PART II

CHAPTER 7
POTENTIAL CONTRIBUTION AND USE OF AGRO-INDUSTRIAL BY-PRODUCTS IN ANIMAL FEEDING

by

M. Chenost and L. Mayer

FAO, Rome, Italy

Resumen

Este documento evalúa las grandes existencias mundiales de los principales subproductos y desperdicios de la agricultura e industrias afines, excluyendo, sin embargo, los materiales que poseen un alto contenido de fibra y los desperdicios animales y de las poblaciones. Expone también las propiedades alimentarias de esos productos y las posibilidades de utilizarlos en los sistemas de alimentación animal.

Los materiales procedentes de la producción o elaboración del azúcar, cereales, almidón, frutas y hortalizas, aceite, bebidas aromáticas y productos pecuarios se presentan en cuatro grupos, a saber: fuentes principalmente de energía, proteínas, energía y proteínas e ingredientes diversos de la alimentación.

Las existencias potenciales de subproductos (por región geográfica y por principales países productores) se calculan por los datos de producción o utilización de la materia prima. Las cantidades teóricamente disponibles pueden diferir mucho de las realmente utilizables para la alimentación animal, a causa de otros usos posibles y de limitaciones como la iregularidad de la oferta y su volumen pequeño o excesivo; los gastos elevados de transporte en relación con el valor del producto; la rapidez de su deterioro y la falta de medios de conservación.

El Grupo I comprende la melaza (de caña y de remolacha), la pulpa de la remolacha azucarera, los frutos cítricos y el café, las bananas de desecho, el salvado de piña y los subproductos de la elaboración de la yuca y la patata que generalmente son ricos en carbohidratos fermentables, y escasos en fibra leñosa y proteínas, ofreciendo por ello, una buena oportunidad para la utilización del nitrógeno no proteínico.

Se mencionan las formas tradicionales de aprovechamiento. Se examinan los recientes trabajos de investigación y desarrollo sobre su utilización como base de nuevos sistemas de alimentación intensiva, particularmente para rumiantes y cerdos. Se llama también la atención acerca de combinaciones idóneas con otros subproductos, resíduos y desperdicios (forrajes voluminosos, subproductos de cereales, excrementos de animales) que pueden hacer a la alimentación en gran parte independiente de las raciones forrajeras clásicas.

Grupo II: Las proteínas de semillas oleaginosas, que constituyen la inmensa mayoría de las proteínas complementarias en el comercio mundial, muestran una distribución regional extremadamente desigual. Los escasos suministros nacionales de tortas oleaginosas, harina de pescado y harina de carne se destinan con frecuencia a animales monogástricos. Se examina la idoneidad específica de las proteínas de semillas oleaginosas y de origen animal para diferentes tipos de alimentación. Se llama la atención acerca de las limitaciones que ocasionan el contenido de aminoácidos, fibra y cenizas, las propiedades de almacenamiento, la contaminación por micotoxinas y la solubilidad en el rumen.

Las legumbres se cultivan extensamente y las semillas desechadas pueden suministrar toda clase de proteínas en la alimentación para la cría y el engorde del ganado porcino. Pueden vencerse los problemas que plantean los inhibidores de la tripsina y las deficiencias de aminoácidos esenciales. El follaje de las leguminosas, las partes superiores de la yuca y la batata, etc., pueden emplearse como fuente única de proteínas para los rumiantes. Se han dado ejemplos de la preparación de proteínas de hojas para los no rumiantes. Las proteínas unicelulares procedentes de la fermentación de los carbohidratos residuales por la levadura y los microhongos pueden constituir la parte principal de proteínas en la alimentación de terneros, cerdos y aves de corral.

Grupo III: Los subproductos procedentes de la molienda del trigo, arroz y maíz se hallan muy extendidos y pueden suministrar proporciones muy importantes de energía y proteínas en todos los tipos de alimentación de los animales. Las limitaciones generales son la calidad de la fibra y las proteínas. Las fracciones de gérmenes y el salvado “cilindro” de arroz han resultado buenas fuentes de precursores de glucosa y de proteínas suplementarias, respectivamente, en raciones de elevado contenido de azúcar y de nitrógeno no proteínico para rumiantes.

Las proteínas y la energía contenidas en los cereales utilizados en destilerías y fábricas de cerveza pueden aprovecharse muy bien por el rumiante a causa de la fibra asociada.

El suero con que tradicionalmente se alimenta a terneros y cerdos puede proporcionar también lactosa y proteínas de elevada calidad al rumiante lactante.

En el Grupo IV se estudian las posibilidades y el valor de algunos desperdicios procedentes de la elaboración de frutas y hortalizas como ingredientes diversos de la alimentación.

Es necesario que la investigación y el desarrollo comprendan el establecimiento sistemático de parámetros alimentarios y las relaciones de entrada y salida respecto a los subproductos utilizados en cuantías diversas en los sistemas de alimentación.

Introduction

The recent food crisis has affected the feed supply universally. However, if one compares the resulting problems for the developing with those of the technically more advanced world, the latter appear to be relatively minor. There, the price ratio of concentrate feed to animal products has narrowed, pointing to the need for some partial substitution of traditional concentrates based on cereal grains.

In contrast, in the developing world, the priority for food production on arable land has made the outlook for feed very serious. At the same time, in most countries, the demand for animal protein is rising steadily (FAO, 1976) owing to the high rates of growth of both the human population and the economy. Rising demands are often met by higher imports of both food and feed, that leading in their turn to ever-increasing trade balance deficits. To counteract such trends, it is urgent to increase animal production from domestic resources. Vast animal resources do, in fact, exist already. Approximately 70% of the world's cattle and buffaloes, and about 60% of its sheep, goats and pigs, are kept in developing countries. In general, however, these animals are so unproductive that they meet only one quarter of these countries' demand for milk and less than half their demand for meat.

A key factor limiting animal production is inadequate nutrition, which cannot be improved by moving on to new land, since land suitable for growing food is becoming scarcer and scarcer, and food production should have clear priority over feed production. Generally, natural grazing land is of low productivity and is either over-or understocked; the development of improved pasture and planted forage is a very slow process, because of limitations imposed by technical, economic and human factors; and supplies of conventional concentrate feeds based on cereal grain are either inadequate or too expensive.

Where, then, can prospects for the expansion of feed resources be found?

Clearly, there is need to integrate animal production into crop production, and the allied processing industries, to ensure that animals take a complementary -- rather than competitive -- part with man in meeting food/feed requirements. Present integration could be greatly improved by the better use of existing resources. This paper is particularly concerned with agro-industrial by-products, the need for whose better utilization is also emphasized by our increasing awareness of environmental pollution problems. The costs of waste disposal might be considerably reduced, or even recouped, through subsequent profits from animal production.¹ The animals that offer the best scope for production using by-products are the ruminants; they can cope with the most varied range of by-products, and their feed requirements are the least competitive with those of man.

