Use of Cassava as Animal Feed

The effect of molasses on quality, feed intake and digestibility
by heifers of silage made from cassava tops 

Ngo Van Man and  Hans Wiktorsson* 

Department of Animal Nutrition, UAF, Thu Duc,
Ho Chi Minh City, Vietnam
Swedish University of Agricultural Sciences, Uppsala, Sweden


Two experiments were carried out to determine the effect on silage quality, and on feed intake and digestibility by growing heifers, of including molasses in cassava tops silage. Four levels of sugar cane molasses: (0, 30, 60 and 90 kg per tonne of fresh cassava tops) and two storage periods (30 and 60 days) were compared.. The design was a 4*2 factorial completely randomized block design with 3 replicates. For the feed intake and digestibility study, the silage was made without or with 60 kg molasses/tonne, using sixteen plastic bags of 1m diameter. Six crossbred Holstein heifers, of 160 - 180 kg live weight, were randomly allocated in a 3 x 2 change-over design to three treatments: Guinea grass ad libitum, 70% of  the ad libitum grass intake with a supplement of non-molasses cassava silage ad libitum, and 70% of the ad libitum grass intake with a supplement of molasses cassava silage ad libitum.  

Based on the colour, smell and mould appearance, all the silages were considered to be acceptable, but with higher levels of spoilage with the highest level of molasses. The silage made with molasses had a lower pH but a similar concentration of lactic acid compared with  the silage without molasses.  Ensiling reduced the HCN and tannin contents of the cassava tops.

The voluntary feed intake per 100 kg live weight of the heifers was 2.59, 2.65 and 2.91 kg DM of Guinea grass, non-molasses cassava tops silage and the molasses cassava tops silage diet, respectively. The apparent digestibility of DM, OM, CP, NDF and ADF decreased in the silage supplemented diets. No significant difference in digestibility was found between the non-molasses and molasses silage diets. The digestibility coefficient (%) of DM, OM, CP, NDF, ADF in non-molasses cassava tops silage and molasses cassava tops silage was 49.4, 52.1, 45.81, 36.6, 27.7 and 49.7, 51.9, 47.55, 28.1, 19.5, respectively.  

It is concluded that cassava tops can be preserved successfully by ensiling with or without molasses additive and that cassava tops silage is a good feed resource for cattle.

Keywords: cassava tops silage, feed intake,  digestibility,  crossbred heifers,  molasses.


Cassava or tapioca (Manihot esculenta, Crantz) is an annual crop grown widely in the tropical regions of Africa, Asia and Latin America. It thrives in sandy-loam soils with low organic matter, and climates characterized by low rainfall and high temperature (Wanapat et al 1997). In Vietnam cassava is cultivated on an area of 231,700 ha (FAOSTAT 1998). It is the third food crop after rice and maize and is mainly cultivated by small-holder farmers in the poorer areas. The main product of cassava is the roots, but considerable quantities of green leaves remain attached to the stem when the roots are harvested. Ravindran and Rajaguru (1988) reported the yield of cassava leaves to be as much as 4.6 tonnes/ha of dry matter taken as a by-product at root harvesting. High contents of crude protein in the cassava leaf have been reported, varying from 17 to 40% in the dry matter (Allen 1984).  FAO (1998) reported that with practices directed only toward harvesting of foliage, up to 6 tonnes of crude protein can be obtained per hectare per year. In Vietnam, cassava leaf residues have been evaluated by Liem et al (1998) in several experiments with monogastric animals. Focussing on the root production, many new high yielding varieties of cassava have been introduced, of which several have a high HCN content in the leaves (Kim 1999). HCN toxicity is considered to be a limiting factor in using a high level of cassava leaves in the diets of monogastric animals. However, ruminants can neutralize the harmful effects of HCN through the activities of rumen microbes and can therefore utilize the leaves more efficiently. In the FAO Tropical Feeds database (FAO 1998), it is stated that cassava leaf meal can be mixed in concentrates for lactating cows at up 35% without any harmful effects.  

