B. Harmison, M.L. Eastridge, and J.L. Firkins
Department of Animal Sciences
Five Jersey cows were used in a 5 x 5 Latin square design to determine the effects of decreasing dietary forage netural detergent fiber (NDF) and different sources of dietary starch on animal performance and nutrient digestibilities. The control diet was balanced for 21% forage NDF (FNDF) and 43% nonfiber carbohydrates (NFC). Four other diets were balanced for 35% NFC and either 16 or 11% FNDF, with either corn or corn and wheat as the sources of starch, arranged factorially. Dry matter (DM) intake decreased linearly as FNDF decreased; however, most of the decrease occurred when FNDF was reduced from 16 to 11%. Yields of milk and 4% fat-corrected milk (FCM) and percentages of milk fat and protein were similar among diets. Digestibility of NDF and acid detergent fiber (ADF) increased as FNDF decreased, and fiber digestibilities decreased with the addition of wheat to the diets. Starch digestibility was similar among diets. Source and amount of starch may be equally or more important than the level of FNDF in maintaining total tract digestibilities of nutrients. Forage NDF in diets for high producing cows can be reduced to 16% when starch sources and levels are adequately balanced.
Provision of adequate fiber is important for ruminants. Although much research on fiber has been done, the precise level of fiber needed in diets remains unclear, primarily because many variables must be considered. Although forages are the major source of fiber, their energy content is low, especially for forages of low quality. As genetic potential increases for milk production, providing sufficient energy and fiber to the dairy cow becomes more difficult.
Highly digestible by-products, such as soybean hulls, are being used to replace portions of forage fiber. The NRC (1989) recommends that 75% of the total dietary NDF be from forages; i.e., 28% NDF x 75% = 21% FNDF. However, Sarwar et al. (1992) replaced FNDF with NDF from soybean hulls and found that 60% NDF from forage was adequate to stimulate ruminal function and milk production when the total diet contained 31% NDF. Forage NDF as a percentage of total NDF, as compared with NDF as a percent of DM, might not adequately reflect the presence of effective fiber when by-products high in fiber are fed. Even when FNDF is used as an index for adequate fiber, particle size and species of forage must be evaluated.
When the forage portion of the diet is replaced with a more digestible source of carbohydrate, the concentration of starch (or NFC) must be considered. Rate of ruminal starch fermentation differs among cereal grains; wheat has a faster rate of fermentation than does corn (Herrera-Saldana et al., 1990; Oke and Loerch, 1991). Poore et al. (1993) proposed that the ratio of FNDF to ruminally degradable starch should be used in diet formulation. Therefore, the dietary level of forage fiber and starch and source of starch may interact to affect animal performance.
The objectives of this trial were to determine the minimum amount of FNDF (21, 16, or 11%) needed to maintain milk production and nutrient digestibilities and to determine the effect of source of starch (corn vs wheat) on ruminal fermentation products, milk constituents, and dairy cow performance.
Five Jersey cows (3 multiparous and 2 primiparous) averaging 55 days in milk were used in a 5 X 5 Latin square design. Periods were 28 days; the first 14 days were used for adaptation to the diets. Ingredient composition of the diets is listed in Table 1. The control diet was balanced at 21% FNDF and 43% NFC. The other four diets contained either 16 or 11% FNDF and corn or corn plus wheat in a 2 X 2 factorial arrangement. Total NDF (% DM) was 29.1% for the control diet and averaged 35.3% for the 16 and 11% FNDF diets (Table 2). Dietary NDF was lower in the control diet because soybean hulls were increased in the 16 and 11% FNDF diets to limit NFC to 35%. All diets were balanced to contain 18% crude protein (CP). Feed was mixed once daily as a total-mixed ration and fed to provide 10% feed refusals.
Daily milk production was averaged during weeks 3 and 4 for statistical analysis. Milk samples were taken (a.m. and p.m.) on days 14, 15, 21, and 22. Samples were analyzed for fat and protein concentrations by infrared spectroscopy (DHI Cooperative, Inc., Powell, OH). Body weights were measured twice weekly on consecutive days.
