The geographical scope

Rangelands are one of the most important land use types in the world. Roughly 47 percent of the earth's land surface is rangeland, about 80 percent of which is at least moderately degraded. Substantially more than half of this area lies in tropical and subtropical areas. By contrast, less than 10 percent of the earth's surface is devoted to arable agriculture. In recent years, degradation has been increasing, and many of the world's major rangelands are at risk. The desertification hazards map, compiled by several United Nations agencies, shows are as that are at greatest risk of desertification because of a continuation of factors such as low rainfall, terrain, soil and vegetation conditions, and high human and animal use. Identifying ways to arrest and reverse these trends is the central concern of this study.

Tropical and subtropical rangeland is usually characterized by rainfall so low or irregular that crop production is difficult unless irrigation systems, which require high capital investment, can be established. Typical vegetation types include grasslands, low-density forests (such as dry tropical thorn forests), and shrublands. The term desert is commonly applied to those rangelands that are the most arid and least productive of vegetation.

Areas at risk of desertification

Rangelands are used to support grazing of browsing livestock. The openrange system of livestock production is characteristically associated with the extensive use of sparse, scattered, and often seasonal or ephemeral natural vegetation. Most of the world's estimated 3 billion head of domesticated livestock are reared on tropical or subtropical rangelands, rather than on highly productive pastures or in feedlots. Most of the red meat consumed by humans is produced on rangelands. Some 135 million people, about 20 percent of the world's population, base their economies and societies on rangeland resources.

The pastoralists' systems of livestock production characteristically require movement of herds and herders over large areas, on either a seasonal basis (transhumant pastoralism) or a basis of perpetual migration (nomadic pastoralism). Pastoral socioeconomic systems can be further distinguished by combinations of other activities with those of herding. Herding, along with some agricultural crop production is a frequent combination, creating what are known as agro-pastoral systems; herding and tree-crop production creates silvo-pastoralism. Important economic activities of other pastoralists involve conducting, supporting, or controlling long-distance overland trade routes.

Reduced vegetative cover is often the result of overstocking, uncontrolled or concentrated grazing, removing woody plants for fuel or shelter, and clearing natural vegetation. Destructive management practices are often related to colonial-period modifications of indigenous management systems: for example, highly regulated land-use systems, such as the herima system of Masina, were widely converted to open-access systems with the imposition of the European law of public domain; the ecological integrity of indigenous pastoral systems was frequently destroyed through the imposition of international boundaries; cattle-based livestock systems were promoted in marginal drylands at the expense of sustainable indigenous systems based on better-adapted forms of livestock; wells projects encouraged concentrated grazing around water supplies; and the new economic order stimulated herd increase and its attendant economic rewards. Population growth in pastoral societies, approximately 1.5-2.0 percent annually (National Research Council, 1983b), is regulated by the observance of restrictive mores and other mechanisms. Hence, internally generated population growth is seldom a problem in pastoral zones. Indeed, declining population, with a corresponding inability to manage large herds, is often a serious problem. Further, the higher annual rates of growth in agricultural societies, characteristically on the order of 2.5-3.0 percent, encourages agricultural expansion into rangelands and forces pastoralists to increase use pressure on the remaining land base. Risks are highest in the more arid rangelands where rainfall is lowest, or least dependable, and where evaporation and transpiration rates are highest.

Soil erosion is another consequence of rangeland mismanagement. Generally, it is a natural and continuous process, but in undisturbed ecosystems with a protective cover of plants, the soil is usually regenerated at the same rate that it is lost. If soil and vegetative cover are not in balance, erosion is accelerated. Under natural conditions, it takes several hundred years to generate a few millimeters of topsoil, but it can be lost to erosion in minutes if no protective vegetation is present.

Rehabilitating rangelands requires a respite from their intensive use. This involves the elaboration of a program of land management or, as a last resort, a program of land preparation, seeding, and planting, followed by careful management of use by livestock. To bring about such a program, alternative supplies of water, fuel, food, and other services may have to be provided to local pastoralists while the improvement is taking place. It is often helpful to develop programs of range improvement with reference to precolonial indigenous adaptations, where applicable, and knowledge of the changes that have occurred in local systems of production through time. Employing local people in seeding and planting efforts must also be a part of the development program. Local people must participate fully in demonstration projects; their understanding and commitment is clearly necessary for success.

