Livestock grazing systems & the environment

Previous section Grazing systems of the arid areas

Grazing systems in the semi-arid and sub-humid tropics⁴

Although these two zones are combined in this discussion because of their similar physical livestock-environment interactions, many of their agro-ecological and socio-economic conditions are different. The semi-arid zones are generally densely populated. Typical areas include parts of the Sahel, the rainfed crop-livestock areas of North Africa, some of the rangelands of Central Asia and the drier areas of the Indian subcontinent. The sub-humid zone savannas have, until recently, been rather sparsely populated because access was impeded by human diseases (such as river blindness, African sleeping sickness) and tick-borne diseases in animals. Typical areas include the savannas of West Africa, eastern Colombia and the Cerrados of Brazil, southern and eastern Africa and eastern areas of the Indian subcontinent. These areas are now becoming the main frontier for agricultural development and worldwide it is where ruminant livestock numbers still grow, as they accommodate, in sub-Saharan Africa an overflow of animals from the drier areas, and, in South America, animals from the Andean Highlands. Livestock interaction involves most components of the environment e.g. land, water and biodiversity.

⁴ Areas with respectively 75-180 and 180-270 days growing season.

State

Land. In the semi-arid and sub-humid tropics, rainfall, and thus feed availability is less variable and therefore livestock are no longer in continuous dis-equilibrium with feed resources. More reliable and more abundant rainfall allow higher population density, more cropping, and more perennial grass and shrub species (which are less resilient) but also implies fewer possibilities for herd flexibility and mobility. As a result, pressure on land resources mounts and land degradation becomes more severe. Although data on a global basis are not available, data from a transect in Mali for example, (Mainguet et al., 1992) showed that land degradation in the 600-800 mm rainfall areas was significantly greater than in the 350-450 mm rainfall area. In the higher rainfall area, the percentage of bare soil increased from 0 to 10 percent over the period 1950 to 1990, whereas in the more arid areas there was no significant change. Irreversible degradation is also reported in the semi-arid regions of North Africa and the Middle East (Sidahmed, 1996). The amount of available fodder from rangeland per hectare declined, indicating long term decline in range productivity. For the former Soviet Union, Gilmanov (1996), estimated that about one-third of the total area of 370 million hectares of rangeland was degraded although, from his figures, no decline in fodder availability per hectare can be detected. Zonn (1995) estimates that for the Kalmyk Republic in Southern Russia 6 million hectares out of a total of 7.6 million is moderately to heavily degraded. Irreversible land degradation in the sub-humid savannas of Africa and South America is still minor.

Water. With a high population density, the development of livestock and human drinking water supplies has been a key aspect of rangeland development in semi-arid zones where surface water is a constraint. Proportionally, degraded lands extend over a larger area in the semi-arid zones than in the arid zones, as a result of higher population densities and greater concentration of water points. As in arid zones, a circular area of degradation of 1-5 km radius around the water point is quite normal, although these “sacrifice areas” are normally quite fertile and are surrounded by bushes at the outer edge of the sacrifice area which balance out the negative effects. Long term studies in Senegal and Sudan (Thomas and Middelton, 1994) did not find a significant expansion of the sacrifice areas of individual water points in these countries.

As the environment becomes more humid, the animal disease threat, especially that imposed by parasites, also increases. In Africa, in particular, control of livestock disease “vectors” such as ticks (carriers of many diseases, such as Anaplasmosis and East Coast Fever) and tsetse flies (carriers of African sleeping sickness in man and trypanosomiasis in animals) is therefore essential for livestock development. Ticks have traditionally been controlled by cattle dips or sprays using organo-chlorines. Several donor-supported projects promoted this technology in the seventies and early eighties. Inappropriate dosing resulted in increased tick resistance, and inappropriate drainage of dip liquid caused water pollution and damage to the fauna. Very little quantitative data are available on the impact of these acaricides on flora and fauna but it can be assumed to be substantial at the local level.

