difference was observed in the TN
content of some of the silages after 7 days of ensilage. However, after 32 days of ensilage,
Table 7
Total nitrogen (TN; g/kg DM), ammonia nitrogen (NH3±N; g/kg TN) and soluble nitrogen (SN; g/kg TN)
content of silage in Experiment 1a
Silage

Control
Mimosa (5 g/kg)
Mimosa (50 g/kg)
Myrabolam(5 g/kg)
Myrabolam (50 g/kg)
Quebracho (5 g/kg)
Quebracho(50 g/kg)
SEM
a

Day 7

Day 32

TN

NH3±N

SN

TN

NH3±N

SN

25.9ac
25.4a
27.3b
26.1c
26.7de
27.0bd
26.2ce
0.16

97a
108b
64c
118d
84e
125f
107b
1.45

376a
368a
279b
393a
326c
377a
287b
8.60

28.1
28.8
28.6
28.3
29.3
29.1
29.5
0.34

126a
203b
125a
225c
124a
173d
138e
2.44

572a
545b
439c
567a
471d
538b
425c
5.38

Means with common letters in columns are not significantly different (p > 0.05); SEM: standard error of
means.

252

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

Table 8
Total nitrogen (TN; g/kg DM), ammonia nitrogen (NH3±N; g/kg TN) and soluble nitrogen (SN; g/kg TN)
content of silage in Experiment 2a
Silage

Control
FA
F
F ‡ FA
MT
MT ‡ FA
QT
QT ‡ FA
SEM

Day 7

Day 14

Day 49s

TN

NH3±N

SN

TN

NH3±N

SN

TN

NH3±N

SN

26.4ac
27.9a
24.0bc
26.4ac
26.5ac
25.3c
27.5a
25.4c
0.6

28.8a
10.3b
19.9c
10.6b
19.4c
13.0d
13.7d
15.2e
0.4

619a
421b
476c
314d
474e
375f
384f
337g
3.5

26.6ac
25.3b
27.5a
25.9cb
26.0cb
26.0cb
24.9b
26.5ac
0.4

32.5a
16.7b
19.9c
14.5d
28.1e
29.9f
34.3g
27.5e
0.3

600a
410b
376c
248d
505e
366cf
510e
352f
4.2

31.6a
29.2bc
30.4ab
29.5abc
27.9c
27.7c
28.4bc
29.7abc
0.7

29.0a
30.5a
26.0b
26.1b
37.0c
34.1d
36.5c
32.9d
0.6

639a
504be
418ce
250d
545b
377c
452e
378c
16.8

a

Means with common letters in columns are not significantly different (p > 0.05); SEM: standard error of
means.

the difference in the TN content of all the silages was not significant. As with silages
opened after 7 days in Experiment 1, significant difference was also observed in the TN
content of the silages opened after days 7, 14 and 49 in Experiment 2. In both
experiments 1 and 2, there was an increase in TN content of the silages as the length of
ensiling increased.
At both sampling times in Experiment 1, the ammonia nitrogen (g/kg TN) content of
silages with the lower level of tannins was higher (p < 0.05) than those of the control
silage or those receiving more tannins. In Experiment 2, the NH3±N content was
significantly higher (p < 0.05) in the control silage than other silages after 7 days of
ensiling. Treatment with a combination of formaldehyde/formic acid or mimosa tannin/
formic acid significantly (p < 0.05) reduced the NH3±N content of the silages in
comparison with treatment with formaldehyde or mimosa tannin alone after 7 days of
ensilage. The NH3±N content increased with increasing time of ensiling. However, the
increase was less in the control silage and silages FA, F or F ‡ FA than in silages
containing tannins after 14 days of ensiling. Between Day 14 and 49 of ensiling, the rate
of increase in NH3±N content of the silages was higher in silages FA, F and F ‡ FA than
in the silages treated with tannins.
Silages made with the higher level of the tannins in Experiment 1 (Table 7) had a
significantly lower (p < 0.05) SN content than the control or silage made with lower
tannins. In Experiment 2 (Table 8), treatment of PRG with all of the additives used here
significantly (p < 0.05) reduced the SN content of the silage. The reduction in SN was,
however, greater (p < 0.05) when the silages were made with a combination of
formaldehyde/formic acid or tannins/formic acid. Thus, reductions in SN of 32, 23, 23,
and 38%, respectively, were obtained in silages FA, F, MT and QT in comparison with
reductions in SN of 49, 39 and 46% recorded in silages F ‡ FA, MT ‡ FA and QT ‡ FA,
respectively after 7 days of ensiling. After 49 days of ensiling, the reductions in SN
content of silages F ‡ FA, MT ‡ FA and QT ‡ FA were 61, 41 and 41%, respectively.
Between the tannins, quebracho tannins alone significantly lowered (p < 0.05) the SN

