Workshop-seminar, 21-24 August 2006,  MEKARN-CelAgrid   Workshop on Forages for Pigs and Rabbits

Citation of this paper

Body of Text

Forages as protein sources for pigs in the tropics

T R Preston

UTA-TOSOLY - Finca Ecológica
Morario, Guapota, AA # 48, Socorro, Santander,
Santander del Sur, Colombia

The original version of this paper was accepted for publication in CABI Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural Resources  ( as of 12 September 2006


The decreasing availability and increasing prices of petroleum-based fuels is already having an impact on availability of feed resources. Research to develop alternative sources of live stock feed thus has a high priority. It is also important that the alternative feed resources can be grown on the farm as this will: (i) directly benefit the poorer farmers, who do not have cash resources to purchase supplements from outside the farm; and (ii) be an active response to the need to localize the farming system as a defense against the decreasing availability and increasing prices of petroleum-based fuels which will drive up the costs of transport.

This review has shown that the leaves from shrubs such as cassava and mulberry, from vegetables such as sweet potato and cocoyam, together with water plants such as duckweed (Lemna spp) and water spinach (Ipomoea aquatica), can be used successfully in diets for pigs to replace at least half the protein usually supplied as soya bean and fish meals.

It is strongly recommended that future research in this area should be with basal diets that are low in protein and in cell wall constituents, as this will facilitate: (i) increasing the percentage of forage protein in the diet; and (ii) decreasing the overall protein level in the diet in accordance with the concept of the "ideal" protein.

Feed resources in the age of declining oil supplies

Conventional feed resources for pigs are cereal grains balanced either by soybean meal or fish meal or combinations of the two. These feed resources promise to follow the trends in fossil fuel, namely a steady increase in the gap between demand and production (Figure 1) and consequent increases in price.

Figure 1: World production and demand for maize (Rameker 2004)

This is because cereal grain is increasingly being used as a source of fuel via its conversion to ethanol (Figure 2).

Figure 2:
Past and future trends for conversion of maize to alcohol in USA
(Pearce Lyons and Bannerman 2001)

The sustainability of ethanol derived from biomass is presently the subject of a fierce debate between: independent scientists (Pimmental and Pazdek 2005) who calculate that it requires more energy to produce than is contained in the resulting ethanol; and government and industry representatives who argue that there is a net increase of 30% relative to the energy costs of production. Nevertheless the reality is that in all parts of the developed world and in advanced developing countries such as Brazil, China, India and Thailand, new ethanol distilleries are being constructed (Figure 3).

Figure 3: Global tendencies in production of alcohol (Berg 2001)

Irrespective of the arguments over its sustainability, there is not enough feedstock (principally cereal grains, sugar cane and cassava roots) for ethanol production to completely replace oil as a transport fuel on a world basis. Thus other solutions must be found. One alternative is the decentralized production of electricity by gasification of fibrous biomass and the use of the electricity to charge the batteries in cars driven partially or wholly by electric motors. This approach has the advantage of generating much employment in rural areas through growing and processing of the biomass.

Implications of ethanol and biomass gasification for live stock production and especially pig production.

One obvious result of the increasing use of grain for alcohol production is that its availability as live stock feed will decrease while prices will increase. Thus there is a need to search for alternatives. By contrast, the promotion of gasification will create opportunities for such alternatives, in the form of crops that produce both fibre and feed.

A secondary result of increasing fuel prices is to favor de-centralized systems of production which are less dependent on transport for the acquisition of inputs. This will also impact strongly on past trends for pig production which have resulted in large scale units (units with 1000 sows and followers are not uncommon) that are completely dependent on importing / transporting their feed supplies. Finally there are environmental issues such as the pollution involved in the disposal of excreta in such large operations and the difficulty (impossibility) of using efficiently this resource as a fertilizer to replace fossil fuel-dependent chemical fertilizers the prices of which will follow those of oil and gas.