Biodiversity. The semi-arid and sub-humid tropics are one of the world's most important repositories of plant and animal biodiversity. Wildlife is an important income earner for African countries, with a large part of the wildlife population outside the protected areas. For example, in Kenya, 70 percent of the large mammals are outside the game parks, and many are under considerable stress (Western, personal communication). Ottichilo (1996) reported a 30 percent decline in these populations over the last 20 years.

In the sub-humid savannas, weed invasion is a major problem threatening biodiversity, although the role of livestock is only secondary here. For example, the grass Imperata cylindrica in the Philippines and Indonesia has now infested more than 5 million hectares. Invasion with broad-leafed plants and shrubs is more common in the savannas of Africa and the Americas. A study using satellite imagery found that 10 percent of the savannas of the Central African Republic, were infested with Chromoleana (“Herbe de Laos”) (Audru, 1995).

Driving forces

Population growth and poverty are the underlying factors determining land degradation and biodiversity loss in the semi-arid zones. The human population in the semi-arid countries of sub-Saharan Africa⁵ has grown at a rate of 3.2 percent over the last decade, resulting in greatly increased demand for food and fuel. Many countries in the semi-arid zone belong to the poorest in the world. Land degradation is thus clearly a result of cropping in fragile areas, fuelwood collection and overgrazing by the prevailing sedentary livestock production systems. Several attempts have been made to determine the relative share of each of these pressures. The UNEP study (Oldeman et al., 1991) estimated that overgrazing causes 35 percent of land degradation whereas agricultural activities and deforestation (or over-exploitation of forest resources) is responsible for 28 percent and 37 percent respectively. For Africa, the estimate is that overgrazing contributes to 49 percent of the total degraded land area. For the USA, tourism and military bases are also important causes of rangeland degradation.

⁵ Botswana, Burkina Faso, Chad, Ethiopia, Kenya, Mali, Namibia, Niger, Senegal and Zimbabwe.

However, with multiple and interlocking forces affecting rangeland degradation, such precise allocation of the causes of degradation seems an over-simplification. It might even be dangerous because it can easily lead to the wrong type of intervention. Nevertheless, it is important to understand the dominant pressure factor in any region in order to be able to identify the correct mitigating measures. For example, quite clearly, to lessen land degradation caused by fuelwood cutting requires different measures than if overgrazing is the main cause of degradation.

A better approach for defining the relative weight of the different causes of land degradation is to define the sustainable human population for each factor. Steeds and Gore (1985) did this for different West African eco-zones (Table 2.1). This table demonstrates that, for the Sahel, fuelwood is the most limiting factor in meeting the subsistence requirements of its human population. For the higher rainfall areas, livestock production remains below that necessary to satisfy subsistence requirements. For all the zones, however, fulfilling food subsistence needs with a combination of livestock and crop production (which is the normal situation) exerts less pressure on the land than satisfying fuelwood requirements.

In the sub-humid savannas, increased cropping leads directly to growth in the livestock population. For example, Bourne and Wint (1994) found that, for every 10 percent increase in cropping, the livestock numbers grew by about three TLU per km². This is a beneficial strategy, mainly brought about by livestock herders' need for markets and crop farmers' need for organic fertilizer. However, heavy livestock grazing in open park land savannas leads to the disappearance of the most palatable grass species where they will be replaced by other herbaceous plants. This leads to increased encroachment of shrubs and bushes which, in the normal vegetation succession, leads to forest regeneration. The forest would not develop into a closed canopy but would, in this system, where bushfire is a normal phenomena, become an open parkland. This cycle usually takes about 30 years (Audru, 1995). Overgrazing in savannas thus leads to initial bush encroachment, with increased biomass production (Caesar, 1992), but, because it can no longer be grazed, the diversity of plant and animal (including game animal) species declines. In some cases the bush encroachment leads to the use of economically and environmentally expensive herbicides.

There are a large number of cases which show increases in plant biodiversity in well-balanced grazing systems, especially those using multi-species. An extensive review of grazing and production data of 236 sites worldwide, including many sites in the semi-arid zone, showed no difference in biomass production, species composition and root development in response to long term grazing in the field (Milchunas and Lauenroth, 1993).

