Influence of Buffer pH and Raw Corn Starch Addition on In Vitro Fiber Digestion Kinetics R. J. GRANT' Bnd D. R. MERTENS Agricultural Research Service, USDA US Dairy Forage Research Center Madison, WI 53706 ABSTRACT
INTRODUCTION
The impact was studied of buffer pH (5.8, 6.2, and 6.8) on in vitro digestion kinetics of NDF from alfalfa hay, bromegrass hay, com silage, and alfalfa and bromegrass hays with raw com starch added to approximate a ration containing 30% NDF. Ash-free NDF was determined at 0,6, 12, 18,24,30,36,48, 72, and 96 h of fermentation. Kinetic parameters were estimated by nonlinear regression using an iteratively reweighted least squares technique. Addition of raw corn starch decreased fiber digestion rate for alfalfa hay and lag for bromegrass hay. Both rate and lag of NDF digestion of all substrates were affected negatively below pH 6.2. Predicted ruminal NDF digestibilities (as percentage of that at pH 6.8 treatment) declined below pH 6.2 for all forages; addition of starch decreased predicted ruminal NDF digestibility by 23% for both alfalfa and bromegrass hays, even at pH 6.8. Results suggest that low pH decreases fiber digestion rate and increases lag and that starch accentuates this effect for some substrates. (Key words: associative effects, artificial rumen, digestion rate, extent of digestion)
The energy requirements for milk production of high producing dairy cows typically are fulfilled with diets containing 50 to 60% concentrates. As concentrate or starch feeding increases, a reduction in fiber digestion can occur that may be associated with decreased ruminal pH. Kaufman (8) and Erdman (3) reported a direct relationship between ruminal pH and lower fiber content of diets when the proportion of concentrate in rations was varied. Dairy cows fed 24 kg/d of a diet containing 32% starch in the concentrate had an average ruminal pH of 6.1 (14). Ruminal pH was below 6.2 for 70 to 80% of a 24-h period in this study. The optimal pH for ruminal microbes ranges from 6.5 to 6.8 (18, 20). Cellulolytic bacteria are especially sensitive to low pH compared with amylolytic species (22). Thus, fiber digestion may be depressed at low pH because of negative effects of pH on cellulolytic bacteria. Mertens and Loften (12) noted the difference between in vitro conditions, in which pH of the system was maintained at 6.8, and in vivo conditions, in which fluctuations in pH below 6.8 often occur. They examined the effect of incremental starch addition on kinetics of forage fiber digestion in vitro at pH 6.8 and discovered that higher starch yielded linear increases in discrete lag prior to fiber digestion with no change in rate. However, their in vitro results could not explain in vivo observations during starch feeding (13), which showed a more dramatic depression in fiber digestion than could be accounted for by small changes in discrete lag that were observed by Mertens and Loften (12). Grant and Mertens (5) developed and described a buffer system that can be used in a batch in vitro system to maintain pH at 5.8 or 6.8 to examine the effects of different pH on fiber digestion. This system can be used to separate the effect of starch addition from pH changes on kinetics of fiber digestion.
Abbreviation key: ARC = alfalfa hay and raw com starch, AR = alfalfa hay, BRC = bromegrass hay and raw corn starch, BR = bromegrass hay, CS = com silage, INDF = indigestible NDF, Kd fractional rate constant of digestion, PED = potential extent of NDF digestion.
=
Received January 8. 1992. Accepted June 10. 1992. lDepartment of Animal Science. Nebraska, Lincoln 68583-0908. 1992 J Dairy Sci 75:2762-2768
University
of
2762
STARCH AND pH IMPACT ON FIBER DIGESTION
The objective of this experiment was to examine the effect of buffer pH and starch addition on fiber digestion kinetics to determine their influence on important aspects of fiber digestion. MATERIALS AND METHODS
Sample Preparation
Substrates were dried at 55·C for 48 h and ground through a I-mm screen using a Wiley mill (Arthur H. Thomas Co., Philadelphia, PA). Alfalfa hay (AH) contained (mean ± SD) 45 ± .6% NDF (ash-free, DM basis), bromegrass hay (8H) contained 67 ± .5% NDF, and com silage (CS) contained 39 ± .8% NDF. Raw com starch was prepared by grinding dry shelled com through a I-mm screen with a Wiley mill and collecting the particles that passed through a 300-llm sieve. The composition of this raw com starch was 90 ± .6% DM, 1.1 ± .2% NDF, 1.2 ± .4% CP, and .24 ± .05% ash. These compositional data were used, and raw com starch was mixed with AH (AHC) and BH (DHC) in proportions to simulate a diet of approximately 30% NDF. Substrate AHC contained 66% AH and 34% raw com starch; BHC contained 44% BH and 56% raw com starch. A 5OD-mg sample of each of the five substrates was weighed into 125-ml Erlenmeyer flasks for in vitro kinetic analyses. In Vitro Procedure
The in vitro procedure was that described by Goering and Van Soest (4) with the following modifications. Buffer solutions of pH 5.8 and 6.2 were obtained by adjustment with 1 M citric acid as described by Grant and Mertens (5). Fermentation times were 0, 6, 12, 18, 24, 30, 36, 48, 72, and 96 h. Neutral detergent fiber residues were measured using .5 g of sodium sulfite per sample and a heat-stable amylase at filtering (A-5426, Sigma Chemical Co., St. Louis, MO). The ruminal fluid inoculum was obtained from a nonlactating Jersey cow fed a medium quality alfalfa and grass hay diet for ad libitum intake. At collection, the pH of ruminal fluid was measured and averaged 6.31 ± .25 (mean ± SD) for both replicates of this experiment. While purged with C02, the inoculum was blended in a
2763
Waring blender (Waring Products Division, New Hartford, CT) for 60 s, strained through two layers of cheesecloth, and filtered through an .ll-mm pore size plastic screen. The pH of flask contents was measured at the end of each fermentation time for each substrate to monitor stability of buffer pH. By 96 h of fermentation, the maximum decline in pH 6.8 flasks was to 6.65, pH 6.2 flasks had declined to 6.15, and pH 5.8 flasks had declined to 5.66. The maximum decline in pH occurred with substrates containing starch. Statistical Analysis
The model for kinetics of fiber digestion was that described by Mertens (10, 11) and Mertens and Loften (12).
where Y Do Kd
L t INDF
= NDF residue at time t, = potentially digestible NDF (percentage of initial DM), = fractional rate constant of digestion (per hour), discrete lag (h), = time (h), and = = indigestible NDF (percentage of initial DM).