Chromic oxide was dosed orally (via gelatin capsules; 5 g of Cr2O3/dose) at 0700 and 1900 hours from days 15 to 27 for determination of apparent total tract digestibilities. Samples were analyzed for DM, organic matter (OM), N, starch, fatty acids, NDF, ADF, and acid detergent lignin. Samples of alfalfa and corn silages, feed offered, feed refusals, and feces were analyzed for neutral detergent insoluble N to correct NDF (NDFN) for protein contamination. Fecal samples were collected in 50-g proportions and mixed with 50 ml of distilled H2O (pH 7); the pH was measured with a portable pH meter (Orion Instruments, Boston, MA) on days 27 and 28. Ruminal samples were taken on day 28 of each period at 6 hours after feeding via stomach tube and analyzed for volatile fatty acids.
Statistical analyses were done with ANOVA using the general linear models and contrast procedures of SAS (1988). Contrasts were corn vs wheat as NFC sources in 16 and 11% FNDF diets, interaction between starch sources (corn vs corn plus wheat) and percentages of fiber (16 and 11% FNDF), and linear and quadratic effects for percentage of FNDF (21, 16, and 11% FNDF with corn). Significance was assumed at P < 0.05, and tendencies were noted at 0.05 > P < 0.10.
Actual FNDF in diets on a NDFN basis were 19.3, 14.6, 10.0, 14.5, and 10.0% for 21% FNDF with corn (control), 16% FNDF with corn, 11% FNDF with corn, 16% FNDF with corn and wheat, and 11% FNDF with corn and wheat, respectively (Table 2). Protein level in the control diet was slightly lower than expected because of the decrease in quality of the alfalfa silage during the trial. Alfalfa silage contained 15.4% CP and 42.7% NDFN, and corn silage contained 8.9% CP and 43.9% NDFN. The control diet had 27% NDF; about 72% of the NDF was from forage, which was similar to the 75% recommended by NRC (1989). Starch percentages were representative of the target levels (43% for control and 35% for the other diets). Calculated NFC averaged 17% higher than the analytical values for starch. The higher values for calculated NFC might be partly explained by the presence of fermentation acids in silages that were not determined by the analytical procedure for starch.
| Table 1. Ingredient composition of diets. | |||||
| Diet1 | |||||
| Ingredient | 21% FNDF
(corn) |
16% FNDF
(corn) |
11% FNDF
(corn) |
16% FNDF
(corn + wheat) |
11% FNDF
(corn + wheat) |
| (% of DM) | |||||
| Alfalfa silage | 22.18 | 16.77 | 11.50 | 16.75 | 11.49 |
| Corn silage | 22.31 | 16.82 | 11.50 | 16.79 | 11.49 |
| Corn, dry ground | 34.58 | 24.28 | 26.66 | 14.89 | 16.78 |
| Wheat, ground | . . . | . . . | . . . | 11.87 | 11.86 |
| Soybean hulls | 1.21 | 23.33 | 30.62 | 23.30 | 30.82 |
| Soybean meal, 44% CP | 12.73 | 10.10 | 11.29 | 4.80 | 6.23 |
| Corn gluten meal | 1.21 | 2.43 | 1.94 | 5.34 | 4.85 |
| Blood meal | 1.62 | 2.00 | 2.00 | 2.00 | 2.00 |
| Tallow | 2.00 | 2.00 | 2.00 | 2.00 | 2.00 |
| Mineral and vitamin supplements2 | 2.16 | 2.27 | 2.49 | 2.26 | 2.48 |
| 1 FNDF = Forage NDF in dietary DM.
