Livestock Research for Rural Development 10 (3) 1998

Citation of this paper

Digestibility and nitrogen retention in fattening pigs fed different levels of ensiled cassava leaves as a protein source and ensiled cassava root as energy source

Du Thanh Hang

College of Agriculture and Forestry, Hue University, Vietnam


This study aimed to evaluate different additives for ensiling cassava leaves and the use of the product as a protein source for pigs. Cassava leaves were wilted and ensiled in plastic bags with 5 or 10% "A" molasses and 10 or 15 % rice bran (fresh basis).   After 7 days of ensiling the leaves the level of HCN had fallen from 177 mg/kg dry matter in the fresh leaf to 80 mg/kg in the ensiled material. There was a further slight reduction in HCN level up to 14 days but no change at 21 days. There were no differences attributable to kind or level of additive.

Digestibility and nitrogen balance trials in a double 4*4 Latin square arrangement were conducted to evaluate substitution of fish meal by ensiled cassava leaf (ECL; levels of 0, 50, 75 and 100 g/day of protein) in diets based on ensiled cassava root (ECR). It was observed that the pigs were reluctant to consume all the ECL, especially when this was at a high level in the diet. Thus the actual intakes of protein from the cassava leaf silage were less than the planned amounts being 36, 63 and 62 g/day compared with the planned quantities of 50, 75 and 100 g/day, respectively. Intakes of fish meal were 134, 92, 69 and 46 g/day and were only slightly less than the planned quantities of 150, 100, 75 and 50 g/day, respectively. The biggest difference between planned and recorded intake was for the intended 100 g/day of protein from ECL with the actual intake being only 62 g/day, thus total protein intake on the highest level of ECL was only 127 g/day compared with the intended level of 150 g/day.  There were no differences in total dry matter intake whereas it was expected that this would have increased with increasing levels of ECL in order to compensate for the lower protein content in this feed compared with fish meal.   There was an indication (P= 0.08) that apparent diet digestibility of dry matter decreased with increasing level of ECL (from 90.1 to 87.4% for 0 to 100 g/day protein substitution). The decrease in crude protein digestibility (from 86.6 to 79.6% from 0 to 100 g/day protein substitution) was highly significant (P=0.001). Nitrogen retention was 14.5, 13.8, 12.0 and 9.91 g/day for ECL0, ECL50, ECL75 and ECL100 diets, respectively. The differences between ECL0 and ECL50 compared with the higher levels of ECL were highly significant (P=0.001). Nitrogen retention values, as percent of N intake and N digested, were high on all diets (ranges were 57-58% and 66 to 61%, respectively) and did not differ among diets, indicating efficient utilization of the absorbed amino acids. There were no indications of cyanide toxicity on any of the diets.

Key words: Ensiled cassava leaves, ensiled cassava root, protein, digestibility, nitrogen retention, growing pigs


In central Vietnam, the main energy sources in pig production are cassava roots, cassava root by-products, rice bran and by-products of crop production. The protein content in these feeds is low and quite high levels of protein supplements are needed to satisfy nutrient requirements. However,  conventional sources of protein meals (fish meal, groundnut cake, soya bean meal) are expensive in Vietnam. It is therefore important to identify local sources of protein especially those that can be produced on the small farm.

There are many research papers about using cassava leaf as a protein source for animal production. Bui Huy Nhu Phuc et al (1996) evaluated levels of substitution of soya bean meal by dried cassava leaves (DCL) and found a depression in N digestibility and retention as the level of DCL increased.  Ensiled cassava leaves were compared with duckweed in diets of sugar cane juice fed to growing Mong Cai pigs (Du Thanh Hang et al 1997). Digestibility of the dry matter of the total diet was 86.0%  and that of  ECL per se was estimated to be 77%.  Satisfactory performance of growing pigs fed final ("C") molasses  supplemented with ECL was reported by  Bui van Chinh et al (1992).  According to Sarwat et al (1988) the inclusion of 15% fresh cassava leaves in the diet had no adverse effects on the performance of growing-finishing pigs. Alhassan and Odoi (1982) and Ravindran (1990) reported linear depression in weight gain and feed efficiency when cassava leaf meals were included at up to  30%  in diets for growing-finishing pigs.

