Chapter 1: INTRODUCTION

In the early 1960s, Professor Max Kleiber expressed a similar concern for these issues and stated that (as quoted by Kronfeld 1982):

" .... metabolisable energy is not a homogeneous en ti ty; instead it represents an assembly of nutrients or metabolites each of which is used with a specific efficiency for a particular purpose."

To this could be added that the availability of these nutrients, and their interactions, affect the efficiency with which nutrients are used.

The misconceptions inherent in any system based primarily on feed analysis are that it is almost impossible to predict:

Whether the feed can support efficient rumen function
The nature or the proportions of the endproducts of fermentative digestion
The potential for rumen escape of nutrients and their digestibility in the small intestine.

For these technical reasons, and also because of differing socioeconomic circumstances, it is proposed that a more appropriate objective for developing countries is to "match livestock production systems with the resources available" (Chapter 1).

This chapter sets out the guidelines for applying these concepts to the development of feeding systems that should make best use of locally available resources and which should be built on traditional practices. Farming systems in developing countries are notoriously difficult to change and innovation must be introduced gradually without inducing excessive risk, which may directly affect the well-being of the family of a poor farmer.

This chapter aims to:

Discuss the validity of the feeding-standard approach under the conditions of tropical developing countries whose feed resources are based primarily on crop residues, agro-industrial byproducts and native pastures
Characterise feed resources, to enable them to be rationally incorporated in feeding systems, taking into account the physical, financial and socioeconomic limitations that affect the functioning of laboratories in most developing countries
Collate the basic information on the digestive physiology and metabolism of ruminants that is required to establish livestock feeding systems w hen the principal resources are crop residues, agro-industrial byproducts and native pasture.

7.1.3 Animal response to nonconventional feed resources

Doubts concerning the usefulness of feeding standards for ruminants in tropical countries surfaced during development work in Cuba (Preston and Willis 1974) in the 1970s when livestock production systems were being established using nonconventional feed resources (ie. diets containing a large proportion of molasses). This research demonstrated that small inputs of bypass protein (Peruvian fishmeal) dramatically increased growth rate and feed efficiency of cattle. However, because of the higher demand for critical nutrients by lactating animals,

high milk production could not be supported on diets in which molasses comprised a large proportion of the total feed intake (Figure 4.17) .

This work demonstrated the high potential yield of animal products from sugar-cane and stimulated research in Mexico, Mauritius and the Dominican Republic that attempted to establish cattle production systems applying the principles developed for feeding molasses (high-sugar diet) (Preston et al. 1976).

Research on the feeding value of whole sugar-cane demonstrated that:

Feed intake was low even though digestibility was high (60-70%)
The animals on this feed apparently needed glucose or glucose precursors because all the sugars are fermented, little propionate is produced and the availability of microbial protein is low because of the presence of a dense population of ciliate protozoa (Bird and Leng 1978).

The implication of these two sets of findings is that rumen function did not provide the required balance of nutrients for productive purposes (Leng and Preston 1976).

Recognition of the role of bypass protein in low-N diets led to research aimed at increasing productivity of cattle and sheep on a range of fibrous feeds (Leng et al. 1977). This is discussed in Chapters 8 and 9. Prior to this work these feeds were considered to have little value other than to support maintenance and were universally referred to as "low quality" fibrous feeds. It was generally believed that the only recourse was to improve their digestibility, in particular by alkali treatment (Jackson 1977, 1978).

The value of alkali treatment was obscured by the fail u re to recognise the fundamental importance of providing nutrients to create an efficient rumen ecosystem and to meet the animal's potential productivity (Leng 1982c; Preston and Leng 1984).

Combining alkali treatment and appropriate supplemen tation has led to practical feeding systems based on rice straw being applied on farms in Bangladesh (Dolberg et al. 1981; Davis et al. 1983) and Sri Lanka (Perdok et al. 1982; Jayasuriya 1984).

Alkali treatment of straws is a difficult operation for small farmers. However, development of principles for supplementary feeding has led to significant improvements in the utilisation of straw as animal feed, particularly in India (see Leng and Preston 1984 and Chapter 11).

The importance of these developments was not so much the use of molasses or straw in animal feeding, since both have been incorporated into diets for ruminants in industrialised countries for many years, but in the magnitude of the contribution of molasses and straw to the total dietary dry matter. In industrialised countries molasses and straw rarely comprise more than 10 to 15% of the diet, the rest of the ration being cereal grains, high-protein forage and oilseed cakes. In contrast, in developing countries the feeding regimes aim to incorporate molasses or straw as the principal component of the diet because they are available locally, and because the use of grain for livestock feeding is restricted for financial, political and socioeconomlc reasons.

Animals fed diets based on sugar, straw or dry pastures have a much larger response to supplementation with small amounts of bypass protein than animals on a diet based on cereal grain (Table 7.1). However, even when these diets are supplemented with bypass protein, feed utilisation efficiency is lower than that obtained on diets of similar metabolisable energy content but with a large part of that energy contributed by cereal grain (Pigden 1972; Redferne and Creek 1972). The inefficiency with which energy is used appears to be most pronounced when the host animal has a high demand for amino acids and glucose precursors (eg. the lactating cow) (Clarke et al. 1972; Chopping et ai. 1970; Perdok et al. 1982).

There are many reasons why cattle when fed molasses or straw as the main components of a diet are more responsive to supplementation and/or manipulation of the rumen when these ingredients are a small proportion of the diet. Some of these differences can be explained on the basis of interactions and associated effects among nutrients and between nutrients and the site of digestion. These are discussed in Chapters 3, 4 and 5.

Feeding standards are almost impossible to apply to the grazing animal because of selective grazing and the difficulties of estimating their intake of pasture.

7.1.4 Nutritive value

In order to accurately predict responses in productivity to supplements with a particular diet, it is necessary to take account of the constraints to metabolism. These relate specifically to the protein, gluc.ogenic energy, VFA energy and long-chain fatty acid (LCFA) energy in the end-products of fermentative and intestinal digestion as a function of total metabolisable energy intake, since the balance of these determines the animal's productivity.

Productivity of ruminants is influenced primarily by feed intake, which in turn is determined by the digestibility of the feed and the capacity of the diet to supply the correct balance of nutrients required by animals in different productive states. Therefore the two major variables that need to be considered are:

The amounts and balances of nutrients required
The quantitative availability of nutrients from the diet.

The balance of nutrients available depends upon:

The amounts of dietary components escaping rumen fermentation that are digested in and absorbed from the intestines (amino acids, glucose and LCFAs)
The rates of production of the end-products of fermentative digestion (which can be highly variable)
Productive functions (pregnancy, lactation, growth, work, maintenance, body weight gain or loss)
Environmental factors (disease, parasitism, temperature and humidity, and other stresses).
The availability of nutrients from a diet is highly dependent on:
The microbial ecosystem in the rumen, which influences the availability of microbial protein, VFA energy and glucogenic energy
The chemical composition and physical form of the diet, which influence the proportion of starch and protein that is not fermented in the rumen and the availability of LCFAs.

A t present it is not possible to predict nutrient requirements and to match these with the availability of nutrients because ofthe many interactions between the animal, its rumen microbial ecosystem and the diet.

Usually the most widely available and cheapest feeds for ruminants in most developing countries are crop residues and, to a lesser extent, agro-industrial byproducts and native pasture. Protein meals, derived from oilseed residues and the processing of animals, fish and cereal grains are expensive and often unavailable.

Table 7.1: Some selected examples of the effects of feeding a bypass-protein meal (EP) to sheep and cattle given different basal diets all supplemented with adequate amounts of fermentable-N
			Growth rate (g/d)
Species	Bypass protein	Basal diet	-BP	+BP	Author
Sheep	Fish meal	Barley grain	230	300	0rskov et al. (1970)
	Fish meal	Sugar / chaff	0	180	Bird et al. ( 1979)
	Pellets (1)	Barley straw	24	100	Bird et al. (1981)
Cattle	Fish meal	Molasses	270	1000	Preston & Willis (1974)
	Fish meal	Cane juice	800	960	Duarte et al. (1981)
	Fish meal	Straw/conc.	180	650	Smith et al. (1979)
	Fish meal	Ammoniated rice straw	100	400	Saadullah et al. (1982)
	Cottonseed meal	Pasture	-320	220	Lindsay & Loxton (1981)
1. Contained formaldehyde-treated cottonseed meal, meat meal and fish meal

In general terms, energy (the basic feed resource) and fermentable N (urea) are relatively inexpensive ingredients, while the sources of amino acids and glucogenic compounds (the protein meals, cereal grains and cereal byprod ucts) are very expensi ve. Since fermentation of carbohydrate provides the energy for microbial growth, and as the feed is often of low digestibility, it is generally desirable to supply fermentable energy on an ad libitum basis. Therefore the basal diet should be freely available.

