R Behnke and C Kerven - Overseas Development Institute, Regent's College Inner Circle, Regent's Park, London NW 14 NS, UK
Introduction
Overall objectives
Project purpose
Results
Activities
Conclusions
References
1 This paper is a slightly modified version of one with the same title which appeared as Natural Resource Perspectives N° 1 published by ODI in November 1994
The arid and semiarid zones cover about one third of the earth's land surface but nearly two thirds of the African continent. The majority of African livestock and possibly 30 million people dependent on livestock reside in these dry zones along with the greatest and most diverse concentrations of wild mammals still in existence (Ellis, 1994). Twenty of the world's poorest countries are found here. Economic considerations, environmental interest, geographical extent and human welfare suggest that African rangelands should be high on the development agenda but they are not. This paper discusses some of the reasons for the neglect and proposes some remedies.
Throughout the 1960s and 1970s the blueprint for African range and livestock development projects was the ranching model (Sandford, 1983). By the early 1980s poor project performance had subverted confidence in this model. A decade of experiments involving large (World Bank pastoral associations) and small (NGO water harvesting, restocking and paravets) donors ensued. This was accompanied by extensive field research (including pastoral systems studies conducted by the then International Livestock Centre for Africa) and theoretical retooling (notably in scientific ecology). Much of this was innovative and practical but it did not provide a framework for assembling new research ideas and intervention techniques into a more adequate policy for rangeland development. This now seems possible.
The ranching model projects of the 1960s and 1970s presumed that enlightened resource management was both intrinsically good and likely to provide economic dividends. Economics and conservation were linked by the presumption of pastoral overstocking (Behnke et al, 1993). Overstocking was supposed to explain why African pastures were degraded, herd output was low and pastoralists were poor. The problem with ranching projects, however, was that they could not deliver lower stocking rates.
The development initiatives of the 1980s, whatever their other successes, proposed few new techniques for adjusting livestock numbers to forage supply on open rangelands. Recent advances: in scientific ecology and pastoral studies have also failed to "solve" the overstocking problem and have not suggested more effective ways of removing surplus animals. They have, however, encouraged a reframing of the problem.
The concept of the four horizontal rows (Table 1, Table 2) depicts relationships among the broad objectives to which a project contributes, the project's immediate purpose, its intended results and the activities it will support. The four vertical columns refer to aspects of this cause and effect sequence - the logical links among intentions and results, measurable indicators of performance, how to obtain data for assessments of performance and the assumptions underpinning each step of the exercise.
There are remarkable continuities and important changes in pastoral and range development projects from 1960 to the present (Table 1, Table 2).
Table 1 African rangelands - the ranching model of the 1960s and 1970s
|
Item |
Intervention logic |
Achievement indicators |
Sources of verification |
Assumptions |
|
Overall objectives |
Rangeland conservation |
Botanical indices |
Assess vegetation condition/trend |
Stocking rate determines vegetation characteristics |
|
Project purpose |
Improve income from livestock |
Animals sales |
National accounts |
Fewer animals = higher output |
|
Results |
Destocking |
Lower stocking rate |
Livestock census |
Rangeland overstocked |
|
Activities |
Marketing, fencing, ranch demarcation, water development, etc | |||
Table 2 African rangelands - tracking and buffering range livestock projects in the 1990s
|
Item |
Intervention logic |
Achievement indicators |
Sources of verification |
Assumptions |
|
Overall objectives |
Sustainable rangeland production |
Outputs/ha, trends over decades |
Historical records, national statistics, modelling |
Episodic ecological change |
|
Project purpose |
Improve income from livestock |
Livestock output in cash/kind |
Improved national accounts |
Income subject to risk |
|
Results |
Track/buffer environmental fluctuations |
Statistical measures of variance |
Climate records, improved national statistics |
High level of environmental variability |
|
Activities |
Service delivery, marketing, tenure rights, water development, etc | |||
Both the former ranching projects and those now proposed share an important characteristic in that they are attempts at natural resource management. In this respect both the very old and the very new project formats are distinct from most pastoral development efforts of the 1980s. Typical projects of the 1980s sought to provide services, improve welfare, improve pastoral incomes or develop pastoral community organizations. They had little success in accounting for how these activities contributed to sustainable resource management and its links with economic development (Table 2).
There are, however, fundamental differences between the objectives of range resource management in the 1960-1970s and the 1990s. These changes are underlined by abandoning of the earlier goal of "rangeland conservation" (Table 1) in favour of "sustainable rangeland production" (Table 2).