In Vietnam, cassava (Manihot esculenta Crantz) is the major crop after rice.  In Hue province, with a total area of 5,009 km² and population density of 207 people/km², the cultivated area of cassava is 5,555 ha compared with 49,490 ha of rice and 823 ha of maize. The average productivity of cassava is estimated to be 3,230 kg of fresh roots/ha (Thua Thien Hue 1997).

According to  Ravindran and Rajaguru (1988) the yield of cassava leaves can be as much as 4.6 tonnes dry matter per ha taken as a by-product at root harvest.  In Vietnam, reported yields are lower:  2.5 to 3 tonnes cassava leaves containing 500-600 kg dry matter and 110 to 130 kg crude protein from 1 ha of cassava prior to root harvesting (Bui Van Chinh et al 1992). Crude protein in the dry matter of cassava leaves is reported to vary from 16.7 to 39.9 (Allen 1984) with an average of 21%.. Almost 85% of the crude protein fraction is true protein according to Eggum (1970). Rogers and Milder (1963) and Eggum (1970) reported that cassava leaf protein is deficient in methionine but high in lysine. Cassava leaves are a good source of minerals, particularly Ca, Mg, Fe, Mn and Zn (Ravindran  and Ravindran 1988). Cassava leaves are also rich in ascorbic acid and vitamin A, and contain significant amounts of riboflavin. But considerable losses of vitamins, particularly of ascorbic acid, occur during processing (Ravindran 1992).

The traditional processing of cassava leaves for feeding to pigs in the Central part of Vietnam is by drying and boiling.  Drying is often difficult due to rain at the time of harvesting and boiling takes time and fuel.  Ensiling the leaves appears to be a more appropriate way of conservation for conditions at the level of the small farm household. This study aimed to determine the digestibility and nitrogen retention in pigs when ensiled cassava leaves replaced up to 50% of  the fish meal protein in diets based on ensiled cassava roots. Two experiments were carried out. The first aimed to evaluate the effects of  molasses and rice bran as additives on some quality parameters of cassava leaf silage.  The second was a digestibility and nitrogen retention study with pigs fed ensiled cassava leaves to replace fish meal in a diet of ensiled cassava roots.

Experiment 1: Effect of molasses and rice bran as additives in ensiled cassava leaves

Material and methods

Ensiling procedure

Cassava leaves were collected from farmers' fields immediately prior to harvesting the roots. The leaves with petioles were separated from the stem  and chopped into small pieces (1-2cm) with a knife. They were then spread out in the shade with cross ventilation and wilted for one day.  The leaves were turned regularly to avoid fermentation and growth of moulds. After wilting, the leaves were mixed with 5 or 10% "A" molasses and 10 or 15% rice bran (fresh basis), put into plastic (polyethylene) bags (about 5 kg per bag) and pressed to exclude the air. The bags were then sealed. 

Chemical analysis

Samples of the freshly processed leaves were taken on the day of ensiling and after 7, 14 and 21 days for analysis of dry matter, crude protein, crude fibre, hydrocyanic acid (HCN) and pH (AOAC 1984).

Results and discussion

wpeB.gif (6047 bytes)
Figure 1: Effect of levels of molasses and of rice bran on HCN content of
ensiled cassava leaves after different periods of ensiling
Chemical composition of ensiled cassava leaves

The change of HCN concentration during the ensiling period is shown in Figure 1. For all the additives  there was a marked reduction in the HCN level after 7 days of ensiling with another slight fall after 14 days and then no further change at 21 days.  The slightly lower levels with rice bran can be explained by the greater diluting effect of the bran compared with molasses. Molasses  contains less dry matter and was used at  lower levels than the rice bran.  Similar values for HCN in fresh cassava leaves were reported by Nguyen Van Lai and Rodriguez (1998); however, in their experiment the level of HCN continued to fall beyond 21 days and up to 56 days of the ensiling period. The values obtained by Bui Van Chinh et al (1992) in cassava leaves ensiled with molasses were 32 - 34 mg/kg DM.

wpeC.gif (4948 bytes)
Figure 2: Effect of levels of molasses and of rice bran on pH 
of ensiled cassava leaves after different periods of ensiling

Changes in pH with time of ensiling are shown in Figure 2. There was an immediate fall from pH 7.1 to pH 4 within seven days of initiating the ensiling process and then a more gradual reduction to pH 3.5 after 21 days. There were no differences due to kind or level of additive. These trends are similar to those reported by Nguyen Van Lai and Rodriguez (1998).