The interaction between wildlife and livestock in these eco-systems is complex. First, there is increasing evidence of a possible complementarity, and only limited competition, of wildlife and livestock in grazing. As already shown by Schwartz and Ellis (1981) the grazing “overlap” between most wildlife species and livestock is limited. Mwangi and Zulberti (1985) and Western (1991) showed that the combination of livestock raising and wildlife management resulted in an equal or better species wealth than either of these activities individually. Furthermore, in national parks in Kenya, such as Amboseli, where livestock are not permitted, biodiversity is decreasing, and there is a corresponding increase in unpalatable species and bush-encroachment (Ottichilo, 1996). On the other hand, as the same author points out, there are many degraded areas in Kenya due to combined wildlife-livestock pressure. The picture is thus not uniform but potential for benign and even synergistic co-existence certainly exists and needs to be encouraged. As previously mentioned, 70 percent of big game animals are outside national parks and, because population pressure is likely to preclude a large increase in the area or number of such parks, the preferred approach must be to maintain a better wildlife habitat outside the park system.

The driving forces leading to losses in animal biodiversity are habitat destruction, species introduction and hunting (WRI, 1994). In an analysis of the threats to mammals in Australia and the Americas, UNEP (1993) argues that pastoral development affects about 30 percent of the species, compared to logging, which affects about 25 percent, and cultivation which affects 55 percent of mammalian species on those continents. Habitat destruction is playing an important role in the developing world, especially in the sub-humid savannas. Road construction, and immigration from the drier areas, leads, in Africa, to the destruction of the habitat of the carriers of African sleeping sickness. In turn, this removes the protection of wildlife which is tolerant to the disease. International agencies, including GTZ and the World Bank, have financed extensive tsetse clearance campaigns in West Africa. Traditionally, these campaigns used a combination of hand sprayers and aerial spraying of insecticides, initially with organo-chlorines, to eradicate the tsetse fly. Nagel (1993) argues that pesticides from this period are still found in some African birds. Since the mid eighties, compounds with shorter residual effect, such as synthetic pyrethroids have been used. These second generation compounds caused substantial initial damage to flora and fauna but permanent effects were not observed with single spraying. For example, in the World Bank funded tsetse clearance project on the Adamaoua Plateau of Cameroon (Muller, 1985), where permanent damage and high residue levels were found, they appeared to be the result of repeated spraying in the border areas. Bush encroachment, as a result of inappropriate grazing management was reported to be the most serious environmental damage. Land use plans have often been advocated as an essential pre-condition for tsetse clearance, and international donors have made the preparation of such plans conditional to the financing of the eradication campaigns. However, experience with the enforcement of such land use plans has been dismal because local authorities lacked the authority and means for their enforcement.

Traditional land tenure practices have tended to be less robust in these more humid areas than in the drier areas. This lack of strong traditional tenure practices has been particularly significant in sub-Saharan Africa where after tsetse clearance operations, uncontrolled settlement patterns developed.

In addition, hunting and culling of wildlife was encouraged in the past, because wildlife in general was considered to be a reservoir of diseases, such as Rinderpest and Malignant Catarrhal Fever and vectors of disease, such as East Coast Fever and Trypanosomiasis (Grootenhuis et al., 1991). However, the control of the above mentioned diseases has improved considerably and there is a much better understanding of which species harbour which specific diseases.

Policy pressures. Inappropriate policies, whether institutional or those governing incentives, often exacerbate the negative effects of livestock on land and biodiversity and undermine some of the potentially positive effects.

The effect of price policies has often been to discourage livestock off-take and thereby increase the pressure on grazing:

• The urban bias. Many African governments maintained overvalued exchange rates and allowed the importation of highly subsidized milk and meat from the EU in order to provide cheap food to urban populations. For the semi-arid and sub-humid grazing systems, especially in West Africa, this dumping of meat and milk reduced off-take and increased grazing pressure. The recent devaluation of the CFA franc almost doubled off-take from the Sahelian herds, thus reducing grazing pressure, and almost eliminated the imports of dumped meat from the EU (Vergriette and Rolland, 1994, and Box 2.7);
• Absence of banking institutions. Similarly, with high inflation rates yielding a low return on deposits and, in many cases, the absence of any form of reliable banking system, livestock have become the main form of investment in many developing countries. This is the case not only for the pastoralists themselves but also for urban investors, who rely on hired labour for herding their animals. Although difficult to quantify, some estimates suggest that as much as 50 percent of the livestock in the Sahelian countries are now owned by outsiders. These outside investors have different objectives, less environmental concern and less respect for traditional tenure rights. A similar phenomena can be observed in South America (Sapelli, 1993);
• Under-investment and market failure in plant and animal biodiversity. Even where there is no deliberate attempt to distort markets, valuable natural habitats are converted to agricultural production, including livestock, because other costs, such as loss of biodiversity, have not been accounted for. Such adverse conversion would be less likely if an appropriate value, including future potential value, was given to the natural vegetation, including medicinal plants and animal wildlife.

The value, costs and returns from wildlife, compared to livestock and agricultural production, is highly variable. At the level of the national economy, the opportunity cost of wildlife biodiversity conservation in protected areas in terms of foregone livestock and agricultural production seems to outweigh the income from tourism and forestry generated by these protected areas. For example, in Kenya, Northon-Griffiths and Southey (1995) estimated the foregone revenue from livestock and agriculture from the parks at US$ 203 million, whilst the revenue from these parks amounted to only US$ 42 million. On the other hand, Engelbrecht and van der Walt (1993) estimated that the Kruger Park in South Africa contributed more than US$ 110 million per year in tourism, compared to a foregone production of only US$ 6 million. A key issue in these estimates is the allocation of benefits. For example, in the Kenyan study, only half of the national revenue from tourism was allocated to the game parks, assuming that the other half applied to other tourist attractions. At household level, the comparative profitability of wildlife and livestock raising varies greatly, according to the ecological conditions and wildlife use (meat, trophy hunting, tourism). Overall, under present conditions of niche markets for game-meat or tourism, wildlife ranching seems financially more attractive (Box 2.8). For the communal areas, however, wildlife cannot provide the multiple functions of producing milk for subsistence and providing traction, fertilizer and investment that livestock can. Without any doubt, the combination of wildlife and livestock is the most appropriate under those conditions.

Inappropriate institutions constrain sustainable wildlife-livestock integration on communal areas. Traditionally there has been a rigid, centralized and regulatory attitude of public institutions in the protection of wild animals. This was especially the case in East Africa, where wildlife management was typically organized by central administrations in a rather military fashion. There was no benefit sharing with the local population, even though wildlife brings high financial cost to the local population because of crop damage, livestock loss to predators and diseases transmitted from wildlife. This has lead to the antagonistic reactions of many herding and farming communities. In addition, the prohibition in many countries of sport hunting and consumption denies important potential benefits from wildlife (Ottichilo, 1996).

Response: Technology and policy options

The prevailing combination of poverty and high population growth, which characterizes the countries of the semi-arid regions of the world, cannot easily be broken. This will mean continued pressure on the remaining trees and shrubs for fuel, on higher potential land for crop production, and therefore more pressure on the grazing resources. Nevertheless, livestock populations can be expected to grow, especially in the sub-humid areas. Physically, these areas are characterized by much more stable climatic conditions, and traditional grazing management concepts of stocking rate control and individual land tenure become more valid as the precipitation increases.

The introduction of market pricing for inputs and outputs will promote more sustainable livestock production. On the institutional side, again, most of the arid land recommendations concerning decentralization and local empowerment apply also here.

Community-based wildlife management is now generally accepted as an important component of sustainable wildlife management in communal areas. Key issues involve:

• the strength and cohesion of the traditional social fabric, as shown by the difference between the successful community management around the Maasai Mara park and the difficulties experienced by a similar initiative in Amboseli where traditional authority is much weaker (Kiss, 1990); and
• achieving an equitable distribution among the local population of the proceeds of wildlife revenue.