This model was fitted initially by subtraction of the 96-h residue from each observation and regression of the logarithm of this value versus time. Discrete lag (12) was determined using the equation
L = (1n Do - In D")/-Kd' where D" = intercept of the equation of In (Y - 96-h residue) over time at t O.
=
The parameter estimates from the natural log-linear procedure were used as starting parameter values for iteration; the nonlinear regression procedure of SAS (19) was used to estimate fiber digestion kinetics. The MarJournal of Dairy Science Vol. 75, No. 10. 1992
2764
GRANT AND MERTENS
TABLE 1. Effect of substrate and buffer pH on NDF digestion kinetics.! Digestion parameters Factor
1NDf"2
PED3
Lag
Rate
(h) Substrate 4 AH AHC BH BHC CS SE
(111)
4.59 b 3.84 b 6.51& 4.48 b 5.23&b
.094& .079b .058c .000c
25.1 b 19.ge 29.6& 16.7 d 12.5 e
.26
.005
1.7
44.1 e 33.6d 55.3 b 45.0" 67.2& 3.2
pH 5.8 6.2 6.8 SE
7.47 a 4.01 b 3.31 b .60
.049 b .073& .074& .003
20.9 20.9 20.5 .1
50.4 50.2 51.0 .1
(%)
.0490
a.b.c.d.eMeans within a factor and column combination differ (P < .05). !Each value is the mean of nine observations. 21ndigestible NDF (percentage of initial DM). 3Potential extent of NDF digestion (percentage of total NDF).
=
4AH Alfalfa hay, ARC = AR plus raw corn starch, BH = bromegrass hay, BHC = BH plus raw corn starch, CS corn silage.
quardt option of nonlinear regression in SAS (19) was chosen, and an iteratively reweighted least squares technique was used in conjunction with nonlinear regression (19). Potential extent of NDF digestion (PED), the percentage of total NDF that is potentially digestible, was determined using the formula PED
= 100
x Do/(Do + INDF).
Predicted ruminal NDF digestibility was calculated using the equation (11)
where RD
= predicted ruminal NDF digestibil-
Kp
= rate of passage, which was set at
ity and
.02/h for fiber particles. Parameters of the digestion model estimated by nonlinear regression were analyzed by the general linear models procedure of SAS (19) using a factorial arrangement of substrate and pH. The model included factors for replicate, Journal of Dairy Science Vol. 75, No. 10, 1992
=
sample, pH, and interactions. Differences among treatment means for significant main effects were detected using Student-NewmanKeuls procedure (19). Unless otherwise stated, statistical significance refers to P < .05. RESULTS AND DISCUSSION
Alfalfa hay had the fastest Kd (Table 1). Com starch addition to AH lowered Kd and had no effect on lag; in BH, raw corn starch addition had no effect on Kd and decreased lag. However, the effects of starch addition to AH or BH were more consistently expressed as a percentage of the treatment without starch. Addition of raw corn starch decreased Kd of AHC and BHC to 84% of AH and BH and also decreased PED of AHC to 76% of AH and PED of BHC to 81 % of BH. Lag of AHC was decreased to 84% of AH; for BHC, it was reduced to 69% of BH. These results suggest that starch addition to AH and BH to achieve a similar NDF concentration in the total substrate may have a proportional, rather than additive, effect on digestion kinetics of fiber for both forages. Indigestible NDF was reduced because starch diluted the NDF in the original sub-
2765
STARCH AND pH IMPACT ON FIBER DIGESTION TABLE 2. Effect of forage. starch, and pH on kinetics of NDF digestion.) Digestion parameters Forage Alfalfa hay
Starch _4
+ + +
Bromegrass hay
+ + +
Com silage
pH 5.8 6.2 6.8 5.8 6.2 6.8 SE 5.8 6.2 6.8 5.8 6.2 6.8 SE 5.8 6.2 6.8 SE
Lag
Rate
(h)
(Ih)
4.81 b 4.83 b 4.l2b 7.Q68 2.65 c 1.8OC
.062c .1I2a .1Q68 .066c .08()bc
.46 9.90'
.005
.Q908b
.05oa .057a .066a .032b .055a .062a .003 .033c
5.05 bc 4.58c 7.01 b 3.59' 3.59' .61 8.56 a 4.68 b 2.47C .77
.060" .045bc .004
INDf2
PED3 (%)
24.4a 25.4a 25.5 a 19.9 b 19.7 b 20.3 b .7 31.1 a 29.3 a 28.5 a l6.6 b 16.3 b 17.l b 1.8
45.9 43.6 43.0 33.8 34.0 33.1 .5
12.7ab 13.8a
67.0 64.0 71.4 .9
Il.Ib .3
53.6 56.5 57.3 44.7 44.0 43.1 1.4
a.b.cMeans within a column and forage combination differ (P < .05). lEach value is the mean of three observations. 21ndigestible NDF (percentage of initial DM). 3Potential extent of NDF digestion (percentage of total NDF). 4Without (-) and with (+) starch.