Experiment 2:  Digestibility and nitrogen retention in fattening pigs fed different levels of ensiled cassava leaves (ECL) as a protein source and ensiled cassava root (ECR) as energy source

Materials and methods

The study was done at the research farm of Hue University which is located in Hue city, Thua Thien Hue Province in Central  Vietnam. There are two distinct seasons: the dry season from April through August, with a characteristic south-west hot wind throughout the season. The hot wind coupled with sunny weather can increase day temperature up to 40 - 41 ºC. The rainy season starts in September and lasts through to March. The season is characterized by an uneven distribution of rainfall which is concentrated from October to December. During this time the average monthly rainfall is around 3,000 mm and the average humidity is between 80 - 85%.


The objective of the study was to evaluate effects on digestibility and nitrogen balance in fattening pigs of  substituting  fish meal by ensiled cassava leaves at levels of 0, 50, 70 and 100 g/day of crude protein.

Treatments and design

The four treatments were:

The experimental design was a double 4*4 Latin square arrangement (Table 1).

Table 1: Allocation of pigs to experimental treatments
Period/Pig no 1 2 3 4 5 6 7 8
1 ECL0 ECL100 ECL75 ECL50 ECL50 ECL75 ECL100 ECL0
2 ECL100 ECL75 ECL50 ECL0 ECL75 ECL100 ECL0 ECL50
3 ECL75 ECL50 ECL0 ECL100 ECL100 ECL0 ECL50 ECL75
4 ECL50 ECL0 ECL100 ECL75 ECL0 ECL50 ECL75 ECL100
Animals and housing

Eight F1 Large-White x Mong Cai castrated male pigs with initial body weight of 48 - 50 kg and age of four months were randomly allotted to the diets according to the design layout (Table 1). The pigs were housed individually in metabolism cages that allowed the separate collection of urine and faeces. The size of the metabolism cage was 1.5 x 0.55 m (length x width) and they were made of galvanized steel with wooden slatted floor. The experimental periods were of 10 days: five days for adaptation period to allow the pigs to become familiarized with the new diet and a five day period for collection of faeces and urine.

Diets and feeding

The cassava leaves (after removing stems and petioles) were chopped into small pieces (1-2 cm) and ensiled with 5 % "A" molasses and stored for 21 days before feeding. The ensiling of the cassava roots (ECR) was according to the method used by Nguyen Thi Loc et al (1997). The roots were washed and chipped by hand and ensiled with 0.5% common salt in plastic bags. The ECR was offered at levels of 2.5 to 3 kg at each of three meals daily. After 40 minutes the uneaten feed was collected and weighed. Amounts offered and refused were weighed and samples taken for analysis. The ECL and fish meal were fed separately following the same procedure as with the ECR.


Urine and faeces of each pig were collected separately and weighed twice daily and stored at - 20 ºC. Urine was collected in a bucket via a funnel below the cage. To prevent nitrogen losses by evaporation of ammonia, the pH was kept below pH 4 by collecting the urine in 50 ml of 25 % sulphuric acid. The urine and feces from each animal were collected for five days and at the end of the period, the faeces were mixed, dried (in a drying over at 60-65 ºC), ground and representative samples taken for analysis. Dry matter of feed offered and refused, dry matter and nitrogen in faeces and nitrogen in urine were determined  according to AOAC (1984).

