Response of Early Lactation Dairy Cows Fed Diets Varying in Source of Nonstructural Carbohydrate and Crude Protein1 DAVID P. CASPER, DAVID J. SCHINGOETHE, and WADE A. EISENBEISZ Dairy Science Department South Dakota State University Brookings 57007-0647 ABSTRACT
(Key words: nonstructural carbohydrates, degradable carbohydrates. degradable crude protein)
Seventy-six high producing Holstein cows were randomly assigned in a 2 x 2 factorial to evaluate two sources of nonstructural carbohydrates (com and barley), which supposedly differed in degradability of starch with two sources of CP degradability (soybean meal and urea) in the concentrate mix during wk 4 through 14 postpartum. Total mixed diets, formulated to be isonitrogenous at 16% CP, contained (DM) 40% com silage, 10% alfalfa hay, and 50% of the respective concentrate mix. Nonstructural carbohydrate degradability was similar for concentrate mixes containing corn or barley. Production of milk (32.2 and 31.8 kg/d) was similar, but production of 4% FCM (29.1 and 27.4 kg/d) and SCM (29.1 and 27.5 kg/d) was decreased for cows fed barley due to lower percentages of fat (3.39 and 3.22) and SNF (8.65 and 8.59). Percentages of protein (3.09 and 3.08) were similar for cows fed com and barley diets. Degradability of CP did not affect production of milk (31.9 and 32.0 kg/d), 4% FCM (28.5 and 28.0 kg/d), and SCM (28.4 and 28.2 kg/d) for cows fed soybean meal and urea. Dry matter intake was lower for cows fed barley (20.7 and 19.2 kg/d), but intakes (20.1 and 19.8 kg/ d) were similar for cows fed soybean meal and urea. Providing an alternative nonstructural carbohydrate source (barley versus com) did not increase utilization of a more readily degradable CP source (urea versus soybean meal).
INTRODUCTION
Received June 9, 1989 Accepted November 1, 1989. 1Published with the approval of the dierctor of the South Dakota Agricultural Experiment Station as Publication Number 2356 of the Journal Series.
1990 J Dairy Sci 73:1039-1050
Varying the source of nonstructural carbohydrates (NSC) and the degradability of those NSC in the diets of lactating dairy cows may maximize rumen microbial protein synthesis and the efficiency of utilizing ruminal undegradable protein. Several studies reported increased utilization of NH3 N for rumen microbial protein synthesis when diets contained greater amounts of NSC (5. 6, 17. 20, 24, 29). Huber and Kung (15) stated the major factor limiting utilization of NPN was a readily available energy source. Casper and Schingoethe (5, 6) demonstrated that including 30% dried whey as a source of readily fermentable NSC in the concentrate mix that contained NPN (urea) increased milk production to the production of cows fed diets containing soybean meal. B~ley is a major cereal grain fed to dairy cows 10 many areas of the world. Barley starch may be fermented more readily in the rumen than is corn starch (14, 18) and would be a more economical source of readily fermentable NSC to increase utilization of NPN for rumen microbial protein synthesis than is dried whey (5). Bacterial N entering the abomasum and ruminal starch digestibility were increased in steers (27) and lactating cows (18) fed barley in place of corn or sorghum. Herrera-Saldana et al. (14) reported a greater rate of starch degradation for barley than corn. Lactating cows fed rations containing a source of more rapidly degradable carbohydrates (barley versus milo) produced more milk (13, 14). In contrast, South Dakota (6) and Illinois (18) researchers observed lower milk production for cows fed barley than corn, and Bilodeau et al. (3) reported no differences in milk production for cows fed corn and barley. The objectives of this
1039
1040
CASPER ET AL.
