Directory UMM :Data Elmu:jurnal:A:Animal Feed Science and Technology:Vol86.Issue1-2.Jul2000:
Animal Feed Science and Technology
86 (2000) 107±115
Short communication
In¯uence of non®ber carbohydrate concentration
on forage ®ber digestion in vitro$
S.G. Haddada, R.J. Grantb,*
a
Department of Animal Production, Jordan University of Science and Technology,
P.O. Box 3030, Amman, Jordan
b
Department of Animal Science, University of Nebraska, Lincoln, NE 68583-0908, USA
Received 22 November 1999; received in revised form 9 May 2000; accepted 25 May 2000
Abstract
We quanti®ed the effect of dietary non®ber carbohydrate (NFC) concentration (30, 35, 40, or
45% of DM) on in vitro digestion kinetics of neutral detergent ®ber (NDF) from alfalfa and corn
silages at pH 5.8 or 6.8. The objective was to simulate the effect of diets differing in effective ®ber
content on the optimal NFC-to-NDF ratio that resulted in maximal ®ber digestion. Ash-free NDF
was determined at 0, 3, 6, 9, 12, 24, 30, 36, 48, 72, and 96 h of fermentation. Kinetic parameters
were estimated using logarithmic transformation and linear regression. The optimal NFC-to-NDF
ratio for maximal NDF digestion differed between the two forages. For alfalfa fermented at pH 6.8,
apparent extent of ruminal NDF digestion was greatest between 30 and 40% NFC, but at pH 5.8
NDF digestion was greatest at 35% NFC. For corn silage fermented at either pH 6.8 or 5.8, apparent
extent of NDF digestion was greatest at 30% NFC. An NFC-to-NDF ratio of 0.70±1.20 maximized
NDF digestion for alfalfa only when fermentation pH was maintained at 6.8. These in vitro results
demonstrate that the optimal dietary NFC content for maximum NDF digestion in the rumen for a
particular forage will be a function of fermentation pH that re¯ects the physically effective NDF
content of the diet. # 2000 Elsevier Science B.V. All rights reserved.
Keywords: Non®ber carbohydrate; Fiber; Ruminal pH; Digestion kinetics
1. Introduction
Carbohydrates comprise the largest single dietary component of rations fed to dairy
cows (up to 75% of DM). The commonly used carbohydrate fractions, namely neutral
detergent ®ber (NDF) and non®ber carbohydrates (NFC), comprised primarily of starch,
$
Published with the approval of the director as Paper Number12831, Journal Series, Nebraska Agricultural
Research Division.
*
Corresponding author. Tel.: 1-402-472-6442; fax: 1-402-472-6362.
E-mail address: rgrant1@unl.edu (R.J. Grant)
0377-8401/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 7 - 8 4 0 1 ( 0 0 ) 0 0 1 6 0 - 7
108
S.G. Haddad, R.J. Grant / Animal Feed Science and Technology 86 (2000) 107±115
sugars, pectin, and b-glucans (Van Soest et al., 1991). The NFC content of a diet or
feedstuff is estimated commonly by subtracting crude protein (CP), NDF, and ether
extract from organic matter with a correction for CP included in NDF (Van Soest et al.,
1991). Although the optimum dietary concentration of NFC for lactating dairy cows is
uncertain, recent research with alfalfa-based diets fed to high-producing dairy cows
suggested that diets should contain >30% NFC (DM basis), with a negligible bene®t of
feeding 42 versus 36% NFC (Batajoo and Shaver, 1994).
Milk production and ruminal ®ber digestion responses to varying dietary concentrations of NFC appear to be a function of ruminal degradability of NFC, ruminal pH, and
®ber source. Previous research indicates that lower dietary NFC was associated with
higher ruminal NDF digestibility when pH increased; however, ruminal NDF digestion
was unaffected by reduced NFC content when ruminal pH was unchanged (Sievert and
Shaver, 1993a, b; Batajoo and Shaver, 1994). During the ®rst 8 h post-feeding, Sievert
and Shaver (1993a, b) measured ruminal pH to be consistently below 6.0 for dairy cows
fed diets containing 42 and 35% NFC in dietary DM. Additionally, the diurnal range in
ruminal pH observed in lactating dairy cows can fall between 6.8 and 5.5, with ruminal
pH below 6.2 for 70±80% of the day for diets containing 50±60% concentrates (Robinson
et al., 1986). Consequently, the effect of NFC content on ruminal pH is crucial for
determining the optimal NFC content in lactation diets.
