Livestock Research for Rural Development 5 (2) 1993

Citation of this paper

Effect of the replacement of steam treated sugar cane bagasse by milo on ruminal fermentation in bovines and in vivo digestibility in sheep

Sergio Raposo de Medeiros and Paulo Fernando Machado*

*Associate Professor - ESALQ/USP - Piracicaba - SP - Brazil - Sponsored by IFS


Two experiments were conducted to determine the effects of substituting milo for steam-treated sugarcane bagasse (STB) on ruminal fermentation and in vivo digestibility. The diet with the greatest proportion of milo was also compared with another similar diet in which STB was replaced by raw sugarcane bagasse (RSB). The following basal diets were fed to 4 fistulated cows, 3 times a day (7 am, 1 pm & 7 pm): 26% RSB + 47% milo; 26% STB + 47% milo; 38% STB + 34% milo and 52% STB + 20% milo (dry matter basis). These diets represented the treatments RB26, SB26, SB38 and SB52, in the same order. They were completed with yeast (15%), molasses (6%), cottonseed meal (3%), sodium bicarbonate (0.8%) and minerals (1.7%). pH were not significantly different among treatments. The acetic acid level was significantly lower (P<0.05) in treatment RB26 than in treatments with STB. Although total volatile fatty acids production in treatment RB26 was numerically lower than in STB treatments, there was no significant difference (p=0.10) among them. Treatment RB26 resulted in highest figures concerning in situ insoluble matter degradation, significantly higher (p<0.05) than in treatments SB26 and SB38 for the 24 hour period of incubation and higher than SB26 for the 48 hour period. The same treatments were given to eight rams, on two latin squares, in order to determine diet digestibilities. The highest values for dry matter digestibility and organic matter digestibility were obtained in treatments that had greater proportions of milo. The ADF digestibility was higher (p<0.05) in treatment RB26 than in treatments SB26 and SB38. Dry matter intake was significantly lower in treatment RB26.

KEY WORDS: Ammonia nitrogen, bagasse, bovine, degradation, digestibility, in situ, milo, pH, ruminal fermentation, sheep, soybean hulls, steam treatment, sugar cane



The use of by-products in animal nutrition is a necessity since it may increase the availability of food for mankind as well as avoid accumulation that contributes to environmental problems. Sugarcane bagasse is the most abundant by-product in Brazil. Poor intake, due to low digestibility and low density, are considered the main reasons for unsatisfactory performance of animals fed this roughage. Steam treatment, known as auto-hydrolysis, objectives the cleavage of the bounds between lignin and the other components of the cell wall, in order to improve its degradability by the enzymes of the rumen microbial ecosystem. Better utilization of the fibrous portion would allow greater use of sugarcane bagasse. The objective of this work was to determine the effect of the substitution of steam treated sugarcane bagasse (STB) by increasing levels of milo on ruminal fermentation, in situ degradability of soybean hulls in bovine and in vivo digestibility in ovine. It was also compared the effects of the substitution of STB by raw sugarcane bagasse (RSB) in the diet with the greatest proportion of milo.


Material and methods

The four diets evaluated are presented in Table 1.


Table 1. Composition of diets (%) (Dry matter basis)


Ingredients RB 26 SB 26 SB 38 SB 52
STB - 26.3 38.9 52.5
RSB 26.3 - - -
MILO 47.0 47.0 34.0 20.0
UREA 0.2 0.2 0.6 1.0
MOLASSES 6.0 6.0 6.0 6.0
YEAST 15.0 15.0 15.0 15.0
COTTON SEED MEAL 3.0 3.0 3.0 3.0
SODIUM BICARBONATE 0.8 0.8 0.8 0.8
MINERALS 1.7 1.7 1.7 1.7

