An e-publication by the World Agroforestry Centre
AGROFORESTRY A DECADE OF DEVELOPMENT 
|
|
An e-publication by the World Agroforestry Centre |
|
AGROFORESTRY A DECADE OF DEVELOPMENT |
|
|
section 3 Chapter 7 H.-J. von Maydell Introduction The challenge to develop agroforestry to solve apparent problems in tropical rural areas led to the establishment of ICRAF in 1977. At that time it was known that a great variety of integrated land-use systems had already been practised throughout semi-arid Africa since times immemorial. They had been empirically evolved and adapted to different environments as well as various ethnic groups. Very little attention was given to these forms of subsistence land use as they were seen to have little or no relevance to colonial and post-colonial, export-oriented, modern agriculture. Moreover, no need was felt to deal with them because no specific problems with regard to natural resources and social developments were recognized. For decades, rural development policies had concentrated on increasing yields from extensive cash-crop plantations, which had high external inputs in terms of technology and capital and aimed at outputs for overseas markets to earn foreign exchange rather than to supply products to African rural and urban populations. Starting from the late 1970s, this attitude gradually changed. World markets for agricultural commodities had shifted. Traditional trade links were losing their previous importance after many states had gained their independence. Growing gaps in the terms of trade forced African governments to re-orientate their land-use policies as well as their general development strategies. With "Africanization", the importance of the hitherto-overlooked values of the African way of life were rediscovered. But, simultaneously, new problems of local food supply, environmental destruction, etc., emerged and conflicts between rural development and industrialization had to be solved. Finally, the severe drought throughout the Sahelian zone and other parts of Africa in the early 1970s had caused a shock among African politicans, agricultural services and the international community, indicating that something would have to be drastically changed in rural development in order to prevent future disaster. Thus, as a first step, at the United Nations Conference on Desertification in Nairobi, Kenya (1977) a "Plan of Action" was formulated which highlighted the role of sustainable-resources management and the till-then-underestimated importance of forestry within integrated land use. The beneficial effects of "non-commercial" trees, shrubs and other woody perennials of the drylands that were recently brought to general awareness unfortunately very soon resulted in extremely high expectations as to what could be achieved by afforestation, reforestation or tree planting. Many of these expectations with regard to dry Africa soon proved to be unrealistic. In reality, multipurpose woody plants, either alone or in small groups and lines rather than closed forests or extensive greenbelts, and local species rather than introduced fast-growing and timber-producing "exotics", turned out to be much better adapted to various sites and demands. The forest services with their limited capacities, however, were not yet organized to change from their standardized programmes of conservation and timber production to the new issue of forestry as a component of integrated land use and "forestry for people". Suddenly, and often under severe time-pressure, they were expected to deal with a wide range of non-forestry subjects, i.e., the co-management of agricultural and pastoral lands. Farmers and herdsmen, as well as government agricultural and livestock services, were simultaneously expected to incorporate forestry components and functions into their management. At all levels, therefore, it was felt that the common bond — integrating, harmonizing and optimizing land-use practices—was missing. This called for immediate interdisciplinary co-operation which became a common term, easy to spell but often difficult to realize in practice. Research and development institutions, therefore, started to add to their previously dominant but isolated topics of commodity, technology and socio-economic research the "new" field of farming systems (Simmonds, 1985). This was a decisive step forward at all levels (national and international, including the Consultative Group on International Agricultural Research, CGIAR). Agroforestry-systems research, as initiated by ICRAF, has played a leading and stiihulating role in this context. Some of the results originating from a special programme, namely, the Agroforestry Systems Inventory, have been published in Agroforestry Systems.
The environment "Dry Africa" includes all parts of the continent with less than about 1,500 mm of annual rainfall. Most of these lands lie within the tropical zone but they do extend beyond the Tropics of Cancer and Capricorn. Dry Africa thus includes a great variety of climates and landscapes and they are inhabited by very different peoples. Therefore all attempts to generalize are of limited value and may lead to controversial discussions. One common feature of these African regions, however, is a more or less pronounced water deficiency. The deficiency is either permanent, seasonal or sporadic with regard to horizontal or vertical distribution, quality and utilization. Rainfall is very variable in most parts. If rains are below average we call it a drought, although drought in its proper sense refers only to rainfall insufficient to support life and satisfy human demand. Plants, animals and people have adapted to the harsh environment of dry Africa (Louw and Seely, 1982). Plants have developed specific mechanisms for drought escape and drought tolerance (e.g., groups of plants such as ephemerals, succulents, xerophytes, halophytes, psammophytes, plants employing the so-called C4-photosynthetic pathway or CAM-plants that can select the C3 or C4 mode of life according to the prevailing conditions, etc.). Adaptation to stress implies also concentration or spatial restriction to small favourable sites and short periods of growth, as well as a high degree of resilience as compared with persistence. Animals and men have developed similar strategies of rather short-term flexible adaptation to changes hi environmental conditions. It is of paramount importance to acknowledge that the carrying capacity of arid and semi-arid lands in Africa is limited. By over-exploitation and/or misuse of natural resources, therefore, critical limits are rapidly reached, and once exceeded very soon result in an irreversible breakdown of the productive potential of a given site. From the vegetation aspect, most ecosystems in dry Africa (excluding deserts and semi-deserts) can be classified under the term "savanna". These are characterized by grasslands with trees, palms and shrubs of varying density from open woodland to single thorny shrubs (see Le Houérou, this volume). Very few sites bear a climax vegetation of dense stands of woody perennials, and these sites are generally preferred for different forms of land use because of their better soils and water supplies. Wherever they remain unoccupied, they form islets important for the survival of both "refugees" and "nomads" among plants, animals and people. Typical sites of this kind are oases and forest islands. The role of such sites, which often support high population concentrations, has to be highlighted for dry Africa. They represent a network within an "ocean" of marginal lands, and their mutual feedbacks and interrelationships are a pre-condition for maintaining and further developing the natural and the socio-economic environment. The role of woody perennials on range and farmlands There is very little archaeological and historic evidence on agricultural — and thus agroforestry — systems of ancient Africa, except for Egypt and parts of north Africa which belonged to the Roman Empire. Most evidence refers to irrigation schemes, to specific crops and their utilization and to sites preferred for cultivation (Hall et al, 1979). The first traces of agriculture in Egypt date back to about 12,500 BC with indications of plant cultivation and animal husbandry, and to 6000 BC for date-palm management. There is hardly any information available oh tree cultivation and on land-use systems of these early times. What is known today has mainly been recorded during times of exploration of the continent, especially from the nineteenth century on and during the colonial period. Trees, shrubs and palms have always played an important role in ecology as weH as in human culture and economy throughout semi-arid Africa. However, the value of these woody perennials was not primarily derived from closed forests, but rather, and even predominantly, from individual specimens which were components of pastoral-, farm- or village-settlement lands. This has impeded understanding between temperate-zone-educated foresters, who think and act in terms of sustainable management of closed-forest systems, and African rural people, who are used to distinguishing specific single trees from which they wish to benefit. Land tenure in these open bush- and woodlands was considered a tribal/community concern in traditional African land use (with private forest ownership being practically non-existent), whereas single fruit trees could well be individually owned and even be inheritable. Forests and forest products were used in a communal or collective way, generally free of charge and with very few restrictions. Forests were generally protected in colonial and modern national legislation and thus did not constitute an accessible resource. Scarcity of supplies, and more often dire need, provoked people to invade these forests in order to continue traditional uses, now classified as illegal, or to convert these forests into non-forest lands. These are some of the reasons why rural people in dry Africa are more closely