study were to test further the hypothesis that diets high in degradable CP and degradable NSC support greater milk production than diets high in degradable CP but with less degradable carbohydrates. MATERIALS AND METHODS
experimental Plan
An ll-wk lactation trial utilized 76 high producing Holstein cows (52 multiparous and 24 primiparous) in a 2 x 2 factorial arrangement involving two NSC sources fed with two CP sources from wk 4 through 14 postpartum. Cows were blocked by parity and calving date and randomly assigned within blocks to one of four treatment diets at approximately wk 3 postpartum. Multiparous cows yielded greater than 27 kgld and primiparous at least 23 kgld milk by the 3rd wk postpartum. Nonstructural carbohydrate sources evaluated were corn and barley. Barley has a greater rate of NSC degradability than corn (14). Crude protein sources evaluated were: commercially available 44% CP soybean meal and urea. Composition of experimental concentrate mixtures is given in Table 1. Total mixed diets were fonnulated to meet or exceed CP. NEJo Ca, and P requirements (19) for a 635-kg mature cow producing 31.8 kg milk with 3.5% fat. Diets varied in solubility and degradability of NSC and N but were isonitrogenous at 16% CPo Diets consisted of 40% (DM basis) corn silage. 10% chopped alfalfa hay. and 50% of the respective concentrate mixture.
Cows were housed in a free stall bam and individually fed total mixed diets once daily for ad libitum intake using Calan feeding doors (American Calan, Inc., Northwood, NH) with amounts fed and refusals recorded daily. Cows were gradually switched from the basic herd concentrate mix of com, oats, and soybean meal to their respective experimental diets the last few days of the 3rd wk postpartum with the 4th wk postpartum as initiation of the ll-wk experimental period. Body weights were recorded 3 consecutive d at the start and conclusion of the experimental period and once each week during the experiment. Sample Collection and Analyses
Cows were milked twice daily with milk weights recorded at each milking. Two 24-h milk samples (p.m. plus am.) were collected from each cow during wk 3 postpartum (pretreatment) and one 24-h samples was taken each week throughout the trial. Samples were analyzed for fat by Milk-O-Tester and protein by Pro-Milk (MIGI, Foss Electric. Hillerod, Denmark) (1) and total solids by Mojonnier (2). Remaining milk in composite samples taken from cows at pretreatment, 3, 6, 9, and 11 wk of the trial were frozen until samples were analyzed for fatty acid distribution as described by Casper et al. (7). Concentrate mix, corn silage. and alfalfa hay were sampled weekly throughout the experiment. Four weekly samples were combined into monthly composites for analysis of DM, CPo ether extract, ash, Ca, and P according to the
TABLE 1. Ingredient content of concentrate mixes1 Concentrate mix Ingredient
C-SBM
C-U
B-SBM
B-U
- - - - - - - - - (%) - - - - - - - - -
59.9
Com. ground shelled
66.0
66.5 73.4 24.7 25.2 17.3 1.0 1.0 5.0 5.0 5.0 5.0 .9.9.7.9 1.9 1.9 2.1 1.9 .5.5.5.5
Barley, rolled Soybean meal, 44% CP Urea. 46% N Molasses. liquid Dicalcium phosphate Limestone Trace-mineral salt
31.8
Iplus 8.818 IU of viuunin A, 1,764 lU of vitamin D, and .9 lU of viuunin E added/kg of mix. C soybean meal, U = urea. B = barley. Journal of Dairy Science Vol. 73,
No.4. 1990
= Com, SBM =
CARBOHYDRATE AND CRUDE PROTEIN SOURCES
procedures of the Association of Official Analytical Chemists (1). Neutral detergent fiber, ADF, and permanganate lignin were determined by the procedures of Robertson and Yan Soest (22). Determination of N solubility and degradability of feedstuffs was estimated using the ficin protease procedure (21). Soluble residue was calculated by the formula of 100 - (CP + NDF + ether extract + ash) to provide an estimate of the amount of NSC (5, 6, 20). Ruminal contents were sampled from each cow three times during the experimental period approximately once every 4 wk. Samples were collected into a 250-ml bottle containing .5 m1 of saturated mercuric chloride 2 to 4 h after feeding by applying vacuum to an esophageal UJbe fitted with a suction strainer. Samples were analyzed for pH, YFA (7), ruminal NH3 (8), and ruminal soluble carbohydrates (6). Samples of jugular vein blood were drawn into heparinized UJbes at the time of rumen sampling and centrifuged for 20 min at 850 x g. The supernatant was decanted and frozen (-20°C) until analyzed for serum urea N (8). Statistical Analyses