As a rule of thumb, 3 g of nitrogen are needed per 100 g offermentable carbohydrate to support efficient microbial growth. It is not always necessary to provide this amount since some feed protein will be fermented to ammonia and some urea-N may enter the rumen in saliva. In addition there is evidence that the rumen microbes need small amounts of amino acids and other nutrients for efficient microbial growth.

The potential of the diet to satisfy the animal's requirements for amino acids and glucogenic precursors depends on the pattern of fermentation and on the proportions of dietary protein and starch that escape fermentation and are digested in the intestines.

The extent to which the protein in a supplement passes through the rumen intact is partly a function of its rate of degradation (solubility) in the rumen, and is likely to be influenced greatly by the rate of flow of fluid and small particles out of the rumen. The latter characteristic is influenced by processing of the feed (physical or chemical), the presence of green forage, the amount of protein reaching the duodenum and external factors such as temperature and exercise or work.

The same factors influence the supply of glucose and glucogenic precursors in terms of the amount of starch that is likely to bypass the rumen to the duodenum. However, the nature of rumen fermentation will have a major influence in terms of supply of propionic acid for glucose synthesis.

7.2 NUTRIENT SUPPLY vs PRODUCTIVE STATE

There is insufficient information to permit the precise quantification of the proportions of the different nutrients required for different productive states (see Chapter 4). Nevertheless, the needs of animals can be approximated. The suggested scheme attaches relative priorities to the groups of nutrients according to the physiological and biochemical processes underlying the expression of the particular productive state (see Table 7.2).

Table 7.2: Relative priorities attached to the requirements for long chain fatty acids (LCFAs), amino acids (AA), oxidation (VFA) energy (C2) and glucogenic energy (Synthesis energy C3:C6) according to the productive function of the animal
	Relative priorities
	C2	C3:C6	AA	LCFAs
Work	xx	xxx	x	xxxxx
Maintenance	xxxx	x	x	x
Late growth /
gestation	xxx x	xxx	xx	x
Early growth	xxxxx	xx	xxx	xx
High milk	xxxxx	xxx	xxxx	xxxxx
Medium milk	xxxxx	xx	xxx	xxxx
Low milk	xxxx	x	xx	xxx
Source: Preston and Leng 1984

The groups of nutrients to be varied for different productive states are:

VFA eneergy
Glucogenic energy
Amino acids
Long chain fatty acids (LCFAs)

Manipulation of the end-products of fermentation and digestion has been discussed in Chapter 5. In this section the relative needs for each class of nutrients are related to the various productive states. Methods of achieving the optimum balance of nutrients are proposed, taking into account the constraints associated with the use of the feed resources that are likely to be available in developing countries.

VFA energy arises from fermentation in the rumen of all types of organic matter, though principally carbohydrates. The principal way to increase VFA energy in a particular feed is to increase the amount of it that is consumed and/or to increase its degradability in the rumen by supplementing it with bypass protein or by treating the feed with alkali (mainly ammonia) or both.

Manipulation ofthe rumen to provide more protein and glucogenic precursors is still at the experimental stage. Dietary supplementation is the most obvious way of manipulating the supply of amino acids, glucose and glucose precursors that can be absorbed.

Most supplements are expensive and their use in ruminant nutrition competes with monogastric animals and humans. However, if the primary feed resource is a product which would have been wasted if it were not fed to ruminants, it can be argued that the ruminant uses the concentrate supplements more efficiently than monogastric animals (see Chapter 1). For this reason the term "catalytic" supplement has been used to describe supplements used in this way (Preston and Leng 1980). Milk, given in small amounts «2 litres daily) as a supplement to calves on a straw-or molasses-based diet, is a good example of a "catalytic" supplement.

It is essential that research should produce response relationships to distinguish economic from biological optima. As a rule of thumb, the role of a supplement ceases to be "catalytic" when it comprises more than about 30% of the diet dry matter. Beyond this point it assumes a major role and substitution occurs.

The productive functions and the need for supplementary nutrients are discussed in order of the least to the most demanding.

7.2.2 Work

Work requires ATP generated from the oxidation of LCFAs, with obligatory requirements for glucogenic compounds and for amino acids (to repair the wear and tear of tissues and to replace protein secretions) (see Chapter 4).

The working animal can often obtain sufficient nutrients from a nitrogen-deficient diet so long as it balances the protein-to-energy ratio needed for tissue turnover by "burning-off''' acetate that is in excess of requirements. However, bodyweight loss may restrict the period of work. If the work is to be prolonged and weight loss is to be minimised, the nutrients available must be balanced so as to meet the needs of the working animal. The digestibility and the intake of the basal diet may also have to be increased by supplementing with urea to correct a deficiency of fermentable nitrogen in the rumen (see Chapter 4). This may be the only manipulation that is necessary, but supplements that are rich in fat and bypass protein could be beneficial, particularly where the animal is in a productive state (eg. pregnant or lactating). If weight loss continues because work is prolonged it may be necessary to increase the degradability of the basal diet, ego by ammoniation (urea ensiling).

The mature, unproductive ruminant does not appear to require nutrients over and above those provided by efficient fermentative digestion.

Since the heavily working animal uses largely LCFAs and glucose (Pethick and Lindsay 1982a; Leng 1985), the supplements used should contain or provide these substrates. This is particularly important in the case of LCFAs, since their absorption and use for fat deposition, and then mobilisation and use for work are much more efficient and require less glucose oxidation than synthesis of fat from acetate and its subsequent use in muscle metabolism.

7.2.3 Maintenance

Maintenance alone obviously requires less energy than work; thus the demand for amino acids relative to energy is proportionately higher than in the working animal. This will always be provided by a rumen system that has sufficient fermentable N. Animals in negative energy balance for an extended period on low-N, roughage-based diets extract more digestible energy from the basal diet when it is supplemented with fermentable N (see Table 9.8).

7.2.4 Growth

Growing animals have a very high requirement for amino acids for tissue synthesis and for glucose for oxidation in specific tissues (eg. brain). In addition, glucose is needed as a precursor of macromolecules. The high demand for glucose in productive animals is discussed in Chapter 4 (see Figures 4.3 and 4.5).

It is imperative to recognise that high growth rates cannot be supported on the products of fermentative digestion and that bypass-protein supplements are essential to take advantage of the VFA energy absorbed.

Many factors influence the level of protein supplementation to be used. Response relationships must be established that relate protein supply to animal productivity for each basal (carbohydrate) resource and for the available protein meals. The response pattern will vary according to the nature of the basal diet and the particular protein supplement. This is demonstrated by data taken from Bangladesh and Cuba. Cattle on ammoniated (urea-ensiled) rice straw, when supplemented with only 50 g of fish meal per day increased their liveweight gain threefold (Figure 8.3). On a molasses-based diet, 450 g of fishmeal per day were needed to raise liveweight gain from 300 to 900 g/d (Figure 8.19).

In some countries, legumes appear to offer a lowcost solution to providing bypass protein on the farm. Based on the premise of using protein economically, the intake of these forages should be restricted and therefore it is preferable to grow the legume in monoculture rather than as a pasture. The amount of bypass protein in a legume (green or dry and used as a supplement) has not been estimated but is not likely to be high when it is used as a small proportion of a diet based on low-digesti bili ty roughage (see discussion of legumes as a supplement). It may be beneficial to grow tannin-rich legumes where these are to be used as supplements to crop residues, whereas, if they are to comprise a large proportion of the diet, lowtannin legumes may be more appropriate (eg. feeding Leucaena with molasses/urea) (Hulman et ai. 1978).

The tree legumes, ego Giiricidia, Erythrina and Leucaena, have great potential as sources of legume fodder because they are high yielding and perennial. They are also deep-rooted and may have access to water and nutrients unavailable to smaller plants.

In animals fed low-N diets supplemented with legumes there is still a need to ensure that ammonia concentrations in the rumen are adequate by supplying fermentable N, usually as urea. It is also important that the protein source, which is usually in short supply, should be partitioned between as many animals as possible.

7.2.5 Reproduction

Increases in fertility brought about through nutrition are usually attributed to increased energy intake. There is, however, evidence that the supply of glucogenic precursors relative to total energy is an important feature of the improved energy status.

Conception and puberty

Recent studies have demonstrated that even when animals are grazing apparently nutritious pasture, the "quality" of the energy can be a limiting factor. At the same metabolisable energy intake (the basal diet was low-N Coastal Bermuda grass hay), animals reached puberty at lower liveweights when glucose availability in the animal was increased (Table 7.3); however, the availability of protein (from the diet that supplied the glucose) is unknown and similar responses could have been elicited by supplemental bypass protein.

The effects of bypass protein on conception rates of cows grazing sub-tropical pasture during the dry season are shown in Table 7.4. A supplement providing fermentable energy (molasses) had much less effect on conception than a bypass-protein supplement, confirming the report of Moseley et al. (1982) that it is the "quality" of the energy (ie. energy in the form of glucogenic compounds) that is critical.