Steam treatment: 13 kgf/cm5 for 8 minutes

The diets were fed to 4 ruminal fistulated cows housed in individual pens; food and water intake were recorded. They were fed 2.1% of their liveweight in dry matter (DM) distributed in three meals at 7:00 am, 1:00 pm and 7:00 pm. At each of the first two meals it was fed 25% of the total and at the last meal it was fed the remaining 50%. Experimental periods lasted 28 days. The last 7 days were used for ruminal fluid sampling and in situ incubation of soybean hulls. Ruminal fluid was collected at 7:00 am, 9:00 am, 1:00 pm, 3:00 pm, 7:00 pm and 9:00 pm, for two days with a vacuum pump and a pipe in which end a DACRON bag was fastened. The pH was measured immediately using a pH meter. Subsequently, it was also determined N-NH3, according to Chaney and Marbach (1962), and VFA, by gas chromatography. In situ degradability (ISD) was determined according to Mehrez et al (1977). In all experimental periods, 5 batches of ISD were made for two periods of incubation: 24 and 48 hours.

The experimental period was repeated 4 times, in order to constitute a 4 x 4 latin square design. Ruminal fermentation measurements repeated on time were analyzed according to a split-plot model by Gill (1986). Data were analyzed using the ANOVA procedure of SAS (1986). ISD data was similarly analyzed, but the two periods of incubation were analyzed separately. Differences among treatments were determined by Tukey's test (p<0.05).

The same diets were fed to 8 rams to determine in vivo digestibility and N-balance. The diets were fed at 6:00 am and 6:00 pm. Maximum intake was previously determined and only 80% of this intake was fed to avoid orts. Experimental periods lasted 28 days. In the last week samples of food, orts, feces and urine were collected. The collections were made concurrently with the meals during 8 days. Samples of feed were taken daily at 10% of the total offered. Daily output of feces were measured and a 20% sub-sample was composed daily and frozen daily. Urine was collected in a bottle with 50 ml of 1:1 HCl solution. Urine was sampled 10%. The samples were kept in a freezer at minus 14 degrees Celsius. The following chemical analyses were determined: dry matter (DM), crude protein (CP), crude fibre (CF), ash (ASH) and acid detergent fibre (ADF). N content were determined in urine samples. ADF was analyzed according to Goering and Van Soest (1970). The remaining analyses were determined according to AOAC (1965).

All digestibility and N-balance data were analyzed using the GLM procedure, SAS (1986) for a replicated 4 x 4 latin square design. The two latin squares were grouped in a latin rectangle according to Mead and Curnow (1986). The GLM procedure was used because 2 rams died during the experiment. The first ram died due to an accident while attempting to escape its metabolic cage. The other one died of no apparent reason and the necropsy revealed an altered liver and a blocked intestinal flow. Since it happened just with this ram, it may not be associated with a suspected STB toxicity. Differences among treatments were determined by Tukey's test (p<0.05).


Results and discussion

As expected, the replacement of STB by milo resulted in varying levels of fiber (Table 2) in the diets. The isonitrogenous diets were accomplished by adding greater proportions of urea in the diets with greater proportions of STB (Table 1).

The lower dry matter intake of diet RB26 (Table 3) was mainly due to an extremely low intake of one of the animals as well as slightly lower intake of the other 3 animals in the same treatment. The great standard deviation (SD) can be attributed to it.

Average ruminal fluid pH for diet SB52 and SB38 were similar to values from works in which great proportions of STB were used along with some effective fiber (Costa 1987, Figueiredo 1990), but they were much higher than the figure obtained by Castro (1989). It may have been the result of an inclusion of 0.8% of sodium bicarbonate and the distribution of the rations in three meals instead of only two meals as performed in the work of Castro. The pH values for diet SB26 remained below 6.2 most of time (Table 3), that is considered the lowest limit to adequate activity of the celullolitic bacteria (Kauffmann et al 1980). Comparison among diets RB26 and SB26, tended to be significant (p=0.17). Similar results were reported by Castro (1989) confirming the remarkable effect of the replacement of RSB by STB on this variable. If it were basically the inclusion of STB the responsible for the decreasing pH values, the diets with more STB should present lower pH. However, pH increases as the proportion of STB increases in the diets. These results indicate that there is an association between the high levels of starch and unsatisfactory physical stimulation of the rumen by STB, leading to lower pH values. They also indicate that when greater proportions of STB were provided it must have stimulated salivation, if not by rumination itself, perhaps by longer mastication periods.