Physically effective NDF has been de®ned as the proportion of NDF that stimulates
rumination, and has been proposed as a major determinant of ruminal pH (Mertens,
1997). The importance of the long particle fraction (measured as physically effective
NDF) at determining ruminal pH is illustrated by diets for which NFC content remains
the same, but forage is either coarse or ®nely chopped. At the same dietary NFC content,
®nely chopped forage has resulted in low ruminal pH and ®ber digestion, while the same
forage, when coarsely chopped, resulted in pH above 6.0 and higher ®ber digestion
(Grant et al., 1990). Consequently, dietary NFC content and fermentability is important in
determining ruminal pH, but forage physical form (physically effective NDF) exerts the
dominant control over pH.
Grant and Mertens (1992a) developed and described a buffer system that can be used
with a batch in vitro fermentation system to maintain constant pH in the range of 5.5 to
6.8 in order to facilitate measurement of the effect of pH on NDF digestion. Using a
buffer system that maintains stable pH over 96 h of in vitro fermentation allows the
effects of NFC content and fermentation pH on forage NDF digestion to be partitioned.
The objective of this experiment was to determine the effect of dietary NFC concentration
on the kinetics of NDF digestion for two common forages at high and low pH,
representative of adequate or de®cient physically effective NDF.
2. Materials and methods
2.1. Substrate preparation
Forage and concentrate substrates were dried at 558C for 48 h and ground through a
1-mm screen using a Wiley mill (Arthur H. Thomas Co., Philadelphia, PA). Substrates
were analyzed for CP (Association of Of®cial Analytical Chemists, 1990), ADF and NDF
109
S.G. Haddad, R.J. Grant / Animal Feed Science and Technology 86 (2000) 107±115
Table 1
Chemical composition (g per 100 g DM) of substrate ingredients
Ingredient
CPa
ADFb
NDFc
Ash
EEd
NFCe
Alfalfa silage
Corn silage
Corn grain
Soybean hulls
Soybean meal
19.0
8.5
10.0
12.1
47.6
35.1
25.0
3.8
50.0
4.9
45.1
46.0
9.0
67.0
10.6
10.8
3.9
2.1
5.1
6.9
3.8
3.0
4.3
2.1
1.5
21.3
34.6
74.6
13.7
33.4
a
Crude protein.
Acid detergent ®ber.
c
Neutral detergent ®ber.
d
Ether extract.
e
Non®ber carbohydrates (100ÿCPÿNDFÿAshÿEE).
b
(Van Soest et al., 1991), ether extract (Association of Of®cial Analytical Chemists, 1990),
and ash (Association of Of®cial Analytical Chemists, 1990). Chemical compositions of
the late-bud stage alfalfa silage, physiologically mature corn silage, soybean hulls, corn,
and soybean meal are listed in Table 1. The individual ingredients were combined in the
proportions shown in Table 2 to create substrates representing isonitrogenous diets
containing either 30, 35, 40, or 45% NFC with either alfalfa silage or corn silage as the
forage. To achieve these calculated NFC concentrations in the samples, soybean hulls
were substituted incrementally for the corn grain.
Table 2
Ingredient and chemical composition (g per 100 g DM) of substrates used for in vitro experiment
Item
Diet
Alfalfa
a
Corn
30
35
40
45
30
35
40
45
Ingredients
alfalfa silage
corn silage
corn grain
Soybean hulls
soybean meal
50.0
±
14.7
25.0
10.3
50.0
±
22.9
16.7
10.4
50.0
±
31.3
8.3
10.4
50.0
±
39.6
±
10.4
±
50.0
±
26.5
23.5
±
50.0
6.1
19.5
24.4
±
50.0
14.6
11.0
24.4
±
50.0
24.8
±
25.2
Compositionb
CP
ADF
NDF
EE
Ash
NFC
NFC-to-NDF
18.5
30.7
41.0
3.2
7.6
29.7
0.72
18.4
27.4
36.9
3.4
7.4
33.9
0.92
18.3
23.5
32.0
3.6
7.2
39.2
1.23
18.4
19.7
26.2
3.8
7.0
44.5
1.69
18.5
27.2
42.3
2.4
4.9
31.5
0.74
18.5
24.0
38.4
2.5
4.7
35.5
0.92
18.6
19.9
33.1
2.7
4.5
40.9
1.24
18.5
15.1
27.1
2.9
4.2
47.1
1.74
a
Percentage of non®ber carbohydrate contained in the substrate.