Results and discussion

Composition of feeds
Table 2: Composition of feeds offered
Dry matter,
% of DM
Crude fibre
% of DM
mg/kg DM
ECL 34.5 4.1 16.8 60.9
ECR 42.0 0.36 2.5 95.9
Fish meal 91.2 6.74

The composition of the feeds is shown in Table 2. The value for nitrogen content of the fish meal indicates that the protein level is about 42% which is below the average for this ingredient which normally has around 60-70% protein in dry matter (Gohl and Speedy 1996).  The low nitrogen content of the ECR shows that there is little protein (<2% in DM) in this energy source which means that responses in parameters of nitrogen metabolism truly reflect the nutritive value of the protein supplement. This can be a problem when cereal grains are used as the basal diet in view of the 9-10% protein in these energy sources. The values for HCN in the ECR and the ECL are less than the suggested maximum  values of this compound in order to avoid toxicity (Gomez and Valdivieso 1985).

Digestibility and N retention

The data in Table 3 show that the actual  intakes of protein from the cassava leaf silage were less than the planned amounts being 36, 63 and 62 g/day compared with the planned quantities of 50, 75 and 100 g/day, respectively. The biggest difference   between planned and recorded intake was for the intended 100 g/day of protein from ECL with the actual intake being only 62 g/day. There were no differences in total dry matter intake whereas it was expected that this would have increased with increasing levels of ECL in order to compensate for the lower protein content compared with fish meal.

Table 3: Mean values for  intake of feed and protein, coefficients of apparent digestibility and N retention for different levels of substitution of fish meal by cassava leaf silage
Treatment ECL0 ECL50 ECL75 ECL100 SE/Prob.
Intake fresh basis, g/day
ECL 0 409 716 701 67.5/.001
ECR 2,250 2,062 2,000 2,000 66.4/0.04
Fish meal 360 240 180 120
Dry matter intake, g/day 1,264 1,226 1,251 1,191 42.0/0.62
Protein intake, g/day

Fish meal




















Protein in DM, %





Digestibility, %
Dry matter 90.1 89.5 87.4 89.6 0.73/0.08
Crude protein 86.6 84.9 80.1 79.6 1.30/0.001
N retention
% of digested N 65.7 68.7 62.9 60.7 2.38/0.13
% of N intake 56.9 60.0 54.0 58.2 2.31/0.31
Per day, g 14.2 13.8 12.0 9.91 0.74/0.001


The dry matter intakes of each of the diet constituents as a percentage of total dry matter intake is shown in Table 4. It can be seen that with free access to the ECL and ECR, but with fish meal restricted, the pigs reached a ceiling in consumption of ECL at a   level of 20% of the total diet dry matter. It is not known if the inability of the pigs to consume the highest levels of ECL was because of low palatability of the ECL or a physical effect due to the fibre content or even the residual level of HCN. The fibre level increased from 3.39 % of DM for the ECL0 diet to 5.17% for the highest level of ECL. A level of 5.17%  fibre in the diet is relatively low and would not be expected to be a constraint  to intake.  It would seem that some other factor is constraining the intake of ECL at levels in excess of 20% of the diet dry matter.

Table 4: Intake of dietary constituents expressede as percentage of total intake (% dry matter basis)










Fish meal






Coefficients of apparent digestibility of dry matter were high with only an indication (P=0.08) of lower values with the ECL.

The decrease in the digestibility of the nitrogenous fraction was more marked (P=0.001) with a linear decrease as levels of ECL increased.  The trends were similar to those reported by Bui Huy Nhu Phuc et al (1996) who worked with ensiled cassava leaf but with dried cassava root meal diets for growing pigs. The decreases in digestibility were more pronounced in the studies reported by Bui Huy Nhu Phuc et al (1996), from 94 to 85% for dry matter and from 88 to 66% for crude protein. 

Daily nitrogen retention decreased significantly (P=0.001) from 14 g/day on the ECL0 diet to 9.91 g/day on ECL100, a similar fall to that reported by Bui Huy Nhu Phuc et al (1996). There were no significant differences between levels of ECL in the nitrogen retained as a percentage of nitrogen consumed and nitrogen digested . This finding is similar to the results of Bui Huy Nhu Phuc et al (1996) although they reported a sharp drop in these values for the highest level of ECL substitution. Furthermore, the values in the present study were some 10 percentage units higher than found by Bui Huy Nhu Phuc et al (1996).