strate; however, this dilution cannot explain the total decrease in INDF when starch was added. If INDF was unaffected by com starch addition, and raw com starch contained no INDF, the INDF for AHC would be expected to be 66% of that for AH (or 16.6), and the INDF for BHC would be 44% of that for BH (or 13.0). Both estimated INDF were lower than that observed for AHC and BHC, which suggests that starch addition may increase the indigestibility of NDF in the in vitro system or contaminate the INDF that is truly indigestible. Averaged over all substrates, buffer pH 5.8 reduced Kd of NDF digestion compared with Kd at pH 6.2 and 6.8 (.049 vs..073 and .074/ h). Lowering pH to 5.8 lengthened lag prior to digestion (7.47 vs. 4.01 and 3.31/h). Buffer pH had no effect on INDF. The detrimental effect of low pH on Kd and L apparently is curvilinear and increases more rapidly as pH decreases below 6.2 for alfalfa and bromegrass. The changes in fiber digestion that occur from low pH may also result from
decreased microbial cellulolytic activity. Cellulolytic organisms were sensitive to pH below 6.2 (22). Because significant sample by pH (P < .074) and sample by starch (P < .06) interactions occurred for Kd and for lag, means for all treatment combinations are given in Table 2. For AH alone, a decline in pH from 6.2 to 5.8 reduced Kd with no effect on lag (Table 2). However, with AHC, a decrease in buffer pH from 6.2 to 5.8 increased lag and decreased Kd. Starch addition appeared to accentuate the negative effects of low pH on the lag component of digestion. In contrast with AH, a decline in buffer pH from 6.8 to 5.8 for BH alone resulted in a longer lag with no significant effect on Kd' although pH 5.8 reduced Kd by 24% relative to pH 6.8. With starch addition (BHC), a pH change from 6.2 to 5.8 led to decreased Kd and increased lag. As with AH, starch addition accentuated the negative effects of low pH on fiber digestion, but starch appeared to have a Journal of Dairy Science Vol. 75. No. 10, 1992
2766
GRANT AND MERTENS
TABLE 3. Illustration of potential impact of starch and pH differences on predicted ruminal NDF digestibilities. 1 RD2 Forage
RD
(+ or _)3
(%)
-
+ + +
5.8 6.2 6.8 5.8 6.2 6.8
31.5 33.6 33.1 22.5 25.8 26.1
94.6 100.8 100.0 86.2 98.7 100.0
94.6 100.8 100.0 67.6 77.4 78.4
+ + +
5.8 6.2 6.8 5.8 6.2 6.8
31.4 37.8 40.1 23.9 30.0 30.3
78.3 94.2 100.0 78.8 99.0 100.0
78.3 94.2 100.0 59.6 74.8 75.6
5.8 6.2 6.8
35.2 43.7 47.0
74.7 92.9 100.0
Starch
Alfalfa hay
Bromegrass hay
Corn silage
pH
(-)
(% of RD at pH 6.8) -
lCalculated using data from Table 2 and a rate of passage of .02Jh for fiber particles. 2Predicted ruminal NDF digestibility = PED x e-KpL x KcJ!(Kd + Kp)' 3Without H or with (+) starch.
greater impact on Kd with declining pH for BH than for AH. Unlike AH or BH, the fastest Kd occurred at pH 6.2 for CS (Table 2). This observation is unexplained but was consistent across in vitro replicates. Additional research is needed to determine whether this result is consistent for CS. As pH dropped from 6.8 to 5.8, lag increased significantly. Indigestible NDF was lowest for pH 6.8. As with AHC and BHC, Kd decreased, and lag increased, when pH was lowered from 6.2 to 5.8. Mertens and Loften (12) speculated that fiber digestion kinetics could be altered by changing 1) Kd' 2) lag, 3) PED, or 4) a combination of these three mechanisms. Their work indicated that the primary effect of starch addition when pH was controlled at 6.8 was to increase lag with little or no effect on Kd or on PED. These results agree with the hypothesis of El-Shazly et al. (2) that microorganisms preferentially use starch before digesting fiber. This hypothesis was confirmed by the work of Hiltner and Dehority (6) and agrees with the work of Russell and Baldwin (15, 16), who observed that some ruminal microbes have greater preferences and affinities for some carbohydrates (glucose, maltose, and sucrose) Journal of Dairy Science Vol. 75. No. 10. 1992
than for others (xylose and cellobiose). Some bacteria, such as Bacteroides succinogenes, utilize both cellulose and starch as substrates and may preferentially digest starch (1, 7). Mertens and Loften (12) concluded that small changes in digestion kinetics associated with starch addition at pH 6.8 could not explain the decreases in fiber digestion that occur in vivo when concentrates are fed. They postulated that the primary mechanism for in vivo depression of fiber digestion was due to the indirect effects of rapid starch fermentation on lowering of ruminal pH. The combined effect of kinetic changes from starch addition and lowered pH is difficult to ascertain without resorting to a model that integrates discrete lag and Kd with rate of passage. The predicted ruminal digestibilities of NDF based on in vitro digestion kinetics given in Table 2 and a rate of passage of .02lh for fiber were calculated using the formula discussed by Mertens (11). When the data are expressed as a proportion of the ruminal digestibility that is expected at pH 6.8 (with or without starch addition), an interesting pattern is evident (Table 3). With no starch addition, NDF digestion of AH apparently is depressed only slightly, even at pH 5.8. Addi-
STARCH AND pH IMPACT ON FIBER DIGESTION