Table 7.4: Liveweight and conception rates of lactating beef cows (with first calf at foot) grazing native pasture and supplemented with 1.86 kg of an energy concentrate (molasses 85%, cottonseed meal 12%, urea 17% and monammonium phosphate 1%) or 1.5 kg of a bypass protein meal (cottonseed meal) during periods when only dry pasture was available. There were 12 cows per group.
	Liveweight	Pregnancy
Supplement	(kg)	(%)
Nil	302	10
Energy	332	20
Bypass protein	343	60
Source: Hennessy (1986).

The growth of the foetus

The growth of the conceptus has little effect on the protein and energy demand of ruminants until the last third of gestation, during which most of the foetal tissues are synthesised. Because of the low demand for nutrients by the pregnant uterus upto the beginning of the third trimester, it appears that nutrients provided by fermentative digestion can support the production of a calf or lamb with a normal birth weight. These animals are generally viable and will survive even where the diet of the dam is of low digestibility, provided that the diet is supplemented with urea. This was shown in studies in which urea was included in the drinking water of ewes on nitrogen-deficient pasture (Table 9.4).

Increases in calf birth weight were recorded when pregnant cattle given a basal diet of hay of low digestibility (45%) were supplemented with urea (Table 9.9). However, it was necessary to provide additional bypass protein to prevent body-weight loss or to promote weight gain of the dam during pregnancy.

It appears that supplementing the diet with urea increases milk production to a level that ensures survival of the offspring, but to allow the young animal to grow at a significant rate, milk yield must be increased further by feeding a bypass-protein meal.

Male fertility

Supplementary feeding has been shown to increase the fertility of male ruminants under extensive grazing conditions. In the extensive grazing systems in the dry tropics of Australia, mating of the beef herd usually starts immediately after the onset of the wet season. Elsewhere, bulls are often placed in the herd only 4 to 6 weeks into the wet season, even though the bulls may have lost 20% of their body weight during the dry season. Lindsay et al. (1982) showed that bulls could be maintained in good condition on low-N spear-grass pasture (which was also oflow digestibility) by providing I kg of a protein supplement daily (Table 7.5). More importantly the circumference of the scrotum decreased considerably when no supplement was fed. There are strong indications that a bull with a smaller scrotal circumference is less fertile and has a lower libido (Blockey 1980). This shows that protein nutrition has an important influence on male fertility.

Table 7.5: The effects of supplementation with 1 kg per day of protected protein (80% formaldehyde-treated cottonseed meal, 10% meat meal, 10% fish meal) on the liveweight change, feed intake and scrotal circumference of bulls fed spear grass pasture hay (Heteropogon contortus) containing 0.4% N.
Initial weight (kg)	433	433
Live weight change (kg)	- 40	+ 14
Feed intake (kg DM/d)
Roughage	5.55	7.74
Total	5.55	8.65
Change in scrotal
circumference (mm)	- 20	0.7
Source: Lindsay et al 1982

The alternative to the establishment of a specialised dairy industry is to encourage milk production from existing animal and feed resources, which are largely in the hands of small farmers. This means dependence on feeds such as crop residues, agro-industrial byproducts and pastures of low digestibility and low N content. It also depends on the use of indigenous cattle, which have often not been selected or developed for milk production, but rather for draught purposes or for survival characteristics.

The constraints to milk production using such resources are a function of the balance of nutrients required for milk synthesis (see Chapter 4)). In summary, it appears that a major constraint to milk production is the availability of glucogenic compounds to provide glucose for lactose synthesis, for oxidation and to provide the N ADPH needed for synthesis of milk fat.

There is good evidence that about 50% of the fatty acids of milk arise from dietary fat in large ruminants (Figure 4.16). This reduces considerably the imbalance between the needs for glucogenic energy relative to total energy or protein in milk and that in the products of rumen digestion. The ewe appears to be disadvantaged in this respect (see Table 4.8). The higher the availability of alimentary LCFAs, the better the ruminant is able to cope with milk production (see Chapter 4). However, a minimum fat level in the diet is required in order to spare glucose for other purposes such as milk lactose synthesis.

For many feeding systems in the tropics the level of fat in the diet could be a primary constraint to milk production, particularly with diets based on molasses or sugar-cane.

The ruminant species that produce high-fat, 10wprotein milk (eg. buffaloes, sheep and camels) are likely to be the most suitable for producing milk on low-protein crop residues and sugar-rich agroindustrial byproducts, provided that fat is available from the diet.

Supplementation of lactating animals, particularly those on diets based on crop residues, sugar-rich agroindustrial byproducts and tropical pastures, should aim to correct the imbalances of nutrients for milk production. Supplementing the diet with bypass protein usually increases feed intake and, as a consequence, increases milk production. But to balance energy quality, fat must be mobilised and glucose diverted from oxidation and tissue synthesis to lactose production; in these circumstances animals tend to lose body weight (0rskov et al. ] 977). Adding dietary fat may reduce this effect (see Chapter 4). Adding a source of bypass starch in such a diet balances the ratio of glucogenic precursors to protein and energy and will tend to prevent mobilisation of body fat. Bypass starch does not stimulate feed intake and milk production is affected only slightly (Table 4.14).

The points to be stressed are that:

Bypass protein, because of its effects on feed intake, almost always increases milk production and, depending on the imbalance in nutrients (fermentation pattern), may cause animals to mobilise body reserves. This can be prevented by feeding high-fat, high-protein meals that supply both protein and LCFAs for post-ruminal digestion
Bypass starch or manipulation of the rumen to give higher propionate production, because it balances the supply of nutrients available for milk production, may prevent mobilisation of body reserves without having a large effect on feed intake or milk production. But because it balances the availability of nutrients for milk production, energy is used more efficiently and body weight is often increased.

7.2.7 Wool or hair production

The effect of nutrition on wool production appears to be dependent almost entirely on the quantity and relative proportions of amino acids absorbed. Therefore, feed intake is the primary limitation to wool or fibre growth although at anyone feed intake, wool growth can be increased by altering the balance of protein relative to energy in the products of fermentati ve digestion (eg. by removing protozoa from the rumen (see Chapter 4). Thus on diets that require fermentative digestion, including those based on sugars or fibre, a bypass-protein supplement will increase wool growth (Table 7.7).

Table 7.7: Goats and sheep on "high quality" carbohydrate feeds do not produce without bypass-protein supplements. Bypass starch appeared to also increase productivity. The animals were given a basal diet which was readily fermented in the rumen (35% oaten chaff, 25% maize flour, 15% molasses, 15% sucrose, 5% barley grain, 4.5% urea and 0.5% complete mineral/vitamin mixture). The basal diet was supplemented with 5% protected casein (formaldehyde-treated) 01' 5% protected casein and 10% cracked rice (bypass protein and starch).
	Basal		Bypass protein		Bypass protein + starch
	Goats	Sheep	Goats	Sheep	Goats	Sheep
Daily gain (g/d)	32	45	68	107	81	119
Patch weight*	0.54	0.74	0.82	1.27	0.76	1.11
Feed intake (g/d)	465	538	604	755	664	736
Feed conversion (g DM/g gain)	14.8	11.9	8.9	7.0	8.2	6.2
Rumen fluid t_1/2 (hr)	16.1	14.1	8.6	9.0	12.1	12.7
Source: Throckmorton et al. (1982). * Wool or hair clipped from a 10 cm² mid-side patch (mg/cm²)d.

7.2.8 Carry-over effects of imbalanced nutrition in early life

Under pastoral conditions with alternating wet and dry seasons, young stock post-weaning are almost invariably S1l bjecied to a deficiency of protein relative to energy in the nutrients absorbed from the digestive tract and the balance of nutrients is aggravated by the low fat content of dry pastures. This results in reduced feed intake and energy deficiency.

In societies that depend on milk as a dietary staple, the young calf and people compete for milk. This is indicated by the purchase of maize by Kenyan pastoralists during periods of low rainfall (and therefore at a time when pastures are deficient in protein and fermentable N). The maize purchases by the family were inversely related with milk off-take (Figure 7.1). Reducing the amount of milk available for the calves can be highly detrimental, particularly when the herd is grazing dry pastures or is being fed on crop residues. The cows will yield less at this time due to the imbalance of nutrients in the feed available. Thus the calf suffers on two counts: an imbalanced basal diet and a reduction in the supply of bypass nutrients from milk.

Figure 7.1: The influence of rainfall (active grass growth) on milk offtake and maize purchases (maize) in OlkarkaT ranch in Kenya. during periods of high milk production, maize purchases by the family are reduced, indicating substitution of cereal by milk in the family's diet. This shows the competition between human needs and those of the calves for available milk supplies (Source: ILCA 1985)

The male stock, because they are less valuable and not generally given supplements, often die from inanition resulting from protein deficiency. This is manifested as a low intake of the available feed, usually straw or grass which is low in fermentable Nand lacks the bypass nutrients that the calf would otherwise have received from the milk, but which is used instead by the family. In countries such as India, female offspring are more prized and are often given supplements of young grass and sometimes byproducts such as cottonseed meal. Their survival rate is much higher than that of the males.