Milk production and composition data were adjusted by analyses of covariance (28) using the milk production and composition during the 3rd wk postpartum as covariates. All data were subjected to least squares analysis of variance for factorial designs (28) by the general linear model procedure (SAS Institute, Cary, NC) with results expressed as least squares means. Whenever significant differences due to NSC source, protein source, time, lactation number (multiparous versus primiparous), and all possible interactions were detected, the Fisher's least significant difference (28) was used to separate least squares means. Nonstructural Carbohydrate Concentration and Degradablllty
Commercial samples (varieties unknown) of com and barley (barley-MW) of Midwest origin were obtained from the South Dakota State University Feed Processing Unit. A commercial barley samples of Northwest origin (barleyNW, variety steptoe) was obtained from Washington State University. In addition, the concentrate mixes of the lactation sUJdy were com-
1041
posited from the monthly samples for analyses. Samples were analyzed for DM, CP, ether extract, and ash (1). Neutral detergent fiber, ADF, and permanganate lignin were determined (22) as well as N solubility and degradability (21). NonstrucUJral carbohydrate concentrations were obtained using the soluble residue calculation (5, 6, 20), "Arizona" starch concentration (14), and by modification of the total nonstructural carbohydrates (TNC) (6) procedures. The TNC procedure was modified by adding 75 m1 of amylog1ucosidase (Diazyme L-200, Miles Laboratories, Elkhart, IN) to the enzyme solution and bringing to 1 L. Arizona and modified TNC procedures estimating NSC concentrations were replicated five times and compared for equality of means using a t test (28) by PROC TTEST (SAS Institute, Cary NC). Comparisons were on individual feeds and concentrate mixes. Because the soluble residue is a calculation from chemical composition and represents a single observation, it could not be included in statistical analysis but was evaluated numerically. Rates of NSC degradation were determined using an enzymatic in vitro system (R. Herrera and J. Huber; personal communication) modified as follows: 1 g of a ground feed sample (to pass a I-mm screen in an ultracentrifuge mill; Brinkmann Instruments Co., Westbury, NY) was incubated overnight at room temperature in a 25D-ml Erlenmeyer flask containing 100 ml of acetate buffer (.1 M, pH 5.0). Samples were then placed on an oscillatory shaker (120 oscillations/min) for 1 h and 1.0 ml of a glucoamylase enzyme (Diazyme L-200, Miles Laboratories, Inc., Elkhart, IN) was added to each flask. Duplicate l.o-ml aliquots were removed after the enzyme addition at 0, IS, 30, 45, and 60 min. The enzyme was inactivated in 1.D-ml aliquots by adding 1.0 ml of a 3% trichloracetic acid solution (Sigma Chemical Company, S1. Louis, MO). Glucose and galactose were then determined by the o-toluidine method (12). The rate of NSC degradation was estimated by measuring the rate of glucose appearance in solution over time. Initial time (t = 0) values were omitted to remove soluble NSC interference. Glucose appearance was regressed upon time to calculate rate of NSC degradation and the intercept was used to estimate initial NSC solubility for 10 replicates. The intercepts and slopes were then subjected to analysis of variance (28) Journal of Dairy Science Vol. 73.
No.4. 1990
1042
CASPER ET AL.
by the analysis of variance procedure (SAS Institute, Cary, NC). Whenever significant differences due to NSC source were detected, Duncan's multiple range test (28) was used to separate means. RESULTS AND DISCUSSION
Chemical Composition of Feeds
Chemical composition of the total diets (Table 3) are calculated from chemical compositions of ingredients given in Table 2. Oiets contained similar amounts of OM, CP, ether extract, lignin, ash, Ca, and P. Soluble and degradable N reflect the proportional consolidation of differences in Table 2. All diets contained less than recommended concentrations of AOF (19) due to the low AOF content of the com silage (Table 2) and to the new modifications of the ADF procedure (22). Concentration of NSe as calculated was higher for com than barley diets, respectively.