This paper seeks to assess the worldwide availability or potential availability of some major by-products and wastes from agro-industries, and to describe some characteristics of such materials and the possibilities for their use in animal feeding systems. Some of the innumerable materials of more local character are dealt with elsewhere in this volume.

Classification of available materials and some constraints on their use

The products dealt with below are those that originate in agricultural cropping and the industrial processing of sugar, cereal grains, starch, fruits and vegetables, oil, beverages and animal products (milk, meat, fish). Excluded are poor-quality roughages and animal wastes since these are the subjects of Parts I, III and IV of this book.

Classification

This very large and heterogenous group of by-products and wastes can be divided into four categories according to their nutrient type (energy/protein), their role in feeding systems or rations, and their quantitative contribution, Byproducts of the first group mainly provide energy and may form the basis or a major part of rations; they derive from sugar cane, sugar beet, citrus fruits, cassava and sweet potatoes, banana, coffee and pineapple. Materials of the second group are mainly used as a source of supplementary protein: oilseed cakes and meals, by-products of animal origin, pulses, aerial parts of cassava and sweet potatoes, forage legumes, leaf protein and single-cell proteins. The products presented in the third group (cereal milling by-products, brewers' and distillers' grains and whey) generally occupy a nutritionally intermediary position between those of the first and second groups. Wastes from the processing of some fruits and vegetables are classified as miscellaneous ration ingredients in the fourth group.

¹ For the future much more could also be done concerning cropping policies to ensure production of crops with the highest overall return per unit area.

To avoid confusion over the terminology “conventional”, “unconventional” and “new feed resources” It must be pointed out that most of the products dealt with are in fact not new; what is new or unconventional, is the role they may play in feeding systems or the different combinations and different levels envisaged for their use.

Constraints

In this paper the theoretical availability of products is calculated from production data of primary commodities or from processing or utilization data, in order to indicate the order of magnitude of the various commodities' possible contribution to feed supply. Especially for developing countries, where precise statistics are scarce, the error involved in such estimates may be very high. In addition, potential availability may differ very greatly from what is, or actually can be, used for feeding, and this for many reasons:

1. lack of economic incentives for animal production, whence alternative use of the products in question, or dumping;

2. lack of knowledge about by-product utilization in animal feeding;

3. uneven local distribution, in conjunction with an underdeveloped physical infrastructure, often combined with fluctuating and unreliable supply;

4. lack of facilities for sufficiently large-scale processing connected with an efficient marketing system;

5. chemical composition or physical properties limiting the suitability of the materials for feed, e.g.:

   • high moisture content, prohibiting transport;

   • high sugar or oil content, causing rapid spoilage by fermentation or oxydation;

   • variations in composition which make ration formulation difficult;

   • poor storage qualities;

   • contamination by mycotoxins;

   • important nutrient deficiencies that must be overcome by complementation or supplementation;

   • toxic content of certain species.

Group I - Energy Sources

Availability and nutritional considerations

Tables 7.1 to 7.4 provide estimates for the world-wide availability of by-products from the processing of sugar, banana, pineapple, citrus and coffee by regions and selected countries. Table 7.5 provides production figures for roots and tubers which, although primary products, are increasingly used as substitutes for cereals in animal feeding in countries not producing cereals.

The composition of the same by-products is given in Table 7.6. The products shown are generally rich in readily fermentable carbohydrate and low in both N and protein. The carbohydrate fraction of molasses consists exclusively of soluble sugars; unripe banana rejects, roots and tubers are especially rich in starch Pulps of beet, citrus and coffee and pineapple bran contain varying amounts of soluble sugars, pectins and cellulose, with a generally moderate degree of lignification; they are more suitable for feeding ruminants than monogastrics, whereas the two former are about equally useful for both ruminants and pigs.

Their high content of readily available energy makes these materials good substrates for the microbial synthesis of protein from NPN in the rumen. Matching ammonia release from NPN with the extraction rates of energy from the carbohydrate source helps to optimize the yield of microbial protein. The most frequent source of NPN is urea, but ammonia, in gaseous or liquid form, can also be used. Its combination with pectin-rich products seems particularly appropriate, since pectins are good binders of ammoniacal nitrogen. The value of any supplementary protein in ruminant diets is largely determined by the degree to which it escapes rumen fermentation. This will be discussed with particular reference to high incorporation rates of NPN into rations.

In high-sugar diets propionic acid, the main precursor of glucose in the process of synthesis, is produced at relatively low levels, and this may easily limit production. Such shortfalls, however, are probably overcome best by inclusion of starchy products in the diet.

Some conventional uses

Molasses has long been used in the feed industry as a binder: rates of incorporation have however seldom exceeded some 10%. When used as an energy supplement to poor-quality roughages at rates up to 20%, it has usually been mixed with urea

¹ Source: FAO Production Yearbook, 1974.
² Approximate figures.

¹ Source: FAO Production Yearbook, 1974.
² Basis: Approx. 20% of production.
³ Basis: 50% of canned products.

¹ Oranges; tangerines; lemons and limes; grapefruits; others.
² Source: FAO Production Yearbook, 1975.
³ Moisture 8%; basis 7.3% fruit for processing (Hendrikson and Kesterson, 1965).
⁴ Moisture 28%; basis 3.78% fruit for processing (Hendrikson and Kesterson, 1965).

¹ Source: FAO Production Yearbook, 1974.
² Source: Bressani, 1974.

Source: FAO Production Yearbook, 1974.

¹ Le Dividich et al., 1976.
² Depending on maturity (first figure: green fruit).
³ Hendrikson and Kesterson, 1965; including citrus pulp, grapefruit pulp, citrus meal.
⁴ Bressani, 1974.
⁵ Hawaii.
⁶ Henry and Morrisson, 1950.
⁷ 10% of total root weight (FAO, 1973).
⁸ 4. It (Dickey et al., 1966), 5.4t (Eskew et al., 1948) per ton of starch produced.

Beet pulp (fresh, ensiled or dehydrated), is traditionally used in temperate countries for winter feeding. It is easily digestible and may be included up to some 80 % in ruminant diets without causing digestive troubles. At such levels, ruminal VFA may be expected to consist of some 70% of acetic acid. It is thus a good choice of feed for milking cows. On the other hand, its use at high levels in Intensive beef-cattle finishing diets is likely to call for supplementation with some cereal grain. Kercher et al. (1965) have, however, obtained good results with diets in which beet pulp replaced all of the cereal in the ration.

Citrus pulp and pineapple bran have been used in animal feeding for several decados, They are used commercially in Florida and Hawaii, after dehydration, as partial substitutes for cereals in concentrate rations for ruminants.

