Feeding strategies for cattle and buffalo used for work

R.A. Pearson

The author may be contacted at the Centre for Tropical Veterinary Medicine, Easter Bush, Roslin, Midlothian, Scotland EH25 9RG, UK.

STRAT�GIES D'ALIMENTATION DES BOVINS ET DES BUFFLES UTILIS�S COMME B�TES DE TRAIT

De plus en plus d'agriculteurs utilisent des animaux polyvalents comme b�tes de trait. La conservation des r�sidus de r�colte et le syst�me d'alimentation par affouragement en vert prennent aussi une importance grandissante. Ces pratiques sont la cons�quence in�vitable de la pression d�mographique et de la r�duction des superficies herbag�res disponibles dans les syst�mes caract�ris�s par de petites exploitations combinant les productions animales et v�g�tales. Les besoins �nerg�tiques journaliers des animaux utilis�s comme b�tes de trait repr�sentent rarement plus du double des besoins �nerg�tiques n�cessaires � leur entretien, et sont davantage fonction de la dur�e de la p�riode de travail que de la nature du travail lui-m�me. Les aliments de base des ruminants de trait sont g�n�ralement de pi�tre qualit� (moins de 9 MJ/kg pour ce qui est de l'�nergie de maintenance), et les animaux n'ont pas la possibilit� d'accro�tre leur consommation journali�re pendant la p�riode de travail afin de satisfaire leurs besoins �nerg�tiques accrus. Les agriculteurs s'attendent g�n�ralement que leurs animaux perdent du poids lorsqu'ils travaillent, � moins qu'ils soient en mesure d'am�liorer la qualit� du r�gime. L'article �tudie les strat�gies d'alimentation des animaux utilis�s comme b�tes de trait en tenant compte de ces facteurs et donne des exemples de projets visant � permettre � ces animaux d'ing�rer davantage d'�l�ments nutritifs.

MODALIDADES DE ALIMENTACION DE LOS VACUNOS Y LOS BUFALOS UTILIZADOS PARA EL TRABAJO

Son cada vez m�s los agricultores que utilizan animales polivalentes para el trabajo. Tambi�n est� aumentando la importancia del almacenamiento de los residuos de las cosechas de un sistema de corte y recogida para la alimentaci�n. Estas pr�cticas son una consecuencia necesaria del aumento de la presi�n demogr�fica y la reducci�n de la tierra de pastoreo disponible en los sistemas agropecuarios de peque�os propietarios. Las necesidades diarias de energ�a para el trabajo raras veces son superiores al doble de las de mantenimiento, y dependen m�s de la duraci�n de la jornada de trabajo que del tipo de labor realizada. Los alimentos de base de los animales de tiro suelen ser de escasa calidad (menos de 9 MJ de EM/kg) y los animales no pueden aumentar su consumo diario durante el per�odo de trabajo para satisfacer las necesidades adicionales de energ�a. En general, los agricultores suponen que sus animales perder�n peso cuando trabajan, a menos que la calidad de la alimentaci�n se pueda mejorar. Se examinan las estrategias de alimentaci�n para los animales de trabajo a la vista de estos factores y se citan ejemplos de proyectos cuyo objetivo era mejorar el consumo de nutrientes de los animales de trabajo.

It has been said that "part of the blame for unsuccessful draught animal power projects can be traced back to inappropriate husbandry levels, especially feeding". One can forgive the inexperience of those people for whom draught animal power is a new technology and who may have had little previous experience in feeding draught animals. For example, crop farmers in sub-Saharan Africa, who have traditionally cultivated their lands manually but who are now in a position to use animal power because of the reduced disease challenge to cattle locally, come into this category. Providing inappropriate information when advising draught animal power farmers in their decision-making is felt to be largely to blame. The aim of this paper is to give some guidelines on the nutrition of ruminants being kept for work that will allow future animal power projects to adopt more successful feeding strategies.