Hay made from cassava leaf and stem was shown to be a good feed resource for ruminants, with a high voluntary feed intake (3.1% of LW) and dry matter digestibility (71%) (Wanapat et al 1997). However, to produce good quality hay from cassava tops requires good weather for drying and special care must be taken  to limit the loss of dry leaves. Ensiling could be a suitable way of preserving the leaves but silage additives should be added to ensure successful fermentation when the ensiled material, such as cassava tops, has a high content of nitrogen and low concentration of water soluble carbohydrates (Petersson 1988). Sugar cane molasses, a common feed ingredient in the tropics, is frequently used as an additive for ensiling tropical forages and improving silage quality.

The present study was therefore aimed at determining the influence of molasses in making silage from cassava tops and evaluating the feed intake and digestibility when the silage was fed to growing crossbred Holstein heifers.

Materials and methods 

Ensiling cassava tops
Silage making

Four levels of sugarcane molasses (0, 30, 60 and 90 kg per tonne of fresh cassava foliage), and two storage periods (2 and 4 months) were evaluated according to a 4*2 factorial randomized complete block design with 3 replicates. A total of 24 plastic bags with 10 kg each of fresh cassava foliage were prepared. Cassava tops were collected in the field immediately after root harvesting in January 1999. Only the tops (leaves, petioles and 40-60 cm of green stem) were taken. The molasses contained 640 g dry matter and 375 g sugars per kg. The tops were chopped into pieces that were 3 to 4 cm  in length. The chopped material was mixed with the molasses and placed in the plastic bags. The contents of the bags were compacted by hand, bound with a string and pressed by placing one sand bag (2 kg) on top of each bag.

Data collection and laboratory analysis

Samples were collected for chemical analysis on two occasions, two and four months after ensiling. The characteristics of the silages; colour, fungal contamination and smell were evaluated. Determinations were made of toluene dry matter (dried and corrected for volatiles according to Lingvall and Ericson 1981), pH (pH-ORION model 420 A), water soluble carbohydrates (WSC), tannins, HCN, N, ash, and EE using the procedures described by AOAC (1984). ADF and ash-free NDF were analyzed according to Van Soest and Robertson (1980). Acetic, butyric and lactic acids were determined by the HPLC technique (Shimadzu, Japan 3081-09202-20ATD-E).  

Feed intake and digestibility study
Silage preparation

Approximately 4000 kg of cassava top residues were collected in the field immediately after root harvesting in February, 1999. Sixteen plastic bags (1.0 m diameter), containing about 260 kg of chopped cassava tops each, were used to make the silages. The cassava tops were chopped into pieces 3 to 4 cm in length and placed in plastic bags in layers of 1.0 - 1.4 m deep. Molasses was mixed with the chopped material at the time of filling and the mixed materials were compacted by two people standing in the bags. After filling, the tops of the bags were bound by plastic string and pressed by placing five 8 kg sand bags on top. After 4 months of storage, the silage was used in the intake and digestibility experiment. 

Animals and experimental management

Six crossbred Holstein heifers, 8-10 months of age and 160-180 kg live weight, were randomly allocated in a 3*2 change-over design (Patterson and Lucas 1962) to three treatments:

Each period included 14 days for adaptation, 5 days for measurement of feed intake, 2 more days adaptation and 7 days for digestibility measurement. At  the beginning of the experiment a 5-day preliminary testing was done to measure the voluntary dry matter intake of the grass diet by each animal in order to determine the ratio of cassava top silage in the experimental diets (planned to be 70: 30 grass: silage on dry matter basis).  During the feed intake measurement the grass supply on the treatments 0MS and 6MS was restricted to 70% of the ad libitum intake (on a DM basis) and cassava silage was supplied ad libitum. For the digestibility determinations the total diet was limited to 85% of the mean DM intake measured during the 5 days of intake studies. During the 9 days of the digestibility study the daily amount of feed was maintained constant. 