Chemical composition of concentrate mixes and forages is in Table 2. Mixes contained similar concentrations of OM and CPo Soluble N was increased within each carbohydrate source by addition of urea to the concentrate Concentration of Nonstructural Carbohydrates mix. The barley and soybean meal mixture Chemical composition of the individual contained more soluble N than the com and soybean mixture. Addition of urea to the barley feeds (Table 4) contained similar concentramix resulted in the highest soluble N in barley tions of nutrients as those listed by NRC (19) plus urea versus all other concentrate mixes. and Satter (25). Comparison of the Arizona and Mixes that contained barley had more degrad- modified TNe procedures for estimating NSe able protein than mixtures containing com. concentrations of feeds and concentrate mixes Why the barley plus soybean meal mixture is given in Table 5. Both procedures estimated contained more degradable CP than barley plus NSC concentrations similarly except for the urea is not known but may be related to ingre- barley plus urea mixture. The reason for differdient formulation (Table I). Several researchers ence between the two procedures for the barley (13, 14, 27) have reported that barley protein is and urea mixture is not known. An overall more degradable than com protein and may be comparison within ingredients and concentrate more degradable than soybean meal (14, 26). mixes indicates good agreement in estimating Differences in NOF, ADF, lignin, ether ex- NSC concentration with either procedure. The and P can be related to ingredi- significant difference in the overall comparison tract, ash, ents used in fonnulation of concentrate mixes for concentrate mixes is due to differences in (Table 1). Mixes containing barley were higher procedures for mixing barley and urea. We in ADF, NDF, and lignin and lower in ether recommend the Arizona procedure (14) as a extract than mixes containing com as would be method of choice because of shorter incubation expected because of barley's greater fiber con- times, fewer sample transfers, and more precitent (19). Mixes containing com have greater sion in the procedure as indicated by the lower concentrations of NSC, as measured by the SE compared with the modified TNC procesoluble residue calculation, than mixes contain- dure. A numerical comparison of the Arizona, ing barley, but within each NSC source, concentrations were higher for mixes containing modified TNC, and soluble residue methods urea because of greater carbohydrate inclusion indicated differences in the ability to predict the concentrations of NSC. For individual feeds, to replace the SBM (fable 2). Com silage and alfalfa hay contained OM, the soluble residue calculation indicated lower CP, soluble N, degradable N, ash, ether extract, concentrations of NSe than the Arizona or and NSC similar to values reported by Casper modified TNC procedures. In regard to concenand Schingoethe (6) and NRC (19) (Table 2). trate mixes, the comparisons were more variaConcentrations of NDF, ADF, and lignin were ble. The inclusion of amyloglucosidase to the lower than values reported by NRC (19), but procedures may be digesting some of the NOF the use of decahydronapthelene inflates AOF fractions of feeds, especially barley, by removvalues by 2 to 3% (22), which is approximately ing glucose from the nonreducing ends. (Table the difference between values in this study and 5). In contrast, the soluble residue calculation NRC (19). assumes all NSC in the various feeds and con-
ea.
JoW1lal of
Dail)'
Science Vol. 73, No.4, 1990
1043
CARBOHYDRATE AND CRVDE PROTEIN SOURCES
TABLE 2. Chemical composition of concentrate mixes and forages for cows fed diets containing corn (C) or barley (B) wilh soybean meal (SBM) or urea (V). Forages Concentrate mix Measurement' DM,% CP Soluble N,2 % of CP Degradable N,2 % of CP NDF ADF Lignin E1her ell.tract Soluble residue3
Ash Ca P
Com silage
Alfalfa hay
.19
41.8
86.8
1.1
.27 1.24 1.33 .33 .15 .06 .05 .51 .16 .08
.5
.02
8.1 49.1 68.7 42.3 24.5 3.2 2.5 42.3 4.7 .2 .2
17.2 45.8 69.5 4S.4 36.9 7.6 1.4 24.2 8.7 1.3 .2
B-V
89.7 ab
B-SBM 9O.l a
21.2 43.6b 73.7 c 12.l b 4.5 c .Sb 2.7 a 5S.6 a 5.5 b 1.3 .5
20.7 37.5c Sl.5 a 18.3a 7.3 a 1.2a 1.6b 52.5 b 6.7 a 1.2 .5
20.S 5l.5a 77.3b IS.3 a 7.oa l.3 a 1.6b 53.6b 5.S b
C-SBM 9O.2 a
C-V
21.0 33.6d 68.8 d 12.4b 5.0b .Sb 2.5 a 5S.4a 5.6b 1.3 .5
SE
89.3 b (% of OM)
a,b,c,dMeans among concentrate mixes wilh unlike superscripts differ (P
(Key words: nonstructural carbohydrates, degradable carbohydrates. degradable crude protein)