affiliated with single trees outside the forest than with forest-land management. The affiliation of trees outside the forests with people is also known from many examples of veneration of trees in African history. In ancient Egypt, for example, sycamore trees (Ficus capensis syn. F. sycomonis) were worshipped as symbols of fertility, suppliers of food for those alive and for the souls of the deceased. Other sacred trees were Balanites aegyptiaca, Oka europaea and Phoenix dactylifera (the date palm). All over semi-arid Africa, various tree, palm and shrub species offering food, medicine, gums and other exudants, or providing shade or an aesthetic sight, were venerated or at least cared for. Most of them grew in open woodlands and savannas, along riverbanks or in some oases, and in parks and gardens. Wood production for fuel, construction, etc., was given less importance because supplies were generally secured from other sources, mostly dead wood. From Egypt, throughout the Sahara to the Sahel and the Sudan savannas, the miombo of East Africa, and open woodlands, savannas and semi-deserts in southern Africa, man's relation to trees was similar. Widely scattered trees formed an essential component of the natural vegetation in man-made environments and in daily life. The mighty baobab (Adansonia digitata) (Figure 1) may be considered a symbol of what a tree could mean to people. Acacia albida is another commonly protected and highly venerated non-forest tree of large areas in semi-arid Africa. Those trees as essential elements of the landscape were used in three different ways: 1. Exploitation Selective exploitation of natural resources took place among "mobile" (nomadic or transhumant people) as well as hi sedentary farming societies. For example, the nomads, when migrating over vast areas, considered the woody vegetation as common property of their tribe or implied everyman's right to utilize what God had provided. No specific limitations had to be observed, since what was not used by them would most likely be used by others. This attitude still prevails with many people hi the region and obviously constitutes a serious constraint to maintaining or regenerating a desirable density of woody vegetation on pastoral lands. Sedentary people generally enjoyed free access to communal/ tribal resources within daily walking distances, i.e., 5-15 km. Collection of fuelwood, poles, and other products of trees and shrubs was unrestricted and considered a common right. As long as population densities remained low these practices could prevail without destructive effects on the environment or on the preferred species' survival. 2. Maintenance Maintenance of selected, naturally grown trees took place with more intensive land use, especially in dry farming. The previous natural vegetation was changed or even removed in favour of field crops and permanent settlements. However, specific trees, palms and shrubs were deliberately left and eventually maintained and carefully protected against livestock and illegal use. Some of these trees were Acacia albida, Butyrospermum parkii, Parkia biglobosa, Tamarindus indica, Ziziphus mauritiana, and there were many others, depending on site, land use and tradition. Some of these trees became private property and could be handed down from generation to generation. The Emirs of Zinder, Niger, at times even passed death sentences upon people who felled Acacia albida trees. 3. Cultivation Cultivation of trees was rarely practised in the past. Egypt was an exception. Parks and tree gardens, such as that in Amun (Nile delta), where 8,000 slaves were employed to maintain the trees, or Matoria, where the Holy Family was reported to have stayed for a while, were established and intensively managed like urban and individual ornamental plantations. Last but not least, the oases with their date palm and other tree or shrub plantations are outstanding examples of early selection, breeding, culture and integrated management of trees in rural environments. It is interesting to note that, whereas the date palm was intensively improved and cultivated, no effort has been made to genetically improve Balanites, originally an equivalent to the date palm and apparently much better adapted to the climate and sites. In other parts of semi-arid Africa (outside oases) tree planting and breeding remained generally unknown until recent times when non-Africans started to propagate the so-called exotics such as cashew (Anacardium occidentale), neem (Azadirachta indica), eucalypts, various fruit trees (e.g., mango, Citrus spp.) and a few ornamental trees such as Khaya senegalensis and Cassia spp. In essence, however, extensive exploitation of natural resources by migrating and "landless" people on the one hand and intensive use by sedentary farming communities or urban people on the other prevailed over many centuries. Both had an important influence on the development of early agroforestry-type land uses. As in other parts of the tropics, these traditional forms developed into three basic categories: silvopastoral, agrosilvopas-toral and agrisilvicultural systems (Nair, 1985). However, dry Africa is extremely heterogeneous. It is not a "khaki-coloured" environment with simple and uniform land uses. The local climate, for instance, may differ even within a few kilometres, especially with regard to quantity and distribution of rainfall. Soils show a remarkable variety, not only in their chemical components and physical structure but also in the micro-organisms essential for plant growth. The number of plant species that exist even under marginal conditions is remarkably high as long as degradation by repeated burning, overgrazing or monoculture cropping has not occurred. Man has adapted his life and his land-use patterns to the heterogeneity of the environment by developing a variety of practices. Consequently, hundreds of land-use systems have developed and many of them are still of great importance. Silvopastoral systems For thousands of years the regions of Africa with less than about 500 mm annual rainfall have been used for migratory animal husbandry. However, only parts of these vast grazing lands were utilized in a way which would allow their classification as silvopastoral agroforestry systems. But browse made up a varying proportion of livestock diet almost everywhere (Le Houérou, 1980, and this volume; IBPGR, 1984). Whether or not traditional forms of pastoral land use in Africa should be considered as agroforestry in its proper sense is a matter of definition and is subject to individual judgement because intentional cultivation or maintenance of fodder trees did not take place in these systems (Figure 2). As a rule, these African dry lands can be classified as marginal, fragile ecosystems, of low carrying capacity and thus limited socio-economic development prospects. The ruling climatic constraint is a single, short rainy season with pronounced variation in quantity and spatial distribution of rainfall, especially in the more arid parts- Quantity and nutritional value of the plant biomass, as well as drinking water available for the animals, are thus subject to unpredictable but foreseeable fluctuations. As a consequence of the climate, the vegetation consists mainly of annual grasses with some perennial species and a varying number of shrubs and trees, attaining between 5 and 20 percent of the area depending on soil type and ground-water availability. The growth period of grasses varies between one and three months. Many, if not most, woody species are deciduous, i.e., leafless over the long dry season. Over thousands of years, animal husbandry in the traditional forms of nomadism and semi-nomadism (transhumance) have proved to be optimally adapted to the carrying capacity of these lands and the rainfall-dependent variation in fodder production (Jahnke, 1982; Lusigi and Glaser, 1984; Galaty et al, 1981). Recent problems have all been caused by unprecedented population growth (the resulting pressure on natural resources exceeding carrying capacities), and by external influences which have increasingly destroyed well-established regional structures. According to Ruthenberg (1980), in nomadism the animal owners do not have a permanent place of residence; they do not practise regular cultivation and their families move with their herds. Nomadism in this pure form was and still is mainly practised in desert and semi-desert environments north and south of the Sahara with less than 300 mm of annual rainfall and in limited parts of eastern and southern Africa. But these systems are less important than the semi-nomadic or transhumant systems in which livestock owners have a permanent place of residence which they maintain over several years. Whereas part of the family travels with the herds over long periods of the year and over long distances, women, children and older men may stay behind on lands which they cultivate over the rainy season in order'to improve the family's food reserves and income.
From the point of view of agroforestry systems, competition, complementarity and mutual dependence of animals, woody perennials and grass/herbs deserve special attention. As long as stocking rates were low, detrimental impacts remained negligible and could easily be compensated for by alternating with more favourable sites. With increasing numbers of livestock over recent decades, however, problems have emerged. Since livestock, except goats, feed selectively on specific plants or parts thereof, the more palatable plants were continuously reduced and weeds invaded. This situation was seriously aggravated by the effects of seasonal grass/bush fires over large areas (Figure 3). The structure of herds also played a role. Many herdsmen prefer to specialize in cattle, sheep, goats, or camels, which leads to a one-sided exploitation of the resources. Mixed herds with cattle feeding on long grass, sheep on short grass and goats and camels browsing shrubs would improve overall carrying capacity.