Culled potatoes are commonly used as sources of starch for both pigs and ruminants in Europe and in some parts of the United States. They are fed raw (whole or chopped), ensiled, sun-dried or cooked.

Dehydrated cassava is extensively used in the EEC countries, where 2 million tons a year are consumed by the feed industry.

Potential uses in new feeding systems

The following examples illustrate different ways in which these products can be used as basic components of rations in new feeding systems, either for specific periods in the life cycle of animals or for a better stratification of livestock enterprises.

The work of Preston and his co-workers in Cuba has shown that molasses can be used as a basic component of intensive diets for ruminants.

A review by Preston (1972) shows how molasses is fed in liquid form separately from the roughage fraction, which is fed restrictively (either cut or grazed) at a level of 2% (fresh basis) of live weight. Molasses may cover 70 to 80% of the energy requirements of growing and fattening beef cattle and 50 to 60% of the energy required by low-yielding milking cows. Up to about 60% of nitrogen may be provided by urea mixed with the molasses. With rising levels of production, however, it must be realized that supplementation of true protein in the “protected” form becomes increasingly important. Among the sources tested are meals of fish, meat, cottonseed, rapeseed, groundnut cake and torula yeast.

The mineral balance is also likely to require correction, particularly with regard to excess K and insufficiency of S to ensure good microbial synthesis.

Positive results have already been obtained in beef fattening operations, for example in Kenya (Redfern, Squire and Creek, 1974). As long as the price of molasses was low, it was economically feasible to feed high levels even when feed conversion ratios were relatively unfavourable.

Molasses can also be the basic component of pig diets (Brooks et al., 1967; Preston et al., 1967; Le Dividich et al., 1974). Rates of incorporation of 40 to 70% have been proposed for growing/finishing pigs. The diarrhea that occurs when molasses is fed at higher levels not having been fully explained, it is not recommended to exceed about 40%. The diarrhea problem is more serious with piglets following weaning at about 5 weeks of age; their diets should therefore not contain more than 20% molasses.

Bananas and tropical roots and tubers can be used as a basis for intensive rations for pigs and ruminants. Research on the use of bananas in the intensive feeding of pigs was initiated during the last decade, particularly in Ecuador, Costa Rica, Colombia and the French West Indies. Because reject bananas suddenly become available in large quantities, preservation is necessary. The principal results (Hernandez and Maner, 1967; Le Dividich and Canope, 1970) show that despite limitations in energy intake, the banana (fresh or ensiled) can be the basic component of the diet of growing/finishing pigs and gestating sows; a saving of concentrates of 50 and 75% respectively can be achieved when it is fed ad libitum (Lehman's method). Generally 1 ton of bananas meets the energy requirements of a pig from birth to slaughter (95 kg), or those of a sow during a gestation period. The same applies to tropical roots and tubers (cassava, yam, sweet potatoes).

Research on the use of whole green bananas as the principal feed for fattening cattle has recently been initiated in the Philippines (Preston, 1975) and at CATIE, Costa Rica. Preliminary results show very good banana intakes and satisfactory gains, even without supplementation of protein but with free access to a liquid mixture of molasses containing 10% urea and to a mineral mixture and with 2 hours' daily grazing of a grass legume pasture.

Experiments with feeding high proportions of banana to goats are also in progress in the French West Indies (Chenost et al., 1976), It would appear that good results are being obtained with growing male goats (ADG, 142 g - FCR, 5:3) and with milking goats fed on silage consisting of 74% banana, 22% wheat bran, 3% bagasse and 1% urea. The addition of some molasses during feeding encourages the intake of such a compound silage. Basic research is still necessary in feeding bananas to ruminants to obtain a better understanding of their efficiency of utilization.

Other research work on the use of bananas and roots and tubers in ruminant feeding has mainly been concerned with the substitution of cereal grain supplements to conventional forage-based diets. Most studies carried out have used these products in dehydrated form, but increasingly they are being investigated either fresh or ensiled, as “succulent concentrates” that could even replace part of the green fodders. This is true of citrus pulp and discarded citrus fruits in Israel and Florida, with coffee pulp in Latin America, with bananas in the French West Indies and Costa Rica, and with pineapple bran in Hawaii, Taiwan and South Africa.

Two main reasons have led research workers to search for new types of diets: the need for simple feeding methods and the seasonality with which many feeds are available. The latter implies the need for preservation, which is increasingly done by ensiling rather than dehydration. Depending on local availability, complete rations may be compounded by using a number of byproducts in varying proportions. Ensiling of such mixtures could help to overcome general ensilation problems caused by high moisture content and an excess of rapidly fermenting sugars. Feeding systems based on such rations could be developed independently of conventional forage-based systems and could also help to solve difficulties during critical periods of nutrient supply.

The following examples show how better use may be made of these types of materials at the small farm level. Bressanl et al. (1975) have underlined the importance of improving the technology of ensiling coffee pulp and, more particularly, ways and means of achieving better intakes by ensiling it with other ingredients such as roughages, molasses and other additives.

Pineapple bran can constitute 65 to 75% of the diet of growing and finishing cattle when ensiled with such other ingredients as molasses, rice straw, ground maize, dried sweet potato chips and urea in various combinations and proportions (O'Donovan et al., 1972). In experimental research being conducted in the French West Indies, various combinations of pineapple bran, bagasse, wheat bran, molasses and urea are ensiled to yield a complete feed.

In Singapore, Muller et al, (1975) have been working on the recycling of animal wastes, using cassava meal and poultry litter, cattle manure, cattle litter and poultry manure in various combinations and proportions. Cassava meal, ranging from 18 to 80% of the silage, supplies most of the dietary energy, whereas the animal wastes chiefly supply protein, fibre and ash. The overall performance of calves fed on these silages from 55 kg to slaughter weight was satisfactory (daily gain 0.75 to 1.2 kg, and FCR about 7:5).

Group II: Protein sources

Some nutritional considerations

In assessing the potential use of source protein supplies and characterizing their specific suitability, due regard must be paid to the requirements (in quantity and quality) of ruminants and non-ruminants. The latter require all their nitrogen through dietary amino acids (AA), while protein quality is mainly a function of digestibility and the pattern of the digested AA. In contrast to this, advantage should be taken whenever possible of the ruminants' capacity to synthesize microbial protein from appropriate sources of NPN and energy; dietary protein may constitute only a part of the AA required at the metabolic level, and this part should be optimized on both a physiological and an economic basis. In a growing or lactating animal, for instance, the contribution from rumen synthesis hardly exceeds around 60%. The difference has to be made up through the diet, the AA being required in ratios similar to those of monogastrics. The usefulness of a dietary protein source for taking over this function depends, in addition to its digestibility and AA composition, on its solubility in the rumen fluid. The significance of solubility seems in turn to be influenced by protein feeding levels. Thus, low solubility means little degradation of the protein by rumen microbes and good availability of the AA for absorption in the intestines. Potential savings in dietary protein input depend largely on the combination of these three factors in a given protein source, so that they may conveniently form the basis for grading the usefulness of proteins as dietary supplements.