Silage

Control
FA
F
F ‡ FA
MT
MT ‡ FA
QT
QT ‡ FA
SEM
a

Dry matter (g/kg)

Nitrogen (g/kg TN)

True protein (g/kg total amino acids)

Rumen

Pepsin-HCl

Intestine

Rumen

Pepsin-HCl

Intestine

Rumen

Pepsin-HCl

Intestine

560a
624b
506cd
486cef
516d
495cde
474ef
469f
10.7

592a
644b
542ce
515d
545c
520de
522cd
515d
11.6

616ac
678b
597a
594a
605ac
599ac
606ac
622c
10.1

857a
815b
771c
634d
753c
763c
768c
725e
12.9

876a
860a
800b
689c
807b
831d
791b
755e
6.53

930a
942a
906b
833c
874d
927a
924ab
928a
8.31

833a
790bd
762bf
630c
800d
681e
757f
659e
13.8

860a
808b
790b
622c
820b
705d
797b
713d
13.2

929ac
949b
933a
932a
921c
907d
918c
947b
3.99

Means with common letters in columns are not significantly different (p > 0.05); cumulative disappearance: incremental disappearance across the gastrointestinal
tract; SEM: standard error of means.

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

Table 9
Cumulative disappearance of dry matter, nitrogen and true protein from grass silage in mobile bags in the rumen, pepsin-HCl and intestine of cowsa

253

254

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

content of the silage more than the mimosa tannins alone after 49 days of ensilage.
Formaldehyde alone only reduced (p < 0.05) the SN content more than MT alone. A
combination of formaldehyde/formic acid also significantly (p < 0.05) reduced the SN
content of the silage compared with the tannin/formic acid combinations after 49 days of
ensilage. The PA content of the silages was positively related with the TN (r ˆ 0.56) and
negatively related with the ammonia nitrogen (r ˆ ÿ0.37) and SN (r ˆ ÿ0.90) content of
the silages.
3.5. Disappearance of dry matter (DM), nitrogen and total amino acid from mobile bags
The apparent cumulative disappearance of DM, nitrogen and true proteins from mobile
bags in the rumen, pepsin-HCl solution and the intestine are presented in Table 9 as g/kg
of the original material. All the silages containing tannins had lower (p < 0.05) DM
disappearance in the rumen than the control silage. Similarly the proportion of rumen
undegradable DM which was lost in the lower tract was higher (p < 0.05) in silages
containing tannins (19.6%) and in silage F ‡ FA (18%) than in the control (11.2%) or
silage FA (9.4%) or F (15.6%). The total cumulative DM disappearance was, therefore,
essentially the same (p > 0.05) between the control silages and all the silages containing
tannins. Dry matter disappearance was highest in silage FA and lowest in silage F ‡ FA.
In comparison with the other silages, the control silage and silage F ‡ FA, respectively,
had the highest (p < 0.05) and lowest (p < 0.05) nitrogen disappearance in the rumen. The
cumulative disappearance of nitrogen after the passage of the mobile bags through the
intestine was comparably higher (p < 0.05) in the control and silages FA, MT ‡ FA, QT,
and QT ‡ FA than in other silages. The actual disappearance of rumen undegradable
nitrogen, in the intestines was highest in silage QT ‡ FA (18.6%) and lowest in the
control (5.9%) silage.
The pattern of true protein (defined as the sum of the amino acids measured)
disappearance in the silages was similar to that of the nitrogen disappearance. The control
silage lost a significantly higher (p < 0.05) proportion of its true protein in the rumen, and
only about 10% of the silage true protein was digested post ruminally. In contrast with the
control silage, a significantly lower percentage of the silage true protein disappeared in
the rumen with all the other silages that were treated with additives. A combination of
formic acid with formaldehyde or tannins significantly reduced (p < 0.05) the
disappearance of true protein in the rumen, and increased the actual disappearance in
the intestine of the rumen undegradable true protein, than formic acid, formaldehyde or
tannins alone. The silage F ‡ FA had the lowest (p < 0.05) disappearance of true protein
in the rumen and also in the pepsin-HCl solution, and the highest actual disappearance
(33.4%) of true protein in the intestine.