2 Minerals and vitamins were provided to meet or exceed NRC (1989) requirements. | |||||
| Table 2. Nutrient composition of diets. | |||||
| Diet1 | |||||
| Item2 | 21% FNDF (corn) |
16% FNDF (corn) |
11% FNDF (corn) |
16% FNDF (corn + wheat) |
11% FNDF (corn + wheat) |
| DM, % | 67.1 | 66.6 | 66.1 | 66.9 | 66.9 |
| (% of DM) | |||||
| CP, % | 16.8 | 17.8 | 18.0 | 17.9 | 18.2 |
| NDF, % | 29.1 | 35.6 | 36.1 | 33.9 | 35.7 |
| NDFN,% | 26.7 | 32.4 | 33.0 | 30.5 | 32.6 |
| ADF, % | 18.0 | 23.1 | 23.5 | 21.1 | 22.3 |
| ADL, % | 2.89 | 2.72 | 2.32 | 2.65 | 2.40 |
| Ash, % | 5.73 | 5.98 | 5.84 | 5.75 | 5.67 |
| Starch, % | 43.3 | 34.2 | 33.1 | 34.3 | 32.9 |
| NFC, % | 47.7 | 39.3 | 39.2 | 41.7 | 39.5 |
| FA, % | 2.79 | 4.08 | 3.58 | 3.67 | 3.68 |
| NEL, Mcal/kg3 | 1.71 | 1.71 | 1.71 | 1.72 | 1.72 |
| Ca, % | 0.83 | 1.03 | 0.80 | 0.90 | 0.84 |
| P, % | 0.40 | 0.41 | 0.44 | 0.39 | 0.38 |
| Mg, % | 0.22 | 0.32 | 0.36 | 0.32 | 0.33 |
| K, % | 1.08 | 1.2 | 1.14 | 1.00 | 0.92 |
| 1 FNDF = Forage NDF in dietary DM.
2 ADL = Acid detergent lignin, NFC = nonfiber carbohydrates determined by calculation: 100 - (NDFN + CP + ash + (FA/0.90)), where NDFN = N-corrected NDF, and FA = fatty acids. 3 Calculated according to Weiss (1993). | |||||
Dry matter intake (DMI; pounds per day and as a percentage of BW) decreased linearly as the amount of FNDF decreased (Table 3). Daily DMI appeared to be more variable with the 11% FNDF diets. The standard deviations in daily DMI for days 1 to 28 among treatments were 1.47 lb/day for the control (21% FNDF), 1.78 lb/day for the 16% FNDF with corn, 2.24 pounds/day for the 11% FNDF with corn, 1.52 pounds/day for 16% FNDF with corn and wheat, and 2.05 pounds/day for 11% FNDF with corn and wheat. The high variability observed with the 11% FNDF diets may have been caused by a lack of effective fiber to stimulate rumination and saliva production or acid accumulation from rapid ruminal fermentation. Yields of milk and 4% FCM, and milk fat and protein percentages, were similar among diets.
Digestibilities of DM and OM tended to respond quadratically as FNDF decreased (Table 4). An interaction (P < 0.02) between source of NFC and percentage of fiber on DM and OM digestibilities was observed. Substitution of wheat for corn in the 11% FNDF diet reduced DM and OM digestibilities; substitution of wheat for corn in the 16% FNDF diet increased digestibilities of DM and OM. Similar to DM and OM digestibilities, an interaction between source of starch and level of FNDF occurred for ADF digestibilities. Digestibilities of NDF and ADF increased as FNDF decreased. The increases in fiber digestibilities were likely caused by the substitution of forage fiber with fiber from soybean hulls, which is highly digestible. When fed to cows, the digestibilities of NDF and ADF were lower in diets containing wheat than in those containing corn. The high rate of digestion of wheat compared with corn probably lowered ruminal pH, thus reducing fiber digestion. Wheat is digested faster than corn in vitro (Herrera-Saldana et al., 1990; Oke and Loerch, 1991), but data are limited to assess the effects of wheat addition on fiber digestibility in dairy cows.