Oilcakes and meals

Oilcake or meal -- a by-product from oilseed crushing -- is by far the largest source of supplementary protein on the world market. Table 7.7 shows 1974 production and exports of major oilseed cakes (in protein equivalents) and fish meal, by commodity and economic region. The oilseed industry as a whole is growing fast, with soybean taking the leading position, owing to greater demand for protein than for oil during the past decade. Some protein sources, such as cottonseed, depend on the production trend of the main product. The developing countries' share in total production is 27%, whereas their share of exports is about 35%. The main trade flow is towards Western Europe and Japan, leading to a very unequal regional distribution and a great shortage of this protein source over most areas.

The order of production of oilseeds and their importance for animal production in the developing countries differs from that in the world as a whole. Table 7.8 provides some production utilization data, as well as principal producers by commodity. It is apparent that in the developing world in the aggregate, groundnuts, coconuts, cottonseed, and oil palm seeds come before soybeans in their quantitative contribution to animal production. Seeds or nuts from sunflower, rape/mustard, olive, castor, linum, safflower, tung and rubber tree, although quantitatively less important on the whole, may well be of major importance on a regional or local basis.

Some properties and use of oilseed proteins and fishmeal

The use of the two nutritionally most satisfactory protein sources -- fishmeal and soybean meal -- is limited in most developing countries by availability or price. Cottonseed is principally a source of protein for ruminants. Usage for non-ruminants is limited for two reasons: (i) cake from undecorticated seeds usually contains less than 30% CP and over 20% fibre (as opposed to 40 to 45% CP and less than 10% fibre from decorticated seeds); (ii) during normal processing, endogenous substances of the group of polyphenolic gossypoi pigments and cyclopropenoid fatty acids are not eliminated. With regard to AA content, lysine (content of which is about two thirds of that of soybean meal) may be critical in broller rations. Cottonseed protein, being relatively soluble, is expected to perform poorly to moderately if it supplies only part of the nitrogen in high energy/NPN diets. However, cottonseed meal performed well when used together with maize germ in a molasses-based beef cattle ration, and all nitrogen was given in the form of protein and not NPN (Redfern, 1973). Preston (1973) also reported that daily gains of 900 g could be obtained from a diet of 60% molasses and 20% whole cottonseed. Cakes and meals from groundnut, sesame, safflower and sunflower, containing no endogenous toxic factors of practical significance, are generally satisfactory sources of supplementary protein for all classes of livestock. With non-ruminants, however, major limitations are set by fibre and ash in products from undecorticated kernels and unhulled seeds. With regard to requirements of monogastrics, groundnut protein is deficient in lysine, methionine and cystine. Sesame, sunflower and safflower proteins, being richer in the S-containing AA but lower in lysine as compared with soybean protein, are suitable complements for the latter. The solubility of these proteins is similar to that of cottonseed, but variations may be caused by different manufacturing processes. A major constraint to the use of groundnut cake is its frequent contamination with aflatoxins, metabolites of fungl, in particular Aspergillus flavus. Groundnuts are especially prone to attack by such fungi, but cottonseed, coconut and oil palm kernels also suffer. Rapeseed protein has a very favourable AA composition, its content of cystine and methionine being higher than that of soybean and its lysine content being equal. Its use with non-ruminants, however, commonly causes palatability, nutritional goitre, growth rate depression and reproductive performance problems that are associated with the presence of substances formed under the influence of the enzyme myrosinase from glucosinolate and isothiocyanate precursors. Attempts are being made in Canada and elsewhere to develop new varieties with lower glucosinolate and crucic acid contents. Heat treatment destroys the enzymes but may also affect protein quality. Recommended rapeseed meal maxima in the ration are: for young pigs, zero; for growing/finishing pigs 5%; for layers 3%; for broilers 10%. Since myrosinase is inactivated in the rumen, ruminants are only affected by initial palatability. However, rapeseed meal is highly soluble and may be expected to perform poorly as a protein supplement is high energy/NPN diets; this has been confirmed experimentally by Preston and Molina (1972). In similar experiments, better results obtained with rapeseed expeller were attributed to the lower solubility of the expeller, owing to the heating received.

¹ Gross weight.

Source: FAO Commodity Review and Outlook, 1975–76.

Copra and palm kernel meals have major limitations as main sources of supplementary protein, their protein and fibre contents being about 20% and 10% respectively. Protein digestibility in both sources is generally relatively low; owing to the frequent overheating of copra meal during processing and the gritty nature of palm kernel meal. According to McDowell and Hernandez (1975), copra meal for intensive beef production in the tropics met with limited success. In the Philippines, satisfactory results were obtained with copra in a banana/urea-based diet (Preston, 1975).

Protein of animal origin

The data given in Table 7.9 certainly over-estimate the present availability of meat, bone and blood meals in the developing countries, where abattoirs are little developed or have no facilities for the processing of by-products. However, the data do not include estimates of recoverable by-products from fallen animals, which may contribute considerably to this source. Thus, in India, fallen cattle are currently estimated at 29 million head annually (Min. Agric. New Delhi, 1974). Meat meal is a well appreciated source of supplementary protein, both for monogastric and ruminant diets. Blood meal is an excellent source of lysine, but its frequent bad odour limits its use to levels not exceeding 10% of the ration. The present problem with these meals is supply in sufficient quantities, but a great deal of scope exists for the development of this source.

Hair and feather meals are rich in unsaturated fatty acids and therefore require stabilization with antioxydants for any period of storage. If they are to provide, for example, one-third of the N required by young chicks, methlonine and lysine should be supplemented. Hydrolized leather meal, which contains about 65% of CP, can be included in broiler diets up to 8%; at higher levels the high chromium content may constitute a hazard.

Fish hydrolysates and silages have been developed in Japan and the Scandinavian countries respectively (James, 1973). This is promising for small-scale production in developing countries, as an alternative to high-cost fishmeal production. Liquid hydrolysates can either be spray-dried or absorbed by such materials as rice bran. Silage, after acidification with a mixture of organic and inorganic acids, can be mixed with other materials, as is done in Denmark with grass meal. Feeding of raw fish is still rare, but in Peru, rations consisting of 50% sweet potatoes and 50% raw fish have been fed to finishing pigs. The presence of thiaminase and transmissions of oil flavour into the meat are, however, two factors limiting its wider application. Sludge from processing of meat and fish is often viewed as a waste disposal problem: according to Grant (1976), 10 000 δ of protein and fat are lost annually in the United Kingdom alone. The recovered protein (by floculation and extraction through ion exchange resin) has a composition similar to that of herring meal and can be incorporated at a rate of 8.5% in pullet diets (Hersted and Hvidsen, 1973).