Male stock that survive are often reared as replacement oxen, and their ultimate body size is important since, in countries where feed resources are scarce, a single ox for work is an obvious advantage in conserving valuable feed resources. However, as body size is related to work capacity, a large animal is needed if the move to a single ox is to be successful.

In some countries there have been suggestions that the mature size of cat tIe and buffalo in the national herd is being lowered because large animals are slaughtered instead of being used as sires. This has often led to special breeding programmes being established to increase the size of animals. As has been emphasised in Chapter 2, "breeding without feeding" can be highly irrational.

There are indications that cattle can be permanen tly stun ted if they are fed imbalanced diets in the pre-and post-weaning periods. The stunting may be due to reduced feed intake, particularly during dry seasons when the level of both fermentable Nand bypass protein limits intake and efficient utilisation of the available feeds (pasture and straw). It may also be due int.er alia to protein deficiency. Two sets of information support this thesis.

Studies by Hennessy (1984) with Hereford cattle grazing native pastures which were low in N in the dry season showed that, if a protein supplement was given during this period the adult body size was increased by 60 kg compared to cattle that were un-supplemented (Table 9.9)
Surveys of body weight of cattle in traditional grazing systems as compared with that of cattle on ranches in Africa indicated that mature body weight was greater in cattle on the ranches. Calf birth weights, weaning weight and liveweight at two years of age were also higher in the animals under ranching conditions (Table 7.8). The competition between calf and human for milk is apparent throughout Africa. The low birth weight of cattle is also indicative of nitrogen or protein deficiency and/or low feed availability to the dam.

Table 7.8: Comparisons between productivity indices for cattle in traditional grazing systems and on ranches in four African countries. The differences between the two groups in all countries is taken as supporting evidence for protein under-nutrition in cattle in traditional grazing system. The evidence for protein supplementation affecting birth weight, weaning weight and mature body size is given in Chapter 9
				Live weight (kg)
Country	Breed	System	Birth	Weaning	2 years	4 years
Mali	Sudanese Fulani	traditional	17	55	125	200
		ranching	21	79	220	280
Nigeria	White Fulani	tradi tional	20	55	140	240
		ranching	24	96	245	350
Ethiopia	Boran	tradi tional	20	55	150	260
		ranching	25	180	265	420
Botswana	Twana	traditional	26	]20	260	300
		ranching	31	]80	360	400
Source: ILCA 1985 (Unpublished)

The two sets of data strongly suggest that inadequate protein nutrition and/or low feed availability at critical periods may lead to permanent stunting of cattle. The small size of cattle in traditional systems may be partly the result of inadequate nutrition in early life.

Little is known about the effects of early undernutrition on subsequent or potential milk production, which t:ould also be adversely affected. There appears to be a need to compare milk yields of stock raised under traditional and improved (supplementary feed) systems.

7.3 PRINCIPLES OF SUPPLEMENTATION

The proposed scheme is empirical but is considered to be appropriate for the conditions of most developing countries.

7.3.1 Balancing the rumen and the animal for critical nutrients

The first step is to select the basal carbohydrate resource according to availability, potential fermentability and pric.e. Supplementary nutrients should then be provided in accordance with their relative priorities (Table 7.9) and costs.

Table 7.9: Priorities for nutritional supplements in diets based on crop residues, sugar-rich agro-industrial byproducts or tropical crops that produce large amounts of biomass

1. Fermentable nitrogen

3g N for every 100 g fermentable carbohydrate

2. Adequate rumen ecosystem

(a) Roughage characteristics

(b) Micro-nutrients

3. Bypass nutrients

(a) Protein

(b) Starch

4. Balance of end-products of digestion in relation to total oxidative energy and animal needs
Supply of:

(a) Amino acids

(b) Glucose and glucogenic compounds

The first supplement to be considered should be a source of fermentable N (usually urea or ammonia) to ensure the level of rumen ammonia is above 150 mg/litre of rumen fluid. The generally recommended level of rumen ammonia for maximising microbial growth is 50 mg/litre. However, this appears to be too low in terms of optimising the rate of degradation of fibrous substrate. Several experiments have shown that the disappearance rate of cellulose and fibre from nylon bags in the rumen is increased when the concentration of ammonia is raised to about 200 mgjlitre (see Figure 7.2; Krebs and Leng 1984; and Chapter 3).

Figure 7.2: Urea infused continuously into the rumen of cattle on a diet of straw and minerals raised rumen ammonia levels and increased straw intake. The disappearance rate from nylon bags in the rumen of ammoniated (with anhydrous ammonia (∆) or urea-ensiling (0)) and untreated straw (□) and cotton wool (●) was increased as rumen ammonia level was increased up to 180 mg ammonia-N/litre of rumen fluid (Source: Perdok 1987).

The rate of dry matter loss from alkali-treated maize cob.s in nylon bags in the rumen increased linearly as rumen ammonia concentration was raised from 30 to 120 mgjlitre of rumen fluid (Alvarez et al. 1983). Similarly, the optimum level of rumen ammonia for maximum rate of fermentation on starchbased diets was above 200 mgjlitre (Mehrez et al. 1977). However, it must be stressed that the rate of breakdown of starch in the rumen is probably never a constraint for the utilisation of grain-based diets. On the contrary, it may well be an ad vantage on such diets to have a lower-than-optimal rumen ammonia level to slow the fermentation rate.

In contrast, rate of degradation is of paramount importance when the diet is based on crop residues, because it is the rate of degradation offibre that eventually limits feed intake and therefore animal productivity.

When rumen ammoma levels are less than 150 mgjlitre it is recommended that the effects of adding urea should be monitored under the prevailing field/farm situation. As a general rule, if deficiency is suspected, urea should be added at the rate of about 1-2% of the organic matter in the diet. It is desirable that supplementation ensures an almost-continuous supply of ammonia-N in the rumen (see Chapter 5).

The second supplement should be a source of highly digestible forage, preferably legume or beet pulp given at about 10 to 20% of the diet. The exact action of this type of supplement on rumen function is not fully understood. In some way it helps to ensure a more efficient rumen environment for the digestion of cell wall carbohydrate (eg. increases fungal biomass).

The third supplement should be an oilseed meal, cereal bran or an animal-byproduct meal (supplying protein and fat) and should be given in amounts not to exceed 20% of the total diet dry matter. The 20% limit is to prevent intake of the supplement from reducing intake of digestible energy of the basal diet or substituting for the basal diet. Lesser amounts may be more economical, and it is imperative that feeding trials be carried out to define response relationships. In this way the amount of supplement can be related to animal productivity. The optimum level (in economic rather than biological terms) and the degree of response to the supplement will depend upon the fermentability of the basal diet.

Supplementation with a source of unreactive LCFAs (eg. calcium soaps) in addition to bypass protein appears to be of great benefit (van Houtert and Leng 1987) but requires more research.

7.3.2 The role of minerals

Introduction

The mineral nutrition of ruminants is a very large subject in its own right and it is beyond the scope of this presentation to do any more than highlight points that fall within the general philosophy.

The identification and correction of trace mineral deficiencies, particularly in grazing livestock, has had dramatic effects and this has focused research on this area. An example was the discovery of cobalt and copper deficiency in coast disease of sheep and their correction by supplementation in Australia (see for review Gardiner 1977). Many of the land areas deficient in minerals have been identified, often mapped and a number of approaches have been taken to correct these deficiencies, including:

Fertilizer application
Provision of licks / supplements to livestock
Administration of slow release bullets that lodge in the rumen (eg. for cobalt, iodine, selenium) or in the abomasum (copper wire)
Routine injections or drenches at yearly intervals or longer.

Many of these methods are described in a book edited by Grace (1983) and are not discussed further.

Almost all elements in the earth's crust have been detected in the animal body, but only a few have been proved to be nutrients and essential in the diet. The quantitative requirements for mineral nutrients are affected by a number of factors such as plane of nutrition, management and nutrient interactions in the diet. Nutrient interactions are often of vital importance. For example, at a given dietary level of copper a sheep could die from hypo- or hyper-cuprosis depending on the level of sulphur and molybdenum in the diet (Underwood 1977).

Management strategies are often associated with imbalances in mineral nutrition. For example, hypocalcaemia in late pregnancy in sheep or early lactation in cattle appears to be a result of failure to mobilise bone calcium and can often be overcome by exposing animals in late pregnancy to periods of low calcium nutrition to help them "learn" to mobilise their calcium requirements from bone reserves (see Remberg 1972). Of great significance is whether the diet is balanced for critical nutrients (other than minerals).