tion of starch to AH appears to reduce NDF digestion by 75 to 78% at all pH. The NDF digestion of grasses (BH, BHC, and CS) was more severely depressed by low pH than alfalfa. However, the addition of starch to BH appeared to depress ruminal digestion of NDF at all pH by a proportion similar to that for AH (76 to 79%). The combined effect of adding starch and lowering pH to 5.8 reduced predicted ruminal NDF digestibility of AH and BH to 60 to 70% of that estimated at pH 6.8 without starch. This magnitude of reduction is closer to the 53% reduction observed in vivo by MacRae and Annstrong (9) than that reported by Mertens and Loften (12) when starch concentration, but not pH, was varied in vitro. When pH was low. and no starch was added, Kd of alfalfa was reduced, and lag for BH was increased. However, when starch was added, there was a consistent trend across all pH for Kd and lag to decrease, except for AH at pH 5.8, for which lag increased. This observation suggests that addition of 34% starch for AHC and 56% starch for BHC may stimulate early fermentation but depress Kd in agreement with the findings of Hiltner and Dehority (6). Furthermore, our results suggest that the effects of pH and starch addition may be somewhat independent and multiplicative. Russell and Dombrowski (17), Russell et aI. (18), and Strobel and Russell (21) have documented that low pH can have a detrimental effect on growth rate and competitive advantage of ruminal microbes. Stewart (20) observed that cellulolytic activity is reduced as pH drops below 6.8. Our results indicate that pH below 6.2 diminished Kd of NDF, although whether this effect was due to decreased microbial populations or enzymatic activity is not certain. Research is needed to confirm the differences in digestion kinetics between legumes and grasses when starch is added to the diet and when ruminal pH is decreased. Our results suggest that legumes are less sensitive to changes in pH than grasses, as measured by predicted ruminal NDF digestibility. Additional research is needed to determine the effects of pH, various starch sources, and processing methods on digestion kinetics of fiber from different sources. This information will be useful to model fiber digestion when grain feeding is practiced and to elucidate the
2767
factors that cause reduced fiber digestion in vivo under these circumstances. CONCLUSIONS
Using a buffer system that maintains a stable pH allows the partitioning of effects on fiber digestion kinetics of starch addition and of other factors, such as pH. As pH of the buffer dropped from 6.2 to 5.8, Kd decreased, and lag remained unchanged for AH. Conversely, for BH, as pH decreased, Kd remained relatively unchanged, but lag increased. For CS, lag increased as pH fell from 6.8 to 5.8, and Kd was greatest at pH 6.2. As pH declined from 6.2 to 5.8, addition of raw corn starch resulted in increased lag and decreased Kd for AH and BH. These changes in in vitro NDF digestion kinetics translated into predicted ruminal NDF digestibilities, which were 60 to 70% of NDF digestibility at pH 6.8 in the absence of starch. These predicted NDF digestibilities compare well with observed in vivo responses to high grain diets. REFERENCES
1 Akin, D. E. 1979. Microscopic evaluation of forage digestion by rumen microorganisms. A review. J. Anim. Sci. 48:701. 2 EI-Shazly, K.. B. A. Dehority, and R. R. Johnson. 1961. Effect of starch on the digestion of cellulose in vitro and in vivo by rumen microorganisms. J. Anim. Sci. 20:268. 3 Erdman, R. A. 1988. Dietary buffering requirements of the lactating dairy cow: a review. J. Dairy Sci. 71: 3246. 4 Goering, H. K., and P. J. Van Soest. 1970. Forage Fiber Analyses (Apparatus, Reagents, Procedures, and Some Applications). Agric. Handbook No. 379. ARSUSDA, Washington, DC. 5 Grant, R. J., and D. R. Mertens, 1992. Development of buffer systems for pH control and evaluation of pH effects upon fiber digestion in vitro. J. Dairy Sci. 75: 1581. 6 Hiltner, P., and B. A. Dehority. 1983. Effect of soluble carbohydrates on digestion of cellulose by pure cultures of rumen bacteria. Appl. Environ. Microbiol. 46:642. 7 Hungate, R. E. 1966. The Rumen and Its Microbes. Academic Press. Inc., New York, NY. 8 Kaufman, W. 1976. Influence of the composition of the ration and the feeding frequency on pH regulation in the rumen and on feed intake in ruminants. Livest. Prod. Sci. 3: 103. 9 MacRae, J. C.• and D. G. Armstrong. 1969. Studies on intestinal digestion in the sheep. 2. Digestion of some carbohydrate constituents in hay, cereal, and haycereal rations. Br. J. Nutr. 23:377. Journal of Dairy Science Vol. 75, No. 10, 1992
2768
GRANT AND MERTENS
10 Mertens, D. R. 1973. Application of theoretical mathematical models to cell wall and forage intake in ruminants. Ph.D. Diss., Cornell Univ., Ithaca, NY. 11 Mertens, D. R. 1977. Dietary fiber components: relationship to the rate and extent of ruminal digestion. Fed. Proc. 36:187. 12 Mertens, D. R., and 1. R. Loften. 1980. The effects of starch on forage fiber digestion kinetics in vitro. J. Dairy Sci. 63:1437. 13 0rskov, E. R., and C. Fraser. 1975. The effects of processing of barley-based supplements on rumen pH, rate of digestion, and voluntary intake of dried grass in sheep. Br. J. Nutr. 34:493. 14 Robinson, P. H., S. Tamminga, and A. M. Van Vuureno 1986. Influence of declining level of feed intake and varying the proportion of starch in the concentrate on rumen fermentation in dairy cows. Livest. Prod. Sci. 15:173. IS Russell, J. B., and R. L. Baldwin. 1978. Substrate preferences in rumen bacteria: evidence of catabolite regulatory mechanisms. Appl. Environ. Microbiol. 36: 319. 16 Russell, 1. B., and R. L. Baldwin. 1979. Comparison of substrate affinities among several rumen bacteria: a
Journal of Dairy Science Vol. 75, No. 10, 1992
possible detenninant of rumen bacterial competition. Appl. Environ. Microbiol. 37:531. 17 Russell, J. B., and D. B. Dombrowski. 1980. Effect of pH on the efficiency of growth by pure cultures of rumen bacteria in continuous culture. Appl. Environ. Microbiol. 39:604. 18 Russell, J. B., W. M. Sharp, and R. L. Baldwin. 1979. The effect of pH on maximum bacterial growth rate and its possible role as a detenninant of bacterial competition in the rumen. J. Anim. Sci. 46:251. 19 SAS~ User's Guide: Statistics, Version 5 Edition. 1984. SAS Inst., Inc., Cary, NC. 20 Stewart, C. S. 1977. Factors affecting the cellulolytic activity of rumen contents. Appl. Environ. Microbiol. 33:497. 21 Strobel, H. J., and J. B. Russell. 1986. Effect of pH and energy spilling on bacterial protein synthesis by carbohydrate-limited cultures of mixed rumen bacteria. J. Dairy Sci. 69:2941. 22 Therion, J. J., A. Kistnor, and 1. H. Kornelius. 1982. Effect of pH on growth rates of rumen amylolytic and lactilytic bacteria. Appl. Environ. Microbiol. 44:428.
