The use of crop residues for livestock feed by small farmers in the Gambia

Sandra L. Russo
Animal Science Department University of Florida Gainesville, Florida 32606, USA

Abstract
Introduction
Methodology
Results
Discussion and implications of results
A framework for farmer acceptance
Conclusions
References

Abstract

The Gambia, one of the smallest countries in Africa, was the location of The Mixed Farming and Natural Resource Management Project from 1981-86. The goals of the project were to intensify and integrate crop and livestock enterprises in existing Gambian farming systems. Cattle production was to be improved through better nutrition during the dry season, by range management techniques including deferred grazing, improvement of rangelands through re-seeding and root transplants, and the storing and feeding of crop residues. Village-level feeding trials were conducted in four villages in 1983 and had increased to 33 villages by 1986. Not all land innovations were used in each village. Socio-economic factors such as the land tenure system, social value of cattle, high labour requirements, limited market access, low land pressure and farmers' goals had a profound effect on acceptance of the innovations. Development and farmer goals must converge before acceptance and adoption will occur.

Introduction

The climate of The Gambia is characterized by a distinct, short rainy season from July-September and a long, dry season in which rain never occurs. Total annual precipitation is 600-700 mm. The sharp dichotomy between rainy and dry seasons is the dominant feature of the agricultural environment. The major cash crop and source of foreign exchange is groundnuts; the preferred crop food is rice.

Gambian farming systems are oriented around production of groundnuts, rice and other cereals. Fallow lands around the village, forests and swamps are considered to be communally owned for grazing as are cultivated lands after harvest. The farming systems are mixed agriculturally, with farmers producing crops and owning livestock. Small stock such as poultry and small ruminants are owned by many and are usually kept in the compound areas. Draft animals (horses, donkeys and oxen) are used extensively, primarily for transport. Everyone aspires to own cattle, even urban dwellers, but only one ethnic group (Fula) historically has experience in cattle management. Cattle are of the N'dama breed which is fairly trypanotolerant.

The major feature of cattle production is the way the herds are managed and their migration patterns. Animals are grouped together into one or more herds and are cared for by a hired herder who is most often Fula. The herder is responsible for keeping the animals away from the croplands during the growing season and for assuring that they have feed and water most of the year. Herders are paid on a per head basis and, when near the village, must share any milk with the owners. Water and forage availability are the most important factors when herders make migration decisions.

Methodology

The Mixed Farming and Natural Resource Management Project (MFP) was a joint project of the Gambian government and U.S.A.I.D (United States Agency for International Development) started in response to the growing pressure of both human and livestock populations on Gambia's limited land resource base. Cultivation of former grazing areas, especially swamplands, led to conflicts for food versus feed production. In 1984-85, for example, cattle population increased by 7.3% while human population increased by 3.4% (Vesseur et al., 1986). Both the Gambian government and U.S.A.I.D. felt that crop and livestock production needed to be integrated.

The project concentrated its activities on bushland and upland rainfed agriculture, working with farmers and livestock owners to determine how an adequate year-round forage supply for their livestock could be produced. The first two years of the project were involved with collecting baseline information on the farming systems, research on maize and forages, on-station livestock feeding trials, range management and seed multiplication. Although the project worked country-wide, range activities were concentrated in the two most eastern divisions, McCarthy Island Division (MID) and Upper River Division (URD). The technologies were designed to be extended easily through a combination of modest governmental support and strong farmer acceptance. Technologies were to be such that farmers could continue to use and adopt them regardless of government support.

Although there continued to be several distinct activity areas for project staff, a serious attempt was made to integrate all of these activities by 1983-84. Interdisciplinary village-level activities became the operative words for the final two years. The range management and forage production programmes operated on both the research stations and in villages. Research station work was directed toward forage introductions, seed multiplication of grass and legume species, intercropping experiments with forages and cereals, and cattle feeding trials using crop residues. Objectives of the village feeding trials were to defer an area from grazing near the village, improve it, encourage the storage of crop residues, and develop a year-round feeding strategy for cattle.