On the other hand, livestock, like game, have always played an important role in promoting the growth of woody perennials by reducing competition from grass for nutrients and water, and by breaking seed dormancy through digestion and subsequent distribution of seeds with their droppings. Thus, over long periods and vast areas in dry Africa, as long as population densities remained low, silvopastoral systems proved to be well adapted to the environmental conditions and to people's demands. Social structures and cultural life were based on nomadism and transhumance. This has changed within the past few decades. Therefore, recent moves to convert these "mobile" forms of "archaic" land use into sedentary, modern management systems is not only a question of technical and economic innovation but requires major socio-cultural change. For most African herdsmen their animals are more than mere farm commodities and production units; they are an integral part of their individual, family and ethnic identity.
There has never been a distinct borderline between mobile, silvopastoral land-use systems in .the more arid parts and sedentary agrisilvicultural systems in the more humid parts of Africa. Overlapping and transitional zones developed due to periodic variability of the climate and migration of different ethnic groups. Therefore, for most parts of the region, traditional agrosilvopastoral land-use systems were a characteristic feature. They were based on agricultural-crop production during the rainy season and pastoral use of the same land during the dry period of the year. In the rainy season livestock were either brought to remote grazing lands by transhumance or, more commonly, driven daily to nearby open pastures next to the villages, the herds returning to the compounds or corrals at night, or even being tethered or kept in stables. Specific trees, shrubs and palms were left standing, deliberately maintained or even planted on the cultivated lands which formed an essential component of the system. The agricultural crop plants included cereals (millet, sorghum, barley, maize), cowpeas, groundnuts and melons in the dryer parts and a greater variety of plants on sites with better water supply and soils. These plants were cultivated mainly for food, their residues being used as fodder in the dry season. Cowpea is an outstanding example, the "straw" in some parts of the Sahel yielding even higher income than the beans. Many, if not most, of these crops were planted year after year until soil fertility was exhausted and the farmer had to shift. A fallow system was then applied for many years, the fallow land, including regrowth of woody plants from root suckers, coppice or invading seeds, being used as pasture the year round. On extensive lands, however, trees and livestock provided for a sufficient recycling of nutrients to allow long-term or even sustainable cultivation. Most of the distinctive agrosilvopastoral systems were on such lands. They are still efficient and offer good prospects for future development. Much more than the annual crops, the livestock and trees of these systems are typical "multipurpose" components. Livestock contribute to the system by providing food (mainly meat and milk) and other products such as hides, skins, hair, horn and sinew. Some supply draught power for cultivation and transport or are used for riding. In parts of Africa, manure is intensively applied to improve soil fertility, both in homegardens and within the fields of the agrosilvopastoral systems. Some dried manure is used as fuel for heating and cooking, e.g., in the highlands of Ethiopia. Livestock represented an important capital asset, as a source of income in addition to the agricultural crops, as an object of value in exchange for goods and services (including during marriages, etc.), and as a "living savings bank". Livestock represented wealth and security of subsistence and, in many societies, prestige. Socially, keeping livestock in agroforestry systems provided employment, especially for unsalaried woirien and children, and specific income and property rights for them. Many of the highly complicated socio-cultural webs and networks were based on groups within the families or villages specializing in the management and use of animals (e.g., men — camels, transport trade; women — goats, food, subsistence). Last but not least, livestock were important in using crop by-products and residues which otherwise would have been of little or no value, and in improving the productivity of marginal lands (i.e., poor sites within the system, as well as the croplands during the unproductive dry season) or fallow lands. Trees, shrubs and palms which originally occurred on part of the land and occasionally regenerated during the fallow periods or survived on unutilized spots, were soon discovered to have multiple uses, many of them as food. Prominent species with edible fruit or seeds, leaves, shoots, flowers and gums are Acacia Senegal, Adansonia digitata, Annona senegalensis, Balanites aegyptiaca, Borassus aethiopum, Butyrospermum parkii, Cordyla pinnata, Detarium senegalense, Dialium guineense, Ficus spp., Hyphaene thebaica, Lannea spp., Parinari spp., Parkia biglobosa, Phoenix dactylifera, Sderocarya birrea, Sterculia setigera, Tamarindus indica and Ziziphus mauritiana. In addition, other species were used to provide leaves for teas, spices and condiments. Usually these trees, shrubs and palms were carefully protected but rarely actively regenerated and maintained. This was different with "imported" species from other regions of the world, many of which quickly became fully integrated. Species such as Anacardium occidentale, Artocarpus spp., Citrus spp., Cocos nucifera, Mangifera indica, Moringa oleifera and Persea americana were planted, bred, selected and grafted, and sometimes fertilized, irrigated and pruned. That is, a more or less pronounced arboriculture was practised even in early times within the agrosilvopastoral systems. Forage-yielding species were given less priority. This was certainly because almost all wild-growing species proved to be palatable, although to varying degrees, and there were more than 100 species providing leaves, pods and bark as fodder. As browsing, mainly during the dry season, was commonly practised by free-roaming livestbck, no specific attempts were developed to maintain or regenerate fodder trees. The management of Acacia albida for pods (a storable forage) can be considered an exception. Trees, shrubs and palms within the system also provided fuel, wood for construction, furniture, fencing (including thorny branches), agricultural and household utensils, transport, etc. The long list of other uses included drugs for human and veterinary medicine, and also tanning materials, dyes, fibres, poisons and repellents. The environmental benefits were only partially recognized. Shade was the main benefit sought, but trees were also used for shelter, windbreaks and soil improvement. From the socio-economic and cultural aspects, some species were maintained and utilized as cash crops, e.g., gum arabic and other gums and resins, for amenity (flowering and evergreen species), or as focal points for assemblies, cultural events or various forms of worship. The outstanding tree species of these agrosilvopastoral systems has been, and still is, Acacia albida, which grows in most parts of dry Africa (Miehe, 1986). The species is almost perfect as a component of agroforestry systems. The large thorny tree keeps its leaves through the dry season and sheds them at the beginning of the wet season. Thus, under the trees, there is no competition for light during the cropping period while during the hot and dry months there is shelter for livestock, which also improve the soil through their droppings. Moreover, the roots penetrate into deep soil layers. They do not compete for water and nutrients with agricultural plants, but they intercept minerals at otherwise inaccessible depths and deposit them on the soil through the leaves which are shed just in time to fertilize the topsoil at the start of the cultivation season. It is estimated that 20^40 trees per hectare can provide nutrients equivalent to the normally required quantity of fertilizer (Figure 5). This was known to most farmers of the zone, and Acacia albida was carefully protected and maintained because the better crop yields underneath were strikingly apparent. The tree also produces large quantities of pods which mature and drop during the dry season. These pods provide a second "dry-season crop" and thus permit one to keep livestock within the system throughout the year. The tree itself has further uses, ranging from firewood and thorny fencing material to tannin, bee forage and a wide range of Pharmaceuticals. 