Measuring the amount of work done by a team of draught animals in the hills of Nepal
Evaluation du volume de travail effectu� par un attelage d'animaux de trait dans les collines du N�pal
Determinaci�n de la cantidad de trabajo realizado por una pareja de animales de tiro en las colinas del Nepal
Photo/Foto: A. Pearson

A farmer ploughing dryland terraces for winter crops in eastern Nepal
Agriculteur labourant des terrasses non irrigu�es destin�es aux cultures d'hiver dans l'est du N�pal
Agricultor arando bancales de tierra seca para cultivos de invierno en el Nepal oriental
Photo/Foto: A. Pearson

CHANGING NEEDS OF FARMERS

The world population is predicted to increase from 5.4 billion to about 10 billion within a few decades, largely in developing countries. To feed this additional population, more land will be needed for crops, consequently reducing that available for pasture and fodder (Sansoucy, 1995). More crop residues and agro-industrial by-products should then become available for feeding livestock. Farmers keeping livestock in these cropping systems, where grazing land is becoming less available, will have to re-evaluate and adopt different feeding systems. In Africa, the number of small ruminants has tended to increase in proportion to livestock as farm size and grazing areas have declined. This is true except for those areas where draught animal power is used for cropping and transport (de Leeuw and Rey, 1995). For example, Coe (1991) found that in the southern portion of the semi-arid zone and in most of the subhumid zone of Senegal, although population pressure is increasing, work oxen still dominate the livestock enterprises of mixed farmers. Where draught animal power has been a feature of mixed farming, the strategy to overcome the reduced quantity and quality of grazing land has been to move to motorized power or to modify animal husbandry practices.
In parts of South Asia that have high population pressures and are farmed intensively, for example in Viet Nam, Thailand and Malaysia, the reduction in grazing land has meant a change over from buffalo to motorized power in many of the rice-growing areas. However, on small farms of less than 3 ha it is unlikely that draught animal power will disappear, since it can compete economically with petrol-driven tractors, both in Asia (Campbell, 1993) and Africa (Panin, 1995). On farms where motorized power is not feasible, many farmers are modifying their herd structure. The use of multipurpose animals for work removes the need to keep work oxen and hence saves on feed requirements. This is increasingly common where grazing land is declining. Surveys undertaken by the Animal Traction Network for Eastern and Southern Africa (ATNESA) have shown a steady increase in the use of cows for work. In southern and East Africa, donkeys are being used increasingly for land preparation and cultivation instead of oxen. They are easier to feed and survive droughts better than cattle. With the exception of places where it is culturally unacceptable to use cows for work or cattle for meat, the modern draught ruminant is now usually truly multipurpose.
Farmers are having to modify their animal husbandry practices in the more intensively farmed and densely populated areas of Asia and Africa. Animals tend to be tethered or kraaled for longer, both for security and because of the reduction in grazing land. They are often fed on stored crop residues, collected fresh forage or purchased feeds. Therefore, more time is spent collecting and carrying animal feed than previously. It has been reported that, as systems become more intensive, farmers who can afford it rely more on crop residues and purchased, high-quality feeds, and there is greater pressure on communally owned grazing resources. Moreover, greater reliance is also often placed on locally reared stock rather than on stock brought in from grazing systems.
The sufficiency of grazing land in the past has generally meant that draught animals have been left to forage on rangelands, particularly out of season, and supervision of feeding has therefore been minimal. The greater control that cut-and-carry methods of feeding entails may involve the farmers in more decision-making. They could decide on the amount and type of feed that each class of livestock will receive at any one time, although sufficient information would be needed on the outcome of this "strategic" feeding to make the best use of the feed resources available. There are now considerable data available to enable the farmers to make informed decisions, and the same data can also be used by local project advisers to produce focused recommendations on feeding strategies.

FEED REQUIREMENTS

In the 1970s it was generally assumed that energy requirements of working animals were dependent on and proportional to the type of work done. Energy used in work is often expressed as a multiple of 24 hours maintenance requirement, to allow comparisons to be made between animals of different weights. Energy costs of work carried out by cattle and buffalo determined in the laboratory and more recently on farms and plantations, complemented by measurements of the work done by draught animals performing different tasks in a variety of countries, has shown that this assumption is incorrect (Pearson, Lawrence and Smith, in press). Animals carrying out "light" cultivation tasks or carting loads over firm, level roads often expend more energy than when they are doing "hard" work such as ploughing. When doing "hard" work, such as ploughing, an animal moves at a slower rate and stops more often. As a result, the animal uses less energy in walking when doing "heavy" work than when doing "light" work and, over the working day, the total energy expended tends to balance out (Lawrence, 1985). The important point is that the type of work does not have a major effect on total energy expenditure in a given time.
Another important observation from studies of work output and energy requirements of draught cattle and buffalo over the last two decades is that energy requirements for work are lower than predicted in the 1970s. Even when oxen work for six to seven hours a day, their total energy expenditure is rarely over twice the maintenance requirements. Some examples of the estimated energy expenditures of draught cattle and buffalo are given in Table 1.