There is therefore high priority to develop alternative feeding and management systems for pigs, as well as other live stock, which can be sustained within their own land base. Apart from the political and environmental benefits of reducing the dependence on oil, and oil-related resources such as cereal grains, there will be social benefits in that land-based farming systems will favor smaller scale units in which all the family can participate, with resultant increases in employment opportunities in rural areas.

Land-based farming systems

The questions to be addressed are:

An early hypothesis, which was proposed as the basis of a research strategy for tropical latitudes, was that sugar cane could be the basis of an integrated farming system in which the juice expressed from the stalk in a simple 3-roll press could be used as the energy source for pigs while the residual bagasse could be the feedstock for gasification to produce electricity (Preston 1989). The advantage seen in such a system was that the juice, composed solely of 100% digestible sugars, would create opportunities for using forages as the protein source, since the absence of fibre in the juice would facilitate the intake of these fibrous feeds.

Unfortunately, the system depended on making efficient use of the bagasse as a source of energy for gasification. This was shown to be technically feasible (Preston T R and Lindgren A 1980, unpublished data) but in an era of cheap oil was not economically viable. So the hypothesis remained unproven and the proposed research on using forages as the protein source was not initiated

However, as indicated in the introductory paragraphs, the world in 2006 is very different from what it was in 1989. Not only is the need to replace oil energy with that from biomass a political issue (Bush W, State of the Union speech 2 006), but the predicted scarcity and high prices of cereal grains will make more attractive the use of alternative sources of energy for live stock. The 1989 hypothesis can now be tested under favourable economic and political conditions.

A basic feature of the original hypothesis was that in order to use efficiently the protein in forages the basic energy source should be low in fibre. There are a number of tropical feeds which satisfy this condition:

When the basal diet is derived from any of the above feeds, it is possible to reduce the protein level, as when the protein is derived mainly from the supplement, the amino acid balance is better and approximates to the "ideal" protein (Wang and Fuller 1989). In this case, it has been shown that total protein levels can be reduced by 30 to 40% compared with the case of diets based on cereal grains (Figure 4).

Figure 4: Requirements for amino acids in sows fed a diet with an "ideal" protein (Wang and Fuller 1989)
compared with the traditional maize-soybean diet (NRC 1988)

Experiments with high levels of protein-rich foliages

Duckweed (Lemna spp.) (Figure 5)

One of the first experiments to evaluate diets in which all of the protein was derived from forage was reported by Rodríguez and Preston (1996). They used young pigs of a local Vietnamese breed (Mong Cai) and crosses of Mong Cai with Large White, giving them free access to sugar cane juice and fresh duckweed (protein content of 35% in dry matter), which had been fertilized with biodigester effluent. Nitrogen retention increased linearly according to the proportion of duckweed in the diet (Figure 6).



Figure 5: Duckweed (Lemna spp)


Figure 6: Relationship between N retention and proportion of duckweed consumed by Mong Cai and Mong Cai crossbred piglets offered a basal diet of sugar cane juice (Rodríguez and Preston 1996)


On-farm results confirmed the potential benefits from the use of duckweed in pig diets (Figure 7).

Figure 7: Supplements of fresh duckweed increased growth rates of pigs fed rice
byproducts on farms in Central Vietnam (Du Thanh Hang 1998)

Cassava (Manihot esculenta) leaves (Figure 8)

Many studies have evaluated use of cassava leaves as partial replacement for soya bean meal and fish meals in conventional diets for pigs (see review by Bui Ngu Phuc 2001). However, the levels used rarely exceeded 15% in the diet, and the emphasis was mainly on methods of reducing the risk of cyanide toxicity from cyanogenic glucosides in the leaves. Sun-drying appeared to be more effective than ensiling in this respect (Bui Huy Nhu Phuc et al 1996).


