6. FORAGE PRODUCTION SEASONALITY - SUPPLEMENTARY FEED SOURCES

6.1 Introduction

Seasonal growth occurs for a number of reasons: low sunlight during rainy overcast weather, different growth patterns of forage species used and flowering or day-length responses which reduce forage yield (Plucknett, 1979). However, the major problem for livestock, in many parts of the tropics, is the irregular seasonal rainfall distribution and its effect on pasture production (Villasmil et al., 1975; Vohnout and Jimenez, 1975).

The problem has been discussed by Faylon (1982) in his description of coconut based livestock production systems in the Philippines: “The situation becomes critical especially during summer months that do not favour the growth of grasses and legumes. During this dry, lean period, the ration should be supplemented with concentrates and crop residues to avoid costly losses in animal weight”. In some areas there is a very pronounced dry season, in others there may be two dry periods separated by the ‘long’ and ‘short’ rains (see Figure 181). If monthly rainfall falls much below 100 mm and certainly below 50 mm, particularly where soils have low water holding capacity, then pasture production will be reduced.

Figure 181. - Mean monthly rainfall Kizimbani, Zanzibar, Tanzania (mean annual rainfall 1848.5 mm).

6.2 Seasonality of forage production

The effect of rainfall on seasonality of forage production is illustrated in Table 119 and Figure 182. In Sri Lanka, Chadhokar (1980c) has indicated that forage production is seasonal because of the bimodal pattern of rainfall resulting in a period of about three to four months mostly between February-April and July-September (March-May and August-October according to Jayawardana, 1992) when forages are in short supply and often animals are underfed to the extent that even to supply forage for animal maintenance becomes impossible. The period varies with country, thus in the Philippines the lean grazing period occurs from November to May (Ranjhan and Faylon, 1992) and in Burundi from May to October (Branckaert, 1993), but as a result animal production is seriously affected and milk production, in particular, drops.

Assuming that the relative feed requirements for livestock, whether in a small backyard feeding system under coconuts or a large commercial coconut plantation, are relatively constant throughout the year, the key issue is to minimize the seasonal effects of feed production (see Figure 183) to carry the maximum number of cattle year round.

6.3 Methods of overcoming the problem of seasonal variations in forage production

There are a number of possible methods that can be used by the coconut grower in an attempt to ensure adequate year round feed supplies for his livestock (Plucknett, 1979; Villasmil et al., 1975; Williamson and Payne, 1978):

1. Stock at, or only just above, dry season feed levels and waste the excess wet season production.

2. Reduce production seasonality by growing a range of species and by judicious use of fertilizer to extend the growing season.

3. Implement seasonal livestock breeding programmes so that the total animal demand for feed achieves a peak at approximately the same time as forage growth is maximal.

4. Use grass-legume mixtures to prolong higher quality of feed into the dry season. Because legumes may be deeper rooted and of higher nutritive value, they could supply better forage.

5. Use legume bushes and trees such as Leucaena leucocephala and Gliricidia sepium as browse sources (Atta-Krah and Reynolds, 1989; Devendra, 1993a).

6. Grow special fodder crops under intensive management in order to provide special pasture or cut-feed for the animals during shortage periods. Such crops could include Napier grass (P. purpureum), Guinea grass (P. maximum), Leucaena (L. leucocephala), Guatemala grass (Tripsacum laxum) and various mixtures. These can be irrigated or grown in open areas adjacent to coconut plantations so that species such as Para grass (B. mutica) might also be grown on swampy land. Paltridge (1957) suggested using Guinea grass as a fodder crop or as a pasture grass under rotational grazing in coconuts in Sri Lanka to provide reserve feed, and Ellewela (1956) recommended Napier and Palisade grass to provide fodder in the dry season. Lane (1981) considers that where management levels are only moderate then fodder grasses such as NB21 P. purpureum × P. americanum hybrid and P. maximum cv. Hamil grown on a cut-and-carry system may allow up to three times the number of stock to be kept compared to grazed stoloniferous pastures. In Venezuela, Combellas and Martinez (1982) evaluated intake and milk production in cows fed chopped Elephant grass and concentrate, while in Colombia, Moore and Buchman (1978) estimated potential beef production on intensively managed Elephant grass. In West Africa, Atta-Krah and Reynolds (1989) introduced the concept of intensive feed gardens (plots of leguminous fodder trees and grasses) for dry season supplementation.

Figure 182. - The effect of rainfall on seasonal pasture yield.

1. Western Samoa under coconuts: Guinea-Centro pastures (Reynolds 1978j; Pottier, 1983).

2. Sigatoga, Fiji: seasonal forage production (Williamson and Payne, 1978).

Table 119. - Seasonal yield of Pangola grass as affected by rainfall and fertilizer applications (Medina et al., 1968; Preston, 1976b)

⁺ kg ha^-1 N 352 P 27.5 K 66

7. Use forage conservation techniques to preserve surplus forage from wet periods. Forage may be preserved as hay, silage, standing hay or artificially dried forage (Anon., 1982d). According to Whiteman (1980) field curing of hay during the wet season growing period of tropical pastures can be hazardous due to the high and frequent rainfall, and silage production from tropical forages is unreliable. However, Chadhokar (1983a) reports that in Sri Lanka cheap trench or pit silos with a ‘kadjan’ (coconut leaf) roof were developed and different types of forage crops tested for ensiling. Various grasses, legumes, fodder trees and paddy straw in combined forms were found to be quite suitable. Additives such as urea and concentrates improved the quality of the resulting silage but quite satisfactory silages were also made and demonstrated to the farmer without additives. Yokota et al. (1992) reported on the nutritional quality of wilted Napier grass ensiled with and without molasses. Grass silage, when fed to Friesian milking cows with Gliricidia and rice polishings, gave very encouraging results. Lane (1981) also indicates that surplus herbage produced during periods of peak growth in Sri Lanka can be ensiled. Ensiling of tropical and temperate forage crops has been reviewed by Wilkinson (1983a + b). In spite of various problems, the use of additives such as molasses (20 kg per tonne at ensiling) and urea (0.5% of fresh crop weight) as well as the development of simple machinery for cutting tropical grass crops to allow for a short period of field wilting before collection should encourage greater quantities of forage to be conserved.

For the large commercial coconut plantation options ii), iii), iv), v) and vi) are probably the areas which hold most promise. For the smallholder the use of deep-rooting species like Napier, Gliricidia, Leucaena, Calliandra and Guinea grown under a cut-and-carry system should enable better quality feed to be produced for dry periods (see Figure 184).