Table 2. Chemical composition of the diets (DM basis)



RB 26

SB 26

SB 38

SB 52

  Avg. sd Avg. sd Avg. sd Avg. sd
DM 51.47 2.18 48.72 0.84 45.38 1.03 42.76 0.83
CP 11.68 1.14 12.29 1.72 12.37 1.10 11.41 1.08
CF 18.99 1.34 14.64 1.01 19.08 0.37 24.29 0.84
ADF 28.44 2.58 25.34 1.37 32.50 1.54 38.05 2.15
ASH 5.33 0.76 5.15 0.16 5.25 0.21 5.33 0.10


The data from STB diets showed similar average ruminal NH3-N concentrations. Higher NH3-N levels corresponded to higher levels of urea present in diets with greater proportion of STB. The NH3-N values are placed among the lowest values from the recommended range for maximum digestion or maximum microbial protein production (Erdmam et al 1986) and hence it may be suspected that the concentrations were not enough for adequate microbial activity. Nevertheless, the wide range, from 0.35 to 29 mg/100ml, do not permit us to be conclusive. The difference between diets SB26 and SB52 tended to be significant (p=0.13). The great SD was also observed by Castro (1989), Silva (1990) and Figueiredo (1990).

Total VFA concentration in ruminal fluid was higher in STB diets. Total VFA concentration for diet SB26 was 29% greater than RB26, what could be a result of a greater supply of fermentable material that have been made available from the steam treatment. Diet RB26 had the lowest values for individuals VFA (acetic, C2, propionic, C3 and butyric, C4) but only C2 was statistically inferior than the value found for diet SB52. Although the C3 concentration for diet RB26 was about half that found for SB26, there was no significant difference because of the great SD (P=0.17). Greater C3 concentration for STB diet is the result of more soluble carbohydrate made available from the degradation of hemicellulose due to the steam treatment (Castro 1989).


Table 3: Average (Avg) and standard deviation (sd) from ruminal variables and intake (%PV)



RB 26

SB 26

SB 38

SB 52

  Avg. sd Avg. sd Avg. sd Avg. sd
PH 6.62 0.28 6.08 0.53 6.26 0.39 6.30 0.41
N-NH3(mg/dl) 4.34 1.90 3.57 1.87 4.76 3.22 6.35 4.53
C2 4.88 0.60b 5.34 0.69ab 5.36 0.92ab 5.71 0.89a
C3 1.82 0.51 3.51 1.53 2.68 0.70 2.72 0.62
Total VFA 7.61 1.31 9.99 1.68 9.25 1.63 9.55 1.59
% C2 64.80 5.96 54.09 6.31 58.08 3.03 59.97 2.69
% C3 23.75 4.03 34.01 10.76 28.91 5.33 28.38 3.66
C2/C3 2.85 0.79 1.82 0.77 2.10 0.51 2.16 0.36
Intake (%PV) 1.78 0.56 2.04 0.13 2.10 0.00 1.96 0.17


Different letters in the same row indicate significant difference by Tukey's test (p<0.05)