The animals were confined in individual stalls in a covered shed open at the sides. They were kept for one month in this location prior to starting the experiment in order to adapt  them to the experimental conditions. They were treated against internal and external parasites.  During the experiment the animals were fed 4 times per day: at 8:30, 11:00, 16:00 and 20:00 h.  Fresh Guinea grass (Panicum maximum 280), cut at six weeks of re-growth was used as the basal feed. Cassava tops silage with or without the molasses additive was taken from the plastic bags once per day, weighed and put into a small plastic bag for feeding the whole day. A mineral supplement (containing salt, dicalcium phosphate, MgSO4, CuSO4, CoCl2, K2SO4, Casein Iodine, MnSO4 and Selenium) was fed at 84g/150-200kg live weight /day. Water was freely available.

Data collection and laboratory analysis

The animals were weighed prior to and after the 5-day feed intake period in the morning, before feeding and watering. The mean weight of the heifer was used in calculating the feed intake per kg live weight. Samples of feeds offered and refused were collected every day for laboratory analysis. During the collection period, refusals were collected at 8:00 h, weighed, mixed, sub-sampled and bulked in bags, one for each animal. During the digestibility study, feces from each animal were collected immediately after defecation throughout the day, and placed in weighed plastic basins until 8:30 h the following morning. The 24-hour fecal output was weighed, mixed and a sub-sample (10% of the daily output) from each individual heifer was stored in a freezer (-20 ºC). The seven samples from each animal during the collection week were de-frosted, mixed, sampled and dried in a forced oven at 60oC for 72 hours for laboratory analysis. Samples were prepared using procedures described by Goering and Van Soest (1970). Feed, refusals and feces samples after oven drying were ground using a laboratory hammer mill with 1mm screen. Dry matter, ash, crude protein, ether extract, ADF, ash-free NDF and permanganate lignin were analyzed using the same methods as described in the ensiling study. Gross energy contents of feed and fecal samples were determined by means of an adiabatic bomb calorimeter and digestible energy was calculated from these results.

Statistical analysis

The data were subjected to an analysis of variance (ANOVA) by using the General Linear Model (GLM) procedure of Minitab (1998). When the F test was significant (P<0.05), Tukey’s test for paired comparisons was used (Minitab 1998). 


Cassava tops ensiled with or without molasses additive
Physical quality of silage

Observations at 2 and 4 months after ensiling (Table 1) showed that the silage colour was pale green to brown yellow at 2 months after ensiling and changed to a yellow to brown colour after 4 months of storage. The molasses additive increased the degree of brown colour in the silage. Moulds were not seen 2 months after ensiling, but increased  with the amount of  molasses additive and the time of storage. The smell was good for all treatments. Based on the colour, smell and mould appearance, the silages were considered to be acceptable without or with the low level of molasses additive for the shorter storage period, and to be acceptable, but with a proportion spoiled after the long-term storage and with the high level of molasses additive. 

Table 1. Classification of cassava tops silage with different amounts of molasses additive and 2 or 4 months storage


Molasses level






Storage mae













































A +sp


A +sp

mae: Months after ensiling.

Gy: Greenish yellow; Yb: yellowish brown; By: browish yellow; Pg: Pale green;

Abs: absent; Ot+number: only on top with cm in thickness

A: Acceptable; A+sp: Acceptable with spoiled portions;


Chemical composition of cassava top silage

The dry matter of the fresh cassava was 25.8% and  only small changes were noted during ensiling on all treatments (Table 2). There were trends towards increased DM content at higher levels of molasses, and to reduced DM content with storage time. The crude protein content of  the cassava tops was around 21% of DM , and in the non-additive silages did not  change after ensiling. As was to be expected, incorporation of molasses (contains less than 4% crude protein in dry matter)  in the silage led to reduced crude protein content, especially at  the higher levels of molasses added. Storage time reduced the crude protein of  the silage somewhat, but the changes were non-significant. The NDF concentration was decreased 8% after ensiling for the cassava tops silage without added molasses. The NDF content was reduced with storage time. 