CP, crude protein; ADF, acid detergent ®ber; NDF, neutral detergent ®ber; EE, ether extract; NFC, non®ber
carbohydrate.
b
110
S.G. Haddad, R.J. Grant / Animal Feed Science and Technology 86 (2000) 107±115
A 300-mg sample of each of the eight substrates was weighed into 50-ml
polypropylene tubes for measurement of in vitro NDF digestion kinetics. The entire
experiment was replicated three times.
2.2. In vitro procedure
The in vitro procedure utilized was that described by Grant and Weidner (1992). The
buffer solution was that of Goering and Van Soest (1970) adjusted to pH 5.8 with 1 M
citric acid or formulated for pH 6.8 as described by Grant and Mertens (1992a).
Fermentation times (at 398C) were 0, 3, 6, 9, 12, 24, 30, 36, 48, 72, and 96 h. Tubes were
swirled gently at inoculation and at each remaining time thereafter.
Ash-free NDF was measured at each time (Van Soest et al., 1991). To ensure complete
starch solubilization with the starch-containing substrates, a heat-stable amylase
(ANKOM Tech. Corp., Fairport, NY) was used during sample re¯ux and at ®ltering.
The ruminal ¯uid inoculum was obtained from a steer-fed medium-quality alfalfa hay. At
collection, the mean pH of the ruminal ¯uid was 6.20 (SE0.11) for the three replicates
of the experiment. Although inoculum source can affect ®ber digestion kinetics measured
in vitro, donor animals were not available for every substrate studied, so alfalfa was fed to
the donor steer because it has been shown to promote rapid NDF digestion rates (Van
Soest, 1994). Jung and Varel (1988) demonstrated that ruminal inoculum from cattle fed
alfalfa hay degrades ®ber fractions of forages more rapidly than inoculum from a wide
range of forage diets including bromegrass, switchgrass, or corn silage.
As tubes were removed from the water bath, the pH of the contents for each substrate
was measured to monitor stability of the buffer pH, which did not decrease more than
0.18 pH unit for any substrate during 96 h of fermentation. Therefore, actual pH for each
treatment remained suf®ciently separated for the results to be biologically meaningful.
2.3. Statistical analysis
The model for the kinetics of NDF digestion was that described by Mertens and Loften
(1980),
Y D0 eÿkd tÿL INDF;
where Y is the NDF residue at time t of in vitro fermentation, D0 the potentially digestible
NDF fraction (proportion of initial DM), kd the fractional rate constant of NDF digestion
(hÿ1), L the discrete lag time (h), t the time of fermentation (h), and INDF the indigestible
NDF residue at 96 h (proportion of initial DM).
The remaining potentially digestible fraction at each time was transformed
logarithmically, and the NDF digestion parameters were estimated by linear regression.
Discrete lag time was calculated using the equation
L
ln D0 ÿ ln D00
;
ÿkd
where D00 is intercept of the equation of ln (Yÿ(96-h) NDF residue) over time at t0.
S.G. Haddad, R.J. Grant / Animal Feed Science and Technology 86 (2000) 107±115
111
Potential extent of NDF digestion (PED) was determined using the formula
PED 100
D0
:
D0 INDF
Predicted apparent extent of ruminal NDF digestion (AED) was calculated using the
equation described by Miller and Muntifering (1985),
AED PEDeÿkp L
kd
;
kd kp
where kp is the fractional rate of ®ber particle passage from the rumen, which was set at
either 0.050/h for particles of NDF or at 0.030/h (Shaver et al., 1986) to simulate faster
and slower particle passage from the rumen as a result of high or low feed intake.
Parameter estimates of the NDF digestion model were analyzed using the general
linear models procedure of SAS (1985), with a factorial arrangement of forage (alfalfa or
corn silage), NFC concentration (30, 35, 40, or 45%), and buffer pH (5.8 or 6.8). The
model included factors for replicate, forage, NFC concentration, pH, and interactions.