Figure 183. - The relationship between season, pasture production and feed requirements.

Figure 184. - Chopped sugar cane, Napier grass, Leucaena and rice bran being fed (as dry season feed) to penned animals in Zanzibar, Tanzania.

6.4 Alternative feed sources

There is another possibility which has been emphasized by Preston (1976b). Because the constraints affecting grazed pastures apply also to a large extent to conventional forage crops such as maize, sorghum and Napier grass (i.e. their active period of growth also occurs in the wet season), Preston has advocated the use of new crops, in particular sugar cane and cassava as a basis for cattle production (see Figure 185).

These crops are available in the dry season with high overall yields (see Table 120) and many of them like banana and sugar cane do not require annual planting. While various systems of feeding livestock have and are being developed using sugar cane, banana, cassava etc. and/or their by-products (see de la Hunte, 1981; Donefer, 1976; Donefer et al., 1973; Foulkes and Preston, 1978b; James, 1973; Leng and Preston, 1976; Muller et al., 1975; Preston, 1972a, 1975, 1976b; Preston and Willis, 1974), the widely practised methods of livestock production based on conventional forage production are likely to remain the norm in most areas, at least in the near future.

However, materials which can be used to supplement the normal forages, particularly during dry periods include:

These include conventional feeds and what have come to be called non-conventional or new feed resources. The rising costs of conventional feeds and reduced grazing areas, require that the accumulating by-products from agro-based industries and other sources be utilized.

At present there is a considerable under-utilization, in terms of animal feeding, of a vast amount of non-conventional feed resources including both the waste material generated from crop and animal production as well as the residues resulting from the processing of food for human consumption (Mahadevan, 1981). As most of these potential feeds are available in coconut growing areas, livestock owners need to be aware of their potential, in particular for use as feed supplements in dry periods. Thus, for example, Ranjhan and Faylon (1992) have summarized the various fibrous crop residues (straws, stovers, etc.) and other crop by-products (brans, milling by-products, etc.) and concentrates available for supplementing cattle and carabao during the dry period November to May. Abu Bakar and Aziz (1993) reviewed alternative sources of roughage available for the dairy cow in Malaysia.

Figure 185. - Seasonal availability of feed for animals (from a variety of crops)
(Preston, 1976).

Table 120. - Maximum yields of some tropical crops (Preston, 1976b; Nestel, 1974)

Several comprehensive reviews of the potential of fibrous agricultural residues in developing countries and of strategies for expanding their utilization have been made; for example, FAO (1985, 1986, 1988, 1992) IDRC and ICAR (1988). Also refer to Chenost and Sansoucy (1991) and Silverio (1983).

6.5 Summary of alternative feed sources for use as supplements

Devendra (FAO, 1988) estimated total availability of by-products from field crops in Asia and the Pacific at some 230,354.3 M mt. and from tree crops at 7,401.1 M mt. More recently, Devendra (FAO, 1992a) increased these figures to 247,666.2 M mt. and 9,136.6 M mt. respectively. In addition, there are various tree leaves, fruit and animal processing wastes and animal excreta, which, if included, would increase substantially the total availability of non-conventional feed resources. Accordingly to FAO (1985a) crop residues account for about 24% of the total feed energy available for ruminant livestock while agro-industrial by-products account for only about 1%. Hutagalung (1981) provided details of the availability, utilization and potential prospects for intensifying the use of feedstuffs from tree crops for feeding livestock. He concluded that while some of those identified had been utilized, the majority were under-utilized or had not even been exploited as animal feeds (see Tables 121, 122, 123).

Table 121. - Main tropical crops, field residues and agro-industrial by-products (Preston, 1982b)

Adegbola (1976) has drawn attention to a general lack of awareness of the possible uses of crop residues and other agro-industrial by-products for animal feeding. Wilson and Brigstocke (1981) suggest that the cause of this situation is due to:

• lack of information on their composition and nutritive value;
• non-availability of adequate handling and processing facilities;
• possible variations in composition;
• logistic problems of availability and seasonality of supply;
• need for transport and processing.

Chen (1989) stresses that agricultural by-products and residues while often plentiful may be very localized, may be found away from the centres of animal production and may be unbalanced in composition.

Wilson and Brigstocke (1981) also indicated that in many tropical countries post-harvest losses of tropical horticultural production are very heavy, approaching or exceeding 50 percent in some cases (Coursey, 1972), yet very little if any of this material is used as animal feed. Preston (1982b) suggested that industrial by-products and crop residues would be utilized with greater efficiency if they were incorporated into production systems which are diversified and integrated. He also suggested that for efficient rumen fermentation inadequacy of any of six factors (fermentable energy, fermentable nitrogen, micronutrients, roughage, by-pass protein and by-pass energy) could affect animal productivity directly via effects on feed intake. Devendra (1992a) has reviewed the current constraints to utilization of non-conventional feed resources and assessed the extent of use of indigenous feedstuffs.

Diaz and Zapata (1990) reviewed the use of agro-industrial by-products for dairy cattle feed in Colombia.

Recent African experience with the utilization of agricultural by-products as livestock feed has been reported in the proceedings of two workshops (ARNAB, 1987; Said and Dzowela, 1989) and Singh and Sehiere (1993) reviewed experience in India in feeding ruminants on fibrous crop residues.

Based on data gathered from Chadhokar (1980c), Devendra (1993; FAO, 1977, 1981a, 1988), FAO (1977, 1978b, 1980, 1982), Gohl (FAO, 1981b), Hutagalung (1981), Moog (1980a), O'Donovan (FAO, 1978b) and Reynolds and Lund (1983), categories of alternative sources of feed can be summarized as follows:

1. Forage grasses and legumes (leaves and stems - Guzman and Allo, 1975; Lane, 1981; Plucknett, 1979; Reynolds and Lund, 1983; Reynolds and Sini, 1976; Rodrigo, 1945; Sahasranaman and Sethumadhava Menon, 1973).