Table 2a. The effect of molasses and storage period on the quality of cassava top silage (g/kg DM except for DM content which is on fresh basis)









Fresh tops








Silage (molasses level)









































Storage time






2 months 








4 months

















Table 2b. The effect of molasses and storage period on the quality of cassava top silage  (mg/100g DM)


A. lactic

A. acetic




Fresh tops






Silage (molasses level)

































Storage time



2 months 






4 months












 pH , water soluble carbohydrates and fermentation products

The mean pH value of the non-molasses silages was 4.39. The molasses additive tended to decreased pH, while storage time significantly decreased the pH. The mean lactic acid content was around 1% of DM, and no effects of molasses additive level and storage period were found. Butyric acid was not detected in any of the silages. Water soluble carbohydrates increased in the silages made with molasses but were reduced by 90 % in the absence of molasses. 

HCN and tannin

The HCN content in the fresh cassava tops was 98 mg per 100g DM and was reduced by 68 %  after 2 months of ensiling. Increasing the storage period reduced the HCN level but the molasses additive had no obvious effect. The tannin content of the cassava tops was reduced after ensiling, but the decrease was small, and no significant change was found due to storage time. There was no apparent effect of the molasses on the silage tannin content. 

Feed intake and digestibility 

Intake of dry matter was highest for the treatment with 6% molasses added to the silage (Table 3) but organic matter digestibility organic matter digestibility (OMD) and crude protein digestibility (CPD) of the silage diets by around 5 and 10 percentage units, respectively, as compared to the grass diet only. There was no difference in OMD and CPD between the 0MS and 6%MS diets. The digestibility of the fiber fractions in the diets, expressed as NDF digestibility (NDFD) and ADF digestibility (ADFD), were also significantly lower in the cassava tops silage diets. The differences were 7 – 11 percentage units for NDFD and 11 – 14 percent units for ADF. No significant difference could be found between the 0MS and 6%MS diets. The energy digestibility (ED) of the silage supplement diets was also lower by around 5 percent units compared to the grass diet. 

Table 3.  Daily dry matter feed intake



Guinea grass,
sole diet

Silage supplemented diets,


0% Molasses

6% Molasses

Feed  (kg DM / animal)


5.29 a

3.37 b

3.34 b



0.00 a

2.06 b

2.39 b







ab Means within rows with different superscript letters are  different (p<0.05).

Replacing Guinea grass with ensiled cassava tops depressed digestibility of most fractions (Table 4). 

Table 4.  Digestibility (%) of the grass and silage supplement diets and the cassava top silages assuming constant digestibility of Guinea grass

Apparent digestibility

Guinea grass

Silage diets



Cassava top silage



0 % Mol

6% Mol

0% Mol

6% Mol

Dry matter

61.6 a

57.0 b

56.6 b


49.3 b

49.6 b


Organic matter


59.9 b

59.1 b


52.1 b

51.8 b


Crude protein

63.2 a a

53.0 b

53.1 b


45.8 b

47.5 b


Ether extract

















60.1 a

48.7 b

45.5 b


27.6 b

19.4 b



66.2 a

59.2 b

55.1 b


36.5 b

28.1 c










abc Means within rows with differing superscript letters are different (P<0.05)


Ensiling cassava tops 

With DM content between 25 and 35% and pH values below 4.5 the silage quality could be considered to be good  (Pettersson 1988). The molasses and storage reduced pH values. Butyric acid is another quality parameter, and silage is considered to be good when the concentration is below 0.1 g/kg fresh material (Lättemäe 1997). No butyric acid was detected in any of the silage samples. The presence of mould in silage is undesirable because it uses silage nutrients and toxins are sometimes produced . There were increasing problems with moulds and spoiled portions with increasing amount of molasses and longer storage period. 

The WSC in herbage is the main substrate for microbial growth. Therefore, concentration of WSC is reduced during fermentation. More variable results were obtained in residual WSC at increased levels of molasses, probably because part of the added sugars was lost in the effluent. It was expected that a higher dose of molasses should result in a higher residual WSC and improve silage quality, as reported  by  Lättemäe et al (1997),  but in our study  high levels of molasses resulted in more molasses in the run-off. Haigh and Parker (1985) suggested that a critical WSC concentration in herbage for successful preservation as silage without additives is 30 g/kg  DM. In legumes, Zelter (1960) suggested a higher WSC level of 120g/kg DM because of low dry matter content at harvest. In our study the WSC content of cassava tops was in the middle of the range reported by these two authors. 