For rangeland managers of the 1960s and 1970s domestic livestock were an intrusive element that destabilized "natural" botanical systems. The notion that domestic herbivores were a foreign intrusion had considerable intuitive appeal in North America and Australia-where stock owned by Europeans had suddenly burst on the scene. The natural "before" and the disturbed "after" were clearly distinguished in historical time and the requirements of industrialized agriculture confronted a nostalgia for a bygone landscape. Concepts of plant community succession and climax provided a powerful theoretical rationale for the conservation of pristine rangeland flora. Traditional range management conceived in these terms was fundamentally botanocentric. The state of the vegetation marked the success or failure of a management regime with botanical indices such as plant population; vegetative mass and species composition providing evidence of range trend, condition and livestock carrying capacity.
Despite its enduring popularity this type of rangeland assessment suffered in Africa from several limitations. The distinction between natural and manmade vegetation caused philosophical problems and operational ambiguities. Humans, their fires, and domestic stock contributed to creating some of Africa's most productive and picturesque savanna landscapes. Parts of North Africa, on the other hand, are undoubtedly degraded by human use but have been so since Greco-Roman times.
The use of predominately botanical indices to assess the performance of a form of agriculture in which plants are not directly used or consumed by humans might also be questioned. Range vegetation must be eaten by animals if profits are to be made. In areas where rainfall is reasonably constant large livestock populations may indeed consume enough vegetation to alter the plant life that they leave behind. Range livestock production - like most forms of agriculture - may alter the natural vegetation to produce food, fibre and other goods for human use.
These botanical changes are not proof of degradation unless agriculture itself is equated with degradation and no distinction is made between agricultural systems capable of producing for prolonged periods and those that are not. The conservation of pristine vegetation is of less concern for agriculturalists than the expected length of time that output can he maintained from altered vegetative states under different management regimes. In short, the objective of botanical immutability is less useful than a workable notion of sustainability. These concerns are reflected by the indicators of project achievement stated as maintenance of livestock product output over an extended time (Table 2).
The common purpose of both old and new kinds of projects is to increase producers' incomes from livestock. What distinguishes the new type from its predecessors is the way livestock income is defined and measured. This is a more significant change than would first be suspected
Livestock income in the older ranching projects was effectively defined as cash from sales of animals for slaughter. This presumed that pastoral development was a matter of technology transfer. Most industrial ranchers supported themselves by marketing carcasses and it was thought that modern African pastoralists should do the same. Such reasoning was an important link in the logical structure, and hence the appeal, of ranching projects. The belief was that animal sales had three beneficial functions: increased pastoral incomes; destocking and conserving rangelands; and supplying urban consumers with an essential commodity (Table 1). There were no uncomfortable trade-offs and everybody was a winner (Kerven, 1992).
Applied research on pastoral economies over the last 10 years has explained why this scenario was too optimistic. This hinges on the volume and kind of produce yielded by pastoral herds. Contrary to the assumptions of the ranching projects, traditionally managed livestock often provide their owners with cash and in kind benefits in excess of those to be derived from additional animal sales (that is, unless meat prices increased). If urban consumers were to eat more meat they had to pay prices high enough to bid against the alternate uses of livestock in rural trade networks, for immediate household use, as inputs into other productive processes or for breeding or growing out. Herd structure studies also suggested that urban consumers (or pastoral cooking pots) were already claiming the categories of animals suitable for slaughter and having few other competing uses. In other words there was no vast underexploited reservoir of meat standing around on the range chewing its cud. Comparative studies also show that pastoral productivity consistently equals, and frequently exceeds, the calorie, protein or cash value of output per unit land area from ranches in comparable ecological situations. Rational pastoralists therefore took their place beside rational peasants in the academic and development literature of the 1980s. The great leap forward in pastoral output and income unfortunately never materialized. It had been based on an illusion all along.
There remains, nonetheless, scope for genuine improvement of pastoral incomes and output through increased commercialization of pastoral systems of production, product disposal and household provisioning. Raising pastoral incomes will therefore remain the purpose of the new generation of projects. The opportunity for these improvements is created by the difference between the calorie and cash terms of trade for livestock products and grain. With certain important exceptions, prices for grain, meat and milk are such that pastoralists obtain more calories by selling livestock produce and buying grain than they could obtain directly from consuming the protein rich products of their herds. For poor pastoralists this means a chance of survival despite reduced per caput herd wealth. For the rich favourable terms of trade provide an opportunity to improve their standard of living or to reinvest surplus earnings in pastoral production.