Agrisilvicultural systems The trees, palms and shrubs just cited were also employed in agrisilvicultural systems, usually located in the wetter parts of the region and on suitable soils. The difference from agrosilvopastoral systems is to be seen in the fact that livestock were not a predominant component and were sometimes completely excluded. Trees and agricultural crop plants were grown together deliberately where intensive sedentary farming and permanent forms of individual land tenure prevailed. However, large numbers of livestock, and especially of cattle and camels, were absent due to health problems (e.g., the tsetse fly). Once more, Acacia albida and various fruit trees were preferred. A great variety of practices was applied (Steiner, 1982), especially in the slightly more humid regions (FAO/ SIDA, 1981). For the dry zones, sorghum, millet and maize-based cropping systems were most important, but other combinations of crop plants, e.g. rice with trees, shrubs and palms, were also used (Figure 6). 
The present If the traditional forms of ssilvopastoral, agrosilvopastoral and agrisilvicultural agroforestry in Africa have been, and partly still are, successful, one may ask why and how the present rural problems have arisen. The answer, evidently, lies in the exponential population growth during the last few decades. Obviously, a simple return to previous subsistence economics after the relatively short period of colonial and post-colonial "errors", cannot solve the existing supply problems of urban and rural areas. Regional development If, in the past, local concerns formed the focus for most activities in rural dry Africa, a regional approach has become dominant in our times. Problems were shifted from the individual family or community level to specific zones within countries or even beyond national borders. The Sahel is one outstanding example. The expansion of decision levels over larger areas, including a variety of heterogeneous environments and different peoples, appears to be one common trend. This implies a tendency to generalize and standardize, which is, in some ways, in contrast to the characteristics of agroforestry and African tradition. Political changes have resulted in the establishment of new states which do not necessarily correspond with natural boundaries and ethnic groups. In many instances this has interrupted basic interrelationships, mutual links and feedbacks. Free migration, essential for nomads and semi-nomads as well as other groups of the population, has been hampered, resulting in problems of supply, employment, and exchange of goods and services. Modern technology has had an enormous impact, changing infrastructures, including transport and communications, reshaping landscapes, e.g., by extensive irrigation and/or hydroelectric schemes, and resulting in the large-scale application of mechanization and chemicals in agriculture. Most important, medical care for both man and animals has made unprecedented progress. This, in turn, has contributed to the well-known and intensively discussed problems of excessive population growth, growth of animal herds, and the consequent escalating pressure on land resources. For the first time in African history, the rate of population growth exceeded that of agricultural production during the past two decades. In many countries this has led to almost permanent dependence on food imports and/or foreign aid. Large parts of dry Africa are now among the so-called least-developed countries of the world, suffering from a situation described by a continuously falling per capita income.
In such an emergency situation, the influence of external partners such as international organizations, bilateral aid, and non-governmental agencies, must necessarily increase and come to exert a strong influence on rural development in Africa. Loss of potential productivity As an immediate consequence of rapid population growth, large areas of potentially productive savanna range and woodlands are degraded by burning, inappropriate land use and exploitation beyond their carrying capacity. Desertification (IUCN, 1986; Baumer, 1987; Rapp et al., 1976) stands for a whole range of detrimental and sometimes irreversible processes which include loss of fertile topsoil by erosion, impairment of soil morphological properties by compaction, salinization of irrigated lands, the extinction of plant and animal species and loss of genetic diversity, and imbalances of water and energy regimes. Pressure on Africa's croplands has escalated in recent years, resulting not only in accelerated deterioration but also in socially explosive competition, e.g., between herdsmen and sedentary farmers and/or urban and rural people in many parts of the continent. Pressure on croplands, in addition, is likely to result in falling yields if, for instance, fallow periods are shortened, cultivation is extended on to land with less fertile soils, marginal climatic or topographic conditions, or when animal dung is burnt for energy for lack of fuelwood. According to FAO (1984) only 20 per cent of all the utilizable lands of Africa were cultivated in 1975. The situation in the arid zones, however, was completely different. In 22 countries classified as severely endangered, 80-100 percent of the land was already occupied. Expansion of crop and rangelands has, therefore, little chance of improving the situation. In the year 2000,29 countries and a population of more than 250 million people will face catastrophe if nothing is changed basically (FAO, 1984). This is where the challenge for agroforestry begins. Can agroforestry really increase the area of productive cropland by replacing long fallow periods with sustainable, permanent production systems — i.e., by introducing multistorey production on the same .piece of land? This could be decisive. To answer these questions, agroforestry research in dry Africa must be given high priority at all levels. Remarkable progress has been made recently in the form of inventories of promising traditional agroforestry systems compiled by ICRAF (Nair, 1987b) and other organizations and institutes, and in analyses of the systems. Much, however, remains to be done in extrapolating the existing agroforestry systems to ecologically similar conditions elsewhere, giving due consideration to changing situations and long-term sustainability. The AFRENA programme of ICRAF is expected to fill many gaps and to strengthen existing national research facilities. Moreover, agroforestry should be developed immediately to intensify production per unit area by recycling of nutrients and nitrogen fixation. Improved application of manure and mulch should be promoted. Lack of nutrients, mainly phosphorus and nitrogen but also of trace elements, is a major constraint. FAO figures (FAO, 1983) indicating an average use of 19.5 kg per hectare or 7.4 kg per capita per year of mineral fertilizer in Africa as compared to 78.5 kg and 25.5 kg as a world average and 114.8 kg and 65.5 kg in industrialized countries, respectively, are significant and alarming. Inevitably increasing demand Within the past 25 years Africa's population has almost doubled, having now reached nearly 500 million people. Within another generation of only 17-25 years there will be 1,000 million Africans. The overwhelming problem of providing adequate food and energy for this population has to be solved. Social changes, such as growing urbanization, the tendency towards smaller household units (as compared with traditional rural entities), technical development and individual or group aspirations for improvement will also have consequences. Demand, however, does not only refer to food but to many other sectors of the rural economy. It is important to note that agroforestry can contribute a variety of goods and services simultaneously, and that agroforestry is suitable for site-specific diversification. This is in remarkable contrast to the above-mentioned tendency towards regional standardization and simplification and indicates the potential for regional solutions and alternatives. At a meeting of some of the International Agricultural Research Centres (see Steppler, this volume) and ICRAF on agroforestry research in Africa, held in Nairobi, Kenya on 22-24 September 1986, the following were identified as issues of paramount importance in the context of further agroforestry development in Africa:
The meeting agreed that all these problems could be alleviated, although it will require time, resources and convincing technologies. Other important problem areas discussed included the lack of knowledge on the proper management procedures for agroforestry technologies which limits the adoption of such technologies by farmers. It was stressed that in all cases farmers' perceptions of the problem were of paramount importance when designing research programmes. This also pointed to the need for more economic studies and proper follow-up of newly introduced agroforestry technologies. African agriculture is 70-90 percent a subsistence economy, and for agroforestry in the dry regions the share may be even higher. A subsistence economy does not promote investment (due to lack of finance) or market-oriented production, however. Although there appears to be a concentration on the development of small-scale subsistence farming, agroforestry should use all possible means to incorporate cash crops with high yields, low risks and a promising market. In parts of East Africa, for instance, the income from tobacco exports would compensate for about 10 times the quantity of maize that could have been harvested from the same area. Similar calculations are available for coffee, tea and other cash crops. Keeping in mind the development aspirations of the farmers, the more recent condemnation of such crops should be thought over and new, "typically African", crops promoted. Nevertheless, agroforestry, in its various forms of traditional systems and practices, is still widespread throughout dry Africa and thus forms a foundation to build upon. Its contribution to a strong subsistence economy is obviously great, although it is generally underestimated due to lack of statistical records. Food At least in dry Africa, food production is at present the main target of agroforestry and this is likely to continue. Local food production has become the main problem for rural development and urban supplies. The situation escalated to catastrophic dimensions during recent drought periods and/ or political stress situations and has affected large parts of the continent. However, the increasing frequency of these catastrophes is just an indication of the ever-increasing instability and weakness of the rural human-ecological systems. Late, but hopefully not too late, it is now recognized that a certain degree of sustainable self-sufficiency in food supplies is essential for the overall development of the African countries. Agroforestry is expected to help in solving almost overwhelming problems. These expectations exist at all levels, but require support and co-operation of all parties involved from international and governmental agencies to the "grassroots", i.e., the individual land user. As mentioned before, the contribution of agroforestry to food production is generally not statistically recorded for a variety of reasons. For example, the significance of the phrase "food from trees, shrubs and palms" is not fully understood. These sources are important in quantity as well as in quality, in being in seasonal complementarity with agricultural crops and animal products, and as emergency supplies in times of drought, as well as part of the regular diet. The latter becomes evident from what is offered at local markets, although this is only a small fraction of what is consumed overall. Moreover, most of these foodstuffs are obtained from "no-man's land", i.e. public lands, free of charge and so far require but a small labour input, mainly by women and children. This social aspect of "tree food" must not be overlooked. Potential food production from trees and shrubs in northern Senegal (Ferlo) and northern Kenya (Turkana, Samburu) has been analysed by Becker (1984). Her results can be summarized as follows: the annual harvestable production of leaves and fruits amounts to about 150 kg per ha in the Saharo-Sahel, 300 kg per ha in the typical Sahel and 600 kg per ha in the Sudano-Sahel. This corresponds with the rule of thumb, derived from various observations in the West African Sahel, that in "normal" ecosystems the annual increment of non-woody biomass from trees, shrubs and palms in kg per hectare roughly equals the average annual rainfall in mm. Results from East and West Africa indicate that about 15 percent of that biomass can be classified as edible. Thus, in the above three ecological zones 23,45 and 90 kg, respectively, of edible material would be available per hectare annually. Correlating these figures with an average population density of 1 per sq km, and assuming a ratio of 4:1 for leaves and fruits, 450-1,800 kg of edible tree/ shrub fruit could be utilized per caput, which corresponds to 1.25-5.0 kg fruit per adult human being daily. In the Ferlo/Senegal, an inventory of baobab (Adansonia digitata) was made in a typical area in a German-Senegalese rural development project. The results showed 6,611 baobab trees and 1,200 people lived on 20 sq km, i.e., 5.5 trees per caput. The leaves of these trees are intensively used as a green vegetable and dried as a powder and have a remarkable nutrient content: 100 g of fresh leaves contain 23 g dry matter, 3.8 g crude protein, 700 mg calcium and 50 mg ascorbic acid. They provide 69 calories. Even more valuable is the fruit pulp which is rich in vitamin Bj and C, and the flour produced from the seeds contains up to 48 percent protein and 2 percent vitamin B! on a dry-weight basis. This is only one example demonstrating the potential food production of woody perennials growing on agricultural and/or pastoral lands in the Sahel. From the many species available, merely a fraction are used for food, and, in normal times, people tend to be very selective about species, varieties, and parts consumed. Plants used by one group of people in one region may not be accepted by other groups in the same region or the same ethnic group in a different region. Moreover, there are pronounced differences in the quality of food, depending on varieties or provenances. Adansonia digitata may again be cited as an example. There are trees with soft and tasty leaves and others with fibrous and bitter leaves. As no cultivation has been practised with these local species, selection and breeding have not yet started but may offer a hitherto unknown potential for improvement. This should be tried with at least 30 food-tree species in the various parts of dry Africa. Food production in agroforestry systems, of course, implies much more than producing and utilizing edible parts of woody perennials. Optimizing agricultural crop and animal production is the main concern and can be achieved as indicated for traditional systems such as those involving Acacia albida. It is interesting to note that, according to different observations in the Sahel and in other regions, and even outside Africa, a combination of trees/shrubs and annual plants (grass, cereals, herbs) is known to result not only in a considerable increase (about 300 percent) of total annual biomass production but specifically in higher yields of the annual plants, both in quantity and in quality. This has been recorded also from millet/sorghum fields with Acacia albida and from pasture lands with about 10 percent of the area covered by a light canopy of savanna trees (the author's unpublished observation). This indicates that Acacia albida agroforestry is a promising form of land use, but more research is needed to further develop the potential based on traditional systems and to clarify the underlying feedback cycles. Forage: Animal husbandry, as stated above, is the main land-use practice in dry Africa and is indispensable for food supply and the functioning of many agroforestry systems. Obviously, successful animal husbandry depends on the availability of sufficient forage. In traditional silvopastoral and agrosilvopastoral systems, a balance existed between stocking rates, natural vegetation and drinking-water supply. Where the carrying capacity was exceeded, people and herds could shift to more favourable sites. Recently this has become difficult due to increased numbers of people and animals and due to the extension of cultivation to lands formerly reserved for pasture. Moreover, desertification and degradation of vast areas are rapidly reducing the productivity of pasture lands. The drought periods of the 1970s and between 1982 and 1985 have clearly indicated that more than relief programmes are needed to ensure the survival of millions of people and their livestock. These relief programmes have largely failed because of a number of misleading assumptions. One of these assumptions is that the productivity of Sahelian pastures is primarily limited by rainfall. Although rainfall undoubtedly has a strong influence on annual phytomass increment, the deficiency of soil nutrients quite often plays an even greater role, especially the availability of phosphorus in the first part of the growing season and of nitrogen in the second part (Breman and de Wit 1983; Penning de Vries and Djitye, 1982). The northern (arid) parts of the Sahel are better endowed with nutrients and thus produce forage with a higher content of digestible protein, vitamins, trace elements, etc. However, the availability of water relative to phosphorus and nitrogen determines quality and quantity of the forage produced. Thus low water availability results in small amounts of nutritious phytomass, while higher water availability gives more phytomass but of increasingly inferior quality. The north of the Sahel is, therefore, more suitable for animal husbandry than the south with regard to forage quality, but the carrying capacity is low and grazing is restricted to between two and four months a year. In addition, drinking water for the animals is usually not available beyond the rainy season. Livestock performance in the south Sahel and in the more humid savanna lands is poor, however, even if in the rainy season and thereafter relatively large quantities of forage are available, because the quality of that forage is low (Breman and de Wit, 1983). This can be partially compensated for by the availability of higher-quality forage from trees and shrubs, either in the form of leaves, bark or fruit. Thus, browse in animal husbandry is less important in the northern than in the southern, generally more humid parts of Africa (Le Houérou, 1980; IBPGR 1984). The availability of browse plays an important part in silvopastoral and agrosilvopastoral systems and, thus in the agroforestry that is practised in most parts of the African dry lands (see Le Houérou, this volume). A second misleading assumption is that livestock herds are generally too large because the nomadic and transhumant herdsmen maintain them for prestige. It is, however, accepted that the mean herd size for minimum subsistence must be 4-5 tropical livestock units (of 250 kg) per member of the family (Breman and de Wit, 1983). On average, not only is this figure not reached, but it has dropped to about 50 per cent during the recent drought periods. To compensate for the risk of losing half of their