Non-additive silage had nearly the same lactic acid concentration compared to the additive silage treatments, suggesting that WSC was not the sole substrate for lactic acid bacteria Starch, the main storage carbohydrate in leaf and stem, may be a substrate after the attack of enzymes in the initial ensiling process, although the majority of lactic acid bacteria do not attack starch  (McDonald et al 1991).  However, the lactic acid concentration, which was in the range of some common tropical herbage silages (Aminah et al 1999), was still low compared to the values for temperate legume silage (Lättemäe et al1997). 

The reduction in HCN content in cassava is due to the action of endogenous linamarase on glucosides following loss of cell integrity or tissue damage. In the ensiling process, chopping and slight wilting during the preparation before sampling, pressing and the initial environment of the aerobic phase resulted in good conditions for reducing the HCN content. When the pH in silage is lowered the enzyme activities are restricted, and the speed of HCN elimination reduced. In this experiment the HCN content of the cassava tops silage was reduced with storage time, but no effect was found of additive level. Similar  results were found by Du Thanh Hang (1998) and Bui Nhu Phuc et al (2000) for cassava leaf silage. 

The tannin content of the silage materials was in the range of common tropical high-protein leaves (Mahyuddin et al 1988; Ahn et al 1989),  and was reduced in the initial period of ensiling. This reduction may have been due to the formation of tannin-protein complexes. Maldonado et al (1995) reported that insoluble tannin and plant leaf protein complexes were established in the pH range 3.5-5.5. In ensiling sorghum with different tannin contents, Rodrigues et al (1998) reported that tannin concentration decreased with increase in the duration of fermentation. No such  reduction with the time of storage and level of molasses additive was found in the present study. 

Feeding value of cassava top silage

Supplementing cassava top silage to the grass diet tended to increase the feed intake, which could have been a result of a stimulatory effect of silage on intake (Aminah et al 1999),  or the effect of the protein in the cassava leaves when added to a low-protein roughage diet (Merkel et al 1999). In the present study, the supplement increased the crude protein intake by 45 to 63% compared with the grass diet. Molasses addition also improved silage quality, which would also result in a higher feed intake (10%) compared with non additive silage. Similar results have been reported by other authors (Pettersson 1988;  Lättemäe 1997).


Using cassava tops hay as the sole feed Wanapat et al (1997) found a DM digestibility of 71%, which is much higher than the value in the present study (50%). Treatment method may explain some of  the difference as water soluble components may be lost in the run-off. In one study Clancy et al (1977) reported that making alfalfa hay by drying can improve the digestibility by 7% compared with silage making. In the present study, digestibility of the cassava silage was determined by difference and the actual result for cassava may not be the same when used as a supplement compared with being the sole feed, as found by Madrid et al (1997).



Financial support from SIDA-SAREC is gratefully acknowledged. The authors would also like to thank Ms.Tran Thi Phuong Dung of the UAF Animal Nutrition Department for help in the analyses, and Mr. Nguyen Van Hiep for help with the field work.


Ahn J H, Robertson B M, Elliot R, Gutteridge R and  Ford C W 1989  Quality assessment of tropical browse legumes: Tannin content and protein degradation. Anim. Feed Sci. Technol. 27: 147-156.

Allen  R D 1984  Feedstuffs ingredient analysis table (USA), P:25-30.

Aminah A, Abu Bakar C and Izham  A 1999  Silages from Tropical Forages: Nutritional Quality and Milk Production. FAO Electronic Conference on Tropical Silage 1999. Rome.

AOAC  1984  Official methods of analysis. 12th Ed. Association of Official Analytical Chemists, Washington, D C 

Bui Huy Nhu,  Ogle B and Lindberg  J E  2000  Effect of replacing soybean protein with cassava leaf protein in cassava root meal based diets for growing pigs on digestibility and N retention. Anim. Feed Sci. Technol. 83: 223-235.

Clancy  M, Wangsness  P J and Baumgardt  B R  1977  Effect of conservation method on digestibility, nitrogen balance and intake of alfalfa.  Journal of Dairy Science. 1977, 60: 4, 572-579.