1
Estimates of the daily energy requirements of draught ruminants as a multiple of maintenance when working under various production systems
Besoins �nerg�tiques journaliers estim�s des ruminants de trait utilis�s dans diff�rents syst�mes de production, exprim�s par leur rapport aux besoins �nerg�tiques de maintenance
Estimaciones de las necesidades diarias de energ�a de los rumiantes de tiro como m�ltiplo del mantenimiento cuando trabajan en diversos sistemas de producci�n

Maintenance energy requirement (MJ/day) = a [b x 0.53 (liveweight/1.08)^0.67 ]
a = 0.71 efficiency of utilization of metabolizable energy for maintenance.
b = 1.10 to account for the increase in metabolic rate on working days.
Energy for work calculated from work done and distance travelled according to Lawrence (1985).

FEED SUPPLY

Location and season determine the feeds to be given to draught ruminants. Most systems make use of the following as staple feeds during the year: weeds and grasses on permanent rangeland, fallows and forested areas; roadside grasses; postharvest stubble; crop residues; and, to a lesser extent, agro-industrial by-products and tree fodder and browse. Farmers in West Java, for instance, used 28 different forages in feeding their draught ruminants throughout the year, of which only seven were described as "good quality" (Santoso, Sumanto and Dharsana, 1993). Studies of feeding practices in other areas of the world reveal a similar picture. Therefore, for most of the year, draught ruminants consume poor-quality forage diets that have a high cell wall content, low nitrogen content and poor digestibility. The metabolizable energy (ME) content of these diets is rarely more than 9 MJ ME/kg and the crude protein, 90 g/kg of dry matter (DM) (Table 2). The start of the cropping season, when draught animals are required to do most work, is usually the time when food stocks are at their lowest. This further exacerbates the problem.

2
Energy and protein contents of crop residues and fodder species used to supplement intake from grazing for draught cattle and buffalo
Teneur en �nergie et en prot�ines de r�sidus de r�colte et esp�ces fourrag�res servant � compl�ter le r�gime herbager des bovins et des buffles de trait
Contenido de energ�a y prote�nas de los residuos de cosechas y de especies forrajeras utilizados como complemento del consumo de pasto de los vacunos y los b�falos de tiro

Sources: McDowell et al. (1974); Thahar and Mahyuddin (1993); Topps and Oliver (1993).

EFFECT OF WORK ON FEED SUPPLY

The effect of work on intake and digestion of feeds by ruminants have been investigated in several different countries and have recently been reviewed (Teleni, 1993; Lawrence and Becker, 1994). When draught ruminants are given high fibre forage diets, low in nitrogen, then food intake and the gastro-intestinal rate of the passage of food are both likely to decrease on working days of five hours or more. Over three months of regular but intermittent work, a gradual increase in average weekly intake has been observed in lactating cows (Zerbini et al., 1993) while steady increases in intake over time on working days as animals adapt have been seen in oxen (Fall et al., 1996).
Significant increases in digestibility as a result of work were reported by researchers in Ethiopia and Nepal, but other workers have reported no effects (Teleni, 1993; Lawrence and Becker, 1994). Fall et al. (1996) suggest that differences in diet quality may contribute to the different responses in digestibility of feed under different conditions.
The major conclusions of these investigations are that any increase in the rate of eating or improvement in digestibility to compensate for reduced time available for eating and ruminating on working days or to satisfy the increased energy demand during working periods are not sufficient to meet all of the additional energy requirement for most types of work when animals are fed low-quality staple diets. In practice, most draught animal farmers expect their animals to lose weight during the working season unless the diet is supplemented with better-quality feed.