herds, the nomads and semi-nomads aim at keeping higher numbers of animals than required for subsistence in normal years. As livestock also form one of the few possibilities for investment and thus serve as a "mobile savings bank", there is a tendency to keep large herds. The tendency is limited, however, by the availability of animals, fodder, and, partially, labour, as well as low animal fertility. Moreover, a clear distinction has to be made as to the animal species concerned. Camels and cattle generally mean more than economic assets to the people, whereas goats and sheep are used for subsistence or cash income. A reason for the large numbers of animals lies also in their limited productivity. The average daily amount of milk that is available for human consumption ranges between 0.5 and 0.751 per person. The mean annual increase in liveweight may reach 20 percent (or 50 kg) per TLU and 30-40 percent for sheep and goats, respectively. This corresponds to a production of 0.3-0.5 kg of animal protein per hectare annually for the nomadic and for the sedentary livestock systems in the Sahel and 0.6-3.2 kg for the transhumant systems, which is even better than under comparable climatic conditions in Australia and North America (Breman and de Wit, 1983). Nevertheless, in parts of dry Africa the number of animals kept already exceeds the carrying capacities of rangeland ecosystems. This results in environmental degradation, low productivity of livestock, and critical supply situations (Figure 7). Much of the present degradation/desertification of former productive rangelands in dry Africa is primarily due to lack of good management rather than overstocking. Moreover, reduction of stock numbers is usually unpopular and difficult, at least as long as no adequate compensation can be offered. Compensation in one sense could be the introduction or further development of complementary cash-crop production in transhumant agrosilvopastoral systems. Thus, once more, agroforestry, mainly with its forms of agrosilvopastoral land use, is expected to improve the situation. However, there are constraints outside the range of technical development. First of all, there is an urgent need for an overall land-use policy in the countries concerned in order to regulate the allocation of lands to pastoral uses and to crop cultivation, respectively. Secondly, tenural rights may need revision, as would market policies for animal and agricultural crop products. Technically, the one-sided support of the veterinary services will have to be changed. Breeding and selection, diversification of herds and better marketing facilities can help to improve yields and reduce risks. Two bottlenecks will have to be overcome: the lack of soil nutrients (phosphorus and nitrogen) and of water to increase forage production. Digging wells other than those that are part of an integrated rural development programme has led to failures, while sustainable water management within agrosilvopastoral systems appears to be promising. In these systems the application of fertilizers may also prove to be economically feasible (which is not the case in pure pastoral management) if combined with optimal use of manure and leguminous woody perennials for soil improvement. Agrosilvopastoral systems could also help to avoid bushfires which are detrimental to soils and plant production. Sustainable management of improved pasture should be the target, a target that has already been achieved in some model projects such as the Ferlo of northern Senegal, and in other parts of the Sahel, North Africa, and East and southern Africa. Obviously, such sustainability depends on optimum grass and herb mixtures with woody perennials. This is one of the challenges to agroforestry in dry Africa. The introduction of more productive fodder shrubs and trees can supplement the presently available edible biomass and thus improve animal nutrition and animal productivity (Figure 8). 
The energy crisis in arid and semi-arid Africa has been intensively discussed in many meetings and publications. Most households and many local small industries and handicraft concerns depend entirely or mainly on fuelwood or charcoal supplies. The crisis, however, is less evident in rural areas, where it is not yet fully recognized, while it has reached a critical stage in urban areas. Most of the fuelwood is free and is collected by women and children. The average minimum demand for cooking appears to be about 1 kg of air-dry fuelwood per person per day, i.e., 360 kg, or about 0.7 m3 of wood annually. In contrast to traditional forms of supply by free fuelwood collection, the increased population has led to the clearing of natural woody vegetation for crop and rangelands, and the "carrying capacity" of natural fuelwood resources has reached, or even surpassed, critical limits, resulting in severe fuelwood deficiencies on the one hand and in environmental degradation on the other. This is a recent development which has to be urgently controlled in order to avoid catastrophe. The great majority of fuelwood plantations in dry Africa (generally sponsored by foreign aid), have proved to be uneconomic in financial terms, the overall cost of production (including high overheads and marketing) far exceeding prevailing market prices, although ecologically they temporarily relieve the immediate pressure on near-by resources. Therefore, producing firewood in agroforestry systems by the landowners themselves is one way to solve the problem (Nair, 1987a). This can be done by managing the natural woody vegetation. The biomass increment of these trees and shrubs is, however, very low. One of the reasons is that, by tradition or by law, only dead wood is harvested. If the rotation period is shortened according to the maximum increment, much more could be harvested on a sustainable basis and presumably in smaller sizes which would yield more energy than (unsplit) thick pieces. Besides, the thick pieces are usually less appreciated because of the type of heat they offer and sometimes higher moisture content. However, even forms of intensive firewood cropping with fast-growing species with a high coppicing potential may well prove to be financially attractive if combined with agriculture, e.g., in forms of alley cropping, or with range management. At present there is very little such activity because free exploitation of natural resources on the one hand and government/external project fuelwood plantations on the other are still believed to be adequate — which they are not. If pure fuelwood plantations by relief projects cannot yield a net revenue, this may be entirely different in agroforestry systems, where fuelwood is only one among a number of crops. This, in combination with energy-saving stoves and techniques, and with the development of alternative sources of energy, should be further promoted. In addition, biogas in agroforestry systems has not yet gained the necessary acceptance in Africa. Experience from Asian countries is available and could be transferred to African farmers. Renewable raw materials Most parts of dry Africa are deficient in exploitable mineral, resources, and due to scarcity of foreign exchange, imports are restricted. The significance of renewable resources that are locally available or can be grown is evident. Agroforestry, providing a wide range of diverse products, deserves special attention. Fuelwood has just been mentioned. Wood for construction, fencing, tools and implements, and many other products including handicrafts can also be produced. Again, people are very selective in the choice of species, and, as the most desirable ones have largely disappeared from the natural vegetation, the conservation of remaining single specimens on agroforestry lands and, in future, plantations of specific trees within agroforestry systems, can help to overcome foreseeable shortages. There are, however, a number of constraints which do not make timber production an attractive proposition within agroforestry. One is the problem of tenural rights, another the legal restrictions in the use of standing trees according to existing forest laws. Furthermore, only a limited number of exotic species are available that promise fast growth and high yields. However, these may not be compatible with other land uses and do not provide for a multitude of desired products and benefits. Therefore more should be done to incorporate multipurpose local species and to offer the landowner certified seeds and plant material and advise on appropriate cultivation techniques. Vegetative propagation (by cuttings) of Euphorbia balsamifera and of Commiphora africana, as well as planting of hedges and live fences using various thorny species, could be further developed. One promising example is Bauhinia rufescens which can be used for thick-set hedges in nurseries, around compounds, etc., as tried in Burkina Faso (Figure 9). Renewable raw materials go beyond wood. They include all the agricultural commodities in addition to food and forage, such as fibres, oils, and insecticidal and repellent extractives. The various gums and resins collected from trees and shrubs should be mentioned — gum arabic from Kordofan, Sudan and parts of the western Sahelian countries, various tanning and dyeing materials and fibres. A variety of pharmacological products that are essential for both human and veterinary medicine are obtained from wild-growing local trees and shrubs. The abundance of these plants has been drastically reduced by monoculture cropping but they are maintained as a free and always available "pharmacy" in most agroforestry systems. Again no statistical data are available, but without doubt agroforestry contributes substantially to health in dry Africa.