Seventy-six high producing Holstein cows were randomly assigned in a 2 x 2 factorial to evaluate two sources of nonstructural carbohydrates (com and barley), which supposedly differed in degradability of starch with two sources of CP degradability (soybean meal and urea) in the concentrate mix during wk 4 through 14 postpartum. Total mixed diets, formulated to be isonitrogenous at 16% CP, contained (DM) 40% com silage, 10% alfalfa hay, and 50% of the respective concentrate mix. Nonstructural carbohydrate degradability was similar for concentrate mixes containing corn or barley. Production of milk (32.2 and 31.8 kg/d) was similar, but production of 4% FCM (29.1 and 27.4 kg/d) and SCM (29.1 and 27.5 kg/d) was decreased for cows fed barley due to lower percentages of fat (3.39 and 3.22) and SNF (8.65 and 8.59). Percentages of protein (3.09 and 3.08) were similar for cows fed com and barley diets. Degradability of CP did not affect production of milk (31.9 and 32.0 kg/d), 4% FCM (28.5 and 28.0 kg/d), and SCM (28.4 and 28.2 kg/d) for cows fed soybean meal and urea. Dry matter intake was lower for cows fed barley (20.7 and 19.2 kg/d), but intakes (20.1 and 19.8 kg/ d) were similar for cows fed soybean meal and urea. Providing an alternative nonstructural carbohydrate source (barley versus com) did not increase utilization of a more readily degradable CP source (urea versus soybean meal).
INTRODUCTION
Received June 9, 1989 Accepted November 1, 1989. 1Published with the approval of the dierctor of the South Dakota Agricultural Experiment Station as Publication Number 2356 of the Journal Series.
1990 J Dairy Sci 73:1039-1050
Varying the source of nonstructural carbohydrates (NSC) and the degradability of those NSC in the diets of lactating dairy cows may maximize rumen microbial protein synthesis and the efficiency of utilizing ruminal undegradable protein. Several studies reported increased utilization of NH3 N for rumen microbial protein synthesis when diets contained greater amounts of NSC (5. 6, 17. 20, 24, 29). Huber and Kung (15) stated the major factor limiting utilization of NPN was a readily available energy source. Casper and Schingoethe (5, 6) demonstrated that including 30% dried whey as a source of readily fermentable NSC in the concentrate mix that contained NPN (urea) increased milk production to the production of cows fed diets containing soybean meal. B~ley is a major cereal grain fed to dairy cows 10 many areas of the world. Barley starch may be fermented more readily in the rumen than is corn starch (14, 18) and would be a more economical source of readily fermentable NSC to increase utilization of NPN for rumen microbial protein synthesis than is dried whey (5). Bacterial N entering the abomasum and ruminal starch digestibility were increased in steers (27) and lactating cows (18) fed barley in place of corn or sorghum. Herrera-Saldana et al. (14) reported a greater rate of starch degradation for barley than corn. Lactating cows fed rations containing a source of more rapidly degradable carbohydrates (barley versus milo) produced more milk (13, 14). In contrast, South Dakota (6) and Illinois (18) researchers observed lower milk production for cows fed barley than corn, and Bilodeau et al. (3) reported no differences in milk production for cows fed corn and barley. The objectives of this
1039
1040
CASPER ET AL.
study were to test further the hypothesis that diets high in degradable CP and degradable NSC support greater milk production than diets high in degradable CP but with less degradable carbohydrates. MATERIALS AND METHODS
experimental Plan
An ll-wk lactation trial utilized 76 high producing Holstein cows (52 multiparous and 24 primiparous) in a 2 x 2 factorial arrangement involving two NSC sources fed with two CP sources from wk 4 through 14 postpartum. Cows were blocked by parity and calving date and randomly assigned within blocks to one of four treatment diets at approximately wk 3 postpartum. Multiparous cows yielded greater than 27 kgld and primiparous at least 23 kgld milk by the 3rd wk postpartum. Nonstructural carbohydrate sources evaluated were corn and barley. Barley has a greater rate of NSC degradability than corn (14). Crude protein sources evaluated were: commercially available 44% CP soybean meal and urea. Composition of experimental concentrate mixtures is given in Table 1. Total mixed diets were fonnulated to meet or exceed CP. NEJo Ca, and P requirements (19) for a 635-kg mature cow producing 31.8 kg milk with 3.5% fat. Diets varied in solubility and degradability of NSC and N but were isonitrogenous at 16% CPo Diets consisted of 40% (DM basis) corn silage. 10% chopped alfalfa hay. and 50% of the respective concentrate mixture.