Cattle owned by a smallholder farmer in the hills of eastern Nepal
Bovins d'un petit exploitant dans les collines de l'est du N�pal
Vacunos pertenecientes a un peque�o agricultor en las colinas del Nepal oriental
Photo/Foto: A. Pearson

FEEDING STANDARDS

Information on feed requirements and food intake of draught cattle is currently being compiled to produce a manual of feeding standards for such cattle (P.R. Lawrence, personal communication). As well as identifying the expected feed requirements for a given amount of work, tables of feed requirements can be used to predict the weight losses that can be expected from a known quality of feed or, conversely, the quality of diet required to ensure the working animal does not lose weight (Fig. 1). Calculations can also be made of the number of draught animals and the feed that would be needed to keep them in order to cultivate a given area of land. The latter is useful for those who might wish to promote animal traction.

Note: Data are based on predictions of energy expenditure during work and energy intake of draught oxen.
Source: Lawrence and Becker (1994).

1
Estimated change in liveweight of different sizes of oxen consuming diets of different quality and working five hours per day
Variation estim�e du poids vif de boeufs de diff�rentes tailles consommant des aliments de qualit� diverse et travaillant cinq heures par jour
Cambio estimado del peso vivo de bueyes de diversos tama�os con una alimentaci�n de distintas calidades y cinco horas de trabajo diario

LIVEWEIGHT AND BODY CONDITION

The amount of work an animal can do is proportional to its liveweight. Theoretically, the larger the animal (regardless of body condition), the more easily it will carry out a particular task and the less stressed it will be when doing so. A large-framed animal may also be better able to respond to an increasing supply of food during a rainy season than a smaller fatter one. However, animals in good condition have "fuel" in reserve, which may be mobilized to make up any shortages in food. Despite the apparent benefits of having animals in good condition at the start of work, there is little conclusive evidence to show that such animals work faster and/or longer than those in poor condition. Nor is there evidence that crop yields are improved when oxen are given feed supplementation in the dry season. Studies in Mali (Bartholomew, Khibe and Little, 1994) and the Niger (Fall, Pearson and Fern�ndez-Rivera, 1996) confirmed that liveweight and not body condition is the main determinant of work capacity in oxen. A similar conclusion was reached by Teleni (1993) studying village buffalo in Asia. He suggested that weight loss in working buffalo is not a problem provided the "critical weight for sustained draught load is not compromised", about 12 to 15 percent liveweight. The implications are to select large-framed animals for draught purposes.
Fall, Pearson and Fern�ndez-Rivera (1996) recommended that a body condition score, as defined by Nicholson and Butterworth (1986), of between 2 and 3 would be a critical score, below which work may irreversibly damage the oxen's health. The ideal body condition score would be between 4 and 6. Oxen with a body condition score of more than 6 may be too fat to move comfortably and would be more susceptible to heat stress than leaner oxen. Moreover, the feeding level required to reach a score of more than 6 is unlikely to be profitable.

Draught oxen feeding at the roadside near Maradi in the Niger
Alimentation des bœufs de trait au bord de la route pr�s de Maradi, au Niger
Bueyes de tiro comiendo al lado de un camino cerca de Maradi, N�ger
Photo/Foto: A. Pearson

A young draught ox eating maize stover
Jeune bœuf de trait ing�rant des tiges de ma�s
Buey joven de tiro comiendo paja de ma�z
Photo/Foto: A. Pearson

Supplementary feeding for a donkey and ox used for transport in the south of the Niger
Alimentation d'appoint d'un �ne et d'un bœuf utilis�s pour le transport dans le sud du Niger
Alimentaci�n complementaria de un asno y un buey utilizados para el transporte en el sur del N�ger
Photo/Foto: A. Pearson

SUPPLEMENTING STAPLE FEEDS

Clearly the economics of feed supplementation, particularly in the dry season, have to be carefully considered. They depend mainly on the productivity expected from the animals.
Where religion prevents the use of cows for work or the rearing of cattle for meat, and draught animals are only expected to provide manure in addition to work, then economic returns of feed supplementation are likely to be small. Similarly, where work periods are short (20 to 30 days), returns from supplementation may also be small, since animals will have ample time to make up any losses during the rest of the year. However, if work is performed over longer periods, then supplementation can be beneficial. The aim should be for the adult animal to have maintained its liveweight over the year, although it may have shown some quite dramatic seasonal fluctuations.
Where animals are to be sold for meat after work, supplementary feeding is worth while, even when the animals are worked for short periods. The aim of supplementation in these cases should be to increase the animals' liveweight over the year.
There is now considerable evidence to show that, unless draught cows are given good-quality feed, particularly during work, the associated weight losses will result in production losses. The aim should be to feed to maintain liveweight. This can be difficult when requirements for pregnancy and lactation as well as for work have to be met. Liveweight loss in working female cattle and buffalos can lead to reduced ovarian activity and longer calving intervals (Winugroho and Situmorang, 1989; Zerbini et al., 1993). Total milk yield can also decrease on working days (Matthewman et al., 1993). The use of cows for work offers the greatest management challenge to farmers. They need to insure that not only is there enough feed to maintain production, but that calving does not coincide with the peak demand for draught power.