Under harsh or marginal conditions there is a general tendency, both ecologically and from the managerial aspect, to reduce heterogeneity of plant and animal species. Having fewer species, however, increases the instability of the system. This is evident from most monoculture crops in dry-zone Africa which can be successfully grown only with the support of substantial external inputs in the form of mineral fertilizers, agrochemicals, irrigation, etc. Agroforestry, by contrast, aims at a natural stability by diversification, multiple use and flexibility. Recently this has turned out to be more important for most rural people than higher yields alone. Combining different species, moreover, reduces the pressure on such limited resources as water and soil nutrients. The function of trees and shrubs as "nutrient pumps", transporting nutrients from deeper soil layers to the soil surface, is increasingly recognized in addition to their role in nitrogen fixation, erosion control, wind protection, and the provision of shade. The role of non-leguminous species in this context is still underestimated and deserves more attention. This includes mulch production and insect repellent effects in some species. Not all trees, shrubs and palms should be incorporated into agroforestry systems uncritically. Species with an extensive lateral root system may be useful for controlling erosion and fixing moving sands, but they will most likely compete with crop plants. Species with a deep-reaching taproot may be excellent "nutrient pumps" but in turn may not stabilize the topsoil. Some species may attract birds, rodents, insects or other animals which can seriously damage agricultural crops or animals (e.g., the tsetse fly depends on woody habitats). Such aspects must be carefully considered. They are generally known to the local rural population, but questions concerning the compatibility of woody species with crops and livestock have to be answered soon in order to avoid failures and disappointment which may lead to a general rejection of agroforestry C9ncepts by rural communities. Socio-economic development This phrase has been somewhat over-used and overloaded with significance. However, many essential aspects of the concept have been overlooked. Quite simply, shade in agroforestry systems results in socio-economic benefits. It is hard to work in the open fields in hot weather, and shade available in or around fields improves working conditions for the farmer. The main argument for planting trees in parts of western dry Africa is to provide shade in and around compounds, villages and fields. A very hot climate also reduces the productivity of animals. More important, however, and as yet largely unsolved, is the problem of how to create more and better jobs in rural areas. At present, many of the most active and skilled people leave their rural habitats and migrate to urban centres, expecting better chances for earning money, better educational opportunities, social security, health care, etc. Old people, women and children stay behind and are faced with the problem not only of surviving but of producing a surplus to feed the city dwellers. Agroforestry is labour-intensive. Does it simply increase the work load for those already burdened, or does it really provide more adequately paid jobs? Oases are one example. Their management is extremely labour-intensive. Many, if not most, oases at present suffer from non-availability of labour, and many of them have been abandoned or are in the process of collapse because the traditional social systems have changed. There are now more attractive job alternatives outside. The problems of manpower, jobs and income have to be solved first and at all levels if agroforestry in the dry zones of Africa is to have a chance in the future. A solution would be better markets, i.e., guaranteed marketing possibilities and attractive prices, and physical and legal market access for agroforestry products. Why should a farmer grow fruit, cereals, vegetables, fuelwood, etc., and produce meat and milk if, when trying to sell this produce on the market, he can find no buyers. This happens frequently due to government policies which favour urban populations for political reasons (e.g., by extremely low food prices) and also because of food-aid programmes. Commonly preference is given to imported food (e.g. wheat instead of locally grown millet and sorghum). These are the most serious socio-economic problems presently retarding the further development of agroforestry in the dry zones of Africa. In addition, the conflicts between traditional and government-land tenurial systems will have to be solved. People who do not own the land they cultivate, or cannot expect an extension of the lease over the period necessary for the rotation of woody perennials, can hardly be expected to invest in agroforestry. Tree planting is even actively prevented by some societies in order to avoid "sedentarization" and land acquisition by private people or by government agencies — because, according to tribal customs, property rights over cropland may be obtained by planting of trees.
At present, in spite of many encouraging projects and remarkable research efforts, agroforestry in the dry zones of Africa is stagnant or on the verge of a standstill. Wherever progress is seen, the impulse primarily comes from outside, i.e., from international organizations, bilateral co-operative agreements or national governments. Very little initiative comes from inside the rural societies, i.e., from the target groups concerned. With their participation the maintenance of age-old practices may continue, but effective innovations are hardly forthcoming. Why is this so? Is there a general resignation within rural societies in Africa or is there no future to be expected for agroforestry? Are there specific constraints to be overcome? Resignation may play a role in some regions. Resignation can be traced back to a variety of reasons. One of them is "familism", i.e., the individual has to subordinate his concerns to the interests of his family/clan. Although providing a certain social security, this hampers innovation, because most decisions are made by elders. Closely linked is "parochialism" or "ethnocentrism", which means orientation towards the interior of a given society and a certain isolation from the outside. Fatalism and superstition are widespread and often in contrast to development efforts, and "status factors", e.g., with regard to specific types of labour (labour division between men and women) as well as socio-cultural restrictions or commitments have to be observed. But people in the dry zones of Africa cannot afford resignation or inactivity in the future. Since the population will double within the next 17-25 years, no time can be lost in creating the conditions for their survival. Remaining at the present levels of extremely low agricultural production on the one hand and increasing environmental degradation on the other will inevitably lead to incredible hardship, if not the starvation of millions of people, within the next generation. An East African proverb clearly defines the challenge: "We have not inherited our environment from our fathers, we have loaned it from our children". Since there is no more free land left to migrate to if the region's own resources are exhausted, people will have to change from the previously highly developed strategies of mobile resilience, however well adapted to the environment, towards a more permanent strategy. Agroforestry in its various forms lends itself to sustainable land use with increased permanence, although it maintains a high degree of flexibility once a site has been chosen. The fact that more people will have to be fed in the future will increase the demand for agroforestry products of all kinds, especially for food, forage, fuelwood and materials. Population growth in other regions may even open additional markets for African rural products, e.g., in southern Asia. To meet these demands, however, a mere exploitation of the natural resources of African dry zones is not sufficient. As in all other sectors of the economy, investments will be needed for agroforestry. Here, however, real bottlenecks exist. First, by tradition and lack of facilities, the majority of people in the dry zones of Africa were not accustomed to invest, except in children and livestock, and there are few resources available to be invested except labour. Moreover, confidence in investment has not increased in recent times. Secondly, and perhaps more seriously, there is escalating competition for power, power based on the disposal of basic investment parameters. Previously, power was mainly attached to control over lands, i.e., over natural resources which are basic for investment. Private land ownership is hardly to be found in the dry zones of Africa. Traditionally the local societies allocated lands to individuals for a limited period and under specific conditions. At present, governments claim that right for themselves, partially overlapping with traditional patterns. Thus the individual land user has very little influence on long-term decisions and, therefore, little interest in long-term investment. Those disposing of the land, however, often do not have the resources and institutional structures for overall rural development. "When land is controlled by government or by a powerful few, the majority has little motivation or possibility to assure continued soil fertility or to maintain perennials" (FAO, editorial of Unasylva, No. 4, 1986). Capital is the second asset for investment. People or groups disposing of money were, and still are, believed to rule others. Rural people in dry Africa certainly do not have the financial resources they would need for investment, and government or even foreign funds are more than limited. The alarming dependence on money supplies, either in the form of grants or credits, will have to be solved in a solid and effective way. Labour, number three of the investment "goods", is generally assumed to be abundant, but this is not true, as indicated above. It is not true with regard to seasonal distribution (only short cultivation periods), and labour is not available at sufficiently low cost. Mechanization as a possible solution would also require money and would remove the decision levels to donors outside the rural systems. Labour, however, includes all human skills, and these depend on both experience and education. As stated, a wealth of experience, especially in resource management of agrosilvopas-toral and silvopastoral systems, exists in the dry zones of Africa, specifically among farmers and herdsmen. This is the reason why it is so important to "rediscover" the existing knowledge and to make it available to all Africans; it is theirs. Education must be restructured and should be based on available knowledge, facts, local experience and applied research. Much progress has already been achieved by international and national institutes engaged in improving land use. ICRAF is contributing substantially to agroforestry development in Africa by its strategic research programme, by its world-wide information services and courses held. But the search for knowledge must also start at the grassroots level, in primary schools, in simple experiments and observations. The question behind all these efforts should -be "What can be improved?" Research must be future-oriented. Our understanding of the feedback between ecological, socio-economic and socio-cultural processes is still so scant that up-grading of our knowledge must be a permanent challenge. This is important for agroforestry. Agroforestry must not be an attempt to return to stone-age practices or to keep people at low levels of development and dependent on external inputs. On the contrary, it has to apply the most efficient technologies available in order to serve people and to maintain the potential of natural resources on a sustainable basis. Agroforestry will have to be opened to dynamic technical progress, and is to be made a land-use concept suitable for spearheading development by its holistic systems approach. In future, for instance, biotechnology (particularly gene technology) can be expected to play a major role. This may range as far as sustainable and large-scale soil improvement by bacteria and breeding of high-yielding multipurpose plants and animals. Efforts should focus on reducing waste and on quality improvement since there are definite limits to quantitative growth. In future it will be unacceptable to produce more waste than target products unless waste can be adequately recycled in agroforestry enterprises. Biotechnology will have to be applied to replace former practices of breeding plants and animals, specifically to shorten the cycles of tree production and reproduction. Modern techniques of cellular and subcellular engineering (such as gene splicing and recombinant DNA, cloning, hybridomas and monoclonal antibodies, protein engineering and nitrogen fixation) will open up hitherto unexploited resources. Simultaneously, however, their potential for misuse will have to be carefully watched.