Figure 8: Cassava (Manihot esculenta)

Despite this concern for cyanide toxicity it is relevant to note that there appear to have been no reported deaths of pigs from this cause. A possible reason that has been suggested for this lack of effect (R A Leng, personal communication) is that the enzymes responsible for the release of HCN from the cyanogenic glucosides are inactivated at the low pH in the pig stomach.

The uncertainly relating to possible risks of HCN toxicity led Du Thanh Hang and Preston (2005) to compare processing of the fresh cassava leaves by washing, chopping and washing; and chopping, washing and wilting, and feeding them to pigs of 25 kg live weight as supplements to a basal diet of cassava roots and rice bran. The cassava leaves were readily consumed providing 38% of the dietary DM and over 70% of the dietary protein with no effect of processing method on total DM intake, which ranged from 27 to 31 g/kg live weight. Levels of HCN were reduced slightly (16%) by washing and substantially (82%) by wilting, resulting in intakes of HCN between 6.0 and 15 mg/kg live weight, levels considerably higher than the range of 1.4 to 4.4 mg/kg live weight, previously reported as safe to avoid toxicity (see Du Thanh Hang and Preston 2005). Following on from these findings, Chhay Ty and Preston (2005a,b) fed fresh cassava leaves as the only supplement to a basal diet of broken rice, comparing this treatment with fresh water spinach foliage or a 50:50 mixture (DM basis) of the two vegetative sources. Feed intakes, growth rates and DM digestibility coefficients were higher when the water spinach was the only supplement or when it was mixed with the cassava leaves, as compared with the cassava leaves alone (Figure 9).

Figure 9: Growth rates of pigs (from 11 to 50 kg) fed broken rice supplemented with fresh cassava leaves (CL),
water spinach (WS) or a mixture of the two (CL-WS) (Chhay Ty and Preston 2005b)

Water spinach (Ipomoea aquatica) (Figure 10)

Water spinach grows equally well in the water or in soil. It is traditionally consumed by people in SE Asia and appears to be devoid of non-nutritional elements. Harvesting this plant from lagoons fertilized with waste water from urban centres is an important source of income for poor peoples in Vietnam, Cambodia and Laos. An important feature of water spinach is its capacity to yield high levels of biomass when fertilized with effluent from biodigesters charged with pig manure (Kean Sophea and Preston 2001).

Figure 10: Water spinach (Ipomoea aquatica)

Like the mulberry plant, consumed by silkworms, it was to be expected that water spinach, consumed as a vegetable by humans, would also have high nutritive value. Research in Cambodia confirms that this is so. Basal diets of broken rice (8% protein in DM) in which water spinach provided 50% of the dry matter and 70% of the protein were highly digestible by pigs and supported higher N retention than similar diets in which cassava leaves were the supplementary protein source (Chhay Ty and Preston 2005a). Growth rates of young pigs on these diets paralleled the N retention data (Chhay Ty and Preston 2005b).

Mulberry (Morus alba) leaves (Figure 11)

The leaves of the Mulberry bush have long been used as substrate for growth of the larvae of the silkworm. It could therefore be expected that they would have potential as a protein source for pigs.

Figure 11: Mulberry (Morus alba)

Chiev Phiny et al (2003) compared sun-dried leaves as partial or complete replacements for rice bran and fish meal in a basal diet of broken rice. Nitrogen retention appeared to be improved with increasing levels of mulberry leaf meal in the diet, when expressed as a percentage of N ingested or N digested. There were no differences in digestibility or nitrogen retention when fresh or sun-dried leaves were included at levels of 30% (DM basis) in a diet in which the energy component was wheat bran.

Sweet potato (Ipomoea batatas) (Figure 12)

A very comprehensive study on sweet potato vines was done by Le Van An (2004). He concluded that "the best options in terms of leaf and stem production were a cutting interval of 20 days and a defoliation of 50% of the total branches. Defoliation reduced tuber production.