Protection of a large number of relatively small rangeland areas is especially valuable to reveal what species are present, to provide for seed production, and to demonstrate the potential for ecological recovery. Demonstrating benefits that can be obtained by restoring the full natural cover and productivity of the vegetation may be a means of persuading local people that relaxing pressure on grazing lands is essential. However, this alone does not mean that long-term viability will be established. Protected areas and other conservation measures may restrict access to fuel, food, forage, and other products. If substitutes are not provided to compensate for the loss of these resources, the local community is likely to frustrate the conservation measures. Compensatory measures might include improving pasture quality, establishing fuelwood or animal fodder plantations, or providing credit, alternative food, supplemental animal feed, fuel, or fiber, as appropriate. If the measures concerned take time to yield results, they must be supplemented by measures that bring immediate benefits. For example, if a protected area or watershed forest is threatened by wood-cutting for fuel, it will be necessary to provide an alternative source of fuel that can be used immediately.

AN IGNORED RESOURCE

Dry grazing lands are largely ignored by governments and technical assistance agencies. Very little agricultural research funding goes to rangeland development. This stems partly from the limited political power of pastoralists or typical rangeland populations, and partly from the high cost of traditional land rehabilitation practices.

As a consequence of this neglect, opportunities have been lost to make vast marginal areas more productive for pastoral people and other rural populations. Further, deteriorating rangelands contribute to flooding; when precipitation interception and infiltration are reduced, runoff correspondingly increases. The problem is compounded by the downstream effects of flooding, variously including damage to crops and settlements. Insufficient research is aimed at finding ways to protect existing rangeland or identifying appropriate new crops that would complement livestock production. Agronomists the world over are taught the principles of agricultural land management, but are unfamiliar with special rangeland techniques. The uninformed conversion of rangeland to cropland has helped destroy rangelands. For example, the moldboard plow destroys the surface vegetation and exposes thin, dry soil to erosion by wind and water. Clearly, abuse of rangelands has accelerated erosion, at times rapidly, at times slowly. As soil erodes, sediments enter water courses and river bottoms, estuaries, and deltas, which have become the sink for nutrient-laden sediments.

CAUSES OF ENVIRONMENTAL DEGRADATION

The causes of land degradation must be understood in developing a strategy for land rehabilitation. Before costly methods are applied to promote range improvement, existing degradation forces should be identified and countered. Although the role of specific social forces is frequently unclear, devegetation - the progressive disappearance of plants from the land - is generally the result of human activities: overexploitation by man and his animals.

Degradation can begin with the depauperization of woodlands adjacent to and within dry rangelands, the encroachment of agriculture, or drought. Improper grazing usually follows these events on constrained range resources. Continued high animal density accelerates the removal of palatable species and the lack of competition permits the growth of species that are less palatable or less capable of supporting livestock on a sustainable basis. In many areas of the tropics and subtropics, the pressure of heavy use has caused ecological deflections from predominantly perennial grasses to annuals, resulting in a sharp reduction of late dry season carrying capacity. Woody shrubs, whose roots compete for soil moisture with the remaining grasses, also increase in number. In many instances, the woody vegetation is cut for fuel and shelter, thereby limiting the fodder contributions of the vegetation. Wind and water assault the exposed soil, causing cyclical flooding and erosion, eventually leaving the land to become bare and sandy. The overuse of vegetation by man and his livestock reduces the possibility of natural reseeding. The complex mixture of native species that are palatable or desirable becomes locally simplified as species become "extinct." In naturally semiarid regions, this leads to what is becoming known commonly as desertification - the simplification of the ecosystem to the point where its biological productivity and diversity are reduced to a minimum, and it can no longer sustain man and his animals in more than a casual, seasonal, and very limited fashion.

In areas such as the West African Sahel, there has been speculation that a progressive ecological downturn in productivity may be occurring in a highly variable climatic region that until recently experienced unusually favorable rainfall when viewed on an annual basis. However, the consensus of informed opinion is that to date there is insufficient evidence for attributing recent droughts to a major climatic change. These climatic cycles are very long term, and reconstruction of paleoclimate is difficult in the absence of historical records of rainfall fluctuation. Climatic cycles are perhaps 400-500 years long, with internal shorter cycles on the order of 70 years, and phases of 25-40 years (National Research Council, 1983b).

On the other hand, there is short-term evidence that the relationship between land abuse by man and his livestock, resulting in ecological deterioration, is more direct and irreversible than is generally appreciated. There is also evidence of low soil fertility, which continues to keep productivity low when combined with low rainfall (Breman and deWit, 1983). Historical review of former levels of vegetation, productivity, and land-use indicators shows that much higher levels of biological productivity were maintained throughout similar past fluctuations in rainfall, until changes in land-use patterns occurred (National Research Council, 1983b). One of the most significant changes was the introduction of large-scale cattle husbandry in colonial Africa, and the subsequent control of the tsetse fly in some are as by insecticides and by removing the vegetation of the fly's habitats.