Hence, productivity determines the need for minerals, particularly the macro elements (eg. calcium, phosphorus, sodium, magnesium). Often there is little need for mineral supplementation where animals are underfed or imbalanced for protein and/or glucose and fat.

Supplementation with minerals is often advocated where animals are just not "doing well" on low feed intakes or imbalanced feeds. It must be strongly emphasised that mineral deficiencies often do not become apparent until a diet is balanced by supplementation and productivity increases.

Soil eating or licking, depraved appetites, and general poor condition in livestock are often taken to indicate mineral deficiencies. Soil eating or licking has not been generally associated with specific mineraI deficiencies, whereas depraved appetite for carcasses is often associated with phosphorus deficiency and botulism in areas where phosphorus deficiency is prevalent. General poor condition of stock in tropical countries appears to be more a problem of imbalance of nutrients but is often compounded by mineral deficiencies.

Mineral supplementation

The concentration of nutritional research effort on diagnosis and correction of mineral deficiencies is highly commendable particularly where livestock on high quality feeds are not producing to expectation (McDowell et ai. 1984). However, this policy is much less relevant when, because of an imbalanced feed animals are at subsistence, or at only moderate levels of production.

The philosophy that has been stressed throughout this presentation is that resources are divided into basal diets and supplements. The supplements are generally the most expensive component of a diet as they are usually high in protein and/or oil and/or starch. Fortunately, many of these are good sources of minerals (ie. high phosphorus content of many cereal and oil milling byproducts). Because compounded mineral supplements are relatively expensive, often not available in small quantities, they are often out of the reach of smallholder farmers, or are difficult for them to obtain. The attitude we have taken in development programmes has been to use local and natural sources of minerals as a means of economising on the need to purchase proprietary mineral mixtures. These include:

7.4 CATEGORISATION OF FEED RESOURCES

In some instances the use of imported materials may be justified, especially where small amounts have dramatic (catalytic) effects. Urea and fish meal used as supplements in diets based on molasses or cereal straws are good examples.

7.4.1 Fermentable carbohydrate

Many naturally-occurring materials can be fed to ruminants but relatively few of these are available in sufficient quantity for them to be used as the principal source of fermentable carbohydrate. The main feed resources that are available in sufficient quantities include:

Pastures (a distinction should be made between pastures during the dry and wet seasons)
Crop residues (eg. straws from rice, wheat, maize, millet and sorghum)
Cut forages and high-biomass crops (eg. sugarcane, elephant grass and other grasses)
Agro-industrial byproducts, of which molasses is the most important. Other byproducts include pulps from the citrus, pineapple and the sisal industries and reject fruit and waste from bananas.

7.4.3 Fermentable nitrogen

A source of fermentable N must be added when the basal diet does not give rise to a sufficiently high level of rumen ammonia. The most important source is urea; animal excreta also falls in this category. The protein of some high-protein forages (eg. sweetpotato foliage) is rapidly degraded to ammonia in the rumen. However, this implies the destruction of protein, which should be avoided wherever possible (eg. it may be better to feed this forage to monogastric animals).

It is emphasised that fermentation (deamination) of protein is not only wasteful of protein but it is energetically inefficient (Chapter 3). A kilogram of protein yields only about 60 g of digestible microbial protein, compared with about 200 g of digestible microbial protein from the same amount of carbohydrate. Because the protein is converted to VFAs and ammonia, feeding a highly soluble protein as a basal source of fermentable organic matter can actually imbalance the protein to energy ratio in the end-prod uets absorbed if none of the protein escapes degradation in the rumen.

7.4.3 Supplements which contribute to creating an efficient rumen ecosystem

The characteristics of a feed that contribute to an efficient rumen are:

Physical factors which influence the motility of the rumen wall, the amount of digesta held in the rumen and its flow to the lower digestive tract. The data in Table 7.10 show that intake decreases (and the volume of rumen liquid increases) as the physical nature of the forage supplement in a molasses-based diet becomes less stimulatory (due to being ground too finely) or is removed completely
Nutrients, other than ammonia, which stimulate microbial growth in the rumen. On fibrous diets, the critical nutrients are those that increase the size of the pool of colonising microbes floating free in rumen fluid
Feeds that either promote fungal growth (eg. sulphur) or decrease the protozoal biomass (eg. w hole cottonseed) or both (see Chapter 5).

Table 7.10: On a liquid molasses-based diet, feed intake decreases and rumen liquid volume increases as the stimulatory value of the supplementary forage is reduced (by fine grinding) or removed completely
	No forage	Plus forage
		Fresh	Dehydrated and ground
DM intake, kg/d	1.8	2.4	2.0
Rumen contents, kg DM	8.3	4.3	5.9
Source: Preston 1972

Many of the above factors appear to be present in green forage (Figure .5.4). Leguminous plants are probably better than grasses in this respect since they may also provide a source of bypass protein. Normally 20% of the dry matter of a diet in the form of green forage is enough to meet the requirement for microbial nutrients. However, even smaller amounts have increased animal performance.

7.4.4 Bypass protein

Once the supply of fermentable N is assured and a small supplement of green forage has been included in a diet, the next limitation to productivity will be the availability of amino acids in the intestine. For many of the feed resources that will be used in tropical countries, the value of bypass protein lies in its effects on increasing efficiency of use of absorbed nutrients and on increasing voluntary intake. This is in addition to its complementary role to microbial protein. Protein that is slowly degradable may provide amino acids and peptides for microbial growth m addition to providing bypass protein.

7.4.5 Bypass starch and glucogenic precursors

Absorption of glucose increases glucogenic energy relative to total metabolisable energy. In addition, energy losses associated with glucose synthesis in the animal and also fermentative losses in the rumen are avoided. Supplements that increase propionic acid relative to the other VFAs increase the amount of glucogenic energy available and also have smaller fermentative losses. The important role of these nutrients is to increase the efficiency with which metabolisable energy is used for productive purposes.

Although all sources of starch are fermented completely in the rumen given enough time, there are marked differences among them in their rates of degradation. Starches from maize, rice, banana and, to a lesser extent, sorghum, appear to have characteristics that permit them partially to escape fermentation in the rumen; in contrast, the starch in cassava and sweet-potato roots are fermented rapidly in the rumen (see Figure 7.3).

7.4.6 Long chain fatty acids

Supplementation with LCFAs appears to have two opposing effects. Increasing the LCFA component of a low-fat diet increases the efficiency of feed utilisation, especially for milk production (Chapters 4 and 5). On high-fibre diets (such as crop residues) more than 5% lipid in the diet depresses fibre digestion. However, recent work with protected fats and soaps has shown that addition of LCFAs in these forms increases feed utilisation for milk production substantially (see Palmquist 1984).

A vail able sources of fatty acids include oilseeds, oilseed residues (particularly expeller cakes), and milling offals or brans and, in some countries, animal byproducts such as tallow. The effectiveness of any lipid source will be enhanced by protection (formaldehyde/protein com plexes -- Ferguson 1975) or by saponification with calcium salts (Palmquist and Jenkins 1982).

7.4.7 Feeds and other materials with a capacity to manipulate the rumen microbial biomass

Manipulation of rumen fermentation using natural feeds is becoming more feasible as knowledge of the processes of rumen digestion develops. The objectives of dietary manipulation are discussed in detail in Chapter 5. In general, manipulation aims at increasing propionate production (G/E ratio), improving the ratio of protein to energy (P /E ratio) in absorbed nutrients and increasing digestibility.

Increasing the proportion of propionate in the products of fermelltation has been associated primarily with adding the chemical monensin to high-quality fattening diets for cattle in Europe and North America. Tropical feed resources such as poultry litter or high oil meals appear to perform a similar role.

There must be considerable potential for using small amounts of plant materials containing elements that affect specific groups of rumen micro-organisms, and therefore act in a similar way to chemical additives. This possibility does not seem to have been explored, but changes in the proportions of VFAs in the rumen have been observed where poultry litter has been added to the diet and has enhanced animal performance (Chapter 8). At present the control of protozoa with secondary plant compounds appears to be one of the possibilities for increasing fibre digestibility in crop residues (see Chapter 5).

7.5 NON-CONVENTIONAL SUPPLEMENTS

The primary limiting nutrients for production on most tropical feed resources are fermentable N, glucogenic precursors, bypass protein and dietary LCFAs. Urea, oilseed cakes, byproducts of cereal milling and animal-byproduct meals are the logical supplements when available. However, there are many situations in which farmers do not have access to these supplements because they are either not available locally or too expensive. In addition, there is often a reluctance to use urea because of the fear of toxicity.

7.5.1 Livestock excreta

Excreta from all types of livestock have been used in livestock rations. It is obvious that excreta in general must be a poor source of fermentable carbohydrate and protein; however, LCFAs (probably as soaps) may build up in litter as small amounts will be present in faeces and they are only slowly degraded by micro-organisms in the litter. Microbial growth in the litter will therefore tend to concentrate the soaps. Excreta from ruminants are high in refractory cell wall carbohydrate with smaller amounts of microbial cells (from the caecum) and some urea if the urine is incorportated with the faeces. Monogastric species produce the most valuable excreta, and especially in the case of poultry there may be considerable contamination with wasted feed grains. Excreta from poultry are rich in N, mostly as uric acid which is hydrolysed to ammonia by rumen micro-organisms.