¹ Based on data taken from TPI (1975).

Protein as a by-product from the manufacture of starch

So-called “fruit water” contains 5% of sollds, almost half of which is CP. Coagulation by heating or by ultracentrifugation (Oosten, 1976) can isolate a good end product, with 8% water and a CP content in DM of 75 to 85%.

Pulses are grown throughout the world, mainly for human food; in the temperate regions, however, a large part is also fed to livestock, in particular in substitution for oilseed meals. The main pulses are dry beans, dry peas, chickpeas, dry broad beans, plgeon peas, cowpeas and lentils. World production by region and commodity is given in Table 7.10. Although it may be assumed that in the developing countries nearly all of it is for human consumption, any local surpiuses and rejects from processing as food (of the order of 5 to 10% of the total supplies excluding husks) could be available for livestock. Indian statistics (Min. Agric., New Delhi, 1974) Indicate that for 1971-72, the estimated availability for livestock was 1.33 million, representing about 10% of total production.

Major pulse species were recently reviewed by Maner (1973) as regards their climatic requirements, nutritional characteristics and role in livestock rations, with emphasis however on non-ruminants. Deficiencies of AA relative to non-ruminants' requirements are common first for methionine, and then for cystine and tryptophan. Except the field bean, most pulse seeds contain trypsin, an inhibitor of protein synthsis in muscles. Cooking of the seeds and supplementation with the limiting AA generally permits incorporation into diets at levels that provide almost all the protein and a major part of the energy from this source. Unheated pulse protein is of relatively high solubility in the rumen. Thus McMeniman (1975) found that 22% of unheated dietary horse bean protein reached the small intestine of sheep undegraded, as opposed to 32% following heat treatment.

Protein from forage

In a number of technically advanced countries progress is being made, using sophisticated technology, in the production of leaf protein from lucerne grown as a main crop. More recently, technologies have been developed at a much lower level, in Europe and elsewhere, for the mechanical separation of fibres and juice of green forage matter to provide a new range of feeds for ruminants and pigs. In the tropics, several examples already exist of how protein from forage byproducts can increase scarce protein supplies. Some of the sources are leaves or aerial parts of cassava, sweet potatoes, Leucaena spp., ramie (Bohemeria nivea), etc., as well as forage legumes grown on the intercropped land of tree plantations.

There are, of course, several ways of exploiting this underutilized potential, but if little or no processing is a necessary condition, then obviously ruminants offer the greatest scope. This also fits into the general situation, where scarce supplies of protein are earmarked for poultry and pigs.

Cassava leaves are a very rich source of protein with a satisfactory AA balance. The protein content, however, varies greatly with age, young leaves containing over 30% CP in the DM, young aerial parts about 20% and mature leaves some 16%. Methionine and tryptophane are the first two limiting AAs. The yields of roots and aerial parts are about equal. Assuming two crops per year, sweet potatoes yield about 30 t of tubers and 25 t of follage. With yearling cattle, Backer (1976) achieved an intake of 2.45 kg DM/100 kg LW and a daily gain of 660 g. When crude protein intake remained at 0.3 kg/100 kg LW, and cull grade tubers and urea were supplemented, daily gain rose to 825 g.

For the more sophisticated utilization of aerial parts from cassava, the original Pro-Xan process for leaf protein extraction has been adapted to cassava forage (Müller et al., 1975). It yields:

1. a high protein/vitamin concentrate suitable for replacing fishmeal, soy protein, etc., in poultry rations;

2. green meal of some 20–25 CP, suitable for pig rations;

3. a fibrous residue containing about 6 to 7% CP suitable for ruminant rations, in particular if complemented with roots and NPN.

In southeast Asia, the legume shrub Leucaena leucocephala has long been recognized for its potential supply of good quality protein and beta-carotene. In the Philippines, Indonesia, etc., leaves containing about 25% CP in DM have been harvested, dried and ground for inclusion in poultry and pig rations at levels of up to 10% and 30% respectively. A pilot undertaking for the commercial production of leucaena meal was recently launched in an FAO/UNDP project

Source : Carcass weights : FAO Production Yearbook, 1975.

¹ Average yield from liveweight as proposed by Society of Meat and Meat Products, American Meat Industry Foundation, 1960. Carcass weight converted to liveweight by utilizing the following average dressing percentages: poultry 75% beef 55 and 50%; pigs 70 and 60%, for developed and developing countries respectively.

² Taking 15 kg, 1.8 kg and 2.8 kg of bone meal per head of cattle, goats and sheep respectively.

¹ SOURCE : FAO Production Yearbook, 1974.

In Malawi. In the future, leucaena could attract attention as a combined source of fuel and protein. Its value as a protein source in ruminant rations has recently been recognized in Mauritius, Mexico and the Seychelles. Research in Mauritius indicates that all the supplementary protein required in a molasses or sugarcane/NPN-based cattle-fattening ration could be supplied through grazing leucaena. It was observed in fact that there were no differences in animal response to supplements of fishmeal and groundnut cake in a molasses/urea-based diet when cattle had access to leucaena grazing (Min. Agric., Réduit, 1973). Its content of tannin complexes was interpreted as having a favourable effect on protein solubility in the rumen.

Tree crops, such as coconut, oil, palm and rubber, may be regarded as producers of by-products not only through their fruits and seeds, but also through allowing interplanting. Land under tree crops, as well as the fallow of tobacco plantations, is usually of good quality and provides suitable conditions for shade-tolerant forage legumes. Müller (1974) estimated the availability of such land in Indonesia at close to 500 000 ha.

Single-cell protein (SCP) from carbohydrate

Vast amounts of low-concentration wastes cannot, for technical or economic reasons, be fed directly. Increasing food and feed prices and a rising concern about environmental pollution have recently led to rapid advances in the fermentation industry. Strains of yeast, bacteria and micro-fungi have been selected that are capable of utilizing the organic matter in wastes as their source of energy and carbon, for synthesis of biomass. The partial recovery of wastes in the form of SCP for both food and feed is thus possible. However, because extensive testing is required prior to use in human diets, it is highly probable that the products will initially be fed to livestock. It seems too early to make a quantitative assessment, since most projects are still on a research or pilot scale. Exceptions are the Swedish Symba-Yeast operation, using potato starch wastes, and a number of operations that use by-product substrates from the chemical and forestry industries. Imrie and Righelato (1976), studying the technical and economic feasibility of low-technology carbohydrate waste fermentation in developing countries, concluded that under certain conditions wastes may be recovered as SCP by commercially viable processes, in particular where wastes are negatively priced. Even operating scales as low as a few hundred tons of SCP per year can be made economic, If the process can be used for most of the year and labour costs are low. Wastes usable as substrates by yeast and micro-fungi are already numerous, and divide easily into two main categories: (i) semi-solid wastes such as spolit fruits, fruit processing residues, and molasses; and (ii) low concentration wastes that usually pollute streams, such as waste waters from the starch processing of cereals, roots and tubers, fruit and vegetable canning, etc. Aspects of SCP production and utilization are covered in greater detail in Linko, below.