Groundwork for the village programmes was laid early in the project by meetings with Livestock Owners Associations (LOA). Villages were asked to set aside at least two hectares of land for deferred grazing and range improvement. One hectare was left unfenced, the other hectare was fenced with wire mesh and barbed wire, using creosote-treated Gmelina arborea posts and steel fence posts. No attempt was made to utilise local fencing materials. Fire lanes were made around each plot. Plots were seeded with Cenchrus ciliaris, Stylosanthes hamata, S. guianensis, S. scabra, Andropogon gayanus, and Panicum maximum. The latter two were also root transplanted. Two weedings the first year were essential for good stand establishment and grazing had to be deferred for two rainy seasons. Shrubs were controlled in some areas with herbicide injections of 2, 4-D. The project provided fencing, seeds, herbicides, and transport while the villages provided labour.

Crop residues have been an underutilised resource in The Gambia. After harvest of the various cereal crops, the residues are left lying in the fields. When the animals return to the village areas after the cropping season, they are allowed to graze on these residues but trampling spoils much of this available fodder. The project encouraged the saving and storing of these residues in a variety of ways. The only crop residue that had been used to any great extent was groundnut hay which was saved for use by draft animals and small ruminants over the dry season. Farmers were encouraged to be more efficient in gathering of the groundnut hay as it was by far the best quality feed for their animals.

Four villages were selected to participate in actual feeding trials in 1984, all of them in the eastern part of the country. These were:

Villagers were asked to contribute two animals each for the feeding trials, one of which would remain with the village herd under normal management conditions, the second of which would participate in the feeding trial. Long, protracted discussions took place concerning what class of animals should be fed. Two types were considered good candidates by the villagers - the old, sick and lame cattle or young heifers. The latter were chosen with a view toward the possibility of calving earlier than the 4-5 year norm. This decision to choose heifers was not without a great deal of project staff input. Almost all of the farmers in these villages, regardless of whether or not they owned cattle or belonged to a Livestock Owners Association, contributed some of their crop residues for the feeding trial.

The feeding trials started in January 1984. The heifers were dewormed, ear-tagged and either weighed (at Boiram) or taped for girth measurement. The "fed" group were placed in the enclosed area where the crop residues had been stored and were fed there for three months. Crop residues were maize, sorghum and millet stovers, rice straw, Andropogon and Panicum hay, and/or groundnut hay. The heifers were either taken to water once a day or water was brought to them. They then went onto the deferred range area for two months and finished the dry season back in the enclosed areas on groundnut hay. By 1986, 33 villages were participating in the feeding trials. Not all villages had every component of the package. In many villages only crop residues were stored and fed without any deferred grazing areas being developed. These latter villages were also participating in the project's maize programme so that maize stover was the primary crop residue saved in 27 of the 28 villages which had only a supplemental feeding programme. Estimates of the amount of maize stover used in 1986 were over 80 tons on a dry matter basis.

Results

Boiram was the only village where we were able to place a scale so quantitative data are scarce. In 1984, heifers in the fed group lost an average of 2 kg over the six month period while the heifers that remained with the herd lost an average of 10 kg, representing 3 and 13% of their body weights. An attempt was made to record girth measurements of the heifers in the other villages. In all of the village programmes, elicited responses from villagers were that the heifers in the feeding trials looked "better" than those not in the trial but these are purely subjective data.

Watering was continually a problem during the feeding trials and methods of watering were modified on a village to village basis. Water was either brought to the enclosed area or animals were trekked to the nearest water source whether it was a well, a swamp, or the river. The watering problem caused the most discussions among the villagers and led to the abandonment, in some cases, of cooperation between villages and a strike, in another case, by village women who refused to draw well water for the heifers. Several villages hired a herder on a part-time basis to take the animals to water once a day.