Cows were housed in a free stall bam and individually fed total mixed diets once daily for ad libitum intake using Calan feeding doors (American Calan, Inc., Northwood, NH) with amounts fed and refusals recorded daily. Cows were gradually switched from the basic herd concentrate mix of com, oats, and soybean meal to their respective experimental diets the last few days of the 3rd wk postpartum with the 4th wk postpartum as initiation of the ll-wk experimental period. Body weights were recorded 3 consecutive d at the start and conclusion of the experimental period and once each week during the experiment. Sample Collection and Analyses
Cows were milked twice daily with milk weights recorded at each milking. Two 24-h milk samples (p.m. plus am.) were collected from each cow during wk 3 postpartum (pretreatment) and one 24-h samples was taken each week throughout the trial. Samples were analyzed for fat by Milk-O-Tester and protein by Pro-Milk (MIGI, Foss Electric. Hillerod, Denmark) (1) and total solids by Mojonnier (2). Remaining milk in composite samples taken from cows at pretreatment, 3, 6, 9, and 11 wk of the trial were frozen until samples were analyzed for fatty acid distribution as described by Casper et al. (7). Concentrate mix, corn silage. and alfalfa hay were sampled weekly throughout the experiment. Four weekly samples were combined into monthly composites for analysis of DM, CPo ether extract, ash, Ca, and P according to the
TABLE 1. Ingredient content of concentrate mixes1 Concentrate mix Ingredient
C-SBM
C-U
B-SBM
B-U
- - - - - - - - - (%) - - - - - - - - -
59.9
Com. ground shelled
66.0
66.5 73.4 24.7 25.2 17.3 1.0 1.0 5.0 5.0 5.0 5.0 .9.9.7.9 1.9 1.9 2.1 1.9 .5.5.5.5
Barley, rolled Soybean meal, 44% CP Urea. 46% N Molasses. liquid Dicalcium phosphate Limestone Trace-mineral salt
31.8
Iplus 8.818 IU of viuunin A, 1,764 lU of vitamin D, and .9 lU of viuunin E added/kg of mix. C soybean meal, U = urea. B = barley. Journal of Dairy Science Vol. 73,
No.4. 1990
= Com, SBM =
CARBOHYDRATE AND CRUDE PROTEIN SOURCES
procedures of the Association of Official Analytical Chemists (1). Neutral detergent fiber, ADF, and permanganate lignin were determined by the procedures of Robertson and Yan Soest (22). Determination of N solubility and degradability of feedstuffs was estimated using the ficin protease procedure (21). Soluble residue was calculated by the formula of 100 - (CP + NDF + ether extract + ash) to provide an estimate of the amount of NSC (5, 6, 20). Ruminal contents were sampled from each cow three times during the experimental period approximately once every 4 wk. Samples were collected into a 250-ml bottle containing .5 m1 of saturated mercuric chloride 2 to 4 h after feeding by applying vacuum to an esophageal UJbe fitted with a suction strainer. Samples were analyzed for pH, YFA (7), ruminal NH3 (8), and ruminal soluble carbohydrates (6). Samples of jugular vein blood were drawn into heparinized UJbes at the time of rumen sampling and centrifuged for 20 min at 850 x g. The supernatant was decanted and frozen (-20°C) until analyzed for serum urea N (8). Statistical Analyses
Milk production and composition data were adjusted by analyses of covariance (28) using the milk production and composition during the 3rd wk postpartum as covariates. All data were subjected to least squares analysis of variance for factorial designs (28) by the general linear model procedure (SAS Institute, Cary, NC) with results expressed as least squares means. Whenever significant differences due to NSC source, protein source, time, lactation number (multiparous versus primiparous), and all possible interactions were detected, the Fisher's least significant difference (28) was used to separate least squares means. Nonstructural Carbohydrate Concentration and Degradablllty