Figure 12: Sweet potato (Ipomoea batata)

There appear to be considerable differences, depending on variety, in the content of crude protein and crude fibre in the dry matter of the foliage of sweet potato, the former ranging from 26.5 to 32.5 % in leaves and from 10.4 to 14.1 % in stems (Woolfe 1992; Ishida et al 2000; Le Van An 2004). For crude fibre the mean values were 11.1 and 20.7 %, respectively (Woolfe 1992). It is therefore important to separate leaves from stems when the aim is to maximize the rate of inclusion of sweet potato foliage in pig diets. According to Le Van An (2004) there are no major differences in nutritive value between fresh, sun-dried or ensiled leaves. Adding synthetic lysine (L) to a basal diet of rice by-products supplemented with sweet potato leaves (SP) increased pig growth rates to a level which did not differ from that on the positive control diet containing fish meal (FM) and which were better than on the diet supplemented with groundnut cake (GC). Daily live weight gains of pigs on the FM, GC, SP and SPL diets were 542, 464, 482 and 536 g, respectively (P<0.05).

Giant Taro (Alocasia macrorrhiza) leaves

This plant appears to have been neglected by live stock researchers despite the fact that it is widely distributed in tropical latitudes, and possesses desirable characteristics such as high protein (24% in DM) and low content of crude fibre (15% in DM) (Gohl 1973). This author described the Giant Taro as a "herbaceous plant with large heart-shaped leavc to 1.5 m long that contain a milky juice. The wild variety has a pungent taste caused by oxalate crystals which also cause the mouth to itch. Cultivated forms do not have this property, and the leaves, stalks and black rhizome can be fed to animals".

Interviews with farmers in Cambodia and Vietnam (Preston T R , unpublished data) failed to elucidate the meaning of "cultivated" as opposed to "wild" forms. Most farmers indicated that they traditionally "boiled" the leaves before feeding them to pigs as in the fresh state the leaves were not readily consumed.

New Cocoyam (Xanthosoma sagittifolium) (Figure 13)

Figure 13: New cocoyam (Xanthosoma sagittifolium)


New Cocoyam, which appears to be native to South and Central America, is similar in appearance to the Giant Taro (Alocasia macrorrhiza). However, in contrast to the Giant Taro, the fresh leaves of the "New Cocoyam" were found to be highly palatable to pigs (Rodríguez et al 2006). At a level of 50% of substitution of the soybean meal protein in a sugar cane juice diet for growing pigs, the growth rates were the same as when the soybean meal provided all the protein (Figure 14).

Figure 14: Growth curves of pigs fed a basal diet of sugar cane juice
supplemented with soybean meal or a 50: 50 mixture (equal protein basis)
of leaves of "New Cocoyam" and soybean meal


The greater part of the soybean meal and maize fed to pigs throughout the world originates in the USA and Brazil. As transport costs rise, driven by decreasing availability of cheap oil, this inevitably will result in increased costs to farmers that have become "addicted" to these feed resources as the basis of their pig production systems.

Research to develop alternative sources of protein therefore has a high priority. It is also important that the alternative feed resources can be grown on the farm as this will: (i) directly benefit the poorer farmers, who do not have cash resources to purchase supplements from outside the farm; and (ii) be an active response to the need to localize the farming system as a defense against the decreasing availability and increasing prices of petroleum-based fuels which will drive up the costs of transport.

This review has shown that the leaves from shrubs such as cassava and mulberry, from vegetables such as sweet potato and cocoyam, together with water plants such as duckweed (Lemna spp) and water spinach (Ipomoea aquatica), can be used successfully in diets for pigs to replace at least half the protein usually supplied as soybean and fish meals.