In the deferred-range areas there were significant improvements in the amount of biomass on offer provided that the fence had remained in place, the livestock had been kept out, that the area wasn't burned accidentally or intentionally or that the land wasn't claimed for crop production. Utilisation of biomass in the deferred grazing areas ranged from 29 - 58% with stocking rates of 6 - 12 TLU (Tropical Livestock Unit)/ha in the 1985 season. In Pinai-Choya a bush fire accidentally set by honey seekers destroyed most of the area four days after the trial started in March 1985. Fortunately, the stored crop residues were not burned.

During the first year of the feeding and deferred range programme it was observed that the weights of the control group did not start to drop until the end of March after the heifers had been in the feeding trials for three months. As a result, the following year's feeding period was changed to start in April. This was an advantage in that less labour was required to store crop residues. The actual manner of feeding the crop residues was also changed by villagers from the recommended method of separate storage and feeding areas with communal feeding of all the heifers once a day. Some villages tethered the heifers inside the enclosure with individual owners placing a quantity of crop residues before their animal once or twice a day. Other villages simply allowed all animals to enter the storage area freely, eating from stocks at will. Another adaptation of the crop residue feeding was that it changed from being a communal effort by several villages to single village efforts to, by 1986, individual efforts. In other words, owners began to save their own residues for their own animals in their own compounds. Farmers expressed surprise that their cattle would eat so much of these crop residues, in particular the maize, sorghum and millet stovers and the rice straw. Utilisation of the stovers was as high as 60% and up to 100% for groundnut hay (Deffendol, 1986).

Discussion and implications of results

The role of Animals in Gambian Farming Systems
Land Tenure Effects on Livestock, Crop, and Forage Production
Forages in Gambian Farming Systems
The success of the Crop Residue Feeding Innovation

The role of Animals in Gambian Farming Systems

Although livestock owners originally agreed that feeding young heifers made sense, as time went by, the older, sick and lame animals found their way into the feeding programme. This is the first example of the conflict between development and farmer goals. Development can be defined as increased total production of crops or animal products, also involving improvements in efficiency, e.g., increased output per unit of labour and increased output per unit of land. The development expert would thus see the long term goal of improved fertility of young heifers as the logical endpoint of a feeding programme. However, the Gambian livestock owner, whose goal is maximisation of total number of animals, sees the survival of all his or her animals right now as the more logical and attainable end. In the event that project staff were right about improved nutrition improving fertility, the owners placed a bull in with those heifers that were being fed additional feed in the feeding trials.

Project interviews of cattle owners at the central abbatoir indicated that owners do not sell their livestock at a seasonally opportune time. Over 3000 observations were made of slaughters in 1985 disaggregated by sex and age (Russo and Spencer, 1988). Male cattle made up 80% of the kill in February when cattle were declining in weight and body condition. In July the cattle kill of males dropped to 50% when cattle were beginning to gain weight. Yearly average kills for 1985 were 63% male and 36% female. Table 1 indicates the contribution of livestock to farmer income in The Gambia. Total income from livestock represents, however, less than 25% of the total farm income. livestock owners in The Gambia as in other West African countries look upon livestock as a form of risk reduction and security. Livestock contribute to farmers welfare in ways other than income or monetary considerations. In a livestock survey in 1985, owners indicated that even with more reliable markets they would be reluctant to sell any animals until cash needs were sufficiently demanding e.g. procurement of food for the family.

Table 1. Contribution of livestock to farmer income, The Gambia, 1985.

Inputs into livestock management are labour, feed and capital, usually drugs (Table 2). Labour per herd is higher for cattle but lower per head. Note that cattle, sheep and goats rarely received any feed "hire draft animals always do. Looking at the economic efficiency measures in Table 3, cattle are indicated as having the highest net earnings per unit of labour and per head for the herds studied in these four villages. A significant reason for the high return per unit of labour is the use of contract herders. Because the herds are taken away from the crop areas during the rainy season, compounds (the farm household unit) incur no labour costs for cattle production. Sheep and goats show better earnings per TLU and 1 Dalasi cost than cattle (see Table 3), yet cattle are still preferred by most Gambians. This suggests that social considerations regarding cattle ownership are more important than monetary values to Gambian livestock owners.

