Contents - Previous - Next


Land. The transformation from tropical forest to crop and pasture land brings about substantial losses of soil fertility and soil erosion. Furthermore, in many tropical rainforest areas, pastures can only be sustained for a short period of up to ten years. Soil nutrients are rapidly depleted after clearing and grasses are soon replaced by less useful vegetation. Natural regeneration of forests is quite difficult, especially in large degraded areas. More than 50 percent of the pasture areas in Amazonia have now been abandoned in a degraded state. For the total area of tropical forest, WRI (1994) estimates that 427 million hectares are degraded, most of it as abandoned pasture or fallow after shifting cultivation.

Plant and animal biodiversity. For the rainforests, data on biodiversity losses are dramatic. Since 1950, about 200 million hectares of tropical forest have been lost, with the result that of some unique plant and animal species, in one of the world's richest sources of biodiversity have become extinct. Forest areas of Central America have declined from 29 to 19 million hectares, since 1950, although, since 1990, the rate of deforestation in this region has fallen. In Central America in the 1980s rainforests disappeared at the rate of 430 thousand hectares per year but this declined to 320 thousand hectares over the period 1990 to 1994. In South America, the deforestation rate in the 1980s was about 750 thousand hectares per year. It is not known, whether this rate has declined over the last years. Table 2.2 gives quantitative information on the rate of degradation, and the current area of remaining tropical rainforest.

Box 2.9 Competition and symbiosis between livestock, forests and people in Africa.

IN THE eastern area of the Central African Republic, the incursion of Fulani pastoralists with their cattle in the tree parklands of the Zande people leads to:

• increased openings in the forest cover
• decrease in animal biodiversity
• degradation of the hunting potential of the Zande, and thus the quality of their diet
• loss of the traditional values of the Zande people.

On the other hand, farmers in the Kissidougou area of Guinea developed a symbiotic rainforest-livestock system. Livestock here are used to clear up undergrowth in the forest, which could then be used for fuelwood and shade for the perennial crops. The population density increased over the last two decades from 10 to 60 persons/km2, but here a positive relationship emerged between population density and forest cover.

Source: Carrera and Toutain, 1996.

Much of the deforested areas in Latin America went into ranching, sometimes after initially being cropped. In Central America, pasture areas have increased from 3.5 million to 9.5 million hectares, and cattle populations have more than doubled, from 4.2 million head in 1950 to 9.6 million in 1992 (Kaimonitz, 1995). In Brazil, about 70 percent of the deforested areas are converted into ranching. In Asia and sub-Saharan Africa, the decline in the forest area is mainly the result of crop expansion and not livestock related.

Table 2.2: Annual deforestation 1980-1990 (million ha and percent) different regions of the world.


Annual change in area (m.ha)

Annual rate of change (%)

Remaining forest area (1990) (m.ha)

Latin America












Source: FAO, 1996.

Driving forces

Land use. Livestock ranching has often been regarded as the driving force behind the deforestation of tropical rainforests. However, there is increasing evidence that livestock ranching in deforested areas is merely the most obvious symptom of a much more complex degradation process with a variety of driving forces. They are detailed below.

Road construction leads to accelerated deforestation, and is possibly the single most important direct cause of deforestation. For example, Ledec (1992) found that in Panama, for each kilometre of an all-weather road, between 400 and 1,000 hectares of forest were cleared. Roadside ranchland was sold in Nicaragua at about three times the price of comparable land one day's walk away (Maldidier, 1993). In Rondonia, Brazil, the paving of an existing road was the main factor in clearing 100,000 km2 of rain forest, and focused, in the early eighties, the world's attention on this relationship (Myers, 1981). Livestock ranching, with little supervision requirement and few bulky inputs became an attractive activity along these new roads.

Arable farming, in the rainforest areas, usually takes the form of "slash and burn" agriculture, especially in the forest areas of sub-Saharan Africa and Asia. Several studies (Bruenig, 1991) describe this as the most important mechanism of deforestation in these areas. Overall the area cropped increased over the last two decades. When population pressure is low (less than 30 people per km2), and the area deforested is relatively small, secondary forest climax vegetation can return, thus limiting longer term soil erosion. If population pressure increases further, a downward spiral of declining soil fertility and crop yields emerges. In Central and South America, the total area under crops has remained stable, but slash and burn cultivation continues at the primary forest frontier, and is compensated by desertion of earlier depleted cropping areas, or their conversion into pasture. This conversion of crop land to pasture is common practice, as many ranchers use a crop of corn to generate income for future ranching.