4. Discussion
One of the main reason for using tannins as silage additives was to see if they would be
able to protect grass/herbage proteins from rapid hydrolysis during ensiling, whilst being
less hazardous than existing alternative treatments. Hydrolysis of proteins is mainly due

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

255

to plant peptidase activities (Carpintero et al., 1979; McKersie, 1981; Heron et al., 1988).
It is optimal between pH 5±7, and continues at much reduced rate at a value below 4
(Henderson, 1993). Not surprisingly, acidifying the PRG with formic acid reduced the
initial pH to below the optimum pH level for plant peptidase activities, and thus reduced
the silage proteolysis (as indicated by reduced SN). Reduction in proteolysis by formic
acid has been documented by Barry et al. (1978a,b), Carpintero et al. (1979), Nsereko
(1996). The lowest pH of 4.6 attained by the control silage after 7 days of ensiling in
Experiment 1 was above the pH (4) below which the activities of plant proteases were
thought to be greatly reduced (Heron et al., 1988). Degradation of plant true proteins to
amino acids by plant proteases was, therefore, thought to be considerable in all the
silages. High levels of proteolysis may increase the buffering capacity of the silages and
this may further delay the attainment of low stable pH (Ohshima and McDonald, 1978;
Papadopoulos and McKersie, 1983; Henderson, 1993). The low final pH achieved in the
control silage in Experiment 2 and all the other silages that have no formic acid treatment
after 49 days of ensilage suggest that good fermentation occurred.
The high WSC content of control silage may have been due to an increased production
of WSC from cell wall carbohydrates (Carpintero et al., 1979; McDonald et al., 1991). In
contrast, the interaction of tannins with and their possible inhibition of cell wall
degradation (Barry and Manley, 1984; Terrill et al., 1989; Reed, 1995) may contribute to
the general reduction in WSC content of the tannin treated silage. However, the higher
WSC of silages treated with a combination of either tannins or formaldehyde and formic
acid on days 7 and 14 may be due to the effect of the rapid reduction in pH associated
with formic acid on the rate of interaction of the tannins or formaldehyde. Low pH has
been reported to reduce the interaction of the tannins with nutrients (Oh and Hoff, 1987;
Field and Lettinga, 1992). The oven DM reported in this study was only corrected for the
added tannins and not for volatile losses. However, lesser DM loss in silages treated with
tannins, formaldehyde or formic acid may be due to reduction in natural fermentation in
these silages. Tannins have been reported to inhibit the digestibility of carbohydrates and
proteins in plants (Barry and Manley, 1984; Terrill et al., 1989; Reed, 1995; PerezMaldonado and Norton, 1996). The higher level of tannins used here has also been
reported to reduce the activities of silage bacteria and moulds (Salawu, 1997), and could
have reduced the conversion of lactate to acetate, ethanol or butyrate. Since, clostridia are
considered to be mainly responsible for the production of butyric acid (Woolford, 1984),
the lower butyrate concentration in silages made with 50 g/kg DM tannins in Experiment
1, may be due to inhibition of clostridia by the tannins. Formic acid alone or in
combination with formaldehyde appears to reduce the activities of lactic acid bacteria.
However, the high concentration of ethanol and butyrate in relation to lactate and acetate
may suggest secondary fermentation in these silages. When formic acid was used with
tannins, lactate fermentation was evident. However, as mentioned earlier, a reduction in
pH by formic acid when used with the tannins was thought to reduce the rate of
fermentation.
The reduction in SN observed with the formaldehyde treated silage is presumably the
result of binding of the proteins by the formaldehyde (Barry et al., 1978a,b). However,
not all of the proteins were bound to the formaldehyde as indicated by the SN content
when formaldehyde was used alone. Also, some of the proteins would have been