2. Fodder trees (leaves, pods and fruit rinds - Anon., 1975c, 1977a, 1979; Chadhokar, 1982; Devendra, 1985, 1993a; FAO, 1992d; Furr, 1965; Hashim, 1994; Manidool, 1992; Nitis et al., 1991a; Thomas and Addy, 1977; Wanapat, 1993). e.g.: Leucaena (L. leucocephala), gliricidia (G. sepium), Sesbania grandiflora, Sesbania sesban, dapdap (Erythrina indica), rain tree (Pithecellobium saman), tamarind (Tamarindus indicus), desmanthus (Desmanthus virgatus - Battad, 1993), jackfruit (Artocarpus heterophyllus), baobab (Adansonia digitata), ana (Acacia albida), lopa (Moringa oleifera) (Stunzner, 1976). According to Blair (1989) the shrubs and trees that can be used as fodder encompass some 74 genera of plants, while Brewbaker (1986) indicates that more than 200 species of tree fix N and are reportedly useful as fodder species. Devendra (1989, 1992) has described many experiments where tree and shrub fodder has been used as supplementary feed, Halim (1991) reviewed experience in South East Asia, Smith (1992) provided a similar review for tropical humid Africa and Speedy and Pugliese (1992) provided a review of global experience. Various multipurpose tree species for small-farm use have been described in Withington et al. (1988).

Table 122. - Estimates of the potential contribution of certain by-product feeds from tree crops for livestock production (Hutagalung, 1981)

Table 123. - Optimum rates of certain feed stuffs from tree crops in diets for cattle and buffalo (Hutagalung, 1981; Devendra, FAO, 1981a)

3. Crop residues and agro-industrial by-products:

   1. Food crops

      1. banana (Musa sp); reject fruit can be fed to livestock (Chenost et al., 1976; Dividich et al., 1978; Ruiz, 1981; Vohnout and Jimenez, 1975), and pseudostems and leaves used as forage (Babatunde, 1992; FAO, 1978a). For a review refer to FAO (1992c).

      2. cereals - rice (Oryza sativa): hulls, bran, millmeal (Dia Ndumbe, 1982; FAO, 1967b) and straw, the latter fed green, dry or treated with NaOH, urea or urine (Doyle et al., 1986b; Florido, 1992; Jackson, 1978; Saadullah et al., 1981b; Schiere and Ibrahim, 1988; Theander, 1981; Trung et al., 1989a) can provide valuable supplementary feed in coconut areas adjacent to rice growing areas; maize (Zea mays): stems, leaves and husks (in Thailand, baby corn waste is popular with dairy farmers - Manidool, 1992); millet (Eleusine coracana): straw; sorghum (Sorghum bicolor): straw.

      3. legumes - cowpea (Vigna unguiculata); common bean (Phaseolus vulgaris, L.); mungbean or green gram (Vigna radiata); pigeon pea (Cajanus cajan); soybean (Glycine max.); winged bean (Psophocarpus tetragonolobus). They can all provide protein rich residues once the pods have been harvested, with soybean being used as human food or as a component in animal feed (Shaw et al., 1980).

      4. root crops - sweet potato (Ipomea batatas): vines and reject roots (Anon., 1980c; Backer et al., 1980; Ruiz, 1981; Ruiz et al., 1980, 1981); cassava (Manihot esculenta): tubers, chips and leaves (Fernandez and Preston, 1978; Hahn et al., 1992; Muller, et al., 1975; Van Eys et al., 1987; Wanapat, 1993). For an overall review see FAO (1992c).

      5. fruit - mango (Mangifera indica); durian (Durian zibethinus); papaya (Carica papaya); pineapple (Ananas comosus); citrus (Citrus sp.); jackfruit (Artocarpus heterophyllus); breadfruit (Artocarpus altilis), guava (Psidium guajava) have been used as: fruit waste, seeds, kernels, chopped fruit, rinds and leaves (Ananthasubramaniam et al., 1977; Chapman et al., 1972; Geoffrey et al., 1984; Gohl, 1978; Guzman, 1976; Idris, 1981a; Kesterson and Braddock, 1976; Devendra and Gohl, 1970).

      6. others - amaranthus (A. spinosus, A. viridis, A. patalas) (Chairatanayuth and Santipong, 1985): reject leaves and stems (Mugerwa, 1961); sago (Metroxylon pagu); sago pith, sago meal and sago refuse have been used in livestock diets (Hew, 1978; Jalaludin, 1987; Yadav and Mahyuddin, 1991); water hyacinth (Eichornia crassipes): used as animal feed as hay and silage, contains about 2.5% oxalic acid - see FAO (1986a, 1992b); Poddar et al. (1990) and Wanapat et al., (1983).

   2. Industrial Crops

      1. major oil crops - coconut (Cocos nucifera): copra, copra cake or coconut meal are widely used in livestock diets (Castillo et al., 1961; Creswell and Brooks, 1971); oil palm (Elaeis guineensis): palm oil, palm kernel meal, palm press fibre and oil palm sludge have all been used in varying degrees as livestock feed (Babatunde et al., 1974; Boer and Sanchez, 1988; Camoens, 1978; Devendra, 1978, 1981; Devendra et al., 1981; Hutagalung, 1978; Idris, 1981b; McIntyre, 1973; Muller, 1978; Tinnimit, 1985; Vadiveloo, 1989). More recently experiments have begun with oil palm trunk (Raman et al., 1990).

      2. other oil crops - babassu palm (Orbygnya cohune); carob (Ceratonia siligua), castor (Ricinus communis), Kapok (Ceiba petandra) neem (Azadirachta indica), sal (Shorea robusta) and tamarind (Tamarindus indica): various seed meals, cakes, kernels and hulls have been used as livestock feed, some requiring detoxification (Patil et al., 1975; Reddy et al., 1977; Roberts, 1975; Sahai and Kahar, 1968).

      3. sugar cane (Saccharum officinarum): chopped whole cane, tops, leaves, derinded sugar cane (comfith), molasses, bagasse and filter mud have been used as livestock feed (Abdalla et al., 1990; Elias, 1988; Donefer, 1976; FAO, 1979, 1988a, 1988b, 1988c, 1992e; Guzman, 1976; Kirk et al., 1982; Mena, 1988; Moog, 1980; Naseeven, 1988; Pigden, 1978; Preston, 1978, 1988; Preston and Willis, 1969; Wadsworth, 1990).

      4. rubber (Hevea brasiliensis): rubber seed meal is a valuable protein source but heat treatment and storage are required to reduce the level of hydrocyanic acid (HCN) when used as a livestock feed (Anon., 1976b; FAO, 1981a; Hutagalung, 1981; Stosic and Kaykay, 1981; Tinnimit, 1985).