TIMING OF FEED SUPPLEMENTATION

During the working season, timing of feed supplementation on working days may be important. Supplementary feeding after work in ruminants minimizes intraruminal heat production while working. Conversely, it has also been suggested that feeding at least two hours before work insures a ready supply of energy-yielding substrates for working muscle. Farmers adopt both these strategies; however, it is not clear whether this is an informed decision or based on traditional practices by those unaware of any alternatives. Labour availability is often overlooked when recommending feeding practices. Children have traditionally been responsible for supervising grazing animals in many areas as well as being involved in cut-and-carry activities in others. Schooling can leave the farmer with a shortage of labour, thereby reducing grazing time and/or the amount of forage that can be cut. In some areas, men work elsewhere and women run the farm. While this increases the household income, it also places constraints on the labour available to manage livestock. Schemes to improve the diet of working animals through the digestibility of the staple feed or by supplementation need to bear in mind these pressures on the labour pool.

STRATEGIES FOR IMPROVING FEED SUPPLY

One of the easiest ways of improving the diet of a draught animal is to save crop residues and bush hays for periods when feed from grazing is of poor quality and in short supply. In many intensively farmed systems in Asia, this is a traditional practice. In parts of sub-Saharan Africa farmers have only recently begun to collect and store crop residues regularly rather than leaving them to be grazed communally by village livestock. It is now well recognized that crop residues are more optimally used when animal-drawn transport is available to allow on-farm storage (de Leeuw and Rey, 1995). However, most crop residues, with the exception of legume hay which is rich in crude protein and highly digestible, are unlikely to supply sufficient energy to maintain liveweight (Table 2 and Fig. 1), although they can be a valuable supplement. Clearly, when weight loss is not advisable, then other strategies have to be considered.
In cut-and-carry systems, strategic feeding can be practised when farmers do not have the means or desire to supplement the staple forage. If animals are offered sufficient feed to allow a good degree of selection of material (150 percent of expected ad libitum intake), then working ruminants can in some cases maintain weight on forages that would normally not be expected to supply sufficient energy for maintenance let alone for production. For example, millet stover leaves, which cattle tend to feed on selectively when given the opportunity, can have a mean crude protein content of 70 to 110 g/kg of DM whereas whole plant stover, including the stem, may contain only 40 g/kg of DM (Powell, 1985). Similarly, Reed, Kebede and Fussell (1988) reported a higher digestible neutral detergent fibre (DNDF) of leaf blades (55.7 to 62.2 percent) than that of the stem (27.6 to 35.2 percent). By allowing working animals to feed on crop residues first thing in the day, farmers can insure that they have the best that is available.
The purchase of cotton seed, groundnut cake, rice bran, maize flour or other concentrate feeds to improve the diet usually requires the lowest labour input but is expensive and therefore not economic for many farmers. Other strategies involve supplementation with locally available cheaper materials or treatment of the basal feeds with urea or alkali. Most practices aim to improve microbial breakdown of the basal feeds in the rumen, thus improving intake and digestibility. The benefits of these techniques are well researched and widely reported but their adoption by farmers is uneven. While they have been successful in some areas (Sourable, Kayouli and Dalibard, 1995), adoption has been poor in others.
Acceptance and adoption by farmers of these methods to supplement or improve the feeding value of basal feeds may often depend on the outputs that are expected from the draught animal. Feeding practices such as urea treatment of crop residues, the production of molasses-urea blocks, the planting of additional forages, supplementation with agro-industrial by-products and the purchase of concentrates all involve extra labour and/or cost. Immediate cash returns from transport, milk or the hiring out of the animals encourage farmers to put time and money into improved feeding. It is more difficult for farmers to accept modifications in their feeding and management practices to enhance nutrient intake when animals are only used in crop production and returns from meat are not so immediate. Economically it may not even be justified. In sub-Saharan Africa, for instance, there are few examples of forage crops being introduced for meat production or to support draught animals. The cost of fencing and sward establishment and the level of management necessary to produce good forage yields make it uneconomic, except in the intensively farmed areas with good markets for milk production (McIntire, Bourzat and Pingali, 1992).