Excreta (often depathogenised with formalin) have been used widely in the developed countries as a component of cereal-grain-based diets, in which their main contribution is as a source of non-protein N and minerals.

In developing countries, only poultry litter has found ready acceptance as a component of livestock feeds. It appears to playa particularly appropriate role in high-molasses diets, in which it complements the readily fermentable sugars and the low levels of fermentable N and of phosphorus in the basal diet. There is also evidence that adding small amounts of poultry litter to a molasses-based diet in some way alters rumen fermentation towards producing a larger proportion of propionate and less butyrate (Table 5.2), and that this is reflected in higher levels of animal performance (see Chapter 8). The data in Table 7.11 indicate that poultry litter is not a source of bypass protein, which is to be expected in view of its chemical characteristics.

7.5.2 Legume forages and foliages from food crops

An alternative resource that can serve as a source of fermentable N and of bypass protein is a forage crop grown on the farm, or produced as a byproduct or residue from a food crop. A legume crop has the additional ad vantage of being able to fix atmospheric nitrogen and thus to reduce the need for fertiliser N.

In the tropics, tree legumes have a special role since they can also be used to provide shade (eg. in coffee plantations), and serve as 'live' fences and as sources of fuel. They are also usually perennial and some can be established easily from cuttings (eg. those used for 'live' posts).

Legume trees which are now used commercially as sources of supplements include leucaena (Leucaena leucocephala) gliricidia (Gliricidia sepium) and erythrina (Erythrina glauca and E. poeppigiana). Cassava, sweet potatoes and, to a lesser extent, bananas also provide valuable forage.

Some results from using these materials as supplements to molasses-based diets are given in Chapter 8.

In temperate countries legumes have long been used as an alternative to nitrogenous fertilisers for increasing pasture production. It has also been recognised that they have a higher nutritive value than grasses, apparently because of their higher protein content. The metabolisable energy in legumes is also used more efficiently for productive purposes than that of grasses of the same digestibility (see Table 7.12).

Table 7.12: More protein (ie. non-ammona-N (NA)) reaches the small intestine (Sf) of sheep when the diet is composed of white clover rather than ryegrass. Total N in the dry matter (DM) in both species was adequate to support efficient rumen fermentation and the implication is that more protein in the clover escapes fermentation. The higher efficiency of utilisation of the metabolisable energy for fattening (K_f) % on the clover compared with the grass confirms work reported elsewhere and implies that legumes are superior to grasses as sources of nutrients
	*Lolium perenne*	*Trifolium repens*
N in DM (%)	2.6	4.2
NAN entering
51 (gjkg DM intake)	30	44
Organic matter
digestibility (%)	82	74
Metabolisable
energy (MJjkg DM)	12.2	11.5
K_f (%)	0.33	0.51
Source: Beever et al 1980; Ulyatt et al 1980

Tropical grasses support lower levels of animal production than temperate grasses, mainly because they con tain less nitrogen and are less digestible (see Minson 1980). The low productivity from tropical pastures has stimulated considerable research aimed at developing grass-legume associations for tropical conditions. The presence of legumes in the sward has led to increases in animal production, but mainly in terms of productivity per unit area rather than per animal (see Mannetje 1981 and Chapter 9).

The present discussion is restricted to the role of legumes as supplements in feeding systems based on low-N crop residues and byproducts.

7.5.3 Attributes of legumes as supplements

In developing countries where competition for land for crops or grazing is high, the area likely to be sown to legumes will almost always be a small proportion of the total area. It follows therefore that the role of a legume must be to increase the efficiency of utilisation of the basal diet (ie. usually a crop residue) at low levels of supplementation (usually less than 20%) and used "catalytically".

As a priority, the legume should have a high protein content to supply both fermentable and bypass protein. There will be additional benefits if it contains other critical nutrients (eg. lipids, minerals, vitamins and other plant compounds) that enhance the rumen ecosystem so as to increase microbial growth, rate of fibre digestion, propionate production and escape of dietary protein (eg. contain tannins).

There are two sets of data that indicate the suitability of legumes as sources of fermentable Nand bypass protein. These arise from comparative studies with grasses and animal response trials.

Temperate legumes

The data in Table 7.12 show that, compared with ryegrass, white clover contains more N in total and provides more protein that is available for intestinal digestion in sheep. This almost certainly indicates that a proportion of the legume protein escapes rumen fermentation. The fact that the efficiency of utilisation of metabolisable energy is higher for a legume than a grass is further evidence that the digestion of legumes provides a better balance of nutrients for productive purposes than is the case with grasses. However, protein from white clover only appears to escape rumen fermentation at high intakes of the clover. Therefore when it is used as a supplement to a fibrous feed it may only provide fermentable N. Clover also provides highly digestible carbohydrate which will increase the digestibility of the basal diet. It also provides lipids which may help to reduce the amount of glucose that is oxidised for synthesis of adipose tissue or milk fat (Chapter 4).

It is likely that legume forages that are rich in tannins will be better sources of bypass protein than low-tannin legumes, since tannins link with proteins during mastication, and appear to reduce microbial degradation of plant proteins (Reid et al. 1974). The high levels of tannins in Lotus pedunculatus, whilst protecting protein from degradation, reduce

In situations in which the fermentable N requirement can be met from other sources (eg. urea or animal excreta) the need is to reduce the degradability of the legume protein so as to increase the proportion of bypass protein. This has been shown to occur when a forage is artificially dried and more so when pelleted (see Table 7.13).

It will rarely be economic to dehydrate or pellet legume forages, and sun drying is the only feasible alternative. Nolan and Leng (1972) showed that about 60% of the protein in sun-dried lucerne apparently escaped rumen fermentation. Sun drying is nevertheless not as effective as dehydration as a means of protecting the protein (see Table 7.14).

As discussed earlier, secondary plant compounds such as tannins are known to protect dietary proteins against microbial attack in the rumen. Thus if a freshly harvested legume given as a supplement is to provide bypass protein then it should be selected for a relatively high content of tannins, even though this will depress fibre digestibility (Reid et al. 1974). This point is illustrated by the data in Table 7.15, which show that, although the tannincontaining legumes (trefoil and sainfoin) were less palatable than lucerne, they supported higher growth rates in heifers. The authors concluded that this was because more of the protein in the tannin-containing legumes escaped degradation in the rumen. Generally these are richer in tannins than are temperate legumes and therefore should function better as sources of bypass protein.

The data in Table 7.16 show that the foliage of the legume trees gliricidia (Gliricidia sepium) and leucaena (Leucaena leucocephala) is highly digestible and that the nitrogen is fermented at a slower rate than the dry matter indicating the likelihood that some of the nitrogen (protein) will escape the rumen fermentation. The use of foliage from legume trees as a source of bypass protein is discussed in Chapter 8.

It must be emphasised that when legumes contain a large proportion of protected protein, some other source of rumen-fermentable N will be required.

7.6 METHODS FOR EVALUATING FEEDS

What is the minimum information required in order to identify and classify feed resources for the functions set out above?

It is quite clear that the classical system of proximate analysis and even the more sophisticated methods for identifying the components of plant cell walls (ie. ADF and NDF) provide little information for the development of feeding systems that aim to use tropical feed resources efficiently. The application of the van Soest ADF and NDF analysis is in basic research to provide understanding of the dynamics of fibre fermentation in the rumen (van Soest 1982).

7.6.1 Sources of fermentable carbohydrates

The factors that determine the value of a carbohydrate source for ruminants are:

The content of soluble sugars
In the case of feeds composed primarily of cell wall components, the degree to which lignin inhibits or physically prevents hydrolysis of cellulose and hemicellulose by rumen microbes.

Chemical analysis for lignin is relatively complex and does not indicate the degree to which the lignin is cross-linked with the other potentially fermentable components, which is the final determinant of the rate of digestibility and total digestibility.

The most appropriate method for obtaining an indication of the nutritional potential of a carbohydrate

resource is the nylon bag technique for measuring digestibility (0rskov et al. 1980). Many modifications have been made to this method, usually with the objectives of standardisation or of describing the rates of disappearance of the dry matter or fibre component.

For practical purposes the two parameters that are most relevant in terms of describing the relative fermentability of a carbohydrate source are:

The loss of dry matter at zero time (due to the soluble materials and very small particles)
The rate of disappearance of non-soluble organic matter (the simplest measurement of this is total dry-matter loss after 48 hours in the rumen).