      5. cocoa (Theobroma cacao); coffee (Coffea arabica): various by-products used as feeds, include cocoa shells, pod husks and bean residue, coffee pulp, hulls, molasses and wastes (Bartley et al., 1978; Bateman and Larrangan, 1966; Bressani et al., 1975; FAO, 1981b; Hutagalung, 1981; Jalaludin, 1989; Jarquin et al., 1976; Smith and Adegbola, 1982; Smith et al., 1988).

4. Animal origin feeds

   e.g. fish meal, animal products - blood, bone and offal, poultry manure: various by-products available from fish processing plants, abattoirs and poultry industry, have been used as livestock feed (FAO, 1980). Optimal use of poultry manure in cattle diets appears to be about 20–30 percent. It is best fed dry following hammer milling, heating or ensiling, the addition of molasses increases palatability (Bhattacharya et al., 1971; Fontenot et al., 1963; Meenawat and Sharma, 1973; Preston et al., 1970; Tinnimit, 1977).

5. Non-protein nitrogen (NPN)

   One of the most limiting factors in feeding crop residues and by-products is their generally low protein content. NPN in the form of urea (46% N) can be used with such feeds to increase their low protein levels. Urea can be used in a number of ways (FAO, 1968, 1981a):

   • sprayed on pastures
   • sprayed on crop residues
   • as a liquid in a trough association with molasses
   • as a block lick
   • included in drinking water
   • as an additive in concentrates.

6.6 Examples of the use of alternative feed sources

The data presented in this section are intended to provide examples for the use of some alternative feeds with both grazed and penned cattle; they should be relevant when cattle are kept under coconut both on large plantations and by smallholders. Also, see Devendra (FAO, 1988, 1992a); FAO (1985, 1988a, 1992c), FAO/IAEA (1989), Jalaludin (1989); Jelan (1991), Lang (1993); Preston and Ogle (1993); Smith, 1993; Speedy and Sansoucy (1991) and Speedy and Pugliese (1992).

6.6.1 Banana

In many banana producing areas a considerable proportion of the crop not only fails to meet the quality standards required for export but is in excess of domestic demand and may be dumped and left to rot. In recent years efforts have been made to utilize reject bananas in livestock production. Observations made by Ruiz et al., (1974) indicated that it is not necessary, although often done, to chop or slice the green fruit for consumption by cattle. Isidor (1973) indicated that intake of banana is not a constraint in its use in animal feeding with total dry matter intake averaging 4.99 kg 100 kg^-1 liveweight day^-1. Herrera and Ruiz (1976) carried out various trials with penned cattle where bananas were substituted for molasses in a molasses, meat and bone meal, sugar cane bagasse and urea diet. Ruiz et al. (1974), Vohnout and Jimenez (1975) demonstrated the effects of banana supplementation on grazing cattle. Ruiz et al., (1974) indicated that an intake of only 5 kg fresh bananas head^-1 day^-1 was necessary to effect the 28 percent maximum improvement in steer weight gain compared to grazing alone. At a stocking rate of just over 10 ha^-1, heifers (average liveweight 240 kg) barely maintained their weight on Guinea grass pastures (Vohnout and Jimenez, 1975), whereas when supplemented with 10 kg fresh green bananas head^-1 day^-1 the daily weight gain was 0.3 kg animal^-1. Ruiz (1981) stressed that the additive effect of bananas allows for an increased output ha^-1 at a constant stocking rate, and the substitution effect allows for an increased stocking rate without changing the output ha^-1. Vohnout and Jimenez (1975) demonstrated that if forage yield is affected by climatic conditions output per area can be maintained close to its maximum by supplying approximately 1.5 kg bananas per 100 kg liveweight for each 10 percent of forage yield reduction. However, if protein becomes a limiting factor, significant departures from a linear relationship may be expected, as forage yield is reduced below 40 percent. The combination of those effects can produce an optimum level of beef production (see Figure 186). If protein is also provided as a supplement, output ha^-1 and animal^-1 should increase significantly. Work in Colombia (Anon., 1983a) demonstrated that because of low protein content digestibility is improved when bananas are supplemented with other feedstuffs. Best results were obtained with a compound feed consisting of 79.2 percent banana and 20.8 percent cereal meal. One ration for penned steers, fed at a recommended level of 21 kg of fresh fruit per 100 kg liveweight, provided weight gains of 1 kg day^-1.

In the Philippines reject bananas have given excellent results as a supplement to molasses-urea diets. McEvoy and Preston (1976) indicated that a daily intake of 3 kg fresh fruit increased the growth rate of 200 kg steers by 20 percent where the basal diet was sugar cane-urea and L. leucocephala. In the Seychelles, Preston (FAO, 1978a) demonstrated liveweight gains of 0.5–0.7 kg day^-1 where banana fruit was used to supplement basal diets of banana forage and sugar cane-urea (with Leucaena forage, meat and bone meal common to both). Wan Zahari et al. (1993) have begun investigations in Malaysia into the use of rejected banana fruit for animal feed.

Figure 186. - Effects of stocking rate and banana intake on beef production from Guinea grass (Vohnout and Jimenez, 1975).

Some work has been done, mainly in Ecuador, on the introduction of banana flour into ruminant diets (Spiro, 1973; Rihs et al., 1975).

Pseudostems and leaves are useful sources of roughage in many tropical countries. In Western Samoa whole banana plants have been used as drought feed for grazing cattle, the leaves having a high digestibility and being readily consumed by cattle (Ffoulkes and Preston, 1978b; Rivera-Brenes et al., 1959). When fed as the only supplement to Zebu steers receiving ad libitum molasses/urea, weight gains were between 0.5 and 0.7 kg day^-1 (Rowe and Preston, 1978). According to Preston (1979a), banana forage (pseudostems and leaves) supplemented with fresh Leucaena forage will support growth rates in cattle of about 0.5 kg day^-1. In Laos banana leaves and pseudostems are fed in September-October when there is a shortage of rice straw (FAO, 1994). Inclusion of banana leaf meal up to 40 percent in the forage ration has been shown to increase weight gains and feed efficiency of Zebu cattle and sheep (Garcia et al., 1973). Geoffry and Despois (1978) reporting on a feeding trial with goats noted a daily dry matter intake of 1.3 kg of banana leaves and 0.66 kg of pseudostem per 100 kg liveweight. In experiments where whole sugar cane and banana tops were compared as the source of roughage for cattle fed molasses ad libitum, weight gains were better for the cattle fed on banana forage than those fed on sugar cane (Fernandez and Preston, 1978; Salais et al., 1977). In another trial when growing bulls were fed with various mixtures of chopped whole sugar cane and banana forage (80% pseudostems, 20% leaves), replacement of 33 percent sugar cane with banana forage increased dry matter intake by about 46 percent and digestible dry matter intake by 103 percent. In Zanzibar, Reynolds and Lund (FAO, 1983b) demonstrated that banana forage could provide maintenance rations in the dry season (see Table 124) as well as reasonable growth in penned animals when combined with other feeds (see Table 125). In India, banana leaves were best fed with rice straw and some protein cake to avoid diarrhoea (FAO, 1981a).