Differences among treatment means for signi®cant main effects were detected using
Student±Newman±Keul's multiple range test (SAS, 1985). Signi®cance was declared at
p
86 (2000) 107±115
Short communication
In¯uence of non®ber carbohydrate concentration
on forage ®ber digestion in vitro$
S.G. Haddada, R.J. Grantb,*
a
Department of Animal Production, Jordan University of Science and Technology,
P.O. Box 3030, Amman, Jordan
b
Department of Animal Science, University of Nebraska, Lincoln, NE 68583-0908, USA
Received 22 November 1999; received in revised form 9 May 2000; accepted 25 May 2000
Abstract
We quanti®ed the effect of dietary non®ber carbohydrate (NFC) concentration (30, 35, 40, or
45% of DM) on in vitro digestion kinetics of neutral detergent ®ber (NDF) from alfalfa and corn
silages at pH 5.8 or 6.8. The objective was to simulate the effect of diets differing in effective ®ber
content on the optimal NFC-to-NDF ratio that resulted in maximal ®ber digestion. Ash-free NDF
was determined at 0, 3, 6, 9, 12, 24, 30, 36, 48, 72, and 96 h of fermentation. Kinetic parameters
were estimated using logarithmic transformation and linear regression. The optimal NFC-to-NDF
ratio for maximal NDF digestion differed between the two forages. For alfalfa fermented at pH 6.8,
apparent extent of ruminal NDF digestion was greatest between 30 and 40% NFC, but at pH 5.8
NDF digestion was greatest at 35% NFC. For corn silage fermented at either pH 6.8 or 5.8, apparent
extent of NDF digestion was greatest at 30% NFC. An NFC-to-NDF ratio of 0.70±1.20 maximized
NDF digestion for alfalfa only when fermentation pH was maintained at 6.8. These in vitro results
demonstrate that the optimal dietary NFC content for maximum NDF digestion in the rumen for a
particular forage will be a function of fermentation pH that re¯ects the physically effective NDF
content of the diet. # 2000 Elsevier Science B.V. All rights reserved.
Keywords: Non®ber carbohydrate; Fiber; Ruminal pH; Digestion kinetics
1. Introduction
Carbohydrates comprise the largest single dietary component of rations fed to dairy
cows (up to 75% of DM). The commonly used carbohydrate fractions, namely neutral
detergent ®ber (NDF) and non®ber carbohydrates (NFC), comprised primarily of starch,
$
Published with the approval of the director as Paper Number12831, Journal Series, Nebraska Agricultural
Research Division.
*
Corresponding author. Tel.: 1-402-472-6442; fax: 1-402-472-6362.
E-mail address: rgrant1@unl.edu (R.J. Grant)
0377-8401/00/$ ± see front matter # 2000 Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 7 - 8 4 0 1 ( 0 0 ) 0 0 1 6 0 - 7
108
S.G. Haddad, R.J. Grant / Animal Feed Science and Technology 86 (2000) 107±115
sugars, pectin, and b-glucans (Van Soest et al., 1991). The NFC content of a diet or
feedstuff is estimated commonly by subtracting crude protein (CP), NDF, and ether
extract from organic matter with a correction for CP included in NDF (Van Soest et al.,
1991). Although the optimum dietary concentration of NFC for lactating dairy cows is
uncertain, recent research with alfalfa-based diets fed to high-producing dairy cows
suggested that diets should contain >30% NFC (DM basis), with a negligible bene®t of
feeding 42 versus 36% NFC (Batajoo and Shaver, 1994).
Milk production and ruminal ®ber digestion responses to varying dietary concentrations of NFC appear to be a function of ruminal degradability of NFC, ruminal pH, and
®ber source. Previous research indicates that lower dietary NFC was associated with
higher ruminal NDF digestibility when pH increased; however, ruminal NDF digestion
was unaffected by reduced NFC content when ruminal pH was unchanged (Sievert and
Shaver, 1993a, b; Batajoo and Shaver, 1994). During the ®rst 8 h post-feeding, Sievert
and Shaver (1993a, b) measured ruminal pH to be consistently below 6.0 for dairy cows
fed diets containing 42 and 35% NFC in dietary DM. Additionally, the diurnal range in
ruminal pH observed in lactating dairy cows can fall between 6.8 and 5.5, with ruminal
pH below 6.2 for 70±80% of the day for diets containing 50±60% concentrates (Robinson
et al., 1986). Consequently, the effect of NFC content on ruminal pH is crucial for
determining the optimal NFC content in lactation diets.