256

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

hydrolysed before the reaction occurred. Rapid proteolysis has been noted to start
immediately after harvesting (Henderson, 1976; McKersie, 1981), and most of the
hydrolysis occurs during the first week of ensiling (Kemble, 1956). When formaldehyde
was used with formic acid, the combined effect of rapid acidification and protein binding
resulted in a much higher reduction in SN concentration (McDonald et al., 1991).
Reduction in SN with formaldehyde or formaldehyde/formic acid means a reduction in
soluble precursors for ammonia production, and hence a reduction in ammonia
concentration of the silage.
Treatment of herbage with tannins also reduced the SN content of the silage. Tannins
protect proteins from enzyme hydrolysis in a similar way to formaldehyde, by binding to
the proteins or free amino acid and rendering them inaccessible for enzyme hydrolysis
(Hagerman and Butler, 1981; Barry, 1989; Wang et al., 1996). Tannins also bind to
enzymes (Kumar and Singh, 1984; Horigome et al., 1988; Makkar et al., 1990; Salawu et
al., 1998). The observation of reduced SN made here with tannins is in agreement with
the findings of Albrecht and Muck (1991), who also showed a strong negative
relationship between tannin concentrations and SN concentrations in legume silages. The
differences in the tannins in the concentration of their active materials, molecular weight
and degree of polymerisation (Field and Lettinga, 1992; Reed, 1995), and pH of the
silages, may account for the differences observed in the ability of the different tannins to
cross link the protein/proteases and thus reduced proteolysis. As with formaldehyde,
when tannin was used in combination with formic acid, there was a combined effect of
the acid (rapid reduction in pH and thus activities of plant proteases) and the tannins
(complexing with the proteins). This is perhaps a little surprising since the reduced pH
may reduce the ability of tannins to complex with the various nutrients (Oh and Hoff,
1987; Field and Lettinga, 1992; Salawu et al., 1997a). The tannins alone were not as
effective at reducing proteolysis as formaldehyde alone. Similarly, a combination of
tannin/formic acid was not as effective as formaldehyde/formic acid. The reason for the
stronger effect of formaldehyde is not known, but the lower mass of this compound may
enable it to gain access to binding sites unavailable to tannins. It is also possible that the
chemical structure, molecular weight and degree of polymerisation of the tannins used
here does not permit effective cross linking with the silage proteins and enzymes.
Quebracho tannins for example have a branched chain structure which is compact and not
very accessible for binding to macromolecules (Hagerman and Robbins, 1993). It should
also be noted that the tannins used here are crude extracts that contain mixture of
compounds. The reason for the differences in TN between silages and with length of
ensilage is not known. It may, however, be partly due to the differences in DM content of
the silages.
In addition to reducing proteolysis during ensiling, both formaldehyde and the tannins
also reduced the disappearance of DM, nitrogen and amino acids in the rumen, and
shifted the point of disappearance of undegraded nitrogen and amino acids to the lower
digestive tract. Reduced intestinal disappearance was not expected with the level of
formaldehyde used here, however, higher rates of inclusion have been reported to be
capable of binding to proteins irreversibly, rendering them unavailable to the animal
(Ashes et al., 1984). Similarly, because of crude nature and the likely changes during
ensilage to the tannins used here, the higher rate of addition (50 g/kg DM) may not