Table 124. - Results of a dry season feeding trial (Reynolds and Lund, 1983)

Poyyamozhi and Kadirvel (1986), Subramanian et al. (1988) and Viswanathan et al. (1989) experimented with dried and ground banana stalk and noted that it was low in crude protein, high in fibre and was more acceptable when fed with concentrates. It was fed to goats as roughage at a level of 20 percent of total dry matter intake and to sheep at up to 50 percent (replacing para grass hay), with no adverse effects, but daily weight gains were low, in the range 27–39 g only. Viswanathan et al. concluded that the nutritive value of banana stalk is comparable with some of the common cereal straws and other crop residues such as paddy straw, sugarcane tops etc., and can be fed to sheep as a roughage (50–125 g day^-1) dry matter without apparent ill effect. Prigge et al. (1980), using whole banana plants, experimented with a chaff cutter and hammer mill to produce press cake for animal feed. Babatunde (1992) reviewed current use of banana and plantain for animal feeding and concluded that the area with most potential was in the use of the pseudostem or stalk and the leaves. Fomunyam (1992) examined the economics of banana and plantain use presenting examples from experience in Cameroon.

6.6.2 Cassava (Manihot esculenta)

Both cassava roots and leaves can be used in livestock feeding. Cassava can be used as a substitute for maize: in India a 50–100 percent replacement of barley with cassava increased the performance of growing calves (Mudgal and Sampath, 1972). It has been noted that cassava was a good energy source for dairy cows (Henke, 1919). It is efficiently used and supports improved milk yields and butter fat contents (Morimoto, 1950; Olaluku et al., 1971). Other tests have found that calves fed on cassava meal may perform better than those on maize meal, (Johnson et al., 1968). Devendra and Lee Kok Choo (1976) fed supplemental feeds based on cassava at the level of 25 percent of total dry matter intake (75% from Napier grass) to Kedah-Kelantan heifers and observed the best liveweight gain (318.7 g day^-1) and feed efficiency (12.4) at the cassava level of 30 percent of the concentrate feed (Khajarern et al., 1979).

Van Eys et al. (1987) used cassava meal as a supplement to napier grass diets for growing sheep and goats and found that animal performance and feed efficiency were maximized when cassava was fed at 30–40 percent of total dry matter. Tudor et al. (1985) concluded that cattle fed high energy diets based on dried cassava tubers can perform well. Muller et al. (1978) suggested that a concentrate consisting of 85 percent cassava root, 6 percent molasses, 8 percent urea and 1 percent mineral supplement could effectively augment tropical forages furnishing all the deficient nutrients required for optimum performance. Fresh cassava roots, when properly processed, may also serve as a basic calorie source for intensive animal feeding. The release of energy from cassava starch and of nitrogen from urea has a very close timing, this is of great importance for the maximum utilization of NPN compounds in cattle diets and for effecting significant savings in protein feedstuffs. In New Caledonia Bregeat (1986) suggested that young steers could be fattened with cassava roots and leucaena forage.

Table 125. - Results of a 40-day feeding trial with Friesian bulls and Jersey-Friesian crosses (Reynolds and Lund, 1983)

* Green fodder consisted of a chopped mixture of: Napier grass 50%, whole sugar cane 25%, banana stems-leaves 25%.

Moore (1976) was the first to report (from studies with beef cattle) the potential value of the aerial part of cassava in sugar cane based diets in Colombia (Devendra, 1981). The value of cassava leaves as a potential protein source has stimulated much research (Alvarez et al., 1977; Meyrelles et al., 1977). Cassava leaves have been fed to cattle in Africa (Guin and Andouard, 1910), in Cuba (Osvaldo, 1966), in Madagascar (Serres, 1969) and as a dry season supplement in Brazil (Gramacho, 1973). Montaldo (1977) reported on the use of the whole cassava plant as feed for ruminants in Venezuela. The present and potential utilization of cassava as livestock feed in Africa was reviewed by Hahn et al. (1992) and a worldwide overview was presented in FAO (1992c).

6.6.3 Cocoa pod husk

Cocoa pod husk has a low crude protein content, but is high in fibre, making it suitable for feeding ruminants. One of the problems with the utilization of cocoa pod husk (CPH) is its rapid deterioration. Feeding fresh pods therefore has disadvantages and in order to reduce the feeding of decomposing CPH, it is common to first dry the husks and grind them before feeding (FAO, 1981a). Pod meal has been fed without toxic effects to cattle in quantities up to 7 kg day^-1 and for dairy cows it seems to be comparable in value to corn-on-cob meal, although having a somewhat lower feed efficiency for beef cattle (FAO, 1981b). The utilization of CPH in ruminant diets has been reported by Aplin and Ellenberger (1927), Haines and Esheverrita (1955), Bateman and Fresillo (1967), Bateman and Larrangan (1966) and Pereira et al. (1984).

In Costa Rica, Bateman and Larrangan (1966) found that cows fed CPH produced a comparable yield of milk to those fed maize. Levels of up to 40–50 percent CPH were used in diets.

Llamosas et al. (1983) evaluated cocoa pods as a possible substitute for elephant grass in the ration for finishing confined steers. The elephant grass was gradually replaced with fresh pods and average liveweight gains of the steers showed no significant difference (P>0.05) as the pod content of the ration increased to a maximum of 100%. Average daily dry matter intake was similar and no harmful effects were noted in the animals.

Smith and Adegbola (1982) reported on two cattle-feeding experiments carried out in Nigeria to determine how much the relatively expensive sources of energy such as maize or sorghum could be replaced with cocoa pods without reducing performance to the point where it became uneconomical. Data from one of the trials is shown in Table 126. An inclusion rate up to 40 percent appeared to be acceptable and was recommended.

Table 126. - Composition of experimental diets and cattle response (Smith and Adegbola, 1982)

Means on the same row with different letters are significantly different (P<0.05).