Physically effective NDF has been de®ned as the proportion of NDF that stimulates
rumination, and has been proposed as a major determinant of ruminal pH (Mertens,
1997). The importance of the long particle fraction (measured as physically effective
NDF) at determining ruminal pH is illustrated by diets for which NFC content remains
the same, but forage is either coarse or ®nely chopped. At the same dietary NFC content,
®nely chopped forage has resulted in low ruminal pH and ®ber digestion, while the same
forage, when coarsely chopped, resulted in pH above 6.0 and higher ®ber digestion
(Grant et al., 1990). Consequently, dietary NFC content and fermentability is important in
determining ruminal pH, but forage physical form (physically effective NDF) exerts the
dominant control over pH.
Grant and Mertens (1992a) developed and described a buffer system that can be used
with a batch in vitro fermentation system to maintain constant pH in the range of 5.5 to
6.8 in order to facilitate measurement of the effect of pH on NDF digestion. Using a
buffer system that maintains stable pH over 96 h of in vitro fermentation allows the
effects of NFC content and fermentation pH on forage NDF digestion to be partitioned.
The objective of this experiment was to determine the effect of dietary NFC concentration
on the kinetics of NDF digestion for two common forages at high and low pH,
representative of adequate or de®cient physically effective NDF.
2. Materials and methods
2.1. Substrate preparation
Forage and concentrate substrates were dried at 558C for 48 h and ground through a
1-mm screen using a Wiley mill (Arthur H. Thomas Co., Philadelphia, PA). Substrates
were analyzed for CP (Association of Of®cial Analytical Chemists, 1990), ADF and NDF
109
S.G. Haddad, R.J. Grant / Animal Feed Science and Technology 86 (2000) 107±115
Table 1
Chemical composition (g per 100 g DM) of substrate ingredients
Ingredient
CPa
ADFb
NDFc
Ash
EEd
NFCe
Alfalfa silage
Corn silage
Corn grain
Soybean hulls
Soybean meal
19.0
8.5
10.0
12.1
47.6
35.1
25.0
3.8
50.0
4.9
45.1
46.0
9.0
67.0
10.6
10.8
3.9
2.1
5.1
6.9
3.8
3.0
4.3
2.1
1.5
21.3
34.6
74.6
13.7
33.4
a
Crude protein.
Acid detergent ®ber.
c
Neutral detergent ®ber.
d
Ether extract.
e
Non®ber carbohydrates (100ÿCPÿNDFÿAshÿEE).
b
(Van Soest et al., 1991), ether extract (Association of Of®cial Analytical Chemists, 1990),
and ash (Association of Of®cial Analytical Chemists, 1990). Chemical compositions of
the late-bud stage alfalfa silage, physiologically mature corn silage, soybean hulls, corn,
and soybean meal are listed in Table 1. The individual ingredients were combined in the
proportions shown in Table 2 to create substrates representing isonitrogenous diets
containing either 30, 35, 40, or 45% NFC with either alfalfa silage or corn silage as the
forage. To achieve these calculated NFC concentrations in the samples, soybean hulls
were substituted incrementally for the corn grain.
Table 2
Ingredient and chemical composition (g per 100 g DM) of substrates used for in vitro experiment
Item
Diet
Alfalfa
a
Corn
30
35
40
45
30
35
40
45
Ingredients
alfalfa silage
corn silage
corn grain
Soybean hulls
soybean meal
50.0
±
14.7
25.0
10.3
50.0
±
22.9
16.7
10.4
50.0
±
31.3
8.3
10.4
50.0
±
39.6
±
10.4
±
50.0
±
26.5
23.5
±
50.0
6.1
19.5
24.4
±
50.0
14.6
11.0
24.4
±
50.0
24.8
±
25.2
Compositionb
CP
ADF
NDF
EE
Ash
NFC
NFC-to-NDF
18.5
30.7
41.0
3.2
7.6
29.7
0.72
18.4
27.4
36.9
3.4
7.4
33.9
0.92
18.3
23.5
32.0
3.6
7.2
39.2
1.23
18.4
19.7
26.2
3.8
7.0
44.5
1.69
18.5
27.2
42.3
2.4
4.9
31.5
0.74
18.5
24.0
38.4
2.5
4.7
35.5
0.92
18.6
19.9
33.1
2.7
4.5
40.9
1.24
18.5
15.1
27.1
2.9
4.2
47.1
1.74
a
Percentage of non®ber carbohydrate contained in the substrate.