Land Tenure Effects on Livestock, Crop, and Forage Production

The land tenure system in The Gambia does not function along Western (capitalistic) understanding of land ownership. Western land ownership is based on the concept of freehold while in The Gambia and much of rural Africa, the landholder has a very different set of rights. Land is to be used for specific purposes with certain restrictions, most notably, the restriction on transfer of rights. The original cultivators of the land have usufruct right that is heritable but not alienable. Moreover, lack of use means the land reverts to the village and can be reassigned by the village chief or elders. The grazing lands are open common to which every Gambian has free access. Grazing land includes almost all national territory which is not built on, cultivated or set aside as a national forest reserve (Eastman, 1986).

Table 2: Inputs utilised in livestock management in four Gambian villages by 24 farmers, 1985-86

^aADE = Adult Day Equivalent

Labour time from males and females 13-59 yrs is 1.0
Other ages is 0.5. Total weighted hours are divided by 8 to calculate ADE

^bIn Dalasis, 1 US = 7 D.

^cOne farmer fed groundnut hay to a single sick cow in June

Table 3: Economic efficiency measures for six livestock types of 24 farmers, 1985-6

^aSee Table 2 for definition

^bTLU = Tropical Livestock Unit
1 cattle =.8 TLU
l sheep =.1 TLU
1 goat =.07 TLU
1 ox =.8 TLU
1 horse = 1.0 TLU
1 donkey =.52 TLU

^cnumbers for draft animals do not include value of draft services

Conflicts over intrusions into crop lands are both frequent and intense yet very few conflicts are reported over grazing lands. Gambians do not believe that the livestock belonging to other Gambians should be excluded from grazing lands. Death losses, while serious in dry years, are still not perceived as unacceptably high by livestock owners. Lack of forage in the late dry season is not seen as a serious problem. Fencing an area that is not cropland to keep cattle out of it goes against their traditional beliefs about open commons yet forage production and rangeland improvement required exclusive control over the land just as crop production does. Experience elsewhere in Africa has shown that controlled grazing is extremely difficult to achieve in democratic societies and is even difficult to achieve with coercion. The majority of farmers disapprove of fencing and controlled grazing because they do not see any benefits accruing directly to them. Livestock ownership therefore assumes the role of land ownership, i.e. the more livestock an individual owns the more the benefits of the community's land accrue to that individual (Hopcraft, 1981). Still, many Gambian villages have grazing areas where they have already established more-or less exclusive customary rights through years of continuous use and would be the type of villages in which to initiate deferred grazing and fencing schemes. Makama Sereh, the very isolated village in these studies, had the most success with deferred grazing because there were no intrusions from outside livestock.

Forages in Gambian Farming Systems

The concept of a crop that is planted solely for use by animals is totally outside Gambian experiences. Livestock owners were always impressed by the numerous types of forages they saw at field days on the research stations but except for less than half dozen owners, most did not care to try forage production for their herds. Every attempt at forage production in the villages, especially in the case of fodder banks, met with misunderstandings and conflicts. Land presumably given for several years to the project for legume fodder banks would get reassigned by the chief for crop production in the second year. Often, livestock themselves would unfasten the gates of the fodder bank and totally defoliate it. The production of forage crops conflicts with food crop production for both land and labour. Improvements of the deferred grazing areas, for example, took place at the same time that farmers were busy with their food crops.