INTRODUCTION
The impact was studied of buffer pH (5.8, 6.2, and 6.8) on in vitro digestion kinetics of NDF from alfalfa hay, bromegrass hay, com silage, and alfalfa and bromegrass hays with raw com starch added to approximate a ration containing 30% NDF. Ash-free NDF was determined at 0,6, 12, 18,24,30,36,48, 72, and 96 h of fermentation. Kinetic parameters were estimated by nonlinear regression using an iteratively reweighted least squares technique. Addition of raw corn starch decreased fiber digestion rate for alfalfa hay and lag for bromegrass hay. Both rate and lag of NDF digestion of all substrates were affected negatively below pH 6.2. Predicted ruminal NDF digestibilities (as percentage of that at pH 6.8 treatment) declined below pH 6.2 for all forages; addition of starch decreased predicted ruminal NDF digestibility by 23% for both alfalfa and bromegrass hays, even at pH 6.8. Results suggest that low pH decreases fiber digestion rate and increases lag and that starch accentuates this effect for some substrates. (Key words: associative effects, artificial rumen, digestion rate, extent of digestion)
The energy requirements for milk production of high producing dairy cows typically are fulfilled with diets containing 50 to 60% concentrates. As concentrate or starch feeding increases, a reduction in fiber digestion can occur that may be associated with decreased ruminal pH. Kaufman (8) and Erdman (3) reported a direct relationship between ruminal pH and lower fiber content of diets when the proportion of concentrate in rations was varied. Dairy cows fed 24 kg/d of a diet containing 32% starch in the concentrate had an average ruminal pH of 6.1 (14). Ruminal pH was below 6.2 for 70 to 80% of a 24-h period in this study. The optimal pH for ruminal microbes ranges from 6.5 to 6.8 (18, 20). Cellulolytic bacteria are especially sensitive to low pH compared with amylolytic species (22). Thus, fiber digestion may be depressed at low pH because of negative effects of pH on cellulolytic bacteria. Mertens and Loften (12) noted the difference between in vitro conditions, in which pH of the system was maintained at 6.8, and in vivo conditions, in which fluctuations in pH below 6.8 often occur. They examined the effect of incremental starch addition on kinetics of forage fiber digestion in vitro at pH 6.8 and discovered that higher starch yielded linear increases in discrete lag prior to fiber digestion with no change in rate. However, their in vitro results could not explain in vivo observations during starch feeding (13), which showed a more dramatic depression in fiber digestion than could be accounted for by small changes in discrete lag that were observed by Mertens and Loften (12). Grant and Mertens (5) developed and described a buffer system that can be used in a batch in vitro system to maintain pH at 5.8 or 6.8 to examine the effects of different pH on fiber digestion. This system can be used to separate the effect of starch addition from pH changes on kinetics of fiber digestion.
Abbreviation key: ARC = alfalfa hay and raw com starch, AR = alfalfa hay, BRC = bromegrass hay and raw corn starch, BR = bromegrass hay, CS = com silage, INDF = indigestible NDF, Kd fractional rate constant of digestion, PED = potential extent of NDF digestion.
=
Received January 8. 1992. Accepted June 10. 1992. lDepartment of Animal Science. Nebraska, Lincoln 68583-0908. 1992 J Dairy Sci 75:2762-2768
University
of
2762
STARCH AND pH IMPACT ON FIBER DIGESTION
The objective of this experiment was to examine the effect of buffer pH and starch addition on fiber digestion kinetics to determine their influence on important aspects of fiber digestion. MATERIALS AND METHODS
Sample Preparation
Substrates were dried at 55·C for 48 h and ground through a I-mm screen using a Wiley mill (Arthur H. Thomas Co., Philadelphia, PA). Alfalfa hay (AH) contained (mean ± SD) 45 ± .6% NDF (ash-free, DM basis), bromegrass hay (8H) contained 67 ± .5% NDF, and com silage (CS) contained 39 ± .8% NDF. Raw com starch was prepared by grinding dry shelled com through a I-mm screen with a Wiley mill and collecting the particles that passed through a 300-llm sieve. The composition of this raw com starch was 90 ± .6% DM, 1.1 ± .2% NDF, 1.2 ± .4% CP, and .24 ± .05% ash. These compositional data were used, and raw com starch was mixed with AH (AHC) and BH (DHC) in proportions to simulate a diet of approximately 30% NDF. Substrate AHC contained 66% AH and 34% raw com starch; BHC contained 44% BH and 56% raw com starch. A 5OD-mg sample of each of the five substrates was weighed into 125-ml Erlenmeyer flasks for in vitro kinetic analyses. In Vitro Procedure
The in vitro procedure was that described by Goering and Van Soest (4) with the following modifications. Buffer solutions of pH 5.8 and 6.2 were obtained by adjustment with 1 M citric acid as described by Grant and Mertens (5). Fermentation times were 0, 6, 12, 18, 24, 30, 36, 48, 72, and 96 h. Neutral detergent fiber residues were measured using .5 g of sodium sulfite per sample and a heat-stable amylase at filtering (A-5426, Sigma Chemical Co., St. Louis, MO). The ruminal fluid inoculum was obtained from a nonlactating Jersey cow fed a medium quality alfalfa and grass hay diet for ad libitum intake. At collection, the pH of ruminal fluid was measured and averaged 6.31 ± .25 (mean ± SD) for both replicates of this experiment. While purged with C02, the inoculum was blended in a
2763
Waring blender (Waring Products Division, New Hartford, CT) for 60 s, strained through two layers of cheesecloth, and filtered through an .ll-mm pore size plastic screen. The pH of flask contents was measured at the end of each fermentation time for each substrate to monitor stability of buffer pH. By 96 h of fermentation, the maximum decline in pH 6.8 flasks was to 6.65, pH 6.2 flasks had declined to 6.15, and pH 5.8 flasks had declined to 5.66. The maximum decline in pH occurred with substrates containing starch. Statistical Analysis
The model for kinetics of fiber digestion was that described by Mertens (10, 11) and Mertens and Loften (12).
where Y Do Kd
L t INDF
= NDF residue at time t, = potentially digestible NDF (percentage of initial DM), = fractional rate constant of digestion (per hour), discrete lag (h), = time (h), and = = indigestible NDF (percentage of initial DM).