For the African savannas, little progress has been made in developing more appropriate land tenure systems, and land use planning is still as ineffective as in the past. Furthermore, many of the areas cleared in the 1980s have been re-invaded by the tsetse fly. Most donors have stopped funding tsetse eradication for this reason and, where tsetse eradication still continues, such as in the EU funded programme in southern Africa, this is done with environmentally more benign methods and a stronger grassroots participation. The impact of those programmes still has to be assessed. In any case, the threat of the tsetse fly is diminishing, as population pressure destroys the habitat of the fly, with or without eradication campaigns. Finally, especially for these areas, there is a great need for alternative investment possibilities, such as savings banks, to divert savings away from investment in livestock.

Before the introduction of technologies intended to rehabilitate degraded land, it is essential that institutional arrangements are in place. For the semi-arid rangelands, once the institutional framework is in place, it should be possible to:

• develop water supplies. Before any development is undertaken, very careful consideration should be given to the possible impact on the distribution of wet and dry season grazing, to settlement patterns and on the efficiency of feed utilization. There is convincing evidence (King, 1983) that increased watering frequency (say from a two-day to a one-day regime) increases the metabolic rate and hence decreases the efficiency of the utilization of scarce feed resources;
• regenerate the vegetation, through resting (deferred grazing, a traditional practice in many Middle Eastern countries), over seeding (normally done with local, adapted species) or planting of adapted fodder and fuelwood shrubs. The latter has been widely practiced in North Africa although with limited success (Msellati, 1995);
• breed for hardiness in indigenous breeds, with selection schemes based on producers' herds;
• introduce multi-species, to encourage a better utilization of the vegetation. There is growing evidence that the combination of different species, including wildlife and livestock has a positive effect on plant and animal biodiversity;
• introduce overall “management” approaches, such as Holistic Resource Management, which establishes clear goals, including environmental ones, and then selects the most appropriate tools for particular site conditions. This approach is strongly based on the principle, that it is not the total grazing pressure, but the continued grazing on the same area, which causes degradation. Timing of the grazing is therefore of critical importance. Initial results with this approach in Chad and Namibia are promising (Pratt et al., in press); and
• improve infrastructure, especially roads and markets, to allow a smooth movement of goods and services and rapid herd off-take, especially in times of drought.

For the sub-humid tropical savannas, the emphasis has to be on the development of multiple uses, combining sustainable farming systems with integrated wildlife-livestock management and the establishment of a few protected areas for key plant and animal species. However, with increasing population pressure, this becomes more difficult. Furthermore, many government agencies in the developing world are too weak and, in the developed world, too dependent on political fashion, to rely exclusively on protected areas for biodiversity conservation, especially for big game of the sub-humid savannas. Eco-tourism and commercial (trophy) hunting would need to be integrated with commercial livestock production and mixed farming.

There is a wide array of technologies available:

• for the management of the natural savannas, regulated burning, i.e., deliberately using fires at different times of the year to affect different classes of plants, can be one of the most efficient ways of maintaining a good tree/shrub/grass mixture. However, regulating savanna burning has not worked in the past because of weak enforcement agencies, and this has led to a general distrust, and often a blanket prohibition, of burning by departments responsible for rangeland and savanna management. This technology may not work if governance is weak unless there is strong participation by the local population;
• for vector control, new and environmentally more benign methods are now available. For tick control, vaccines are available for the diseases spread by a number of sub-species, although not yet for East Coast Fever. Long duration tick sprays, which can be applied on individual animals (pour-on) now exist, and these have a negligible environmental effect. A number of environmentally benign control methods also exist for tsetse control (World Bank, 1995); and
• CIAT has developed a means of improving savanna in South America with improved grass (Andropogon) and legume species in rotation with rice (CIAT, 1994) , and ILRI (1994) has proposed improved fodder production using leguminous plants.

Research needs. A number of research needs emerge. More research is needed on the institutional framework, especially regarding access to land and land improvement since these semi-arid and, in particular, sub-humid eco-systems offer more potential for increased productivity. Better adapted fodder species and techniques, water harvesting and livestock breed improvement are some of the technologies to be explored. A campaign to encourage the use of multiple species of both livestock and wildlife, and on combinations which improve species wealth, would be required to capitalize on the environmental potential of these zones.

Next section Grazing systems and tropical rainforests