Commercial investment in pastoralism is needed since traditional production systems are well adapted to the demands of their natural environment and output per hectare is already high. Increased output is thus dependent on the use of new industrial inputs. If producers are to obtain the cash to purchase these inputs commercial livestock production is unavoidable. Indeed, it is not only unavoidable but is happening. Recent field research has shown that small scale commercial innovations are continuously undertaken in most pastoral economies. These spontaneous changes are the exact opposite of the carefully engineered leaps envisaged in the ranching model. The more modest pastoral development efforts of the 1980s also successfully promoted incremental changes.
Maintaining a constant low stocking rate was the primary intended result of the ranching project. In hindsight, however, there was little scientific evidence that destocking programmes would fulfill their environmental objectives or increase total livestock output (Table 1). Lower stocking densities could, in fact, actually damage pastoral incomes.
Stocking rates low enough to ensure that forage shortages never occurred would be uneconomic to maintain in very dry environments with widely fluctuating rainfall. Economically optimal stocking densities also vary according to the kinds of products yielded by herds, the breeds and species kept and the husbandry techniques employed. Pastoral stocking densities may - be too high to maximize beef production per hectare or to meet the botanical standards of professional observers trained in a ranching environment. These densities may, nonetheless, maximize the combined output of live animal products such as milk, traction, manure and fibre. Destocking these "overstocked" rangelands would probably depress both individual and aggregate pastoral incomes. This probably explains the almost universal rejection by pastoralists of enforced destocking programmes.
Environmental benefits of conservative stocking regimes are as dubious as their alleged economic benefits. Irreversible changes in plant life occur episodically and not incrementally in climatically unstable environments. The dominant variables driving these ecological changes are physical factors, such as rainfall, outside management control. In these event dominated systems it is unrealistic for managers to try to forestall environmental change by tinkering with a single dependent biological variable such as livestock numbers. Managers who cannot control their environment must quickly adapt to it if they are to minimize the consequences of unpredictable rainfall fluctuations. This opportunistic approach to rangeland exploitation demands temporary but sudden and very substantial adjustments in livestock feed demand in response to precipitous changes in feed supply (Table 2).
Flexible strategies of resource exploitation must be profitable as well as environmentally beneficial if producers are to adopt them. In other words, environmental concerns must dominate project objectives but economic concerns define project purpose (Table 2). These dual intentions require both biological and economic indices of project success -tracking and buffering of environmental fluctuations. Tracking refers to the biological phenomenon of prompt adjustment of livestock forage demands to fluctuating levels of primary production. Buffering of environmental fluctuations refers to the economic phenomenon of shielding of pastoral incomes from the worst effects of violent climatological and biological fluctuations.
Quantifiable measures of project success are not the same for biological tracking and economic buffering. Parallel changes in livestock feed requirements and supply provide evidence of successful biological tracking. Economic buffering dampens the effects of environmental variability by creating more stability in the income from livestock than in rainfall or primary production levels. Range livestock development needs activities, or combinations of activities, that simultaneously produce high coefficients of variation for feed demand and low ones for the value of output. This is "opportunistic" rangeland management or the attempt to maintain large, healthy and productive herds that allow, when conditions dictate, removal of as many animals as necessary as quickly and profitably as possible.
Opportunistic resource management is not new to African pastoralists. Official endorsement of opportunism does not, therefore, demand the radical reform of existing husbandry systems. It does, however, bring government and donor management objectives into line with customary practices, anticipates evolutionary rather than sudden economic change and, belatedly, adds pastoral development to the growing list of participatory or client oriented forms of development. Responsible project design must, however, balance local and national interests, and match community priorities with wider policy concerns. Environmental tracking and economic buffering are among these concerns and provide criteria for screening local initiatives.
Field workers, often in conjunction with NGOs, have developed many new techniques for delivering services to pastoralists over recent years. These include improved systems for primary animal health care, water harvesting and storage, design of drought and famine early warning systems and postdrought restocking. All range livestock development initiatives already have a core of tested field techniques to draw on, reject or modify in the light of local circumstances and the policy framework (Table 2).
There is also scope for modifying existing project methods. Project components urgently in need of revision are those that are expected to perform new functions under opportunistic management. In the older project framework drought was an emergency - an unexpected catastrophic event outside the parameters of normal planning. In the present framework erratic rainfall (that is, drought of varying severity) is viewed as a continual hazard. Incorporating drought into the concept of normal climatic variability demands a rethinking of how pastoral relief and development is to be achieved.
Famine relief
Emergency sales force many animals to market during droughts, forcing prices down when poor harvests and grain scarcities are causing cereal prices to rise. If market forces set food prices under these adverse conditions some pastoralists may starve unless they receive relief provisions. Development agents are therefore confronted with an apparent dilemma in that they either let human and animal populations "track" environmental fluctuations and people suffer or they "buffer" pastoral incomes from environmental stress but foster dependency.