More recently, Smith et al. (1988) fed cattle on 50 percent cocoa-pod-based diets and demonstrated a linear relationship between growth rate (y) and dietary level of cocoapod (×), y = 0.82 - 0.009 × with a negative correlation of r = -0.97. (see Figure 187). Concluding that the material needs to be treated in some way to improve performance on diets containing fairly high levels, forage supplementation with gliricidia was tried but did not improve cocoa-pod utilisation because of low intake of gliricidia. However, treatment with cocoa-pod ash (CPA), a caustic material, improved cocoa-pod digestibility and appeared to be as effective as NaOH solutions of equivalent alkalinity. It was suggested that more experimentation is required to determine optimum CPA solution concentration required and treatment ratios. Adebowale et al. (1991) reported on the use of ash of CPH as a source of alkali for upgrading agricultural residues such as straw and maize stover.

Figure 187. - Relationship between growth rate in cattle and dietary cocoa-pod level
(Smith et al., 1988).

In Malaysia Mohd. Sukri Hj. Idris and Rosmawati Othman (1981) reported on the performance of local Indian Dairy × Red Dane male calves fed with Napier grass at one percent liveweight and supplemented with three rations consisting of 30, 45 and 60 percent cocoa pod over a period of 187 days. The group that was offered 30 percent CPH ration showed the highest liveweight gain of 0.59 kg day^-1, followed by groups that were fed 45 and 60 percent (0.44 kg and 0.40 kg day^-1 liveweight gain respectively). The gross margin after deducting feed cost was highest in the group fed the 30 percent CPH ration, at M $ 336.38 per animal.

Bacon and Anselmi (1984) used CPH (35–50%) as the main constituent of a feedlot ration in a cocoa plantation in Sabah. 50% CPH was found to be too high. Other ingredients included 15% palm kernal cake, 5% maize, 35% mixed fodder, 5% molasses and 5% feed additives. An average weight gain of 0.54 kg day^-1 for steers and 0.38 kg day^-1 for buffaloes was obtained.

Later work reported by Jalaludin (1989) indicated that Kedah × Kelantan × Brahman steers on a 50 percent CPH ration gained 0.5 kg day^-1 compared with a control (no husk) group of only 0.26 kg day^-1. He also suggested that from studies with sheep CPH had potential as a feed for dairy cows.

In Malaysia, work with sheep demonstrated the value of molasses not only as a carrier of the by-product for utilization, but also in enhancing palatability (Devendra, 1978b). For more intensive use of CPH it was suggested that approximately 50–70 percent of the total diet of ruminants should be made up of CPH plus molasses.

Wong and Hassan (1988) compared fresh and dried cocoa pod husk in sheep feeding trials and noted no significant difference in intake.

6.6.4 Copra cake, coconut meal or poonac

Dried copra from which the oil has not been removed is used in some rural areas as a livestock feed supplement, although too expensive to use as a common food. Henman (1977) reported on its use as a replacement for more expensive imported manufactured feeds. Generally, it is the coconut meal produced after oil has been expelled from the copra which is used. In Sri Lanka coconut poonac has been particularly used for feeding dairy cattle (Silva, 1980).

Coconut meal is a relatively good source of energy and protein for use in livestock diets. It contains about 18 percent crude protein, 8 percent fat, 12 percent crude fibre and between 6.3 and 7 KJg^-1 metabolizable energy (Hutagalung, 1981), and is widely used as a supplement in coconut producing areas. For ruminants it is often fed intact after soaking or mixed with one or two other dietary ingredients, for example, with citrus meal in a 1:1 ratio. It can be used at fairly high levels, but animals which are not used to it are at first somewhat reluctant to feed on it. To overcome this difficulty, it should be introduced slowly in the feed. Lekule et al. (1986) concluded that for pigs the optimum level of inclusion was 10%. The maximum safe amount for dairy cows seems to be about 1.5–2 kg day^-1; larger amounts may result in tallowy butter. Beef cattle can consume much more without affecting carcass quality (FAO, 1981b).

In the Philippines, Castillo et al. (1961) reported significant increases in milk butterfat content of dairy cows fed high levels of coconut meal. Much higher levels (up to 92% in the total diet) have been reported to give between 0.39–0.40 kg daily liveweight gain in Local Indian Dairy calves in Malaysia (Devendra and Sivarajasingam, 1975). Abdullah Sani Ramli and Basery Mohamed (1982) showed that animal performance under coconuts in Malaysia was greatly improved by supplementation with copra cake. An improvement of 33.5 and 16.6 percent respectively was recorded for daily liveweight gain and final weight of Local Indian Dairy × Australian Milking Zebu yearling bulls when copra cake was supplemented at 20 percent of the daily feed intake. Lanting et al. (1989) reported on the use of copra meal as a concentrate supplement fed to milking cows with urea-treated rice straw-based rations in Philippines.

In Fiji, McIntyre (1973) fed supplements of coconut meal and molasses to grazing dairy cows and demonstrated significant increases in milk yields (see Table 127) with the coconut meal supplement producing top yields. Feeding of both supplements at prevailing prices was economic.

Galgal et al. (1993) reported that a growth rate of up to 0.93 kg day^-1 of liveweight can be achieved from Brahman-cross steers grazing native pastures under coconuts in Papua New Guinea when supplemented with copra expeller pellets at a rate of 30% of estimated daily dry matter intake. Supplementary feeding trials in Australia using copra meal and cotton seed meal have shown that even though copra meal contains only half the level of protein compared with cotton seed meal (21% vs. 40%), copra meal gave a similar growth response to the same amount of cotton seed meal (Hennessy et al., 1989; Gulbransen et al., 1990). In the study by Hennessy et al. (1989), weaner steers on low quality pasture given 500 g day^-1 of copra meal made similar liveweight gains to those fed 500 g day^-1 of cotton seed meal. When 30 g of urea was added to 500 g of copra meal, weaner weight gain was half way between those fed 500 and 1,000 g of copra meal (see Table 128).

Ehrlich et al. (1990) noted that although dairy cows found copra meal somewhat unpalatable it could be used to replace sorghum grain as a supplement. Copra meal intake averaged 2.5 and 3.1 kg day^-1 for cows offered 3 and 6 kg daily. Milk yields averaged 12.4 kg day^-1 for unsupplemented cows and 13.0 kg day^-1 for supplemented cows, with butterfat content of milk significantly increased where copra meal was fed.