Commercial samples (varieties unknown) of com and barley (barley-MW) of Midwest origin were obtained from the South Dakota State University Feed Processing Unit. A commercial barley samples of Northwest origin (barleyNW, variety steptoe) was obtained from Washington State University. In addition, the concentrate mixes of the lactation sUJdy were com-
1041
posited from the monthly samples for analyses. Samples were analyzed for DM, CP, ether extract, and ash (1). Neutral detergent fiber, ADF, and permanganate lignin were determined (22) as well as N solubility and degradability (21). NonstrucUJral carbohydrate concentrations were obtained using the soluble residue calculation (5, 6, 20), "Arizona" starch concentration (14), and by modification of the total nonstructural carbohydrates (TNC) (6) procedures. The TNC procedure was modified by adding 75 m1 of amylog1ucosidase (Diazyme L-200, Miles Laboratories, Elkhart, IN) to the enzyme solution and bringing to 1 L. Arizona and modified TNC procedures estimating NSC concentrations were replicated five times and compared for equality of means using a t test (28) by PROC TTEST (SAS Institute, Cary NC). Comparisons were on individual feeds and concentrate mixes. Because the soluble residue is a calculation from chemical composition and represents a single observation, it could not be included in statistical analysis but was evaluated numerically. Rates of NSC degradation were determined using an enzymatic in vitro system (R. Herrera and J. Huber; personal communication) modified as follows: 1 g of a ground feed sample (to pass a I-mm screen in an ultracentrifuge mill; Brinkmann Instruments Co., Westbury, NY) was incubated overnight at room temperature in a 25D-ml Erlenmeyer flask containing 100 ml of acetate buffer (.1 M, pH 5.0). Samples were then placed on an oscillatory shaker (120 oscillations/min) for 1 h and 1.0 ml of a glucoamylase enzyme (Diazyme L-200, Miles Laboratories, Inc., Elkhart, IN) was added to each flask. Duplicate l.o-ml aliquots were removed after the enzyme addition at 0, IS, 30, 45, and 60 min. The enzyme was inactivated in 1.D-ml aliquots by adding 1.0 ml of a 3% trichloracetic acid solution (Sigma Chemical Company, S1. Louis, MO). Glucose and galactose were then determined by the o-toluidine method (12). The rate of NSC degradation was estimated by measuring the rate of glucose appearance in solution over time. Initial time (t = 0) values were omitted to remove soluble NSC interference. Glucose appearance was regressed upon time to calculate rate of NSC degradation and the intercept was used to estimate initial NSC solubility for 10 replicates. The intercepts and slopes were then subjected to analysis of variance (28) Journal of Dairy Science Vol. 73.
No.4. 1990
1042
CASPER ET AL.
by the analysis of variance procedure (SAS Institute, Cary, NC). Whenever significant differences due to NSC source were detected, Duncan's multiple range test (28) was used to separate means. RESULTS AND DISCUSSION
Chemical Composition of Feeds
Chemical composition of the total diets (Table 3) are calculated from chemical compositions of ingredients given in Table 2. Oiets contained similar amounts of OM, CP, ether extract, lignin, ash, Ca, and P. Soluble and degradable N reflect the proportional consolidation of differences in Table 2. All diets contained less than recommended concentrations of AOF (19) due to the low AOF content of the com silage (Table 2) and to the new modifications of the ADF procedure (22). Concentration of NSe as calculated was higher for com than barley diets, respectively.