A more attractive approach to providing relief would attempt to make sure high levels of grain were available through normal commercial channels. This could be achieved by bulk sales of relief supplies at concessionary prices. Hoarding and speculation would be controlled by adjusting the volume of external supply relative to the strength of internal demand. It would be necessary to maintain drought early-warning systems to provide information on the geographical extent and severity of a crisis in order to estimate these supply-demand factors but in order to influence these factors there would need to be adequate transport infrastructure and, possibly. transport subsidies to ensure that food moved in the desired direction. Consistent with the anticipated results of the project (Table 2) these arrangements would promote both buffering and tracking since pastoralists could maintain their incomes only by selling stock during droughts.
Livestock marketing
Tracking and buffering would also require improved livestock marketing systems. Livestock marketing was previously seen by project designers as a mechanism for maintaining continuous high offtake and low steady stocking rates to prevent overstocking and deaths of animals during droughts. Ecological research suggests, however, that livestock population crashes are unavoidable when rainfall is erratic. A more realistic project goal under these conditions is not to design marketing systems that can forestall fluctuations in throughput but to design systems capable of absorbing such fluctuations. Low cost techniques of meat preservation, improved transport infrastructure, access to the largest possible consumer market for meat and elimination of subsidized international competition may be components of this effort. Withdrawal of government regulatory agencies or marketing monopolies that add to the covert costs of trading, stifle competition and depress producer prices may be equally important.
Land tenure
Project managed land tenure reforms in the past attempted to limit herd growth by confining herds to restricted areas. Managers viewed permeable territorial boundaries as undesirable since livestock owners were thereby allowed to escape the negative effects of overstocking.
Opportunistic strategies of resource exploitation turn this reasoning on its head. Shortage of feed is often localized due to the erratic distribution rather than the total absence of rain. If adjacent grazing areas experience asynchronous productivity flushes, herd mobility - and non-exclusive tenure arrangements that allow it - are a cost effective way for animals to walk away from temporary local imbalances in stock numbers and feed supply. The practical question for project design is not how to eliminate non-exclusive tenure systems but how to ensure that pastoralists can take advantage of them.
Rangelands must be comanaged by local communities and government authorities. Government cannot intensively administer rangelands because their output is generally low and erratic and will not pay the costs of direct administration. The only economic solution is for users to bear the costs of resource management. They will be willing and capable of doing this only if they have proprietary rights. In pastoral areas, however, administrators cannot expect to allocate resources once and for all. These are environments where rainfall and forage productivity are fleeting resources. Human and livestock populations must rearrange themselves in space on a seasonal and interannual basis. By conferring basic property rights on producers and local communities impartial intercession will still be required to sort out the conflicting, shifting and multiple entitlements implied by these rights. In order to be able to do this local government authorities must establish their neutrality, institute procedures for conflict resolution and enforce their decisions.
Cultivated forage
Past forage development programmes concentrated on improving yields from cultivated fodders. Opportunistic management would put less emphasis on the search for yield increases and concentrate on production of forage when it was most needed in low rainfall years. Measures of success would be indicated by yields which were less variable than those from surrounding natural vegetation and improved profits from livestock resulting from a greater total feed supply.
A review of existing research (Scoones, 1994) has produced recommendations for project activity in the areas discussed in this paper. Field research on problems of opportunistic management would undoubtedly produce more precise recommendations or increase confidence in those already proposed.
In the short term what is left is a mixed picture. Some activities such as paraveterinary programmes, water harvesting and famine early warning systems are ready for large scale use. Other potential project components such as famine relief, land tenure, livestock marketing and forage production still need research and field experimentation but are likely over the long term to improve the tracking and buffering of environmental variability in Africa's rangelands.
Behnke RH, Scoones I and Kerven C (eds) 1993. Range ecology at disequilibrium: new models of natural variability and pastoral adaptation in African savannas. Overseas Development Institute: London, UK
Ellis J. 1994. Ecosystem dynamics and economic development of African rangelands: theory, ideology, events and policy. In: Environment and agriculture: rethinking development issues for the 21st Century. Winrock International: Morrilton, USA.
Kerven C 1992. Customary commerce: a historical reassessment of pastoral livestock marketing in Africa. Overseas Development Institute: London, UK.
Sandford S. 1983. Management of pastoral development in the Third World. John Wiley & Sons: Chichester, UK.
Scoones I (ed) 1994. Living with uncertainty: new directions in pastoral development in Africa. Intermediate Technology Publications: London, UK.