A mule carrying forage collected by children near Toluca in Mexico
Mulet transportant du fourrage r�colt� par des enfants pr�s de Toluca, au Mexique
Mulo transportando forraje recogido por ni�os cerca de Toluca, M�xico
Photo/Foto: A. Pearson

Tree fodder (Ficus auriculata) cut to feed draught cattle in eastern Nepal
Arbre fourrager (Ficus auriculata) coup� en vue de nourrir des bovins de trait dans l'est du
Corte de forraje de un �rbol (Ficus auriculata) para alimentar vacunos de tiro en el Nepal oriental
Photo/Foto: A. Pearson

FEEDING STRATEGIES

There have been several projects to raise the nutritional status of work animals. Some have been more successful than others. The following examples have been chosen to illustrate some of the factors which have a bearing on the success or failure of these feeding strategies.
In Mangwende, an area which is representative of the high-potential communal areas of Zimbabwe and which has an annual rainfall of about 800 to 1 000 mm, animals tend to be in poor condition when needed for work. Maize stover is one of the main feeds available to supplement rangeland grazing. Traditionally, farmers leave the stover to be grazed in situ; others collect and store it on the farm. However, during the late dry season the stored stover is piled into the night kraal so it is available to all animals. An improved management system was tried (Chikura, 1994), whereby farmers harvested and stored the maize stover and then, in the late dry season, fed it separately to their work oxen; approximately 6 kg/day between their finishing work and being returned to the herd. A block lick (24 percent crude protein and 5 percent urea, Rumivite Economy) was also made available in each pen for the hour allowed for supplementary feeding. During the three-month working season (November to January), there was no significant difference in the rate of liveweight gain between the supplemented group (200 g/day) and unsupplemented animals (170 g/day), nor was work output or frequency of work any different between groups. In this case, supplementation had no benefit on outputs. Farmers were mainly interested in ploughing their fields as early as possible and would do so whatever the condition of their oxen and whatever the stress. Farmers knew the animals would recover later in the season and could see little benefit in adopting the modified feeding practice (Chikura, 1994).
In three departments of the Niger (Tillaberi, Dosso and Maradi), urea treatment of roughages (rice straw, sorghum and millet stalks) to increase digestibility has been adopted by many farmers. A drought had resulted in increasing the concentration of cattle in these important crop producing areas. Poor growth and health of the livestock were identified as problems by the farmers, who normally had a reliable market in which to sell their cattle for meat in the end. Unlike the previous example in Zimbabwe, farmers were therefore receptive to feeding strategies to improve the condition and health of their animals. Uptake has been good, mainly because the technique proved to be cost-effective - especially since the draught animals had a greater capacity for work - fattening was less time-consuming, fewer supplements were needed and the final market price was higher (Sourable, Kayouli and Dalibard, 1995).
In Bangladesh, Saadullah et al. (1994) reported that the provision of urea-molasses blocks to draught cows fed ad libitum on rice straw and allowed to graze two to three hours daily was associated with an increase in feed intake, a daily liveweight gain for dams (79 versus 44 g/day) and calves (316 versus 160 g/day), higher calf birth weight (16 versus 14 kg) and reduced calving interval (685 versus 691 days). Daily milk yield (530 versus 243 g) and lactation length (275 versus 230 days) were also increased. Productivity was compared with that of cows supplemented with wheat bran in place of the molasses block. Supplementation was seen by the farmers as a necessary consequence of using cows for work and most were concerned that the most effective supplement be used. As a result they tended to prefer the urea-molasses block.
In West Java, producers of draught animals face a severe decline in forage quality and quantity in the late dry season (October to November) as well as a severe labour shortage for collecting forage in the early wet season (November to January). These shortages are responsible for poor weight gains as well as stretching family labour resources. The introduction of suitable forages was seen as a relatively inexpensive method of overcoming these constraints and one which farmers were likely to adopt (Petheram, Goe and Astatke, 1989). Early trials, with farmers selecting areas to plant the forages on the farm and assessing the usefulness of various species, were promising. However, practical difficulties, such as which species to use, survival and the management required, particularly in the early stages of the forages, meant that considerable input had to be made before the benefits were apparent. As a result, adoption has been low. A similar delay in benefit is often seen in schemes to introduce tree fodders as supplementary feeds although, once selected, planted and protected, little extra labour input or cost is required.
The main conclusion to be drawn from these and other projects is that recognition of the considerations that influence the feeding and managing of draught animals on specific farms is essential. This allows recommendations on feeding strategies to be proposed that are likely to meet with the approval of farmers and therefore have a high adoption rate. This may be obvious, but it is surprising how often it is overlooked at the planning stage. These considerations and the involvement of farmers in the identification of the appropriate feeding intervention are the key to developing successful feeding strategies for working cattle and buffalos.