It must be emphasised that the nylon bag method should be used for comparative purposes. It is essential that researchers use internal standards to relate the degradability of a new feed resource to one that is known. The method is location specific and should be used to measure the relative fermentability of feed and to provide information concerning the efficiency of the rumen ecosystem or the capacity of the feed to escape the rumen.

Feed resources that have an organic-matter degradability of less than 40% in 48 hours have little potential for supporting growth and lactation but they may have application in a diet for working animals. To be used as a feed resource for milk and meat production such feeds would need to be submitted to some physical or chemical treatment to increase their susceptibility to fermentation by rumen micro-organisms.

7.6.2 Fermentable nitrogen

It is useful to know the total content of N in a feed, and most feeds have been analysed and their N contents are known within relatively narrow limits.

The N content of a feed is not always a reliable indicator of (i) its capacity to provide ammonia to meet the requirements of rumen micro-organisms or (ii) the amount of bypass protein that it will provide. For example, some high-protein meals are only slowly degraded in the rumen and ammonia levels are too low to meet the requirements of efficiently growing

The solubility of protein (measured as N) when the feed is macerated in artificial saliva or a buffer solution (pH 6.5) gives a useful indication of the bypass protein content of a meal. However, the simplest and most appropriate method for determining whether the N requirements of microbes are met by a particular meal is to measure the level of ammonia in rumen fluid of animals receiving that feed.

Rumen ammonia levels on most feeds probably ought to exceed 150 mg/litre. On fibrous diets ammonia levels apparently need to be considerably in excess of the generally accepted level of 50 mg/litre (Figure 7.2). As a guide, if rumen ammonia levels are less than 100 mg/litre of rumen fluid some 6 hours after the animal has begun to feed/graze, then a source of fermentable N should be added to the diet, usually at a rate of about 10-20 g N /kg of degradable organic matter. The most available sources of fermentable N are urea, animal excreta and, to a lesser extent, green forage.

The data m Figure 5.3 show that high levels of cottonseed meal in a diet based on molasses and straw did not provide the minimum levels of ammonia needed in rumen fluid. This was probably because of solvent extraction and pelleting of the meal -- processes that depress solubility. In this basic diet the addition of urea was much more effective in raising the level of rumen ammonia to the desired level, at the same time allowing the protein meal to be used more efficiently as a bypass nutrient.

7.6.3 Rumen function and feed resources

The nylon bag technique can be used to determine if the rumen ecosystem provides the optimum conditions for fermentation. In most cases the objective will be to maximise the rate of degradation of the cell wall components, principally cellulose and hemicellulose. For this purpose, the basal diet should be fed to animals fitted with rumen cannulae. Placing pure cellulose (eg. cotton wool) or some other source of ground fibre in the nylon bags and measuring rates of degradation will indicate if the microbial ecosystem is optimal for the digestion of a fibrous feed resource (see Figure 7.2).

As has been already indicated, with ruminants fed crop residues the improvement in the rumen ecosystem is most readily brought about by supplementation with small amounts of green forage. Other useful supplements in this respect may be those supplying minerals (particularly Sand P), vitamins and co-factors, as appears to be the case with poultry litter and molasses.

Providing urea will also improve the rumen ecosystem when there is a deficiency of N in the basal diet (Chapter 3). Figure 5.4 illustrates how this method can be used to demonstrate the stimulatory effects of green forage on the rumen ecosystem in sheep receiving a basal diet of sisal pulp. The digestibility of cellulose in 48 hours was increased by a factor of almost 2 when 25% of the sisal pulp was substituted by green African Star grass. There was no further increase in rate of digestion of the cellulose when the amount of green forage in the diet was increased to more than 25%.

The proportions of the rumen VF As and the populations of protozoa, bacteria and fungi (free in liquid and on particles) are indicative of the fermentation pattern and also indicate possible deficiencies in protein and/or glucogenic energy (Chapter 4).

7.6.4 Bypass protein

Characterising a supplement according to its content of bypass protein can only be done in a feeding trial. Generally two experiments will be required. In the first, growing lambs or young cattle (recently weaned) should be fed the basal diet already supplemented with fermentable N and a small amount of green forage (in case of known deficiencies of these two factors). The different sources of bypass protein should then be added to the basal diet in fixed quantities equivalent to approximately 15 g N /100 kg liveweight.

In the second experiment, the most promlsmg source of bypass protein should be included in a surface response trial in which increasing amounts of the supplement are given in the range of zero to 20 g N /100 kg liveweight. In all cases the basal diet must be adequately supplemented with fermentable Nand factors for optimising the rumen ecosystem. According to the nature of the response (measured in terms ofliveweight gain or wool growth in the case of sheep) it will be possible to determine the level of protein supplementation that gives optimal economic results.

It is important to emphasise that the economic otimum normally will be at a lower level of protein supplementation than that which provides the maximum biological response. A promising method (Leng et al. 1984) uses the relative responses of wool growth in mature sheep as a criterion of rumen bypass and relates the response to a standard protein (in this case formaldehyde-treated casein). The advantage of using wool growth, rather than milk production or

liveweight gain, as an index of the amount of bypass protein in the meal is that consecutive experiments can be done with the same animals.

7.6.5 Glucogenic precursors

There are essentially two effects that will stimulate the proportion of glucogenic substrate in the nutrients absorbed from a diet. These are:

Stimulation of propionic acid production at the expense of the other VFAs in the rumen
Escape of starches undegraded from the rumen. These will be mainly from cereal grains.

Supplements that stimulate ruminal propionate production generally reduce methane production; these include cereal grains and seeds with high oil content, particularly cottonseed. Feeding a source of oil in addition to altering the glucogenic ratio in VFAs also reduces protozoal numbers in the rumen on forage-based diets (see Table 7.17). It is possible that both propionate and protein are increased in the nutrients absorbed when an oilseed is included in a forage diet. However, a high oil content in a foragebased diet reduces digestibility. As the influences of supplements in the rumen are difficult to predict, the only recourse for assessing a supplement's effect is to measure VFA proportions in rumen fluid taken by stomach tube.

Table 7.17: The effects of supplements of cottonseed oil or whole cottonseed (20% oil) on fermentative patterns and protozoal numbers in the rumen of sheep given a basal diet of wheat straw or oaten chaff respectively. The potential "beneficial" effects on VFA proportions and protozoal numbers in the rumen are counteracted by a decreased fibre digestibility.
	Cottonseed oil (g/d)		Whole cotton seed (g/d)
	0	9	0	75	150
Rumen fluid
VFA (mM)	82	92	78	90	76
VFA proportions
HAc	74	71	70	65	64
HPr	18	23	20	23	26
HBu	7	5	7	9	7
Protozoal Nos.
(x 10^-⁴/ml)	15	2	74	18	1
Source: Bird and Dicko (1987).

The extent to which starch escapes fermentation will depend on many factors including the basal diet, intake, rate of solubilisation and fermentation of the starch. This latter will be affected by its crystalline structure and associated feed materials (eg. waxes). There appears to be no way, at present, of predicting the quantity of starch from a supplement that may escape from the rumen. In this respect the grains most resistant to degradation in the rumen appear to be maize, sorghum and rice. Broken rice in rice pollard appears to be an important component. In unpublished research from Mexico, where the broken rice in rice polishings was substituted (was adulterated illegally) by lime stone chips, cattle growth rates on chopped sugar cane supplemented with this byproduct were markedly depressed compared with animals given rice pollard containing broken rice. The depression in production was due to inefficient utilisation of feed.

An indication of the potential for starch to bypass the rumen can be obtained using the nylon bag technique, with fistulated cattle or sheep on a standard diet. The rate of loss of starch from a nylon bag in the rumen compared with starch loss from a barley sample (which is known to be almost totally digested in the rumen) will give an indication of the likelihood and extent of rumen bypass.

7.6.6 Long-chain fatty acids

An indication of the amount of lipids that a feed contains appears to be relatively easily obtained by simple Soxhlet extraction. However, variable quantities of other materials may be present in such an extract and if research continues to demonstrate a critical role for the LCFAs as means of increasing the efficiency of utilisation of a feed, the components of the ether extract will have to be characterised more carefully in terms of the component fatty acids.

As LCFAs may be present in some supplements as insoluble soaps-largely calcium-(eg. in poultry manure), acid hydrolysis prior to solvent extraction is recommended.

7.6.7 Minerals

This is a most diffuse area to cover, because in practice the mineral composition of most feedstuffs is variable and dependent on the minerals in soil and the manufacturing process. The practice of providing minerals from available resources has to take into consideration many interacting factors. The skill of the nutritionists (and the intuition and experience of the farmer) must be paramount in decision-making. The nutritionist must have an in depth knowledge of mineral nutrition on which he must base many of his decisions.

Is mineral deficiency the first consideration in the development and improvement of livestock production strategies? The answer is clearly no; the first step is to maximise the utilisation of the least-cost basal diet by supplementation and/or manipulation to secure the optimum balance of organic nutrients. Following production trials, problems of mineral deficiencies (either acute or chronic) may become apparent and then steps may be taken to rectify these by either changing the supplements or adding to them a particular blend of minerals.