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

257

adversely affect microbial metabolism in the rumen. This will, however, require
verification in a feeding trial. Thus, in both the formaldehyde and tannin treated silages,
the reductions in rumen degradation were compensated for by increased disappearance in
the intestine. It should however, be noted that contrary to our expectations, most of the
nitrogen in the tannins and formaldehyde treated silages disappeared in the rumen. This
suggests that both the tannins and formaldehyde were able to protect herbage proteins
from plant/microbial enzyme hydrolysis during ensiling, but were unable to sufficiently
protect the protein from hydrolysis in the rumen. As mentioned earlier, the low final pH
of the silage may weaken the strength of tannin-protein or formaldehyde-protein
complexes thereby making them more susceptible to loss in the rumen. This characteristic
has also been found with tannin-casein and tannins-bovine serum albumin complexes
(Muhammed, 1997; Salawu et al., 1997a).

5. Conclusions
The fermentation and quality of silages obtained in Experiment 1 with or without
tannins were generally poor. However, the results obtained in Experiment 2 suggest that
tannins might be used as silage additives. The most obvious effect of tannins in both
experiments is in the reduction in SN and as such proteolysis during ensilage. The tannins
were also able to reduce degradation of silage true proteins in the rumen, and protected
proteins were digested in the intestine without any adverse effect on the overall DM and
nitrogen digestion. The tannins used here were not as good as formaldehyde in protecting
silage proteins, and were unable to enhance silage quality like formic acid. However, the
wide variety of tannins available, and the extremely hazardous properties of
formaldehyde, justify further work to establish the potential for the use of a variety of
tannins at differing concentrations and conditions of use as silage additives.

Acknowledgements
The financial support of the Commonwealth Scholarship Commission and the
European Union through project ERBTS3 CT930211 is gratefully acknowledged.

References
Albrecht, K.A., Muck, R.E., 1991. Proteolysis in ensiled forage legumes that vary in tannin concentration. Crop
Sci. 31, 464±469.
Agricultural Research Council (ARC) 1984. The nutrient requirements of ruminant livestock. Supplement No. 1.
Report of the protein group of the ARC workshop party. Commonwealth Agricultural Bureau, Farnham
Royal, UK.
Ashes, J.R., Mangan, J.L., Sidhu, G.S., 1984. Nutritional availability of amino acids from protein cross-linked to
protect against degradation in the rumen. Brit. J. Nutr. 52, 239±246.
Association of Official Analytical Chemist (AOAC) 1984. Schedule 43.236-43.264.(Amino acid analysis).
Official Method of Analysis, AOAC, Washington, DC.