This model was fitted initially by subtraction of the 96-h residue from each observation and regression of the logarithm of this value versus time. Discrete lag (12) was determined using the equation
L = (1n Do - In D")/-Kd' where D" = intercept of the equation of In (Y - 96-h residue) over time at t O.
=
The parameter estimates from the natural log-linear procedure were used as starting parameter values for iteration; the nonlinear regression procedure of SAS (19) was used to estimate fiber digestion kinetics. The MarJournal of Dairy Science Vol. 75, No. 10. 1992
2764
GRANT AND MERTENS
TABLE 1. Effect of substrate and buffer pH on NDF digestion kinetics.! Digestion parameters Factor
1NDf"2
PED3
Lag
Rate
(h) Substrate 4 AH AHC BH BHC CS SE
(111)
4.59 b 3.84 b 6.51& 4.48 b 5.23&b
.094& .079b .058c .000c
25.1 b 19.ge 29.6& 16.7 d 12.5 e
.26
.005
1.7
44.1 e 33.6d 55.3 b 45.0" 67.2& 3.2
pH 5.8 6.2 6.8 SE
7.47 a 4.01 b 3.31 b .60
.049 b .073& .074& .003
20.9 20.9 20.5 .1
50.4 50.2 51.0 .1
(%)
.0490
a.b.c.d.eMeans within a factor and column combination differ (P < .05). !Each value is the mean of nine observations. 21ndigestible NDF (percentage of initial DM). 3Potential extent of NDF digestion (percentage of total NDF).
=
4AH Alfalfa hay, ARC = AR plus raw corn starch, BH = bromegrass hay, BHC = BH plus raw corn starch, CS corn silage.
quardt option of nonlinear regression in SAS (19) was chosen, and an iteratively reweighted least squares technique was used in conjunction with nonlinear regression (19). Potential extent of NDF digestion (PED), the percentage of total NDF that is potentially digestible, was determined using the formula PED
= 100
x Do/(Do + INDF).
Predicted ruminal NDF digestibility was calculated using the equation (11)
where RD
= predicted ruminal NDF digestibil-
Kp
= rate of passage, which was set at
ity and
.02/h for fiber particles. Parameters of the digestion model estimated by nonlinear regression were analyzed by the general linear models procedure of SAS (19) using a factorial arrangement of substrate and pH. The model included factors for replicate, Journal of Dairy Science Vol. 75, No. 10, 1992
=
sample, pH, and interactions. Differences among treatment means for significant main effects were detected using Student-NewmanKeuls procedure (19). Unless otherwise stated, statistical significance refers to P < .05. RESULTS AND DISCUSSION
Alfalfa hay had the fastest Kd (Table 1). Com starch addition to AH lowered Kd and had no effect on lag; in BH, raw corn starch addition had no effect on Kd and decreased lag. However, the effects of starch addition to AH or BH were more consistently expressed as a percentage of the treatment without starch. Addition of raw corn starch decreased Kd of AHC and BHC to 84% of AH and BH and also decreased PED of AHC to 76% of AH and PED of BHC to 81 % of BH. Lag of AHC was decreased to 84% of AH; for BHC, it was reduced to 69% of BH. These results suggest that starch addition to AH and BH to achieve a similar NDF concentration in the total substrate may have a proportional, rather than additive, effect on digestion kinetics of fiber for both forages. Indigestible NDF was reduced because starch diluted the NDF in the original sub-
2765
STARCH AND pH IMPACT ON FIBER DIGESTION TABLE 2. Effect of forage. starch, and pH on kinetics of NDF digestion.) Digestion parameters Forage Alfalfa hay
Starch _4
+ + +
Bromegrass hay
+ + +
Com silage
pH 5.8 6.2 6.8 5.8 6.2 6.8 SE 5.8 6.2 6.8 5.8 6.2 6.8 SE 5.8 6.2 6.8 SE
Lag
Rate
(h)
(Ih)
4.81 b 4.83 b 4.l2b 7.Q68 2.65 c 1.8OC
.062c .1I2a .1Q68 .066c .08()bc
.46 9.90'
.005
.Q908b
.05oa .057a .066a .032b .055a .062a .003 .033c
5.05 bc 4.58c 7.01 b 3.59' 3.59' .61 8.56 a 4.68 b 2.47C .77
.060" .045bc .004
INDf2
PED3 (%)
24.4a 25.4a 25.5 a 19.9 b 19.7 b 20.3 b .7 31.1 a 29.3 a 28.5 a l6.6 b 16.3 b 17.l b 1.8
45.9 43.6 43.0 33.8 34.0 33.1 .5
12.7ab 13.8a
67.0 64.0 71.4 .9
Il.Ib .3
53.6 56.5 57.3 44.7 44.0 43.1 1.4
a.b.cMeans within a column and forage combination differ (P < .05). lEach value is the mean of three observations. 21ndigestible NDF (percentage of initial DM). 3Potential extent of NDF digestion (percentage of total NDF). 4Without (-) and with (+) starch.