The use of yeast, bacterial and fungal protein in diets for poultry, pigs and calves has been studied extensively. General recommendations regarding inclusion levels are of little help, considering the great variations in SCP characteristics due not only to the type of organism but also to the strain, mutant and substrate. However, the general assessment of SCP as partial replacement for fishmeal, soybean meal and dried skim milk is good. The first limiting AA is usually methionine, often followed by the S-containing AAs in aggregate. Reporting on a nutritional evaluation programme of Symba-Yeast grown on potato starch wastes in diets for young animals, Skogman (1976) emphasized that very satisfactory results were obtained with yeast replacing 40% of the skim milk powder fed to calves from two days of age. If performance is not to be adversely affected, replacement of about 40% of the skim milk in the diets of very young calves appears to be a general upper limit.

Group III - Mixed energy/protein sources

By-products from the processing of cereal grains

In contrast to the temperate regions, production of feedgrains has not expanded considerably in the tropics; malze is an exception, although in most countries this is primarily a human food. The justification for such production may, in fact, be questioned wherever in the tropics land is capable of supporting production of crops with much higher yields. Nevertheless, processing of food cereals (in particular wheat, rice and maize) also yields a large, valuable and widely underutilized source of nutrients for animal feeding. Table 7.11 gives estimates of the regional availability of wheat, rice and maize and of their protential by-products. Estimates of by-products from maize milling are difficult to make for most regions, since utilization data are lacking.

The variety of materials in this group of by-products, and their degree of uniformity, depends very much on the level of sophistication of the milling process. Thus, simple one-stage milling usually yields all by-product fractions in one mixture, the use fulness of which is commonly limited by the amount of hulls and, in the case of rice, oil it contains. At a high level of milling, however, the main by-product fractions can be obtained in the following proportions:

1. Wheat - screenings, germ and brans about 30% of original grain;
- feed flour: 10 to 40% of white flour;

2. Rice - broken grain, bran and polishings: 15 to 25% of grain without hulls;

3. Maize - dry milled: germ and bran: about 30% of original grain;
- wet milled: germ, maize gluten meal (fine fibre and gluten) and gluten feed (coarse fibre and steep-water dehydrated): about 40% of original grain.

Generally these by-products are incorporated, in greatly varying proportions, into concentrate mixtures for supplying energy, protein, vitamins and minerals to all livestock.

Limitations are frequently imposed by fibre in brans, by protein quality and palatability in maize gluten fractions, and by unsaturated and oxidized fatty acids, particularly in rice and maize. Deoiling of the bran and germ fractions permits higher inclusion levels, particularly in pig rations, and avoids development of rancidity. Le Dividich et al. (1975) found that up to 45% of fine wheat bran in the rations of growing/finishing pigs was perfectly acceptable in terms of performance and carcass characteristics.

In ruminant nutrition, certain fractions may take the role of supplementary protein in high-energy/NPN diets, or that of glucose precursors in high-sugar diets. Research on the use of sugarcane and urea for cattle feeding in Mexico (Leng and Preston, 1976) has shown that animal performance response to rice polishings as the sole supplementary protein was excellent, better than, for example, blood, cottonseed, coconut and groundnut meals, as a result of their balanced content of essential AA. It is known that high-sugar diets result in a relatively low ratio of propionic/butyric acid in the rumen. With growing, pregnant or lactating animals in particular, this may critically limit liver synthesis of glucose from its principal precursor, propionic acid. In such cases, good response in animal performance has been obtained from supplementation of small quantities of grain, germ meal and bran, provided however that the AA supply was adequate (Redfern, 1973; james, 1973; Silvestre et al., 1976).

Brewery and distillery by-products

This sector offers an example of byproduct recovery being quite a profitable undertaking for the industry; In fact, data from countries with a developed marketing system provide evidence that this source is already fairly well utilized. Table 7.12 gives estimates of regionally recoverable brewers' spent grains and yeast. Similar calculations, for distillers' grains and solubles, could be made if adequate data from the distillery industry were accessible. In 1967, for instance, the US industry used about 1.1 million tons of grains (maize, rye, malt, wheat, sorghum and barley) for distilling spirits (inglett, 1970). The estimated recovery of spent grains and solubles, at one third of availability, was 0.36 million. This is, in fact, less than the USDA (1975) estimate of availability for feed use between 1969 and 1973, 0.43–0.46 million tons annually.

Nutritive values vary greatly and depend a great deal on the raw material, the alcohol production process and the degree of combination of yeast and solubles with the spent grain. Since only the starch of the grain is used for alcohol production, the remaining nutrients are in a higher concentration. Yeast grown during fermentation adds particularly to the protein and vitamin values. For example, typical ME values of distillers' dried grains with solubles (from maize) are some 2 900 kcal/kg for cattle, 2 800 for pigs and 2 600 for poultry. Dried solubles and brewers' yeast are best known as useful ingredients for monogastrics' rations, and ME values of the former are close to those for ruminants. With spent grains, however, a high fibre content (usually between 12 and 20%) may result in the ME value for monogastrics being only half of that for ruminants. A widely preferred method of utilization has been in dairy cattle rations. As an alternative to drying spent grains, ensiling with preservatives such as molasses and organic acids has proved very satisfactory.

Whey

This is the major by-product of the cheese-manufacturing industry; production of each kg of cheese in general, and each kg of cottage cheese in particular, yields 8 and 5 kg whey respectively (Coton, 1976). Table 7.13 gives the distribution of the 70.7 million tons available in 1973.

¹ Source: FAO Production Yearbook, 1974.
² = Production minus seed and waste plus/minus international trade.
³ Basis: Percentage of availability (40% N. & C. America; 35% Europe, U.S.S.R., Oceania; 25% Africa, Asia. S. America).
⁴ Basis: 10% bran; 3% polishings.
⁵ Food processing by-products only. Basis: Percentage of availability (30% N. & C. America, 25% S. America).

¹ Source: Yearbook of Industrial Statistics 1965-74.
² Basis: 10 t dry spent grain per 1 000 hl of beer (Ben Gera and Kramer, 1968).
³ Basis: 400 kg dry yeast per 1 000 hl of beer (Ben Gera and Kramer, 1968).

Fresh whey contains 93% water, 0.9% CP (86% of which is true protein), 0.3% fat, 5.1% lactose and 0.6% ash. Removal of protein through ultra-centrifugation results not only in a lower DM content but in decreasing levels of lactose and minerals.