Chemical composition of concentrate mixes and forages is in Table 2. Mixes contained similar concentrations of OM and CPo Soluble N was increased within each carbohydrate source by addition of urea to the concentrate Concentration of Nonstructural Carbohydrates mix. The barley and soybean meal mixture Chemical composition of the individual contained more soluble N than the com and soybean mixture. Addition of urea to the barley feeds (Table 4) contained similar concentramix resulted in the highest soluble N in barley tions of nutrients as those listed by NRC (19) plus urea versus all other concentrate mixes. and Satter (25). Comparison of the Arizona and Mixes that contained barley had more degrad- modified TNe procedures for estimating NSe able protein than mixtures containing com. concentrations of feeds and concentrate mixes Why the barley plus soybean meal mixture is given in Table 5. Both procedures estimated contained more degradable CP than barley plus NSC concentrations similarly except for the urea is not known but may be related to ingre- barley plus urea mixture. The reason for differdient formulation (Table I). Several researchers ence between the two procedures for the barley (13, 14, 27) have reported that barley protein is and urea mixture is not known. An overall more degradable than com protein and may be comparison within ingredients and concentrate more degradable than soybean meal (14, 26). mixes indicates good agreement in estimating Differences in NOF, ADF, lignin, ether ex- NSC concentration with either procedure. The and P can be related to ingredi- significant difference in the overall comparison tract, ash, ents used in fonnulation of concentrate mixes for concentrate mixes is due to differences in (Table 1). Mixes containing barley were higher procedures for mixing barley and urea. We in ADF, NDF, and lignin and lower in ether recommend the Arizona procedure (14) as a extract than mixes containing com as would be method of choice because of shorter incubation expected because of barley's greater fiber con- times, fewer sample transfers, and more precitent (19). Mixes containing com have greater sion in the procedure as indicated by the lower concentrations of NSC, as measured by the SE compared with the modified TNC procesoluble residue calculation, than mixes contain- dure. A numerical comparison of the Arizona, ing barley, but within each NSC source, concentrations were higher for mixes containing modified TNC, and soluble residue methods urea because of greater carbohydrate inclusion indicated differences in the ability to predict the concentrations of NSC. For individual feeds, to replace the SBM (fable 2). Com silage and alfalfa hay contained OM, the soluble residue calculation indicated lower CP, soluble N, degradable N, ash, ether extract, concentrations of NSe than the Arizona or and NSC similar to values reported by Casper modified TNC procedures. In regard to concenand Schingoethe (6) and NRC (19) (Table 2). trate mixes, the comparisons were more variaConcentrations of NDF, ADF, and lignin were ble. The inclusion of amyloglucosidase to the lower than values reported by NRC (19), but procedures may be digesting some of the NOF the use of decahydronapthelene inflates AOF fractions of feeds, especially barley, by removvalues by 2 to 3% (22), which is approximately ing glucose from the nonreducing ends. (Table the difference between values in this study and 5). In contrast, the soluble residue calculation NRC (19). assumes all NSC in the various feeds and con-
ea.
JoW1lal of
Dail)'
Science Vol. 73, No.4, 1990
1043
CARBOHYDRATE AND CRVDE PROTEIN SOURCES
TABLE 2. Chemical composition of concentrate mixes and forages for cows fed diets containing corn (C) or barley (B) wilh soybean meal (SBM) or urea (V). Forages Concentrate mix Measurement' DM,% CP Soluble N,2 % of CP Degradable N,2 % of CP NDF ADF Lignin E1her ell.tract Soluble residue3
Ash Ca P
Com silage
Alfalfa hay
.19
41.8
86.8
1.1
.27 1.24 1.33 .33 .15 .06 .05 .51 .16 .08
.5
.02
8.1 49.1 68.7 42.3 24.5 3.2 2.5 42.3 4.7 .2 .2
17.2 45.8 69.5 4S.4 36.9 7.6 1.4 24.2 8.7 1.3 .2
B-V
89.7 ab
B-SBM 9O.l a
21.2 43.6b 73.7 c 12.l b 4.5 c .Sb 2.7 a 5S.6 a 5.5 b 1.3 .5
20.7 37.5c Sl.5 a 18.3a 7.3 a 1.2a 1.6b 52.5 b 6.7 a 1.2 .5
20.S 5l.5a 77.3b IS.3 a 7.oa l.3 a 1.6b 53.6b 5.S b
C-SBM 9O.2 a
C-V
21.0 33.6d 68.8 d 12.4b 5.0b .Sb 2.5 a 5S.4a 5.6b 1.3 .5
SE
89.3 b (% of OM)
a,b,c,dMeans among concentrate mixes wilh unlike superscripts differ (P