CP, crude protein; ADF, acid detergent ®ber; NDF, neutral detergent ®ber; EE, ether extract; NFC, non®ber
carbohydrate.
b
110
S.G. Haddad, R.J. Grant / Animal Feed Science and Technology 86 (2000) 107±115
A 300-mg sample of each of the eight substrates was weighed into 50-ml
polypropylene tubes for measurement of in vitro NDF digestion kinetics. The entire
experiment was replicated three times.
2.2. In vitro procedure
The in vitro procedure utilized was that described by Grant and Weidner (1992). The
buffer solution was that of Goering and Van Soest (1970) adjusted to pH 5.8 with 1 M
citric acid or formulated for pH 6.8 as described by Grant and Mertens (1992a).
Fermentation times (at 398C) were 0, 3, 6, 9, 12, 24, 30, 36, 48, 72, and 96 h. Tubes were
swirled gently at inoculation and at each remaining time thereafter.
Ash-free NDF was measured at each time (Van Soest et al., 1991). To ensure complete
starch solubilization with the starch-containing substrates, a heat-stable amylase
(ANKOM Tech. Corp., Fairport, NY) was used during sample re¯ux and at ®ltering.
The ruminal ¯uid inoculum was obtained from a steer-fed medium-quality alfalfa hay. At
collection, the mean pH of the ruminal ¯uid was 6.20 (SE0.11) for the three replicates
of the experiment. Although inoculum source can affect ®ber digestion kinetics measured
in vitro, donor animals were not available for every substrate studied, so alfalfa was fed to
the donor steer because it has been shown to promote rapid NDF digestion rates (Van
Soest, 1994). Jung and Varel (1988) demonstrated that ruminal inoculum from cattle fed
alfalfa hay degrades ®ber fractions of forages more rapidly than inoculum from a wide
range of forage diets including bromegrass, switchgrass, or corn silage.
As tubes were removed from the water bath, the pH of the contents for each substrate
was measured to monitor stability of the buffer pH, which did not decrease more than
0.18 pH unit for any substrate during 96 h of fermentation. Therefore, actual pH for each
treatment remained suf®ciently separated for the results to be biologically meaningful.
2.3. Statistical analysis
The model for the kinetics of NDF digestion was that described by Mertens and Loften
(1980),
Y D0 eÿkd tÿL INDF;
where Y is the NDF residue at time t of in vitro fermentation, D0 the potentially digestible
NDF fraction (proportion of initial DM), kd the fractional rate constant of NDF digestion
(hÿ1), L the discrete lag time (h), t the time of fermentation (h), and INDF the indigestible
NDF residue at 96 h (proportion of initial DM).
The remaining potentially digestible fraction at each time was transformed
logarithmically, and the NDF digestion parameters were estimated by linear regression.
Discrete lag time was calculated using the equation
L
ln D0 ÿ ln D00
;
ÿkd
where D00 is intercept of the equation of ln (Yÿ(96-h) NDF residue) over time at t0.
S.G. Haddad, R.J. Grant / Animal Feed Science and Technology 86 (2000) 107±115
111
Potential extent of NDF digestion (PED) was determined using the formula
PED 100
D0
:
D0 INDF
Predicted apparent extent of ruminal NDF digestion (AED) was calculated using the
equation described by Miller and Muntifering (1985),
AED PEDeÿkp L
kd
;
kd kp
where kp is the fractional rate of ®ber particle passage from the rumen, which was set at
either 0.050/h for particles of NDF or at 0.030/h (Shaver et al., 1986) to simulate faster
and slower particle passage from the rumen as a result of high or low feed intake.
Parameter estimates of the NDF digestion model were analyzed using the general
linear models procedure of SAS (1985), with a factorial arrangement of forage (alfalfa or
corn silage), NFC concentration (30, 35, 40, or 45%), and buffer pH (5.8 or 6.8). The
model included factors for replicate, forage, NFC concentration, pH, and interactions.
Differences among treatment means for signi®cant main effects were detected using
Student±Newman±Keul's multiple range test (SAS, 1985). Signi®cance was declared at
p