Many people are searching for a paradise which in reality never existed on earth. Remarkably, however, paradise was not described as open fields and grazing lands or closed forest. Paradise was a highly diversified garden, in some way a perfectly functioning "agroforestry" system, in which valuable trees played an essential role. Misuse of fruit trees finally led to the expulsion of man from that paradise. It is worthwhile to reflect from time to time upon the meaning of old symbolic wisdom. It is not only worthwhile but absolutely necessary to know more about the wisdom of rural people in the arid zones of Africa in order to further develop agroforestry to the year 2000 and beyond. But likewise it is important to elaborate new strategies by combining knowledge from inside the target groups involved and from external research organizations. This is the type of investment which is likely to yield the highest rates of interest in material and intellectual terms. For the future of agroforestry, however, more than research, technical development and funds will have to be mobilized. An overall national land-use strategy is required for all countries, based on realistic facts and targets, interdisciplinary and co-ordinated efforts towards a holistic systems approach, the harmonization of human needs and the environmental carrying capacity. The African CILSS-countries have taken some very important steps into this direction. Moreover, these strategies will have to be effectively applied in practice. The responsibility for the future, therefore, lies not only with the farmers, scientists and technicians but first of all with the policy makers. This has to be stated quite clearly. Above all, agroforestry is not a panacea of land-use methods. It may be useful or even optimal under specific conditions but indifferent or even unsuitable elsewhere. Agroforestry may not be acceptable and practicable for all people, and some groups may be more active and successful in developing agroforestry than others. The challenge is thus to concentrate on the most promising sites and systems and to develop these like islands within other forms of land use. The islands, hovever, should be mutually linked to form a network in order to increase the region's carrying capacity — a web requiring minimum material inputs and providing, above its traditional elasticity, sustainability and the development potential that is needed for the livelihood of future generations.
Bajaj, Y.P.S. (ed.). 1986. Biotechnology in Agriculture and Forestry —
1.
Trees. Berlin: Springer Baumer, M. 1987. The potential role of agroforestry in combating desertification and degradation of the environment. Wageningen Technical Centre for Agricultural and Rural Co-operation (CTA) (ACP-EEC Lome Convention). Becker, B. 1984. Wildpflanzen in der Ernaehrung der Bevolkerung afrikanischer Trockengebiete: Drei Fallstudien aus Keniaund Senegal. Inst.f.Pflanzenbau u.Tierhyg.i.d.Trop. u.Subtrop, Goettingen: Goettinger Beitrge zur Land- und Forstwirtschaft in den Tropen und Subtropen, Heft 6. Breman, H. and C.T. de Wit. 1983. Rangeland productivity and exploitation in the Sahel. Science 221: 1341-1347. FAO. 1983. Fertilizer yearbook, Vol. 32. Rome: FAO. ——.1984. Land, food and people. Rome: FAO. FAO/SIDA. 1981. Agroforesterie Africaine. Une étude préparée par la Faculté des SciencesAgronomiques de l'Etat Section forestierè des pays chauds. Gembloux, Belgique. Rome: FAO. Galaty, G.J., D. Aronson and P.C. Zalzman. 1981. The future of pastoral peoples. Proceedings of a conference held in Nairobi, Kenya, 4-8 August 1980. Ottawa: IDRC-175e. Hall, A.E.. G.H. Cannell and H.W. Lawton. (eds.). 1979. Agriculture in semi-arid environments. Berlin: Springer Verlag. Le Houérou, H.N. 1980. Browse in Africa: The current state of knowledge. Addis Ababa:International Livestock Centre for Africa (ILCA). IBPGR. 1984. Forage and browse plants for arid and semi-arid Africa. Kew, U.K.: Royal BotanicalGardens, and International Board for Plant Genetic Resources (IBPGR). IUCN. 1986. The IUCN Sahel report: A long-term strategy for environmental rehabilitation. Gland, Switzerland: International Union for Conservation of Nature and Natural Resources (IUCN). Jahnke, H.E. 1982. Livestock production systems and livestock development in tropical Africa. Kiel, FRG: Kieler Wissenschaftsverlag. Louw, G. and Seely, M. 1982. Ecology of desert organisms. London/New York: Longman. Lusigi, W.J. and G. Glaser. 1984. Désertification et nomadisme: une étude pilote en Afrique orientale. Nature et Resources 20 (1): 21-31. Miehe, S. 1986. Acacia albida and other multipurpose trees on the Fur farmlands in the Jebel Marra highlands, Western Darfur, Sudan. Agroforestry Systems 4: 89-119. Nair, P.K.R. 1985. Classification of agroforestry systems. Agroforestry Systems 3: 97-128. ——.1987a. Agroforestry and fuelwood production. In D.O. Hall and R.P. Overend (eds.), Biomass: Renewable energy. Chichester: John Wiley. ——.1987b. Agroforestry systems inventory. Agroforestry Systems 5: 301-318. Penning de Vries, F.W.T. and M.A. Djitéye (eds.). 1982. La productivité des pâturages sahéliens: Une étude des sols, des végétations et de I'exploitation de cette ressource naturelle. Agricultural Research Reports 918. Wageningen: Centre for Agricultural Publishing and Documentation. Rapp, A., H.N. Le Houérou and R. Lundholm (eds.). 1976. Can desert encroachment be stopped? A study with emphasis on Africa. Ecological Bulletin No. 24. Stockholm: NFR/ UNEP/ SIES. Ruthenberg, H. 1980. Farming systems in the tropics. 3rd ed. London: Oxford University Press. Simmonds, N.W. 1985. Farming systems research: A review. World Bank Technical Paper 43, Washington: World Bank. Steiner, K.G. 1982. Intercropping in tropical smallholder agriculture with special reference to West Africa Schriftenreihe der GTZ. No. 137. Eschborn: GTZ. United Nations. 1977. Round-up, Plan of Action and Resolutions, United Nations Conference on Desertification, 29 August-9 September 1977. New York: United Nations. |