strate; however, this dilution cannot explain the total decrease in INDF when starch was added. If INDF was unaffected by com starch addition, and raw com starch contained no INDF, the INDF for AHC would be expected to be 66% of that for AH (or 16.6), and the INDF for BHC would be 44% of that for BH (or 13.0). Both estimated INDF were lower than that observed for AHC and BHC, which suggests that starch addition may increase the indigestibility of NDF in the in vitro system or contaminate the INDF that is truly indigestible. Averaged over all substrates, buffer pH 5.8 reduced Kd of NDF digestion compared with Kd at pH 6.2 and 6.8 (.049 vs..073 and .074/ h). Lowering pH to 5.8 lengthened lag prior to digestion (7.47 vs. 4.01 and 3.31/h). Buffer pH had no effect on INDF. The detrimental effect of low pH on Kd and L apparently is curvilinear and increases more rapidly as pH decreases below 6.2 for alfalfa and bromegrass. The changes in fiber digestion that occur from low pH may also result from
decreased microbial cellulolytic activity. Cellulolytic organisms were sensitive to pH below 6.2 (22). Because significant sample by pH (P < .074) and sample by starch (P < .06) interactions occurred for Kd and for lag, means for all treatment combinations are given in Table 2. For AH alone, a decline in pH from 6.2 to 5.8 reduced Kd with no effect on lag (Table 2). However, with AHC, a decrease in buffer pH from 6.2 to 5.8 increased lag and decreased Kd. Starch addition appeared to accentuate the negative effects of low pH on the lag component of digestion. In contrast with AH, a decline in buffer pH from 6.8 to 5.8 for BH alone resulted in a longer lag with no significant effect on Kd' although pH 5.8 reduced Kd by 24% relative to pH 6.8. With starch addition (BHC), a pH change from 6.2 to 5.8 led to decreased Kd and increased lag. As with AH, starch addition accentuated the negative effects of low pH on fiber digestion, but starch appeared to have a Journal of Dairy Science Vol. 75. No. 10, 1992
2766
GRANT AND MERTENS
TABLE 3. Illustration of potential impact of starch and pH differences on predicted ruminal NDF digestibilities. 1 RD2 Forage
RD
(+ or _)3
(%)
-
+ + +
5.8 6.2 6.8 5.8 6.2 6.8
31.5 33.6 33.1 22.5 25.8 26.1
94.6 100.8 100.0 86.2 98.7 100.0
94.6 100.8 100.0 67.6 77.4 78.4
+ + +
5.8 6.2 6.8 5.8 6.2 6.8
31.4 37.8 40.1 23.9 30.0 30.3
78.3 94.2 100.0 78.8 99.0 100.0
78.3 94.2 100.0 59.6 74.8 75.6
5.8 6.2 6.8
35.2 43.7 47.0
74.7 92.9 100.0
Starch
Alfalfa hay
Bromegrass hay
Corn silage
pH
(-)
(% of RD at pH 6.8) -
lCalculated using data from Table 2 and a rate of passage of .02Jh for fiber particles. 2Predicted ruminal NDF digestibility = PED x e-KpL x KcJ!(Kd + Kp)' 3Without H or with (+) starch.
greater impact on Kd with declining pH for BH than for AH. Unlike AH or BH, the fastest Kd occurred at pH 6.2 for CS (Table 2). This observation is unexplained but was consistent across in vitro replicates. Additional research is needed to determine whether this result is consistent for CS. As pH dropped from 6.8 to 5.8, lag increased significantly. Indigestible NDF was lowest for pH 6.8. As with AHC and BHC, Kd decreased, and lag increased, when pH was lowered from 6.2 to 5.8. Mertens and Loften (12) speculated that fiber digestion kinetics could be altered by changing 1) Kd' 2) lag, 3) PED, or 4) a combination of these three mechanisms. Their work indicated that the primary effect of starch addition when pH was controlled at 6.8 was to increase lag with little or no effect on Kd or on PED. These results agree with the hypothesis of El-Shazly et al. (2) that microorganisms preferentially use starch before digesting fiber. This hypothesis was confirmed by the work of Hiltner and Dehority (6) and agrees with the work of Russell and Baldwin (15, 16), who observed that some ruminal microbes have greater preferences and affinities for some carbohydrates (glucose, maltose, and sucrose) Journal of Dairy Science Vol. 75. No. 10. 1992
than for others (xylose and cellobiose). Some bacteria, such as Bacteroides succinogenes, utilize both cellulose and starch as substrates and may preferentially digest starch (1, 7). Mertens and Loften (12) concluded that small changes in digestion kinetics associated with starch addition at pH 6.8 could not explain the decreases in fiber digestion that occur in vivo when concentrates are fed. They postulated that the primary mechanism for in vivo depression of fiber digestion was due to the indirect effects of rapid starch fermentation on lowering of ruminal pH. The combined effect of kinetic changes from starch addition and lowered pH is difficult to ascertain without resorting to a model that integrates discrete lag and Kd with rate of passage. The predicted ruminal digestibilities of NDF based on in vitro digestion kinetics given in Table 2 and a rate of passage of .02lh for fiber were calculated using the formula discussed by Mertens (11). When the data are expressed as a proportion of the ruminal digestibility that is expected at pH 6.8 (with or without starch addition), an interesting pattern is evident (Table 3). With no starch addition, NDF digestion of AH apparently is depressed only slightly, even at pH 5.8. Addi-
STARCH AND pH IMPACT ON FIBER DIGESTION