Table 127. - The effect of coconut meal and molasses supplements on milk yields in Fiji
(McIntyre, 1973)

* Adjusted by covariance on previous lactation.
Grazing comprised Batiki blue grass (Ischaemum aristatum), Koronivia grass (B. humidicola) and Para grass (B. mutica).
Stocking rate of 1 cow ha^-1.
Coconut meal fed at a rate of 2.7 kg head^-1 in the first two weeks; from the third week amount fed depended on milk yield in previous week - the cows which produced 9.1 kg milk or more received 1.8 kg of coconut meal day^-1 plus an extra 0.9 kg for every extra 4.6 kg of milk.
Molasses fed at a rate of 4.1 kg head^-1 in the first two weeks; from the third week amount fed depended on milk yield in previous week - the cows which produced 9.1 kg milk or more received 2.7 kg of molasses day^-1 plus an extra 1.4 kg for every extra 4.6 kg of milk.

Table 128. - Effect of supplementary feeding on weight gain by beef cattle fed or grazed on low quality tropical pastures

6.6.5 Gliricidia and Leucaena

The advantages and beneficial effects of using various tree leaves as ruminant fodder were reported by Devendra (FAO, 1992, 1993a) and Speedy and Pugliese (1992) provided a global review of the use of fodder trees and shrubs as protein sources for livestock. Manidool (1985) reported on the use of Leucaena, Sesbania, Gliricidia, Erythrina and Pigeon Pea as protein supplements in rations consisting of low quality, fibrous crop residues. Withington et al. (1988) reviewed multi-purpose tree species for small-farm use.

Gliricidia (G. sepium/maculata)

As reported in Chapter 3 Gliricidia is a high protein (25–30%) deep rooted browse plant which is a valuable source of supplementary green feed during the dry season, with annual yields slightly lower than for Leucaena (Chadhokar, 1983a). However, it has a major advantage over Leucaena in that it can grow on quite acid soils. Feeding of Gliricidia has been reported from many countries (Skerman, 1977; Whyte et al., 1953), although in India Mahadevan (1956) indicated that cattle do not like G. maculata. In Sri Lanka it was reported to be quite palatable to both cattle and sheep, even when fed in large quantities over long periods (Chadhokar and Kantharaju, 1980; Chadhokar and Lecamwasam, 1982). In Zanzibar, Tanzania, Gliricidia was readily eaten by cattle in feeding trials (Lund, personal communication).

Chadhokar and Lecamwasam (1982) fed Jersey milking cows with a mixture of Gliricidia and B. brizantha with no adverse effects on their health or milk production (see Table 129). In another trial, Meuse-Rhine-Ijssel milking cows were fed a diet consisting of equal proportions of Gliricidia, paddy straw and 4.3 kg rice polish (10–12%) for two months. Milk yield and composition were comparable with milk produced by cows on a mixture of fresh grass and 4.3 kg concentrates (see Table 130). Similarly, Gliricidia with paddy straw in equal proportions and 1.9 kg rice polish was fed to crossbred heifers to compare with fresh grass and 1.9 kg concentrates. Average total weight gains of 123 and 136 kg animal^-1 respectively were obtained over a feeding period of 322 days with no adverse effect on breeding (Chadhokar, 1982). Jaysuriya (1984) noted that a supplement of gliricidia increased milk production in Surti buffaloes with either untreated or urea-treated rice straw (see Table 131). Gliricidia leaves are less palatable than leucaena, but livestock become accustomed to them (Manidool and Chantkam, 1986). Liyanage and Jayasundera (1988) reported that Gliricidia loppings mixed with B. miliiformis in a 50:50 ratio and fed to cross-bred heifers resulted in an average liveweight gain of 0.7 kg head^-1 day^-1. In a pasture/cattle coconut integrated system a mixture of Gliricidia and Leucaena fed to heifers at a rate of 6 kg along with pasture in the dry season, produced average liveweight gains of 0.3 kg head^-1 day^-1.

Table 129. - Effect of Gliricidia maculata in a mixture with Brachiaria brizantha on milk yield and its composition (Chadhokar and Lecamwasam, 1982).

Leucaena (L. leucocephala)

There are many reports of Leucaena being grown with grass species such as Guinea grass and elephant grass (see Figure 188), stocked at high rates and giving high daily liveweight gains (Foster and Blight, 1983; Furr, 1965; Hill, 1971; Jones, 1979; Partridge and Ranacou, 1974; Plucknett, 1970; Thomas and Addy, 1977). In dairy cows higher milk production and fat content from cows on Leucaena than from similar cows on pasture and concentrates was reported by Henke and Morita (1954).

Table 130. - Effect of various feeds on milk yield in MRY (Netherlands) cows
(Chadhokar, 1983a)

Table 131. - The performance of Surti buffaloes on straw based diets (Jaysuriya, 1984)

Figure 188.- Leucaena and elephant grass on a smallholder property in Vanuatu.

Although Leucaena is much relished by cattle there have been many reports of reduced levels of animal production because of the presence, in Leucaena, of the toxic amino acid mimosine (Blunt and Jones, 1977; Jones et al., 1976). However, work by Jones (1981a) has shown that ruminal metabolism of mimosine and its derivate DHP differs in places like Hawaii, India, Brazil and Philippines compared with Australia. Whereas in north Australia the feeding of high levels of Leucaena to ruminants has resulted in toxicity characterized by hair loss, excessive salivation, loss of appetite, low weight gains, enlarged thyroid glands, low serum thyroxine levels and death of newborn animals, no such effects have been reported from Hawaii or elsewhere. Probably if Leucaena is restricted to 30 percent or less of the diet the effects will be minimized. (See Chapter 3, Legumes vii).