Two conflicting forces operate against including forages in either small pastoral/semi-pastoral productions systems (McIntire and Debra, 1986). One is land abundance in rural areas "here herds are highly mobile, such as in The Gambia. Forages compete with rangeland and crop residues which can be grazed with very little labour input. The second force operating against the inclusion of forages into the system is land scarcity near urban areas, such as Dakar, Senegal, or in rural areas where land is scarce due to high population pressures, such as Western Kenya. There, forages compete with food crops for land while more crop residues are available. Table 4 indicates these conflicts and interactions and includes the use of crop residues. One interaction is that crop residues can be transported to urban areas and sold; another interaction is that the herds can move to the rural areas. In The Gambia, the advantages of livestock production as practiced currently in terms of labour saving alone make it difficult to convince farmers of the value of forage production per se.

Finally, many villagers looked at the imported fencing materials and imported seeds and saw that this technology was beyond their reach. Not only was it capital intensive in that fencing 10 hectares cost approximately US $2000 but also no attempt was made to use locally available fencing materials or local forages. Thick, strong fencing of garden plots, compounds and mosques is quite common, for example. Herdsmen were asked (Lawry, 1987) if they thought it was possible to have deferred grazing areas without fencing and they were very doubtful, citing the tradition of the open commons as the reason. As for the plant species used for renovations of the sites, Andropogon is the only locally available species, all of the other species are imported.

One of the project's goals was to develop a forage production programme in the Ministry of Agriculture, both in agronomy and range management. As discussed above, forage production in The Gambia faces a number of obstacles. It will occur if farmers see it as providing benefits to them and their livestock. That it will do so is not yet apparent to Gambian farmers. The Mixed Farming Project was mandated to begin a forage agronomy programme which the agronomists started in the usual way, e.g., plant introductions, variety testing and fertilizer trials. This has been termed the notional stage of technology evaluation and is the norm for the beginning of most projects (Menz and Knipscheer, 1981). While such research definitely needs to be done in The Gambia, it should not have been the sole focus of the forage programme because improved forages and pastures do not fit into current Gambian farming systems. Nor, as discussed earlier, will forages fit into many other African farming systems with abundant land resources.

Table 4. Food Vs. Forage Vs. Crop residues

Adapted from McIntire and Debra (1986)

The success of the Crop Residue Feeding Innovation

The only well-accepted innovation was the adoption of crop residue storage and feeding. As mentioned earlier, farmers have their groundnut hay for use by draft animals and small ruminants during the dry season. Therefore. it was not a major technological leap, complicated management decision, nor sociological blunder to ask them to save other crop residues. Four years of research station feeding trials using every type of crop residue in several combinations showed that weights of growing animals could at least be maintained, if not increased, over the dry season (Hedrick and Bojang, 1983; Russo and Ceesay, 1986). This was corroborated in the village studies by livestock owners who could see these benefits immediately. Furthermore, in 1985, the rains were late so that only those villages with stored crop residues had feed for their animals. This made a deep impression on livestock owners and probably contributed to the adoption of this innovation that year.

The major problem encountered was the amount of labour required to store the crop residues. People are busy harvesting crops but could not take the time to stack the crop residues. One solution to this problem was through community work days, a common practice in this Muslim country. A second problem was the actual manner of storage, whether on platforms, in the field, fenced, etc. Field staff were trained in storage techniques, a handbook was written for extension staff (Russo, 1985), and many discussions with farmers took place. Once it was explained that the crop residues had to be stacked so as to avoid termite and small ruminant access, villagers soon came up with their own local modifications and ways of storage. In late 1985 when technical input by both the project and the government was minimized, a survey was done of all the participating villages and surrounding villages to see what farmers were continuing to do without project input. Overwhelmingly, the crop residue storage portion of the deferred grazing scheme was the only part that was being continued and picked up by other villages which had not been in the programme (Russo and Patrick, 1986, unpublished data). These adopting villages were all Fula, the only ethnic group with experience in livestock management. This raises an interesting question as to the importance of training crop farmers in animal nutrition and management. An independent survey in 1987 of villages participating in the feeding trials (Lawry, 1987) also showed that the principles of crop residue feeding have been widely adopted in both eastern districts.