258

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

Association of Official Analytical Chemist (AOAC) 1990. Official Method of Analysis, AOAC, Washington,
DC.
Barry, T.N., 1989. Condensed tannins: Their role in ruminant protein and carbohydrate digestion and possible
effects upon the rumen ecosystem. In: Nolan, J.V., Leng, R.A., Demeyer, D.I. (Eds.), The Roles of Protozoa
and Fungi In Ruminant Digestion. Penambul Books, Armidale, NSW, Australia, pp. 153±169.
Barry, T.N., Manley, T.R., 1984. The role of condensed tannins in the nutritional value of Lotus pedunculatus for
sheep (Quantitative digestion of carbohydrates and proteins). Brit. J. Nutr. 51, 493±504.
Barry, T.N., Cook, J.E., Wilkins, R.J., 1978a. The influence of formic acid and formaldehyde additives and type
of harvesting machine on the utilization of silages 1. The voluntary intake and nitrogen retention of young
sheep consuming the silages with and without intraperitoneal supplements of DL-methionine. J. Agric. Sci.
Cambr. 91, 701±715.
Barry, T.N., Cook, J.E., Wilkins, R.J., Beever, D.E., 1978b. The influence of formic acid and formaldehyde
additives and type of harvesting machine on the utilization of silages 2. Changes in amino acid composition
during ensiling and their influence on nutritive value. J. Agric. Sci. Cambr. 91, 717±725.
Carpintero, C.M., Henderson, A.R., McDonald, P., 1979. The effect of some pre-treatments on proteolysis
during the ensiling of herbage. Grass Forage Sci. 34, 311±315.
Field, J.A., Lettinga, G., 1992. In: Hemingway, R.W, Laks, P.E., (Eds.), Toxicity Of Tannic Compounds To
Microorganisms. Plenum Press, New York, pp. 673±692.
Hagerman, A.E., Butler, L.G., 1981. The specificity of proanthocyanidin-protein interactions. J. Biol. Chem.
256, 4494±4497.
Hagerman, A.E., Robbins, C.T., 1993. Specificity of tannin-binding proteins relative to diet selection by
mammals. Can. J. Zool. 71, 628±633.
Henderson, H.E., 1976. Acidity and proteolysis as factors affecting the nutritive value of corn silage. J. Anim.
Sci. 42, 208±218.
Henderson, N., 1993. Silage additives. Anim. Feed Sci. Technol. 45, 35±56.
Heron, S.J.E., Edwards, R.A., Phillips, P., 1988. The effect of pH on the activity of ryegrass proteases. J. Sci.
Food Agric. 46, 267±277.
Horigome, T., Kumar, R., Okamoto, K., 1988. Effects of condensed tannins prepared from leaves of fodder
plants on digestive enzymes in vitro and in the intestine of rats. Brit. J. Nutr. 60, 275±285.
Kemble, A.R., 1956. Studies on the nitrogen metabolism of the ensilage process. J. Sci. Food Agric. 7, 125±130.
Koehler, L.H., 1952. Differentiation of carbohydrate by anthrone reaction rate and colour intensity. Anal. Chem.
24, 1576±1579.
Kumar, R., Singh, M., 1984. Tannins, their adverse role in ruminant nutrition. J. Agric. Food Chem. 32, 447±
453.
Makkar, H.P.S., Dawra, R.K., Singh, B., 1990. In vitro effect of oak tannins on some hydrolytic and ammonia
assimilating enzymes of the bovine rumen. Ind. J. Anim. Nutr. 7, 207±210.
McDonald, P., Henderson, A.R., Heron, S.J.E., 1991. The Biochemistry Of Silage. Chalcombe, Marlow, UK.
McKersie, B.D., 1981. Proteinases and peptidases of alfalfa herbage. Can. J. Plant Sci. 61, 53±59.
McNabb, W.C., Waghorn, G.C., Barry, T.N., Shelton, I.D., 1993. The effect of condensed tannins in Lotus
pedunculatus on the digestion and metabolism of methionine, cystine and in organic sulphur in sheep. Brit. J.
Nutr. 70, 647±661.
Minitab Incorporation 1995. Minitab for Windows, Release 10.51 Extra. Enterprise Drive, State College, PA,
USA.
Muhammed, S., 1997. Antinutrient effects of plant phenolic compounds. Ph.D. Thesis, University of Aberdeen,
Scotland, UK.
Nsereko, V.L.J., 1996. Peptidase inhibitors as additives for ensilage: Effect on silage characteristics with
reference to peptide nitrogen. Ph.D. Thesis, University of Aberdeen, Scotland, UK.
Oh, H.I., Hoff, J.E., 1987. pH dependence of complex formation between condensed tannins and protein. J. Food
Sci. 52, 1267±1270.
Ohshima, M., McDonald, P., 1978. A review of the changes in nitrogenous compounds in herbage during
ensiling. J. Sci. Food Agric. 29, 497±505.
Papadopoulos, Y.A., McKersie, B.D., 1983. A comparison of protein degradation during wilting and ensiling of
six forage species. Can. J. Plant. Sci. 63, 903±912.