In practice, calcium, sodium, phosphorus, sulphur, magnesium and the trace elements cobalt, copper, selenium, zinc and iodine are the minerals most likely to be in short supply.

A most important aspect of supplementation is that the commodities used to balance the metabolic nutrients may overcome a primary mineral deficiency. The mineral composition of the most important supplements is given in Table 7.18.

Table 7.18: Levels of minerals in some supplements that can be used to correct essential mineral deficiencies. Cobalt is notably absent from this list and its content in plants is variable but the green, leafy vegetables and leaves of trees, legumes and browse plants and grasses are relatively rich sources which contain 0.2-0.6 ppm Co (Underwood 1 977). Legumes are often richer in Co than grasses but this difference is less marked when soils are deficient in this element. It appears rational to suppose that deep rooting trees, particularly legumes may be important sources of Co in deficient areas.
	Poultry manure
	Cage	Floor	Molasses cane	Rice bran/ polishings	Wheat bran	Cottonseed meal	Bone Meal
% in DM
Crude protein	29	25	3	11	15	41	26
Crude fat	1.7	2.3	0	12	4.0	2	5
Ash	27	14	8.1	11	6.4	6.4	5.9
Calcium	7.8	2.5	0.8	0.04	0.1	0.3	23
Phosphorus	2.2	1.6	0.1	1.4	1.2	1.0	10
Sodium	0.4	0.4	0.9	0.1	0.6	0.4
Potassium	1.4	1.8	2.4	1.2	1.2	1.2
Magnesium	0.6	0.4	0.4	0.7	0.6	0.4	.4
Sulphur	?	?	0.5	0.2	0.2	?	.2
ppm in DM
Manganese	291	?	5	?	100	21	30
Iron	200	?	100	160	170	90	4000
Copper	61	23	18	13	10	16	19
Zinc	325	343		30	95	57	200
Selenium	?	?		?	0.75	0.9	?
Source: Allen 1982

The following examples indicate the approach that should be taken to correct mineral inadequacies by using locally available feeds.

Poultry manure

Poultry manure at about the 10% level is likely to supply all deficient minerals in straw and dry pasture, especially if there is additional fortification with molasses.

Byproducts of cereal grain milling

Rice polishings and wheat bran are high in phosphorus (1.2-1.4%) and if fed at between 10 and 15 % of the diet they should correct imbalances of this mineral in most crop residues and byproducts.

Molasses

This feed is widely available. It is low in phosphorus but is rich in calcium and trace elements, with the

possible exception of cobalt. Usually the sulphur level is sufficient (0.3-0.4%) to balance added urea in order to maintain the recommended 10:1 N:5 ratio.

Foliage from legume trees

Leaves from most plants appear to be good sources of cobalt and other trace elements.

Fertilizers

These can can often be used in small quantities particularly to supply phosphorus.

7.7 A STRATEGY FOR DEVELOPING FEEDING SYSTEMS

Estimate, using the nylon bag technique, the soluble component and the 48-hour loss of dry matter of the available carbohydrate resource.

Interpretation: If the 48-hour loss converted to a standard of known digestibility is 55-65% then the feed has a high potential to support growth and milk production. If the apparent digestibility in nylon bags is between 45 and 50% the potential is low and chemical treatment may be considered. Below 40%, there is little likelihood of the feed resource being useful for productive purposes without treatment to increase digestibility.

2. Give the basal feed resource to a group of four animals (preferably weaned calves of about 150 kg and 9-12 months of age); measure the intake and note the general performance of the animal in terms of the symptoms of digestive or metabolic upsets that could indicate the presence of toxic components in the feed.

Interpretation: If the daily intake of dry matter is less than 2 kg/IOO kg liveweight, it is likely that there are deficiencies of nutrients for rumen organisms or that there is an imbalance between the amount of amino acids and the amount of energy absorbed.

3. Take a sample of rumen fluid (use a stomach tube with non-fistulated animals) and measure the level of ammonia.

The rate of liveweight gain should be estimated by regression analysis of fortnightly liveweights (Y = kg) and time on experiment (X = days).

Simultaneously with the above trial, the effect of green forage on the rumen ecosystem can be assessed (using the nylon bag technique) and the level of ammonia in the rumen fluid can be determined at intervals over 24 hours.

Interpretation: The results from the above trials should indicate whether it is beneficial to supplement the basal diet with:

5. Having established the need for supplementary green forage and bypass protein, the final step is to evaluate the response curve to different sources of bypass protein and LCFAs (as soaps or protected fat). However, before recommendations on the use of fat can be made, considerable research is needed.

The most difficult step is to assess the effectiveness of the feeding strategy that is developed in terms of the balance of the final end-products of digestion. Feed intake is one of the best guides. However, acetate clearance rate (see Chapter 4) appears to reflect directly the complete assembly of nutrients for a given productive state and can be measured in the intact animals at the end of, or during, a feeding or grazing trial.

Table 7.3: Feeding monensin to growing heifers on a basal diet of Bermuda-grass hay and concentrates increased propionate and decreased butyrate proportions in the rumen VFAs. Puberty was accelerated as evidence by the greater proportion of heifers cycling by the end of the test period.
	Control	Monensin
Liveweight (kg)
Initial	219	219
Final	313	319
Feed intake (kg)	8.0	7.7
Rumen VFA (% molar)
HAc	74	69
HPr	19	26
HBr	6	3
Total VFA (mM/litre)	65	67
Fertility (% heifers cycling)	58	92
Source: Moseley et al 1982

Table 7.6: Higher proportions of propionic acid in the rumen VFAs of growing bulls as a result of supplementation with the rumen manipulator "Lasalocid" are associated with greater testicular development and reduced both age and live weight at puberty.
	Control	Lasalocid
Rumen VFA (% molar)
HAc	65	60
HPr	21	32
HBr	15	7
Total VFAs (mMjlitre)	87	83
Increase in scrotal
circumference (cm)¹	3.1	5.3
Testicular volume (cm³)	57	91
Age at puberty (d)	471	437
Weight at puberty (kg)	379	366
Source: Neuendorff et al 1982 ^1. From29 to 175 days

Table 7.11: Comparison of sunflower cake with poultry litter as supplements for grazing cattle
	Supplement
Poultry litter	Sunflower cake	Live weight gain, g/d
1.3	0	480
1.0	0.12	580
0.64	0.31	680
0	0.60	740
Source: Delgado et al 1979

Table 7.13: Preparation of clover before feeding to sheep markedly affects the amounts of amino acids flowing to the small intestine (51). Freezing 01' artificially drying and pelleting both apparently increase dietary protein avoiding fermentation in the rumen. maturity (Source: D Dellow and J V Nolan, unpublished data).
	Fresh	Frozen	Dried/pelleted
Amino acid intake, g/d	127	127	124
Amino acids entering SI, g/d	80	133	175
Sources: MacRae and Ulyatt (1974); Beever et al. (1971).

Table 7.14: Effect of drying temperature on solubility and digestibility of nitrogen and nitrogen balance in lambs fed dried lucerne
Temperature (°C)	Soluble N (%)	N digestibility (%)	N retention (g/d)
65	43	49	6.0
130	40	68	7.4
160	40	66	6.9
180	34	52	3.4
Source: Goering and Waldo 1974


Figure 7.4: Relationship between ruminally undegraded (bypass) protein and organic-matter intake in sheep fed clover 01' ryegrass cut at two stages of maturity (Source: D Dellow and J V Nolan, unpublished data).	Figure 7.5: Relationship between microbial nitrogen flow to the abomasum and organic-matter intake in sheep fed clover 01' ryegrass cut at two stages of maturity (Source: D Dellow and J V Nolan, unpublished data).

Table 7.15: Relative palatability of four legume forages and the live weight gains of heifers grazing on them as the sole source of feed.
Legume forage	Relative palatability	LW gain (g/d)
Lucerne	66		0.67
Birdsfoot trefoil	54		0.81
Sanfoin	55		0.80
Cicer milk vetch	4		0.42
Source: Marten and Ehle 1984

Table 7.16: Rates of loss of dry matter (DM) and nitrogen (N) from Leucaena and Gliricidia foliage incubated in nylon bags in the rumen of a goat fed a mixture (60:40 basis) of Gliricidia and Leucaena foliage. Mean values for 10 bags.
	Components of foliage
	Leaflets	Petiole	Green bark	Green bark
DM loss in 24 h, %
Leucaena	44	23	-	44
Gliricidia	69	29	69	41
N loss in 24h, %
Leucaena	31	12	-	21
Gliricidia	43	50	75	67
Source: IFE 1984

Chapter 7

GUIDELINES FOR FEEDING SYSTEMS

7.1 Introduction

7.1.1 Limitations to "conventional" feeding standards

7.1.3 Animal response to non­conventional feed resources