tion of starch to AH appears to reduce NDF digestion by 75 to 78% at all pH. The NDF digestion of grasses (BH, BHC, and CS) was more severely depressed by low pH than alfalfa. However, the addition of starch to BH appeared to depress ruminal digestion of NDF at all pH by a proportion similar to that for AH (76 to 79%). The combined effect of adding starch and lowering pH to 5.8 reduced predicted ruminal NDF digestibility of AH and BH to 60 to 70% of that estimated at pH 6.8 without starch. This magnitude of reduction is closer to the 53% reduction observed in vivo by MacRae and Annstrong (9) than that reported by Mertens and Loften (12) when starch concentration, but not pH, was varied in vitro. When pH was low. and no starch was added, Kd of alfalfa was reduced, and lag for BH was increased. However, when starch was added, there was a consistent trend across all pH for Kd and lag to decrease, except for AH at pH 5.8, for which lag increased. This observation suggests that addition of 34% starch for AHC and 56% starch for BHC may stimulate early fermentation but depress Kd in agreement with the findings of Hiltner and Dehority (6). Furthermore, our results suggest that the effects of pH and starch addition may be somewhat independent and multiplicative. Russell and Dombrowski (17), Russell et aI. (18), and Strobel and Russell (21) have documented that low pH can have a detrimental effect on growth rate and competitive advantage of ruminal microbes. Stewart (20) observed that cellulolytic activity is reduced as pH drops below 6.8. Our results indicate that pH below 6.2 diminished Kd of NDF, although whether this effect was due to decreased microbial populations or enzymatic activity is not certain. Research is needed to confirm the differences in digestion kinetics between legumes and grasses when starch is added to the diet and when ruminal pH is decreased. Our results suggest that legumes are less sensitive to changes in pH than grasses, as measured by predicted ruminal NDF digestibility. Additional research is needed to determine the effects of pH, various starch sources, and processing methods on digestion kinetics of fiber from different sources. This information will be useful to model fiber digestion when grain feeding is practiced and to elucidate the
2767
factors that cause reduced fiber digestion in vivo under these circumstances. CONCLUSIONS
Using a buffer system that maintains a stable pH allows the partitioning of effects on fiber digestion kinetics of starch addition and of other factors, such as pH. As pH of the buffer dropped from 6.2 to 5.8, Kd decreased, and lag remained unchanged for AH. Conversely, for BH, as pH decreased, Kd remained relatively unchanged, but lag increased. For CS, lag increased as pH fell from 6.8 to 5.8, and Kd was greatest at pH 6.2. As pH declined from 6.2 to 5.8, addition of raw corn starch resulted in increased lag and decreased Kd for AH and BH. These changes in in vitro NDF digestion kinetics translated into predicted ruminal NDF digestibilities, which were 60 to 70% of NDF digestibility at pH 6.8 in the absence of starch. These predicted NDF digestibilities compare well with observed in vivo responses to high grain diets. REFERENCES
1 Akin, D. E. 1979. Microscopic evaluation of forage digestion by rumen microorganisms. A review. J. Anim. Sci. 48:701. 2 EI-Shazly, K.. B. A. Dehority, and R. R. Johnson. 1961. Effect of starch on the digestion of cellulose in vitro and in vivo by rumen microorganisms. J. Anim. Sci. 20:268. 3 Erdman, R. A. 1988. Dietary buffering requirements of the lactating dairy cow: a review. J. Dairy Sci. 71: 3246. 4 Goering, H. K., and P. J. Van Soest. 1970. Forage Fiber Analyses (Apparatus, Reagents, Procedures, and Some Applications). Agric. Handbook No. 379. ARSUSDA, Washington, DC. 5 Grant, R. J., and D. R. Mertens, 1992. Development of buffer systems for pH control and evaluation of pH effects upon fiber digestion in vitro. J. Dairy Sci. 75: 1581. 6 Hiltner, P., and B. A. Dehority. 1983. Effect of soluble carbohydrates on digestion of cellulose by pure cultures of rumen bacteria. Appl. Environ. Microbiol. 46:642. 7 Hungate, R. E. 1966. The Rumen and Its Microbes. Academic Press. Inc., New York, NY. 8 Kaufman, W. 1976. Influence of the composition of the ration and the feeding frequency on pH regulation in the rumen and on feed intake in ruminants. Livest. Prod. Sci. 3: 103. 9 MacRae, J. C.• and D. G. Armstrong. 1969. Studies on intestinal digestion in the sheep. 2. Digestion of some carbohydrate constituents in hay, cereal, and haycereal rations. Br. J. Nutr. 23:377. Journal of Dairy Science Vol. 75, No. 10, 1992
2768
GRANT AND MERTENS
10 Mertens, D. R. 1973. Application of theoretical mathematical models to cell wall and forage intake in ruminants. Ph.D. Diss., Cornell Univ., Ithaca, NY. 11 Mertens, D. R. 1977. Dietary fiber components: relationship to the rate and extent of ruminal digestion. Fed. Proc. 36:187. 12 Mertens, D. R., and 1. R. Loften. 1980. The effects of starch on forage fiber digestion kinetics in vitro. J. Dairy Sci. 63:1437. 13 0rskov, E. R., and C. Fraser. 1975. The effects of processing of barley-based supplements on rumen pH, rate of digestion, and voluntary intake of dried grass in sheep. Br. J. Nutr. 34:493. 14 Robinson, P. H., S. Tamminga, and A. M. Van Vuureno 1986. Influence of declining level of feed intake and varying the proportion of starch in the concentrate on rumen fermentation in dairy cows. Livest. Prod. Sci. 15:173. IS Russell, J. B., and R. L. Baldwin. 1978. Substrate preferences in rumen bacteria: evidence of catabolite regulatory mechanisms. Appl. Environ. Microbiol. 36: 319. 16 Russell, 1. B., and R. L. Baldwin. 1979. Comparison of substrate affinities among several rumen bacteria: a
Journal of Dairy Science Vol. 75, No. 10, 1992
possible detenninant of rumen bacterial competition. Appl. Environ. Microbiol. 37:531. 17 Russell, J. B., and D. B. Dombrowski. 1980. Effect of pH on the efficiency of growth by pure cultures of rumen bacteria in continuous culture. Appl. Environ. Microbiol. 39:604. 18 Russell, J. B., W. M. Sharp, and R. L. Baldwin. 1979. The effect of pH on maximum bacterial growth rate and its possible role as a detenninant of bacterial competition in the rumen. J. Anim. Sci. 46:251. 19 SAS~ User's Guide: Statistics, Version 5 Edition. 1984. SAS Inst., Inc., Cary, NC. 20 Stewart, C. S. 1977. Factors affecting the cellulolytic activity of rumen contents. Appl. Environ. Microbiol. 33:497. 21 Strobel, H. J., and J. B. Russell. 1986. Effect of pH and energy spilling on bacterial protein synthesis by carbohydrate-limited cultures of mixed rumen bacteria. J. Dairy Sci. 69:2941. 22 Therion, J. J., A. Kistnor, and 1. H. Kornelius. 1982. Effect of pH on growth rates of rumen amylolytic and lactilytic bacteria. Appl. Environ. Microbiol. 44:428.