A number of studies have been conducted to evaluate the efficiency of Leucaena as a supplement to sugar cane-based diets in ruminants. Work by Alvarez et al., (1977, 1978) indicated that Leucaena can effectively substitute a large proportion of rice polishings or other by-product feeds from grains (Hutagalung, 1981). When Leucaena was the only supplement for the sugar cane-urea-based diet, loss of weight and reduced milk production resulted. However, Sobale et al. (1978) indicated that the performance of growing bulls fed Leucaena ad libitum was superior to those fed on a restricted amount of forage. According to Hulman et al. (1978), the optimum rate of fresh Leucaena for maximising gain and molasses intake was 2 percent of liveweight daily. Fresh Leucaena fed at 2 percent of liveweight in association with molasses/urea has given liveweight gains of 0.7 to 0.8 kg day^-1 (Hulman, personal communication). Cheva Isarakul and Potikanond (1986), Devendra (1983), Mendoza (1983), Moog (1981) and Perez (1976) reported on the use of ipil-ipil leaves (Leucaena) as a supplement to rice straw (see Tables 132–133) and Perdok et al. (1983) used leucaena leaves with treated rice straw as feed for lactating Surti buffaloes. A supplement of leucaena leaves alone slightly increased milk yield (to 2.73 kg day^-1) while leucaena plus coconut cake substantially increased milk yield from 2.41 to 3.36 kg day^-1. (A supplement of Gliricidia increased milk yield to 2.60 kg day^-1). Devendra (1989a) presented some of the economic benefits of tree fodder supplementation, principally resulting from reduced cost of feeding.

Other trees which have been used as sources of fodder for ruminants (in addition to those mentioned on page 280) include:

Acacia albida
Acacia angustissima
Acacia auriculiformis
Acacia farnesiana
Acacia mangium (see Figure 189)
Acacia nilotica
Acacia tortilis
Albizia lebbek
Albizia amara
Albizia chinensis
Albizia falcataria
Albizia procera
Albizia siamea
Artocarpus heterophyllus
Bauhinia variegata
Calliandra calothyrsus (see Figure 190)
Cassia siamea
Dalbergia sissoo
Enterolobium cyclocarpum
Erythrina lithrosperma
Erythrina poeppigiana
Erythrina variegata
Eugenia jambos
Ficus bengalensis
Ficus glomerata
Hardwickia binnata
Inga laurina
Hibiscus sinensis
Leucaena diversifolia
Flamingia congesta
Samanea saman
Sesbania formosa
Sesbania grandiflora
Sesbania sesban (see Figure 191)
Tamarindus indica

(Awang and Taylor, 1993; Baggio and Heuveldop, 1984; Blair et al., 1988; Catchpoole and Blair, 1990; Ella et al., 1989; Evans and Rotar, 1987; FAO, 1987a; FAO, 1989, 1992b, 1992d; Ford, 1987; Gutteridge and Akkasaeng, 1985; Halim, 1991; Hashim, 1994; Holm , 1972; Jambulingam and Fernandes, 1986; Kang and Mulongoy, 1987; Liyanage and Jayasundera, 1987; Lindblad and Russo, 1986; Maasdorp and Gutteridge, 1986; Nair et al., 1984; Panjaitan et al., 1985; Panjaitan, 1988a, 1988b; Petaia and Ash, 1992; Seibert, 1987; Wiersum and Dirdjosoemarto, 1987.

Table 132. - Feedlot growth performance of bulls on rice straw and dry ipil-ipil leaves
(Perez, 1976)

I - 60% rice straw + 40% concentrate
II - 60% rice straw + 40% ipil-ipil leaves
III - 10% rice straw + 90% dry ipil-ipil leaves.

Table 133. - Feedlot performance of bulls on rice straw, concentrates and fresh ipil-ipil leaves (Perez, 1976).

I - 50% rice straw + 50% concentrate
II - 35% rice straw + 35% dry matter from ipil-ipil leaves + 30% concentrate.

Figure 189. - Acacia mangium growing in southern Vietnam.

Figure 190. - Calliandra calothyrsus growing in trial plots in Vanuatu.

Figure 191. - Sesbania sesban nursery seedling in W. Samoa.

6.6.6 Oil Palm (Elais guineensis)

Oil palm products which can be used as livestock feed include: crude palm oil, palm oil mill effluent (POME), palm kernel cake, oil palm sludge and palm press fibre (Devendra, 1981; Devendra et al., 1981). Palm oil was used as a supplement at the rate of 2 to 8 percent with an increased milk yield and milk fat content (Storrey et al., 1968). Under intensive management systems, supplementation of palm kernel meal to grass-molasses-based diets improves the daily liveweight gains of growing Zebu-Holstein dairy bulls compared with those receiving either grass or grass-molasses mixtures (Camoens, 1978). As a maintenance ration for growing dairy bulls, palm press fibre could be used as a roughage to substitute for grass when supplemented with molasses, urea, minerals and vitamins. The main potential of press fibre lies in its use in feedlot type operations within or close to oil palm plantations. Muller (1978) reported on the use of palm press fibre at the rate of 3 percent in combination with 78 percent pineapple waste, 10 percent poultry litter, 5 percent molasses and 4 percent other agro-industrial wastes and chemicals. Idris (1981b) concluded that a simple feedlot ration containing 30 percent palm press fibre was the most economical for beef steers. Pillai and Tan (1976) indicated improved liveweight gains when feeding palm oil sludge to cattle. Wan Mohamed et al. (1989) evaluated various oil palm by-products in prime lamb production, Mustaffa-Babjee et al. (1984) used palm kernel cake as a sole ration for fattening beef cattle (daily gains of 0.7 kg), Jelan et al. (1991) used palm kernel cake on a smallholder model feedlot over a three-year period, Halim (1993) demonstrated that early supplementation with palm kernel cake was important for optimum benefits on a smallholder beef farm and Vadiveloo (1989) reported on the feeding to goats of leucaena supplemented with dehydrated palm oil mill effluent (POME), the major by-product from the extraction of palm oil, as an energy supplement. As a concentrate supplement for growing sheep fed low quality forage, palm kernel cake and POME are best used in 50 : 50 combination (Boer and Sanchez, 1988). Dalzell (1977) studied the utilization as animal feeds of the by-products of palm oil production with buffalo and cattle either grazed or intensively stall fed over a three year period. The economic analysis of the stall fed animals provided a very positive indication that palm oil products should be utilized close to the palm oil mill (Devendra, 1981).

Miyashige (1989) has emphasized that ongoing research in Malaysia is focusing on improving the digestibility of palm press fibre and developing an efficient and economical feeding menu for ruminants using oil palm by-products as basal feedstuffs together with other by-products such as molasses. Experiments have also been conducted to process oil palm trunk as ruminant feed (Oshio et al., 1989; Raman et al., 1990) and Abu Hassan et al. (1993) reported on the use of oil palm frond silage as a source of roughage for milk production in Sahiwal-friesian cows.

Wong and Wan Zahari (1992) suggested levels of inclusion (see Table 134) of oil palm by-products in ruminant rations.

Table 134.- Suggested levels (%) of inclusion of oil palm by-products in ruminant rations (Wong and Wan Zahari, 1992)