Crude protein digestion increased as FNDF decreased. Digestibility of starch was similar among diets and averaged 94.5%. The NFC digestibilities were about 16% lower than the starch digestibilities. This difference could be caused by the inherent problems with calculation of NFC as a residual from other analyses and production of metabolic fecal NFC. Digestibility of NFC for the 16% FNDF and corn diet was lower than for control and the 11% corn diets. By substituting wheat for corn in the diets, NFC digestibilities increased (P < 0.03), primarily because of the low NFC digestibility for the 16% FNDF diet with corn. An interaction between source of starch and FNDF percentage was observed for NFC digestibility; the highest digestibility occurred with the 16% FNDF diet with wheat.
| Table 3. Production responses by cows fed diets differing in percentage of forage NDF (FNDF) and source of starch. | ||||||||||
| Diet | Contrast1 | |||||||||
| Item | 21%
FNDF (corn) |
16%
FNDF (corn) |
11%
FNDF (corn) |
16%
FNDF
(corn +wheat) |
11%
FNDF (corn + wheat) |
SE | L | Q | Corn vs wheat | INT |
| (P) | ||||||||||
| DMI, pounds/day | 38.9 | 38.5 | 34.8 | 38.3 | 35.0 | 1.3 | 0.04 | 0.25 |
0.97 |
0.86 |
| DMI, % of BW | 4.24 | 4.17 | 3.89 | 4.23 | 3.96 | 0.13 | 0.03 | 0.31 |
0.63 |
0.96 |
| BW, pounds | 926 | 928 | 891 | 917 | 917 | 9 | 0.05 | 0.15 |
0.39 |
0.08 |
| Milk, pounds/day | 46.9 | 46.0 | 45.3 | 44.9 | 45.3 | 1.8 | 0.52 | 0.97 |
0.76 |
0.74 |
| Milk fat, % | 4.66 | 4.75 | 4.56 | 4.67 | 4.59 | 0.19 | 0.79 | 0.68 |
0.88 |
0.76 |
| Milk protein, % | 3.93 | 3.96 | 3.80 | 3.93 | 3.92 | 0.07 | 0.25 | 0.28 |
0.55 |
0.30 |
| 4% FCM, pounds/day | 50.6 | 50.4 | 48.4 | 49.1 | 49.3 | 1.5 | 0.26 | 0.59 |
0.89 |
0.50 |
| 1 Contrasts: Linear (L) and quadratic (Q) effects for percentage of forage NDF in corn diets with 21, 16, and 11% FNDF, corn vs wheat in 16 and 11% FNDF diets, and interaction (INT) between percentage of FNDF and source of starch. | ||||||||||
| Table 4. Total tract apparent digestibilities (%), fecal pH, and concentrations (moles/100 moles) of ruminal acetate, propionate, and butyrate for cows fed diets differing in percentage of forage NDF (FNDF) and source of starch. | ||||||||||
| Item2 | Diet | Contrast1 | ||||||||
| 21% FNDF
(corn) |
16% FNDF
(corn) |
11% FNDF
(corn) |
16% FNDF
(corn + wheat) |
11% FNDF
(corn + wheat) |
SE | L | Q | Corn vs wheat | INT | |
| (P) | ||||||||||
| DM | 68 .5 | 66.9 | 69.2 | 69.8 | 66.4 | 1.1 | 0.53 | 0.09 | 0.98 | 0.02 |
| OM | 68.8 | 67.2 | 69.6 | 70.2 | 66.8 | 1.1 | 0.43 | 0.06 | 0.92 | 0.02 |
| NDFN | 50.6 | 52.0 | 58.0 | 51.6 | 50.2 | 2.2 | 0.02 | 0.23 | 0.08 | 0.11 |
| ADF | 53.2 | 53.9 | 59.1 | 54.2 | 48.0 | 1.8 | 0.01 | 0.11 | 0.01 | 0.01 |
| CP | 68.0 | 69.8 | 70.7 | 72.1 | 69.9 | 1.2 | 0.04 | 0.62 | 0.52 | 0.22 |
| Starch | 94.0 | 94.1 | 93.6 | 95.2 | 95.6 | 1.1 | 0.85 | 0.83 | 0.20 | 0.69 |
| NFC | 79.4 | 77.5 | 79.1 | 82.9 | 79.1 | 1.1 | 0.03 | 0.01 | 0.03 | 0.03 |
| FA | 59.8 | 76.1 | 68.6 | 67.2 | 67.6 | 4.7 | 0.30 | 0.14 | 0.31 | 0.41 |
| Fecal pH | 6.44 | 6.31 | 6.23 | 6.38 | 6.26 | 0.05 | 0.04 | 0.71 | 0.38 | 0.73 |
| Acetate | 63.3 | 63.3 | 61.4 | 63.0 | 61.4 | 0.7 | 0.21 | 0.40 | 0.86 | 0.81 |
| Propionate | 21.9 | 22.5 | 26.4 | 25.3 | 26.5 | 1.2 | 0.08 | 0.29 | 0.25 | 0.30 |
| Butyrate | 11.44 | 11.12 | 9.32 | 8.71 | 8.99 | 0.61 | 0.07 | 0.29 | 0.05 | 0.12 |
| 1
Contrasts: Linear (L) and quadratic (Q) effects for percentage of
forage NDF in corn diets with 21, 16, and 11% FNDF, corn vs wheat in 16
and 11% FNDF diets, and interaction (INT) between percentage of FNDF and
source of starch.