In the areas where this occurred (sub-Saharan West Africa, the Horn of Africa, Botswana in southern Africa), a previously complex ecosystem comprising trees, shrubs, and grasses, with a wide variety of wildlife and livestock (camels, goats, sheep, and cattle), has become progressively simplified. The remaining perennial browse trees and shrubs and grasses are subject to overgrazing.

The consequences of devegetation are most stark in arid and semiarid areas, where they lead rapidly to erosion, elimination of many useful species, and a drastic drop in carrying capacity, resulting in death of animals and the drift of refugees to other areas. In wetter agricultural areas, environmental consequences may not be so dramatic, but the economic impact can be substantial because the erosion-droughtflood cycle interferes with cropping.

In recent years, an additional factor has been the rising cost of petroleum fuels, which has led to a growing use of fuelwood. Much of the fuelwood has been taken from areas surrounding agricultural regions for local domestic use. However, fuelwood and charcoal are increasingly being sold in towns to peripheral urban dwellers for cooking as well as to commercial establishments for various uses. Little of this additional demand has been met by increased planting of commercial or state fuelwood lots (National Research Council, 1983a). The impact of the deforestation is being felt as reduced moistureholding capacity and increased exposure of soil and crops to wind and rain cause lower yields of crops and loss of topsoil.

THE NEED FOR REHABILITATION

Rangelands are easy to destroy, but difficult to restore. However, because of their size and potential for increased productivity, their improvement can bring many benefits to developing countries. If left undisturbed, rangelands will, in most cases, regrow and reclothe themselves with vegetation, but it may take more than 50 years to return to a stable state. Opportunities should be seized that emphasize vegetation for useful forage and conservation of species.

Improving rangeland is generally a costly process because of low economic returns, relatively low productivity, fragility, site specificity, and often capricious precipitation. For privately owned land, it makes no sense to invest $50 to renovate a hectare of land worth $5. To halt the situation of dams, harbors, and rivers, and to slow the vast loss of topsoil and avoid threat of devastating flash floods, in most cases will require programs that encompass large areas and are supported both by the government and by the local community structure.

To overcome future starvation and malnutrition, either marginal lands must produce more food or there must be a decrease in the human populations relying on such lands. Revegetation is the critical key to the renovation of many rangelands. A cover of vegetation keeps the soil and water in place and decreases evaporation and runoff. Vegetation creates a living barrier to runoff and erosion by increasing percolation and soil storage of rainfall. It shelters the soil from scouring and provides a barrier that slows runoff. Its leaf litter absorbs water and protects the soil surface from rain and wind. Roots and soil organisms help break up the soil, making it porous so that water can infiltrate.

Revegetation is possible both by such natural processes as seed dispersal by wind, birds, and other animals, or carried by seasonal streams and floodwaters; planting by human activity, deliberate or unplanned also promotes new growth. Planned revegetation has two components: (1) direct intervention by seeding or planting trees, shrubs and grasses, using a variety of techniques; and (2) management interventions that serve to strengthen the role of natural revegetation processes.

Strengthening natural regenerative systems, in addition to direct planting methods, can be achieved in several ways. The basic objective is to add extra genetic material adapted to the ecosystem to provide a greater range of material for multipurpose use, and to assist the system to regain its former productivity and resiliency. For example, existing distribution mechanisms can be used, such as planting a mixture of perennial legumes around degraded well-heads, from which animals will spread them naturally (National Research Council, 1981). Nomadic pastoralists supplied with seed will be able to replant fodder shrubs and improve oasis systems with windbreaks or living fences composed of fodder legumes.

Where grazing pressure can be reduced and regenerative mechanisms strengthened, substantial regrowth is likely. The usual way of approaching this objective is to encourage herders to control their grazing. However, herders are usually more sensitive to the management of their grazing resources than are outsiders, and if overgrazing occurs, it is because other conditions make misuse of resources necessary for survival.

Animals can be used to upgrade rangelands if their use is carefully managed. Cattle, sheep, goats, and other livestock prefer somewhat different forages, so they can be grazed together or selectively in sequence. Grazing animals may encourage the dominance of unpalatable species depending on how grazing occurs. This selectivity is especially notable with some trees and shrubs. Improved grazing management may not eliminate these unpalatable species; therefore, other methods for removing them must be found.

Demands of the urban areas also generate pressures on rangelands. For example, many countries encourage the development of crop production in marginal areas to increase the amount of food available to feed a burgeoning urban population. Ways should be found to raise the efficiency of food and fodder production in areas where high productivity is possible, rather than expanding into marginal are as where crop productivity is low but grazing is more appropriate and dependable.

This report addresses issues of socioeconomic context, regional assessment and site evaluation, approaches to management, and criteria for plant selection in intensive rehabilitative efforts. The case studies provide further information regarding these issues, as well as descriptions of projects that have succeeded as a result of broadbased analysis and sensitivity to environmental and social context.