In Indonesia, transmigration projects (moving people from the over populated areas of Java to the lesser populated outer islands) have often been cited as the principal cause of deforestation. Again, cropping is the first land use and livestock are introduced quite late in the farming system and mainly in environmentally friendly stall feeding systems. Animal agriculture is therefore not one of the leading causes of deforestation. In sub-Saharan Africa, crop production, and especially the expansion of permanent plantation crops, such as oilpalm and rubber, have been major causes of deforestation. Very little tropical rainforest has been converted into ranches on this continent. Forest overexploitation is also an important factor in deforestation, especially in Asia and Africa, where about 20 percent of the areas are over-exploited. Such over-logged areas are then easily converted into shifting cultivation areas. Logging is not important as a cause of over-exploitation in South America (Sharma et al., 1993).

Relative importance of different pressure factors in deforestation. As in the case of desertification, there have been several quantitative estimates on the individual importance of these factors. However, unlike in the case of desertification, the direct cause is easier to ascertain, making these estimates more relevant. Nevertheless, there are some overlaps, especially as logging is often a precursor for cropping, which in turn often precedes ranching. This means that even in tropical rainforests the cause of environmental damage is hard to apportion. For the record, Table 2.3 provides some estimates. Overall, "slash and bum" agriculture is said to cause 60 percent of the deforestation (Bruening, 1991).

Policy pressures. Probably more than for any other livestock- environment interaction, inappropriate incentives, land tenure and institutional policies have played a major role in deforestation. land degradation and erosion of biodiversity in the humid tropical areas.

Figure 2.1: Population density and forest coverage in the humid zone.

Source: Unasylva, 1993.

Table 2.3: Some estimates of the main causes of deforestation (percent of total deforestation).




Forest exploitation

South America


44 (70 in Brazil)




Negligible (Philippines & Indonesia to some extent)






Source: Bruenig, 1991.

Inappropriate incentives from governments and international agencies in support of livestock development, played an important role in the destruction of the Latin American rainforests. Especially in the seventies, the livestock sector secured a disproportionate share6 of credit at subsidized rates and with lenient re-imbursement conditions and control. Most of these loans were invested in land or deforestation, and in turn raised land prices even further. However, it appears that the role of subsidized credit is limited. For example, Ledec (1992) showed that only 7 to 10 percent of the deforestation in Panama could be attributed to subsidized credit. A similar picture emerges from Brazil. The large ranches greatly benefited from investment subsidies (up to 75 percent of the investment costs), tax holidays, and subsidized interest rates. However, the large ranches caused only a small part of the deforestation (up to 30 percent in some states, mostly less) and most ranching in Amazonia is in the hands of medium to small size ranchers, who have had less easy access to these subsidies (Hecht et al., 1992).

6 Between 25 and 45 percent of the total agricultural loans, more than double the share of livestock in the sector. This benefited only the bigger ranchers (Kaimonitz, 1995).

Beef exports to the USA, frequently quoted as one of the main causes for ranch development in Latin America (the "hamburger connection" was coined on that premise), were important in Central America in the 1960s and early 1970s, when international beef prices were high (Myers, 1981). But the Amazon area never produced more than 5 percent of the total beef supply of Brazil, and that production was not exported. Furthermore, declining global prices in the eighties, and increased protection in the USA market, especially following NAFTA, have reduced exports from the other Latin American countries. However the fall in prices seems to have had only a limited effect on deforestation. It may have pushed traditional livestock ranches into cropland, but it has not halted expansion in the frontier zones. Ranchers have had very few alternative opportunities.

Regarding land tenure, the key phenomenon was the land grab in rainforests, which occurred as a result of a general increase in land prices in Central and South America over the last decades (Kaimonitz, 1995). "Money laundering, labour remittances and a hedge against inflation" kept prices above the productive value (Shearer et al., 1993). The almost permanent price increase made land a safe investment and led to increased speculation. Finally, recent government settlement programmes and the establishment of internationally funded protected reservation areas, raised prices further. A large part of the expansion of pasture land may therefore have had more to do with land speculation than with cattle raising per se. Deforestation for ranching thus became more a titling strategy than one based on an economic activity (Jones, 1990). Indigenous users and tenure rights, although existing, played practically no role in the protection of the tropical savannas and forests.

Institutional requirements concerning land titling procedures reinforced these trends. Several Latin American countries actually prescribed occupation, and thus deforestation, as a condition for giving out ownership titles. As clearing land for ranching was much easier than establishing proper cropland, ranching became the preferred occupation to fulfill the title requirements. Furthermore, land clearing discouraged squatters. For example, Edelman (1992) reported that practically all 13 squatter invasions that occurred in one particular area of Costa Rica, took place in heavily forested areas, and not in ranching country. Ranchland was psychologically better protected. Finally, some governments even encouraged deforestation for livestock production by colonization schemes.