The use of these vegetative protein sources is facilitated when they are accompanied by energy sources that are low in both fibre and protein. A low content of fibre in the energy source creates "space" in the digestive tract of the pig for fibre present in the leaves. In contrast to leaf proteins, which are mainly present as enzymes having a balanced array of essential amino acids (Table 1), the readily digestible energy in plants acts as a reserve supply of nutrients to promote germination or recovery in growth after a period of stress, such as happens in the dry season. With the exception of rice, the protein associated with the energy in cereal grains is also a store of nutrients but one in which the amino acid array is usually imbalanced (eg: the deficiency of lysine in maize). By contrast, the nutrient reserves in most tropical crops (rice being the exception), contain little or no protein (eg: the roots of cassava, tubers of sweet potato and taro, the fruit of bananas and plantains, and the stalks of sugar cane). They also are low in fibre. Thus there is potential for a natural synergism between sources of energy and protein in tropical latitudes.

When energy sources low in protein (eg: sugar cane juice, cassava roots, sweet potato tubers or banana fruit) are combined with leaves which normally have well balanced arrays of amino acids, it is possible to reduce the overall level of protein in the diet since the amino acid array in such a combination of feeds will more closely approximate to that in the "ideal" protein (See Wang and Fuller 1989).

Unfortunately much of the research with protein-rich forages has been with conventional energy sources (maize and rice bran). As these feeds contain fibre and are imbalanced in essential amino acids, the potential to use high levels of protein-rich foliages is compromised. It is strongly recommended that future research in this area should be with basal diets that are low in both protein and in cell wall constituents, as this will facilitate: (i) increasing the percentage of forage protein in the diet; and (ii) decreasing the overall protein level in the diet in accordance with the concept of the "ideal" protein.

There is also a need to establish more accurately the levels of nutrients in protein-rich leaves, especially the proportions of the most limiting essential amino acids. An attempt to summarize some of these data (Table 1) revealed a great deal of variation with contradictory values being reported by authors even within the same publication. Part of the problem appears to be the confusion when authors report only levels of methionine, which may or may not also include values for cystine.

Table 1:  Major essential AA in the “ideal protein”, soybean meal and leaves of selected protein-rich leaves


Ideal protein(1)

Soybean meal (2)

Water spinach (3)

Cassava leaves (4)

Sweet potato leaves (5)

Duckweed (7)

New cocoyam (6)



g AA/kg N*6.25














































As proportion of lysine = 100






























Composition of fresh leaves, g/kg fresh matter












Composition, g/kg DM



Crude protein








Crude fibre


















(1)Wang and  Fuller  1989; (2) Martin 1990 ; (3)  Phiny et al 2003; Phiny  2006, personal communication; (4) Eggum 1970; (5) Woolfe 1992; (6)Rodríguez et al 2006; (7) Le Thi Men 2006


What appears to categorise the most useful sources of leaf protein, as identified by animal trials in this review, are a relatively low level of crude fibre and a ratio of sulphur amino acids relative to lysine close to that in the ideal protein. The wide differences found in many reports for amino acid levels in cassava leaves and sweet potato leaves (the most reliable values for cassava leaves were done 36 years ago) emphasize the need for a coordinated research effort in which common samples of the most useful protein-rich leaves are distributed for analysis to several laboratories where the necessary equipment and expertise are available.


An L V 2004 Sweet potato leaves for growing pigs: Biomass yield, digestion and nutritive value. Doctor's dissertation. ISSN 1401-6249, ISBN 91-576-6750-0

Berg C 2001
World Ethanol Production Fo Licht's International Molasses and Alcohol Report .

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

Chhay Ty and Preston T R 2005aEffect of water spinach and fresh cassava leaves on intake, digestibility and N retention in growing pigs. Livestock Research for Rural Development. Vol. 17, Art. #23. Retrieved June 30, 2005, from


Chhay Ty and Preston T R 2005b: Effect of water spinach and fresh cassava leaves on growth performance of pigs fed a basal diet of broken rice. Livestock Research for Rural Development. Volume (17) 7. Retrieved , from


Du Thanh Hang 1998 Ensiled cassava leaves and duckweed as protein sources for fattening pigs on farms in Central Vietnam. Livestock Research for Rural Development. Volume (10) 3. Retrieved , from


Du Thanh Hang and Preston T R 2005The effects of simple processing methods of cassava leaves on HCN content and intake by growing pigs. Proceedings of Workshop-seminar "Making better useof local feed resources" MEKARN-CTU, Cantho, 23-25 May, 2005.