A framework for farmer acceptance

An early examination of Gambian farming systems which could have included socio-economic information and a more thorough investigation of secondary sources of data about similar projects would have led to a somewhat different design for this project and a much different approach to the field work. It is the current vogue to examine the "failures" of livestock development projects in Africa but concrete suggestions to prevent such failures are few. The most blatant design flaws are an oversimplification of the farming system and strong disciplinary biases. Focus on only one problem, for example, inadequate feed supplies during the dry season, by only one discipline such as range management inevitably leads to perceived failures by both the development experts and the farmers.

The project tried to introduce four innovations through these village-level feeding trials with varying degrees of success. Why did it fail dismally with some innovations and succeed with others? The answers lie not in the biological feasibility of each innovation because all were possible. The blame cannot be laid on project personnel, political reasons such as researcher vs. extension conflicts, nor weather as all impinged equally on all innovations. Rather, answers must be looked for in socio-economic areas. Borrowing heavily from social science researchers, a methodology is proposed that can be used in developing a framework for designing farmer/user level programmes. This framework focusses on the potential goals; conflicts that may arise between the so-called development experts (both national research and extension staff and expatriate staff) and farmers. Development is not simply the study of responses to an innovation by a group of local people but a study of the critical interactions between at least two groups (Galaty, 1981). Use of this framework on the four innovations is described by Table 5. It can be seen that while development goals and farmer goals are similar, i.e., production of more feed, farmers concerns in 3 out of 4 areas prevent them from adopting the innovation. These concerns are all socio and/or economic. Some of the farmer concerns are beyond their control, such as the land tenure system; while some could be alleviated by cash alone. The framework simply asks what logical conclusion can be reached concerning farmer acceptance based on the constraints he or she faces. There may be possible conflicts between government planners and donors who have specific concepts and goals, local staff with their own agendas, and the farmers who have a socio-economic system which is perceived as being flawed by the former two groups.

This framework is used in Table 6 to explore other possible innovations associated with livestock production in a mixed agricultural system like The Gambia. The innovations are rotational grazing, imposition of a grazing fee or setting livestock quotas, practicing selective feeding of certain classes of livestock, and increasing offtake. With rotational grazing, goals are similar in that more feed is a desired product but farmers cannot or do not want to devote labour to increased herd management. A grazing fee goes against farmer goals to maximize herd size and against an almost universal goal to avoid taxation of any kind. Farmer acceptance of the first two innovations could be modified by a government and about which farmers can do nothing. In the case of selective feeding, again, a change in government policy, in this case, the price structure, allowing more to be paid for better quality meat might encourage more sales. Selective feeding is a system requiring much labour and contact hours per animal. In the Gambia, livestock owners are rarely the managers; over 90% of the input cost in livestock production is for contract labour. The last innovation, increased offtake, could only occur if nutrition were improved, leading to improved calving and calf and adult survival. Data show that there are very few surplus males, that the calving interval is 1.5 years, and that calf mortality is 16% (Vesseur et al, 1986). For the latter two innovations, farmer acceptance is questionable but not necessarily negative.

The major problem with this framework, or with any other, is getting people to use it. Using the framework may seem unwiedly or too "soft" an approach for a biological scientist but it is a rapid way to test which innovations might be more acceptable to farmers and therefore, more likely to succeed. Researchers who are willing to try a new fertilizer or vaccine should be equally as willing to use the tool of another discipline such as social science, to improve their programmes.