M.B. Salawu et al. / Animal Feed Science and Technology 82 (1999) 243±259

259

Perez-Maldonado, R.A., Norton, B.W., 1996. The effects of condensed tannins from Desmodium intortum and
Calliandra calothyrsus on protein and carbohydrate digestion in sheep and goats. Brit. J. Nutr. 76, 515±533.
Porter, L.J., Hrstich, L.N., Chan, B.C., 1986. The conversion of procyanidins and prodelphinidins to cyanidin
and delphinidin. Phytochemistry 25, 223±230.
Reed, J.D., 1995. Nutritional toxicology of tannins and related polyphenols in forage legumes. J. Anim. Sci. 73,
1516±1528.
Salawu, M.B., 1997. The nutritive value of the leguminous browse Calliandra calothyrsus and the role of tannins
in ruminant feeds. Ph.D. Thesis, University of Aberdeen, Scotland, UK.
Salawu, M.B., Acamovic, T., Stewart, C.S., Hovell, F.D. D., 1997a. Proanthocyanidin content of Calliandra
leaves and the effect of polyethylene glycol and pH on complexes formed between Calliandra tannins and
Bovine serum albumin, Proc. Brit. Soc. Anim. Sci. Scarborough, UK.
Salawu, M.B., Acamovic, T., Stewart, C.S., Hovell, F.D.D., McKay, I., 1997b. Assessment of the nutritive value
of Calliandra calothyrsus: In sacco degradability and in vitro gas production in the presence of quebracho
tannins with or without Browse Plus. Anim Feed Sci Technol. 69, 219±232.
Salawu, M.B., Acamovic, T., Stewart, C.S., 1998. Calliandra calothyrsus leaf extracts' effects on microbial
growth and enzyme activities. In: Garland, T., Barr, A.E., (Eds.), Toxic Plants and Other Natural Toxicants.
CAB International, Wallingford, UK, pp. 509±513.
Salawu, M.B., Acamovic, T., Stewart, C.S., Hvelplund, T., Weisbjerg, M.R., 1999. The disappearance of dry
matter, amino acids and proanthocyanidins in the gastro intestinal tract from different fractions of Calliandra
leaves. Anim. Feed Sci. Technol. 79, 289±300.
Sigma-Aldrich 1988. In: Lenga, R.E., (Ed.), The Sigma-Aldrich Library of Chemical Safety Data, vol. 1. SigmaAldrich, Milwaukee, pp. 1733±1735
Steele, R.G.D., Torrie, J.H., 1980. Principles and Procedures Of Statistical Analysis. MacGraw Hill, New York.
Terrill, T.H., Windham, W.R., Hoveland, C.S., Amos, H.E., 1989. Forage preservation method influences on
tannin concentration, intake and digestibility of Sericea lespedeza by sheep. Agron. J. 81, 435±439.
Waghorn, G.C., Ulyatt, M.J., John, A., Fisher, M.T., 1987. The effect of condensed tannins on the site of
digestion of amino acids and other nutrients in sheep fed on Lotus corniculatus L. Brit. J. Nutr. 57, 115±126.
Wang, Y., Waghorn, G.C., McNabb, W.C., Barry, T.N., Hedley, M.J., Shelton, I.D., 1996. Effect of condensed
tannins in Lotus corniculatus upon the digestion of methionine and cysteine in the small intestine of sheep. J.
Agric. Sci. Cambr. 127, 413±421.
Woolford, M.K., 1984. The Silage Fermentation. Microbiology, Series 14. Marcel Dekker, New York.