CONCLUSIONS

There is now good information on the feed requirements of animals used for work and on the effects that work has on nutrient intake. Research and farmers' experience have shown the need to supplement the basal feed of most work animals. The exceptions to this are the non-productive animals that are only kept for work and used for short periods in the year (20 to 30 days each year) and those which have access to an abundance of good-quality feed throughout the year. While there is a range of technologies to improve nutrient intake of animals fed largely on low-quality forages, the key issues in the feeding of animals used for work are how much of what and when? Little benefit has been reported of supplementation during the dry season on subsequent work output in semi-arid areas of sub-Saharan Africa, but studies in the Indian subcontinent and Southeast Asia report more favourably on the benefits gained. This is clearly a subject which requires further study. The challenge for the researchers and extension officers is to develop, in consultation with farmers, acceptable and workable modifications to traditional feeding practices which will enable them to maintain fit and healthy draught animals, in an environment of increasing population pressure, decreasing grazing land and often declining labour levels.

Bibliography

Bartholomew, P.W., Khibe, T. & Little, D.A. 1994. Effect of liveweight and body condition on work output from draft oxen. Trop. Anim. Health Prod., 26: 58-64.
Campbell, R.S.F. 1993. Draught animals in the AAAP zone and their economic future. In W.J. Pryor, ed. Draught Animal Power in the Asian-Australasian Region, p. 10-16. Workshop with 6th AAAP Societies Congress, Bangkok, Thailand, 23-28 November 1992. ACIAR Proceedings Series No. 46. Canberra, ACIAR.
Chikura, S. 1994. Improving the management of feed resources for draft animals in Mangwende, Zimbabwe. In P. Starkey, E. Mwenya & J. Stares, eds. Improving Animal Traction Technology, p. 162-163. Proceedings of First ATNESA Workshop, Zambia 1992. Wageningen, the Netherlands, CTA.
Coe, M.R. 1991. Le b�tail trypanotol�rant pour la traction en Afrique occidentale et centrale. Bulletin de liaison sur le b�tail trypanotol�rant, 4: 5-10.
de Leeuw, P.N. & Rey, B. 1995. Analysis of current trends in the distribution patterns of ruminant livestock in tropical Africa. World Anim. Rev., 83(2): 47- 59.
Fall, A., Pearson, R.A. & Fern�ndez-Rivera, S. 1996. Nutrition of draught animals in semi-arid West Africa. III. Effect of body condition prior to work and weight losses during work on food intake and work output. Anim. Sci. (submitted)
Fall, A., Pearson, R.A., Lawrence, P.R. & Fern�ndez-Rivera, S. 1996. Nutrition of draught animals in semi-arid West Africa. II. Effect of work on intake, digestibility and rate of passage of food through the gastro-intestinal tract in draught oxen fed on crop residues. Anim. Sci. (submitted)
Lawrence, P.R. 1985. A review of the nutrient requirements of draught oxen. In J.W. Copland, ed. Draught animal power for production, p. 59-68. ACIAR Proceedings Series No. 10. Canberra, ACIAR.
Lawrence, P.R. & Becker, K. 1994. The effects of draft power on feed utilisation and reproduction in cattle and buffaloes. In K. Becker, P. Lawrence & R. �rskov, eds. Sustainable Small-scale Ruminant Production in Semi-arid and Sub-humid areas, p. 101-121. Proceedings of an International Workshop, University of Hohenheim, Stuttgart, Germany.