Table 5: Innovations introduced by the Mixed Farming Project, 1981-86

LS = Livestock systems
CS = Cropping systems

Table 6: Possible innovations in rangeland management for The Gambia

LS = Livestock systems

Conclusions

The struggle to design meaningful, useful and appropriate research is not new and, in fact, has been the subject of many meetings, conferences and panels. Concerned individuals and institutions continue to believe that there must be a better way and that it must be found soon in the light of the very real problems faced by Africa. Recently, several researchers have been developing frame workers and models to link technological innovations and the agricultural system that is both under study and the target for development. These frameworks vary in their methodological approach, some using local research and extension staff, secondary sources of data and farmers as key components in design (Patrick and Russo, 1988) while others rely heavily on research staff, experimental data and computers (Hart, 1987; ILCA; 1987). The former, into which this particular framework can be fitted, is useful in farming systems where little quantitative data is available yet much is known about constraints in the system. It could, indeed, be seen as the first step for the more quantitative frameworks and models. The more analytical frameworks and models are useful in agricultural systems which have had the benefit of physical and biological experimentation and where a decision needs to be made concerning which of many alternatives to choose.

In many projects, potential development impact is not quantitatively considered in the design and evaluation of the technology selected (Hart, 1987). There are too often serious gaps in knowledge and time lags between the notional stage of technology evaluation which is the start of a project and actual field implementation by farmers which could be ten years down the road. Use of any of these frameworks to facilitate an integrated development process is not only indicated, but essential. Most planners seem to remain unaware and insulated from the findings of experienced field workers that the most important aspect of development is involving local populations in the planning, design, implementation, and evaluation of development activities intended for their benefit in light of both farmer and development goals that the likelihood of success increases.

References

Deffendol, S. 1986. Final report: Range ecology component. MFP Tech. Rep. 14. MANR/GOTG/CID/CSU. Ft. Collins, CO. 39 pp.

Eastman, C. 1986. Traditional Gambian land tenure and the requirements of agricultural development. MFP Tech. REp. 19. MANR/GOTG/CID/CSU. 36 pp.

Galaty, J.G. 1981. Organisations for pastoral development: contexts of causality, change and assessment. In: J.G. Galaty, D. Aronson, P.C. Salzman and A. Chouinard, (Eds.) The future of pastoral peoples. IDRC, Ottawa. pp. 284-293.

Hart, R.D. 1987. An analytical framework for the design and evaluation of crop and livestock technology. Plenary paper presented at the Farming Systems Symposium, Fayetteville, AR. 23 pp.

Hedrick, D. and Bojang, M. 1983. Final report of the forage agronomist, Mixed Farming Project. MFP Tech. Rep. 2. MANR/GOTG/CID/CSU. 45 pp.

Lawry, S.W. 1987. Report of an assessment of deferred grazing schemes in The Gambia. DAHP/GARD. Banjul, The Gambia. 11 pp.

McIntire, J. and Debra, S. 1986. Forage research in smallholder and pastoral production systems. Forage Network in Ethiopia Newsletter. No. 14: 5-11. ILCA, Addis Ababa.

Menz, K.M. and Knipscheer, H.C. 1981. The location specificity problem in farming systems research. Agri. Systems 7: 95-103.

Patrick, N.A. and Russo, S.L. 1988. A proposed framework for designing livestock development projects in West Africa: The Gambia as an example. J. Agr. Human Values. 4(2&3): 105 - 110.

Russo, S.L. 1985. Handbook on the use of crop residues for livestock feed. MFP/DAHP. Abuko. 11 pp.

Russo, S.L. and Ceesay, B. 1986. Research in forage agronomy by the Mixed Farming Project, 1983-1986. MFP. Tech. Rep. 12 MANR/GOTG/CID/CSU.

Russo, S.L. and Spencer, W.P. 1988. Goals conflicts: Cattle production in The Gambia. Paper presented at Intl. Rangeland development Symp. Corpus Christi, Texas, pp. 65-80.

Vesseur, P., Hurkens, E.J., Eckert, J., Crouch, J.C. and Mol. L. 1986. The Gambia cattle herd: a survey report. MFP Tech. Rep. 13. DAHP/GTZ/MFP/MOA. Banjul, The Gambia. 24 pp