The lower C3 percentage of diet RB26 was probably the result of high pH values for this diet. The STB determines higher rates of fermentation, fast passage rates and low pH. These are unsuitable conditions for methanogenic bacteria, and as a result the production of methane decreases. To keep the fermentative balance there is a change in the pattern of fermentation that consists in the increase of C3 production. The C2:C3 ratio did not significantly differ among treatments despite the great numerical difference between diets SB26 and RB26. The diets with more fiber or more effective fiber, as diet RB26, had greater pH and so were more appropriate to the development of structural carbohydrate degrading (SC) bacteria, which main final product of fermentation is C2. Conversely, diets with less fiber, had lower pH, inadequate to SC bacteria, and as a result, the bacteria that ferments non-structural carbohydrates (NSC) predominated. The main final product of the NSC bacteria is C3. The results from ISD indicate the diets that had greater ruminal fiber degradation potential (Table 4). ISD was greater for diet RB26 than for diets SB26 and SB38 for 24 hour period of incubation (p<0.05). Diet RB26 was also greater than diet SB26 for the 48 hour period (p<0.05). Data analyses clearly show the importance of pH in the degradation of fiber. The pH values for diet SB26 remained most of the time below the lowest limit recommended for fiber degradation and consequently presented the poorest results regarding ISD. Conversely, diet RB26 had the highest values for pH and the best results for fiber degradation in the rumen. Among STB diets, the ISD increases as the proportion of this by-product increases, as for pH. Besides the lower pH, it also could be suspected that the high quantities of non-structural carbohydrate (NSC) represented by starch would have hindered degradation (Hoover 1986). Results from diet RB26 indicate that it is unlikely to be correct since this diet had the same amount of starch as diet SB26 and had the best ISD results. Although considering the increase in NSC determined by the steam treatment, the main effect caused by that is a faster and greater pH, decreasing due to a more intense fermentation. Similar results were found for diets SB38 and SB52 and since NH3-N was the only parameter that presented considerable difference, although not significant (4.76 and 6.35, respectively) it could be presumed that higher NH3-N concentrations would not have improved ISD. However, we have to consider that, despite the fact of similar fermentative activity, these two diets had different energy levels that can affect the microorganism requirements for NH3-N and the NH3-N concentration itself.

Table 4. Results from soybean hull in situ degradation



RB 26

SB 26

SB 38

SB 52

Incubation Avg. sd Avg. sd Avg. sd Avg. sd
24h 38.6 3.7a 26.8 3.0b 30.4 3.8b 33.1 1.0ab
48h 55.2 9.4a 38.1 3.1b 42.5 4.8ab 43.9 2.8ab


Different letters in the same row indicate significant difference by Tukey's test (p<0.05)

The digestibility coefficients and intake values are presented in Table 5. Intake from STB diets were all similar. Thus the comparison among them is consistent. However the RSB diet presented much lower intake (p<0.01), in most part due to the unusually low intake by one animal in the last experimental period (29%). Therefore it should be taken in account when comparing STB to RSB diets since intake has great effect on digestibility (Tyrrel and Moe 1975).

Values determined for digestibility are in accordance with values from other authors (Pate 1982, Conceicao et al 1986, Bem 1990). Dry matter and organic matter (OM) digestibility for diet RB26 were greater than diets SB38 and SB52 (p<0.05). For the same variables diet SB26 was superior to diet SB52 (p<0.05). Those differences were expected since the diets that had higher digestibility had greater proportions of concentrate. Inversely digestibility data concerning the fibrous portion from diets tended to increase as the proportions of roughage increased. As described by Woodford et al (1986) the increase of fibrous material in the diet decreases the digestibility of all components of the diet except the fiber, that tend to increase. The best digestibility values for CF and ADF were those from diet RB26, being in accordance with results from ISD and with the lower intake found for this diet.


Table 5: Treatments effect on digestibility coefficients and on intake of digestible dry matter



RB 26

SB 26

SB 38

SB 52

  Avg. sd Avg. sd Avg. sd Avg. sd
DM 73.1 8.3a 70.4 1.7ab 66.5 3.0bc 63.7 5.4c
MO 75.1 8.1a 72.2 1.9ab 68.3 2.6bc 65.4 5.1c
CP 69.8 5.0 64.4 7.16 66.0 5.2 65.5 5.3
CF 60.7 13.8 53.3 6.0 54.8 5.6 59.1 6.3
ADF 60.7 13.5a 49.2 4.4b 51.2 5.2b 51.1 8.7ab
INTAKE(g) 934 225b 1196 123a 1196 112a 1118 125a

Different letters in the same row indicate significant difference by Tukey's test (p<0.05)


The comparison between diets RB26 and SB26 shows that the fibrous portion of the bagasse was better utilized when no treatment was performed. Nevertheless, if the intake was similar this difference would certainly decrease.

N-balance data are in Table 6. There were no significant differences between diets. Due to problems with urine collection, the SD was very high, compromising the results.

Table 6. Effect of treatments on N-balance


  RB 26 SB 26 SB 38 SB 52
  Avg. sd Avg. sd Avg. sd Avg. sd
N Bal (g) 9.2 6.8 11.3 6.2 11.1 5.1 8.8 7.6




The STB replacement by increasing levels of milo worsens the ruminal conditions for fiber degradation, decreasing the digestibility of the fibrous portion of the diet, but increase the DM digestibility.