Duong Thanh Liem, Bui Xuan An, Ngo Van Man  Nguyen Can Hao  Duong Du Dong, Bui Hnu Ngo Phuc  and Nguyen  P  Loc  1998  Production of leaf meal from leguminous leaves. Agriculture and Food Industry, Journal of Science, Technology and Economic Management, ARDM, Hanoi. 4-98, 144-146.

Du Thanh Hang 1998  Digestibility and nitrogen retention in fattening pigs fed different levels of ensiled cassava root as energy source. Livestock Research for Rural Development. 10 (3):

FAO 1998 From the original book Tropical Feeds, by Bo Göhl, database by Andrew Speedy and Nick Waltham Version 8. Rome.

FAOSTAT 1998 Food and Agriculture. Organization Statistics, 1998. Rome.

Goering  J K  and Van Soest  P J  1970  Forage Fiber Analysis. Agricultural Hand Book No. 379. Agricultural Research Service,  United States Department of Agriculture, Washington DC. pp. 1-30.

Haigh  P M  and  Parker  J W G 1985  Effect of silage additives and wilting on silage fermentation, digestibility and intake, and on live weight change of young cattle. Grass and Forage Science, 40: 429-436. 

Kim  H   1999  Cassava cultivars selection results: Cassava variety KM98-1. IAS The 8th Cassava workshop proceedings, HCM City, March, 1999, 62-78.

Lingvall  P  and Ericson B  1981  Dry matter determination of silage. (Mimeo. Swedish Univ.  of Agric. Sci., Dept. of Anim. Nutr., Report 45. Uppsala.

Lättemäe  P 1997  Ensiling and Evaluation of Forage Crops. Dissertation, Swedish Univ. of Agri. Sci., Report 32. Uppsala.

Madrid  J,  Hermandez F,  Pulgar M A and Cid  J M  1997  In vivo digestibility of treated and untreated barley straw: results of direct and by-difference digestibility trials. Anim. Feed Sci. Technol.  65: 129-138.

Mahyuddin  P, Little  D A  and Lowry  J B 1988  Drying treatment drastically affects feed evaluation and feed quality with certain tropical forages species. Anim. Feed Sci. Technol. 22: 69-78.

Maldonado  R A P, Norton B W  and Kerven G L 1995  Factors affecting in vitro formation of tannin-protein complexes. J. Sci. Food Agric. 69: 291-298.

McDonald  R, R A  Edwards and J F D  Greenhalgh  1991  Animal Nutrition (4th edition). Longman  Scientific & Technical, John Wiley and Sons Inc., New York.

Merkel  R C, K R  Pond, J C  Burns and D S  Fisher  1999  Intake, digestibility and nitrogen utilization of three tropical tree legumes. II As protein supplement. Anim. Feed Sci. Technol.,  82:107-120.

MINITAB 1998 MINITAB release 12.21, Minitab Inc., 3081 Enterprise Drive, State College, PA 16801-3008, 814-238-3280, USA.

Patterson  H D  and Lucas  H L  1962  Change-over Designs. North Carolina Agricultural Experiment Station, USA. Tech. Bul. No. 147.

Pettersson  K 1988  Ensiling of forages. Factors affecting silage fermentation and quality. Dissertation, Swedish Univ. of Agri. Sci., Report 179. Uppsala.

Ravindran  V and  Rajaguru  A S B 1988  Effect of stem pruning on cassava root yield and leaf growth. Sri Lankan Journal of Agricultural Science. (25) 2:32-37.

Van Soest  P J  and Robertson  L B 1980 Systems of analysis for evaluating fibrous feeds. In: W J  Pigden, C C  Balch and M  Graham (Editors). Standardization of Analytical Methodology for Feeds: Proceeding of  a Workshop held in Ottawa, Canada, pp.49-60.

Wanapat  M , Pimpa  O,  Petlum A and  Boontao  U 1997  “Cassava hay” A new strategic feed for ruminants during the dry season. Livestock Research for Rural Development. (9) 2:

Zelter  S 1960  Fermentation behaviour of Lucerne ensiled by different methods. Proc. 8th Intern. Grassl. Congr., Reading, 505-510.

 Go to top