In 1973, 47% of the whey in the USA, and 40% in the UK, was used in animal feeding, but substantial quantities are estimated to be wasted throughout the world. Conventionally, fresh whey has been consumed mainly by pigs. Following dehydration, its lactose and high-quality protein contents make it a valuable ingredient for milk replacers. Interest in feeding whey to ruminants has steadily increased, and whey or permeate is acceptable both fresh or dehydrated. Following a short adaption period, lactating cows, for example, drink daily up to some 90 1, fattening bulls 40 1 and weaned calves 15 1. Lactose may thus cover up to 30% of energy requirements.

Group IV - Miscellaneous ration ingredients

Vegetable and fruit wastes

These consist of solid wastes (trimmings, peels, leaves, stems, discarded crops) and liquid wastes containing sugars, starch and other organic components leached from the product; their quantities, which vary according to the technology of processing, are difficult to assess, for lack of data on the quantities of fruits and vegetables actually processed.

On the basis of US data (NDC, 1971) given in Table 7.14 it may be assumed that the percentages of potentially available semi-solid by-products from the industrial processing of vegetables and fruits are 25% and 30% respectively. Assuming the total world production were processed, the potential amounts of solid wastes would be 70 million t for vegetables and 74 million t for fruits.

A main example of the use of vegetable by-products in animal feeding is that of tomato processing wastes, particularly in the United States; nutritional evaluation, however, is still insufficient to allow firm conclusions. (See however Ben Gera et al., 1968; Chapman et al., 1958; Wie Han Kwie et al., 1968; Ammerman et al., 1965; Abou Akkada et al., 1975.) Dried tomato pulp (whole culled tomatoes) or pomace (residue after pressing the juice) have a high N content (24% DM) but are of rather low digestibility. They may replace part of both the energy and roughage in ruminant diets, and have also been used at levels of 3 to 6% to replace lucerne meal in poultry diets.

The world production of olives, 7.7 million t, is confined mainly to the Mediterranean countries and the Near East. The residues of oil extraction traditionally consist of pellicula, pulp and, if the whole fruit is pressed, stone; in the Acapulco process, stones are removed prior to pressing.

Assuming that the olive pulp represents 30% of the original weight of the olives, world production can be estimated at about 2.3 million t.

Apart from some minor problems with the presence of the stone, more serious difficulties in using olive pulp in animal feeding are caused by its content of tannin and other substances with a depressive effect on rumen activity. Even after solvent extraction of fat, during which the tannin content is also reduced, the digestibility of olive pulp remains low. Its incorporation in the diet also depresses the digestibility of other ration ingredients, as well as the retention of N (Thériez and Boule, 1970).

Reject dates and date stones may make a major contribution in several countries, but more research is needed for a better assessment of their value.

Other by-products

The total volume of residues from the pyrethrum industry is difficult to assess. Their DM content is 78% and they are rich in ash (high in P and Cu, low in Ca), their CF is 22 to 24% and their DCP about 8%. Their TDN value has been estimated at 66%. According to Were (1975), they can replace 50% of the roughages in beef cattle diets and can constitute 30% of the ration of milking cows.

Bakery products, consisting of bread, cakes, etc. which remain unsold, are rich in energy, and their protein content usually exceeds 10%. They can replace dried skim milk in pig rations at levels of 5 to 20%. They are excellent substitutes for maize in feeding steers, and can be incorporated at 30% of the concentrate for milking cows fed a maize silage (Wing, 1965).

Concluding remarks

All the material discussed in this paper, as well as many products of local importance that are not discussed, have been used to some degree in animal feeding. In most cases it has been impossible to make any estimate of the extent of their current use. However, it is undoubtedly true that there exists an enormous untapped potential for feed from agro-Industrial by-product sources. Of the many materials mentioned here, there is hardly any aspect of their suitability for use as feed (chemical and physical properties, technology of processing, intake, biological value, toxicity, economics of use) that has not been investigated somewhere, although most of the research work has been isolated, fragmented and piecemeal. Nevertheless, several examples of a more systematic nutritional and economic evaluation of by-products through feeding systems exist for such feeds as molasses, reject bananas and coffee wastes, as well as (on a non-by-product basis) for sugarcane and cassava. More research is still required on materials that have so far been poorly tested, with regard to the establishment of nutritional parameters, requirements for processing and input/ output relationships under well-defined conditions of feeding and management. Information from particularly satisfactory systems could well lead to adaptive research work and the subsequent development of systems in regions with similar potential.

¹ Source: Coton, 1976.

¹ United States only. Source: National Industrial Pollution Control Council, 1971, p. 77.

References

Abou Akkada, A.R., A. Khalil, M.A. Kosba and M.M. Khalifah. 1975. Effect of feeding residue of tomato canning on the performance of laying hens. Alex. J. Agric.Res., 23(1), 9–14.

Ammerman, C.B., R.H. Harms, R.A. Dennison, L.R. Arrington and P.E. Loggins. 1965. Dried tomato pulp, its preparation and nutritive value for livestock and poultry. Florida Agric.Exp.Stn., Bull. 691, 5–19.

Backer, J. 1976. Utilización integral de camote (Ipomoea batatas (I., Lam)) en la producción de carne. M.S. Tesis Univ. Costa Rica - CATIE, Turrialba.

Ben-Gera, I. and A. Kramer. 1968. The utilization of food industry wastes. Acad. Press Inc. N.Y. Advance In Food Research 17, 77–152.

Bressanl, R., M.T. Cabezas, R. Jarquín and Beatriz Murillo. 1975. The use of coffee processing wastes as animal feed. Proc. Conf. Animal Feeds of Tropical and Subtropical Origin, April 1974. Trop. Products Institute (TPI) 56–62 Grays Inn Road, London.

Brooks, C.C. and I. Iwanaga. 1967. Use of cane molasses in swine diets. J. Anim. Sci. 26, 741.

Chapman, H.L., C.E. Haines, J.R. Crockett and R.W. Kidder. 1958. Dried tomato pulp for fattening steers on pasture. Florida Univ.Agr. Exp. Stn. Everglades Sta. Mimeo Rept., 59 - 3.

Chenost, M., M. Candau, F. Geoffroy and P. Bousquet, 1976. Possibilities of using bananas for the feeding of ruminants in humid tropical regions. J. Univ. Puerto Rico (in press)

Coton, S.G. 1976. Recovery of dairy wastes. In: Food from Waste, Editors: C.G. Birch et al., Applied Science Publishers Ltd., London.

Dickey, H.C., H.H. Brugman, B.E. Plummer and M.E. Highlands. 1966. The use of by-products from potato starch and potato processing. Proc. Intern. Symp. Utilization and Disposal of Potato Wastes, Frederiction, Canada.