McDowell, L.R., Conrad, J.H., Thomas, J.E. & Harris, L.E., eds. 1974. Latin American Tables of Feed Composition. Gainsville, USA, University of Florida Press. 509 pp.
McIntire, J., Bourzat, D. & Pingali, P. 1992. Animal fodder interactions. In J. McIntire, D. Bourzat & P. Pingali, eds. Crop-livestock interactions in sub-Saharan Africa, p. 103-133. Washington, DC, World Bank.
Matthewman, R.W., Merritt, J.B., Oldham, J.D. & Horgan, G.W. 1993. Effect of exercise on milk yield, milk composition and blood metabolite concentrations in Hereford x Friesian cattle. Asian-Australasian J. Anim. Sci., 6: 607-617.
Nicholson, M.J. & Butterworth, M. 1986. A guide to condition scoring of zebu cattle. Addis Ababa, ILCA. 36 pp.
Panin, A. 1995. Empirical evidence of mechanisation effects on smallholder crop production systems in Botswana. Agric. Syst., 47: 199-210.
Pearson, R.A., Lawrence, P.R. & Smith, A.J. The Centre for Tropical Veterinary Medicine (CTVM) pulling its weight in the field of draught animal research. Trop. Anim. Health Prod. (in press)
Petheram, R.J., Goe, M.R. & Astatke, A. 1989. Approaches to research on draught animal power in Indonesia, Ethiopia and Australia. Graduate School of Tropical Veterinary Science, James Cook University, Townsville, Australia. 115 pp.
Powell, J.M. 1985. Yields of sorghum and millet and stover consumption by livestock in the subhumid zone of Nigeria. Trop. Agric. (Trinidad), 62: 77-81.
Reed, J.D., Kebede, Y. & Fussell, L.K. 1988. Factors affecting the nutritive value of sorghum and millet crop residues. In J.R. Reed, B.S. Capper & P.J.H. Neate, eds. Plant breeding and the nutritive value of crop residues, p. 233-249. Addis Ababa, ILCA.
Saadullah, M., Sarker, D.R.D., Huq, M.A., Samad, A. & Islam, S. 1994. Study on the effect of replacement of wheat bran by urea molasses block lick in straw-based diet on the performance of draught cows. Bangladesh J. Anim. Sci., 23: 163-168.
Sansoucy, R. 1995. Better feed for animals: more food for people. World Anim. Rev., 82(1).
Santoso, Sumanto & Dharsana, R. 1993. Draught animal systems in Indonesia. 1.4. West Java. In E. Teleni, R.S.F. Campbell & D. Hoffmann, eds. Draught animal systems and management: an Indonesian study, p. 21-31. ACIAR Monograph No. 19. Canberra, ACIAR.
Sourable, K.M., Kayouli, C. & Dalibard, C. 1995. Le traitement des fourrages grossiers � l'ur�e: une technique tr�s prometteuse au Niger. World Anim. Rev., 82: 3-13.
Teleni, E. 1993. Energy expenditure and nutrient requirements of working animals. In W.J. Pryor, ed. Draught Animal Power in the Asian-Australasian Region, p. 93-99. Workshop with 6th AAAP Societies Congress, Bangkok, Thailand, 23-28 November 1992. ACIAR Proceedings Series No. 46. Canberra, ACIAR.
Thahar, A. & Mahyuddin, P. 1993. Feeding and breeding of draught animals. 2.1. Feed resources. In E. Teleni, R.S.F. Campbell & D. Hoffmann, eds. Draught animal systems and management: an Indonesian study, p. 41-54. ACIAR Monograph No. 19. Canberra, ACIAR.
Topps, J.H. & Oliver, J. 1993. Animal Foods of Central Africa, 2nd ed. Technical handbook. Zimbabwe Agric. J.
Winugroho, M. & Situmorang, P. 1989. Nutrient intake, workload and other factors affecting reproduction of draught animals. In D. Hoffmann, J. Nari & R.J. Petheram, eds. Draught animals in rural development. ACIAR Proceedings Series No. 27. Canberra, ACIAR.
Zerbini, E., Gemeda, T., Franceschini, R., Sherington, J. & Wold, A.G. 1993. Reproductive performance of F₁ crossbred dairy cows used for draught: effect of work and diet supplementation. Anim. Prod., 57: 361-368.