Eggum B O 1970 The protein quality of cassava leaves. British Journal of Nutrition, Volume 24, Page 761


Göhl B 1971 Tropical Feeds Database based on the original book by by Andrew Speedy and Nick Waltham Retrieved from

Ishida H, Suzano H, Sugiyama N, Innami S, Tadokoro T and Meakawa A 2000
Nutritive value on chemical components of leaves, stalks and stems of sweet potato (Ipomoea batatas Poir). Food Chemistry 68: 359-367.

Kean Sophea and Preston T R 2001 Comparison of biodigester effluent and urea as fertilizer for water spinach vegetable. Livestock Research for Rural Development 13 (6)


Le Thi Men 2006 Genetic and nutritional diversity of animal and plant resources as animal feed and their utilization in swine husbandry in the Mekong Delta of Vietnam. Doctoral thesis. Tokyo University of Agriculture and Technology, Japan.


NRC 1988 Nutrient Requirements of Swine (9th Ed.). National Academy Press, Washington, DC.


Pearse Lyons T and Bannerman J 2001The US Fuel Alcohol Industry from 1980 to 2001; Lessons for other markets . In: A time for Answers. Proceedings of Alltech's 15th Asia-Pacific Lecture Tour 115\


Phiny C, Preston T R and Ly J 2003 Mulberry (Morus alba) leaves as protein source for young pigs fed rice-based diets: Digestibility studies; Livestock Research for Rural Development (15) 1. Retrieved , from


Phuc B H N, Lindberg J E, Ogle B and Thomke S 2001 Determination of the nutritive value of biomass products for monogastrics using rats. Part 1. Comparison of eight forage species at two level of inclusion in relation to a casein diet. Asian Australasian Journal of Animal Science. 14, 896-993


Pimental D and Patzek T W 2005 Ethanol production using corn, switchgrass, and wood; biodiesel production using soybean and sunflower. Natural Resources Research 14(1):65-76.


PrestonT R 1989 Food, fuel and jobs from sugar cane and tree legumes. In: New Crops for Food and Industry (edited by G E Wickens, N Haq, P Day) 1989, 444 p. Springer Publishers.


PrestonT R and Rodríguez Lylian 2004: Production and utilization of cassava foliage for livestock in integrated farming systems. Livestock Research for Rural Development. Vol. 16, Art. #28. Retrieved , from


Rameker J 2004 Grain production, supply and demand outlook and trends in livestock feed industry. Proceedings 11th Animal Science Congress, Kuala Lumpur, Malaysia, September 5-8, 2004


Rodríguez Lylian and Preston T R 1996 Comparative parameters of digestion and N metabolism in Mong Cai and Mong Cai*Large White cross piglets having free access to sugar cane juice and duck weed. Livestock Research for Rural Development (8) 1. Retrieved , from \


Rodríguez Lylian, Lopez D J, Preston T R and Peters K 2006 New Cocoyam (Xanthosoma sagittifolium) leaves as partial replacement for soya bean meal in sugar cane juice diets for growing pigs. Livestock Research for Rural Development. Volume 18, Article #91. Retrieved September 11, 2006, from


Speer V C 1990 Partitioning nitrogen and amino acids for pregnancy and lactation in swine: A review. Journal of Animal Science 68: 553-561.


Wang T C and Fuller M F 1989 The optimum dietary amino acid pattern for growing pigs. British Journal of Nutrition. 62, 17-89


Woolfe J A 1992 Sweet potato: an untapped food resource. Cambridge University Press, Cambridge, 643 pp.


Go to top