Eskew, R.K., P.W. Edwards and C.S. Redfield. 1948. Recovery and utilization of pulp from white potato starch factories. U.S. Dept.Agric., Eastern Regional Research Laboratories, AIC-204.

FAO. 1976. Commodity Review and Outlook 1975–1976.

Grant, R.A. 1976. Protein recovery from meat, poultry and fish processing plant. In: Food from Waste. Editors: C.G. Birch et al. Applied Science Publishers Ltd., London.

Hendrikson, R. and J.W. Kesterson. 1965. By-products of Florida citrus. Univ. of Florida. Agric. Exp. Stn. Bull. 698, 1–74.

Hernandez, J. and J.H. Maner. 1967. Evaluación del banano verde, maduro y verde cocido en crecimiento y engorde de cerdos, INIAP, Ecuador. Experimento SDGP 2.23.

Henry, W.A. and F.B. Morrisson. 1950. Feeds and Feeding, 21st Edition, Morrisson Publishing Co., Ithaca, N.Y.

Herstad, O. and H. Huidsen. 1973. Protein recovered from industrial waste water as feed for chicks. Acta Agric. Scand. 23 (3), 151–156.

Highlands, M.E. 1967. Potatoes and potato pulp for livestock feed in potato processing. Editors: Talburt and Smith. The Avi Publishing Company, Inc., 540–550.

Imrie, F.K.E. and R.C. Righelato. 1976. Production of microbial protein from carbohydrate wastes in developing countries. In: Food from Waste, Editors: C.G. Birch et al. Applied Science Publishers Ltd., London.

Inglett, G.E. 1970. Fermentation and distilling industries. Corn: Culture, processing, products. (Ed. G.E. Inglett). The Avi Publishing Company. Inc.

James, D.G. 1973. The fishing industries of Denmark and Greenland, CSIRO Fd. Res. 33, 49–58.

James, L.A. 1973. Comfith in rations for livestock. Proc. CIDA seminars on sugar cane as livestock feed Barbados, Jan. 1973.

Kercher, C.J., W. Smith and L. Paules, 1966. Barley, corn and beet pulp in low fibre rations for fattening cattle J. Anim. Sci., 25, 904.

Kwee, W.H., I. Ben Gera and A. Kramer. 1968. Processing tomato cannery waste for food and feed. Mimeo., Univ. of Maryland.

Le Dividich, J., I. Canope, F. Hedreville and E. Despois. 1975. Une nouvelle source énergétique et azotée pour l'alimentation du porc en finition aux Antilles: le son fin de blé. Nouv. Agron. Antilles Guyane, 1, (4) 284–292.

Le Dividich, J., F. Geoffroy, I. Canope and M. Chenost 1976. Surplus and discarded bananas as a potential animal feed resource. Wld.Anim.Rev. 20.

Leng, R.A. and T.R. Preston. 1976. Sugar cane for cattle production: present constraints, perspectives and research priorities, Trop. Anim. Prod. 1, 1–22.

Maner, J.H. 1973. Investigation of plants not currently used as major protein sources. Alternative sources of protein for animal production. National Academy of Sciences, Washington, D.C.

McDowell, R.E. and A. Hernandez-Urdaneta 1975. Intensive systems for beef production in the tropics. J.Anim. Sci. 41, 1228-37.

McMeniman, N.P. 1975. Aspects of nitrogen metabolism in the ruminants. Ph.D. Thesis, Dept. Agric. Biochem., Univ. Newcastle upon Tyne.

Min. Agric, New Delhi. 1974. Rep. of Committee on Livestock Feed and Fodder.

Min. Agric, Réduit. 1973. Unpublished data.

Müller, Z. 1974. Livestock nutrition in Indonesia. Technical Report No. 2, UNDP/ FAO Project INS/72/009.

Müller, Z., K.C. Chou and K.C. Nah. 1975. Cassava as a total substitute for cereals in livestock and poultry rations. Proc. Conf.Anim. Feeds of Trop. and Subtrop. Origin, April 1974. Tropical Products Institute (TPI) 56–62, Grays Inn Road, London.

National Indust. Pollution Control Council. 1971, Waste Disposal Control in the Fruit and Vegetable Industry. Pollution Technology Review 1., Noyes Data Corporation, 1973.

O'Donovan, P.B., M.C. Chen and P.K. Lee. 1972. Conservation methods and feeding value for ruminants of pineapple bran mixtures. Trop. Agric. 49 (2), 135–141.

Oosten, B.J. 1976. Protein from potato starch mill effluent. Food from Waste. Editors: C.G. Birch et al. Applied Science Publishers Ltd. London, 196–204.

Philips, T.P. 1974. Cassava utilization and potential markets. IDRC-020^e, Ottawa, Canada.

Preston, T.R. 1972. Molasses as the major energy source for beef cattle. Wld. Rev. Nutrit. Dietet. (ed. Bourne, G.H.), Karger's Basle.

Preston, T.R. 1973a. Unpublished data.

Preston, T.R. 1975, Unpublished data.

Preston, T.R., and A. Molina. 1972. Rapeseed meal in molasses/urea-based diets for fattening cattle. Rev. Cuba cienc, agric. (Engl ed.) 6.

Preston, T.R. and R.E. Valdez. 1976. The nutritive value of sugar cane for ruminants. Mimeo, Report, CIEG, Chetumal, Q.R., Mexico.

Preston, T.R., N.A. McLeod, L. Lassot, M.B. Willis and M. Velasquez. 1968. Sugar cane products as energy source for pigs. Nature 219, 727.

Redfern, D.M. 1973. Feeding trials using molasses based rations for intensive beef production. Working paper UNDP/ FAO Project KEN/72/008.

Redfern, D.M., H.A. Squire and M.J. Creek. 1974. The adaptation and evaluation of high level molasses feeding for use in Kenya. AGA, SF/KEN 72/008, Tech. Report 4.

Silvestre, R., N.A. McLeod and T.R. Preston. 1976. The effect of increasing levels of maize grain on the performance of steers fed chopped sugar cane and urea or sugar cane and molasses/urea with a fixed amount of protein supplement. Trop. Anim. Prod. 1: (in press) Abstr.

Skogman, H. 1976. Production of Symba yeast from potato wastes. Food from Waste. Editors: C.G. Birch et al. Applied Science Publishers Ltd., London.

Thériez, M. and G. Boule. 1970. Valeur alimentaire du tourteau d'olive. Ann. Zootech. 19 (2), 143–157.

TPI. 1975. Research needs and priorities in relation to certain agricultural commodities. Chapter 2: The oilseed economy. Tropical Products Institute, 56–62 Grays Inn Road, London.

Wing, J.M. 1965. Preliminary evaluation of dried bakery products in dairy cattle rations. Feedstuffs, 37 (19), 38.