The higher values for nitrogen retention as percentage of nitrogen digested in the present study, compared with the results of Bui Huy Nhu Phuc et al (1996), indicate that the dietary amino acid profile was better balanced in the diets used here.  It has been shown that the limiting amino acid in cassava leaf is methionine (Eggum 1970; Rogers and Milner 1963) but that it is rich in lysine and threonine.  Soya bean is also low in methionine while fish meal is a good source of this amino acid, thus a combination of cassava leaf and fish meal, as used in the present study, would be expected to be superior to a combination of cassava leaf and soya bean meal, which was the complementary protein source used by Bui Huy Nhu Phuc et al (1996).


Ensiling cassava leaves with 5% molasses or 10% rice bran is an effective method of conserving this feed resource and reducing cyanide levels to non-toxic proportions. There appear to be no advantages from using higher levels of either molasses or rice bran.

Ensiled cassava leaves can be used at up to 20% of the dry matter in diets of ensiled cassava roots for growing-fattening pigs thus providing  some 40% of the total dietary protein equivalent in a cassava root diet of about 50% of the supplementary protein. These levels assume that the complementary protein source is fish meal.


The author wishes to thank the Danish Embassy in Hanoi and the University of Tropical Agriculture Foundation (UTA) for financial support for this study which was submitted as partial requirement for the Master of Science degree in Sustainable Use of Natural Renewable Resources of the UTA. The SAREC research fund (Vietnam) also provided partial funding. 


Alhassan W S and Odoi F 1982 Use of cassava leaf meal in diets for pigs in the humid tropics. Tropical Animal Health and Production (14) :216-218

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

AOAC 1984 Official methods of analysis. Association of Official Agricultural Chemists, Washington, DC.

Bui Huy Nhu Phuc, Preston T R, Ogle B and Lindberg J E 1996 The nutritive value of sun-dried and ensiled cassava leaves for growing pigs. Livestock Research for Rural Development (8) 3:26-33

Bui van Chinh, Le Viet Ly, Nguyen Huu Tao and Do Viet Minh 1992. Using "C" molasses and ensiled cassava leaves for feeding pigs. Results of research 1985-1990. Agricultural Publishing House 1992, p46.

Du Thanh Hang, Nguyen Van Lai, Rodríguez Lylian and Ly J 1997 Nitrogen digestion and metabolism in Mong Cai pigs fed sugar cane juice and different foliages as sources of protein. Livestock Research for Rural Development (9) 2:45-49

Eggum O L 1970 The protein quality of cassava leaves. British Journal of Nutrition (24) :761-769

Gohl B and Speedy A W 1996  Tropical Feeds Database. Version 6. FAO:Rome

Gomez G and Valdivieso M 1985 Cassava foliage: chemical composition, cyanide content and effect of drying on cyanide elimination. Journal of the Science of Food and Agriculture. (36) :433-441

Nguyen Van Lai and Rodríguez Lylian 1998 Digestion and N metabolism in Mong Cai and Large White pigs having free access to sugar cane juice or ensiled cassava root supplemented with duckweed or ensiled cassava leaves. Livestock Research for Rural Development (10) 2:HTML format

Nguyen Thi Loc, Preston T R and Ogle B 1997 Cassava root silage for crossbred pigs under village conditions in Central Vietnam. Livestock Research for Rural Development (9) 2:12-19

Ravindran V and Ravindran G 1988 Changes in the nutritional composition of cassava (Manihot esculenta Crantz) leaves during maturity. Food Chemistry (27) :299-309

Ravindran V 1990 Feeding value and digestibility of cassava leaf meals for growing pigs. Proceedings Fifth Australasia Animal Production Congress. Volume 3: 20.

Ravindran V 1992  Preparation of cassava leaf products and their use as animal feed. In " Roots, tubers, plantains and bananas in animal feeding". (Editors: D Machin and AW Speedy). FAO Animal Production and Health. Publication 95: 111-126.

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

Rogers D J and Milner M 1963 Amino acid profile of manioc leaf protein in relation to nutritive value. Economic Botany. (17) :211-216

Sarwat S V, Kakala S N and Kategile  J A 1988  Performence of growing - finishing pigs when fed diets containing fresh cassava leaves and roots. East African Agricultural and Forestry Journal 53: 111-115.

Thua Thien Hue 1997 Annual Statistics. Hue Province

Received 10 December 1998

Go to top