2 NDFN = NDF corrected for nitrogen contamination, NFC = nonfiber carbohydrates = 100 - (NDFN + CP + ash + (FA/.90)), and FA = fatty acids. | ||||||||||
Fecal pH decreased linearly as FNDF decreased (Table 4). These results suggest that starch digestion in the rumen might have been reduced and that fermentation was carried out in the hindgut. Ireland-Perry and Stallings (1993) observed decreased fecal pH as fecal starch content increased, and the increase in fecal starch content was attributed to lower dietary ADF. Molar proportion of acetate was similar among diets, and propionate tended to increase linearly as FNDF decreased (Table 4). Butyrate decreased as FNDF decreased and was lower with diets containing wheat than with those containing corn.
Using soybean hulls to replace FNDF can be cost effective when forage is expensive. In the current trial, the feed costs used were (expressed per ton; as-fed basis); $66 for alfalfa silage, $24 for corn silage, and concentrates at $159 for 21% FNDF, $143 for 16% FNDF with corn, $139 for 11% FNDF with corn, $150 for 16% FNDF with corn and wheat, and $144 for 11% FNDF with corn and wheat. Costs per cow for the daily ration were $2.60, $2.63, $2.35, $2.70, and $2.37 for 21% FNDF (control), 16% FNDF with corn, 11% FNDF with corn, 16% FNDF with corn and wheat, and 11% FNDF with corn and wheat, respectively. Income over feed costs was calculated based on actual DMI and milk production, feed costs, and milk price ($0.15/pound). Daily income over feed costs among treatments were $4.44, $4.28, $4.46, $4.04, and $4.43, respectively. Longer studies are needed to determine the economics of such feeding strategies because of probable interaction of diet with BW change, body condition, and ultimately milk production.
Herrera-Saldana, R. E., J. T. Huber, and M. H. Poore. 1990. Dry matter, crude protein, and starch degradability of five cereal grains. J. Dairy Sci. 73:2386.
Ireland-Perry, R. L., and Stallings, C. C. 1993. Fecal consistency as related to dietary composition in lactating cows. J. Dairy Sci. 76:1074.
National Research Council. 1989. Nutrient requirements of dairy cattle. 6th rev. ed. Natl. Acad. Sci., Washington, DC.
Oke, B. O., and S. C. Loerch. 1991. Effects of chemical treatments of cereal grains on in vitro dry matter disappearance and steer feedlot performance. Anim. Feed Sci. Technol. 34:163.
Poore, M. H., J. A. Moore, T. P. Eck, R. S. Swingle, and W. H. Brown. 1993. Response of lactating Holstein cows to diets varying in fiber source and ruminal starch degradability. J. Dairy Sci. 76:2235.
Sarwar, M., J. L. Firkins, and M. L. Eastridge. 1992. Effects of varying forage and concentrate carbohydrates on nutrient digestibilities and milk production by dairy cows. J. Dairy Sci. 75:1533.
SAS STAT® User's Guide. Release 6.03 Edition. 1988. SAS Inst., Cary, NC.
Weiss, W. P. 1993. Predicting energy values of feeds. J. Dairy Sci. 76:1802.