The comparison data among diets with STB and RSB indicates that, despite poorer ruminal fiber degradation conditions, STB diet was more digestible since, even though diet RB26 intake was much lower, there was no difference between their digestibilities.



AOAC 1965 Association of Official Analytical Chemists. Official methods of analysis. Tenth edition Washington D.C.

Bem C H W 1990 Efeito do bicarbonato de sˇdio e/ou lasalocida sobre digestibilidade de dietas com bagašo de cana-de-aš˙car (Mestrado - Escola Superior de Agricultura "Luiz de Queiroz"/USP)

Castro F B 1989 Avaliašao do processo de digestao do bagašo de cana-de-aš˙car (Saccharum sp L) auto-hidrolisado em bovinos. Piracicaba, 123pp (Mestrado - Escola Superior de Agricultura "Luiz de Queiroz"/USP)

Chaney A L and Marbach E P 1962 Modified reagents for determination of urea and ammonea. Journal of Clinical Chemistry 8:130-132

Conceicao M N, Lacorte M C F, Burgi R and Bose M L V 1986 Determinašao da digestibilidade do bagašo de cana-de-aš˙car auto-hidrolisado com carneiros. IN: REUNIAO DA SOCIEDADE BRASILEIRA DE ZOOTECNIA, Campo Grande pp119

Costa L R O 1987 Suplementašao de urÚia em dietas de bagašo de cana-de-aš˙car (Saccharum sp L) auto-hidrolisado para ruminantes. Piracicaba 102pp (Mestrado - Escola Superior de Agricultura "Luiz de Queiroz"/USP)

Figueiredo M P 1990 Cana-de-aš˙car e bagašo de cana tratado Ó pressao de vapor na dieta de ruminantes: efeitos sobre a fisiologia e microbiologia do r˙men. Piracicaba 103pp (Mestrado - Escola Superior de Agricultura "Luiz de Queiroz"/USP)

Gill J L 1986 Design and analysis of experiments - in the animal and medical sciences, Volume 2. The Iowa State University Press 304pp

Goering H K and Van Soest PJ 1970 Forage fiber analyses (Apparatus, reagents, procedures and some applications). Washington, D.C. Agricultural Research Service USDA 19pp. Agriculture Handbook, 379

Hoover W H 1986 Chemical factor involved in ruminal fiber digestion. Journal of Dairy Science 69:2755-2766

Kauffmann W, Hagmeister H and Dirksen G 1980 Adaptation to changes in dietary composition, level and frequency of feeding. IN: Ruckebush Y, Thivend P, Digestive physiology and metabolism in ruminants. Connecticut, Avi Publishing Company, Inc p. 587-602

Mead R and Curnow R N 1986 Statistical methods in Agriculture and biology. Chapman & Hall London 335pp

Mehrez A Z, Orskov E R and Mcdonald I 1977 Rates fermentation in relation to ammonia concentration. The British Journal of Nutrition 38(3):437-43

N.R.C 1984 Nutrient Requirements of Beef Cattle. Sixth Revised Edition Washington, D.C.

Pate F M 1982 Value of treating bagasse with steam under pressure for cattle feed. Tropical Agriculture, Trinidad 59(4):293-297

SAS 1986 Institute Inc. SAS/STAT USER'S GUIDE. ed. Cary, NC.

Silva S C 1990 Efeito do bicarbonato de sˇdio e ou lasalocida sobre os parÔmetros ruminais de bovinos alimentados com bagašo de cana-de-aš˙car tratados Ó pressao e vapor. Piracicaba 130pp (Mestrado - Escola Superior de Agricultura "Luiz de Queiroz"/USP)

Tyrrel H F and Moe P W 1975 Effect of intake on digestive efficiency. Journal of Dairy Science 58(8):1151-1163

Woodford J A, Jorgensen N A and Barrington G P 1986 Impact of dietary fiber and physical form on performance of lactating dairy cows. Journal of Dairy Science 69:1035-1047