Response: Technology and policy options

Many causes for ranching encroachment in rainforest areas have disappeared. Subsidized credit for ranching has been phased out almost everywhere and the overall volume of credit to the sector has declined. In addition, some of the more exogenous factors turned against ranch expansion. Meat exports from Central and South America have declined dramatically and world market prices for beef have dropped over the last decade. All these measures reduced investments by large and urban investors in ranching, and brought down the overall rate of deforestation. However, these changed incentives had little effect on the small and medium producer, and deforestation for crops and small-holder ranching continues. A new look at the policy instruments for these farmers is thus required, as additional action is necessary to halt the still on-going conversion of rainforest into ranches. The following measures have been suggested (Kaimonitz, 1995):

Discourage road construction. Road construction is the single most important factor in deforestation. Discouraging road construction would probably be the most powerful deterrent for ranch and farm establishment and even though it may be politically difficult, this should seriously be considered for highly valuable forest areas;

Establish protected areas, and expand the rights and protection of indigenous land use rights. While effective protection of large areas may be difficult to implement under the prevailing situation of understaffed, underpaid and poorly motivated government staff, the most valuable areas can be set aside. Testing of revenue-sharing mechanisms of the proceeds of rainforests, involving all stakeholders (and not only the indigenous occupants, but also the ranchers) could be tested, to improve the incentives for forest conservation;

Introduce taxation for pasture and crop land to discourage conversion of forest to crop and pasture land. This is an attractive concept although, with poor or even non-existing property registry service in many of the rainforest areas, it will be very difficult to enforce;

Promote more research and extension. Kaimonitz (1995) especially stresses these needs, as he sees slow technological change as one of the key factors favouring extensive livestock production. Indeed, the basic ranching system has changed very little since the early 1960s, and no agricultural alternatives to ranching have been developed. However, intensification of livestock production in one area can certainly reduce the pressure in another area, if well connected (see Box 2.10), although the validity of this principle cannot be universally assumed.

Grazing systems in temperate zones

Driving forces
Response: Technology and policy options

Grazing systems in the temperate zones cover about 440 million hectares, or 13 percent of the world's pasture lands. They contain about 2 percent of the world's cattle and 10 percent of the small ruminants. They are mainly based on natural permanent pastures, and occur mostly in China, northern USA and South America and the CIS. They contribute about 5 percent of the global beef supply, 12 percent of mutton supply and 3 percent of the global milk supply.

Box 2.10 Does intensification reduce grazing pressure?

MANY DEVELOPMENT projects have been based on the assumption that intensification reduces 9 razing pressure. However, there is no empirical evidence that this is the case. Social forestry projects in indict, for example, promoted stall feeding with improved crossbred cows, but there is no evidence that this decreased the number of traditionally managed animals in the forest. Similarly, stratification of production in, for example, North Africa and the Middle East, has not shown a clear reduction in grazing pressure in the young lamb production areas. It appears that intensified production alone does not reduce grazing pressure, and institutional and incentive changes (such as access to common grazing resources) need also to be changed.


Most grazing areas in the temperate zones are in relatively good vegetative condition. They were originally formed by the grazing of wild animals and, because livestock grazing is not fundamentally different, there is no convincing evidence that the change from wild to domestic grazing has, in general, had any long term detrimental effect. Long term stocking trends both in the USA and Australia show an initial high stocking rate in the late 19th century and early 20th century, falling off after a period of about ten years, to a lower level which is then sustained over the following decades. This indicates a fairly stable eco-system. Indeed, independent observers argue that rangelands in the western United States are now in a better ecological condition than at any other time in this century (General Accounting Office, 1988). For example, in 1987 there was half as much "poor" rangeland but double the area of "good" rangeland as in 1964 (USDA, 1988). In this equilibrium environment, there is ample experimental evidence (Blackburn et al., 1982) that there is no difference in erosion and water infiltration between light, moderately grazed and un-grazed areas. In addition, there are some excellent examples of how grazing with domestic stock can improve biodiversity (Box 2.11) and how grazing can reduce fires and promote seedling establishment by reducing biomass accumulation. Box 2.12 demonstrates how livestock can be used to reduce the amount of petroleum based herbicides.

However, there are regions where the vegetative condition of temperate grazing land is less satisfactory, notably in the CIS and Asia, western Europe and the western United States. Grasslands in the CIS are often heavily overgrazed. This is caused by the significant preference given under the previous command economy to industrial, mechanized production, with a strong dependence on feed grains. Meadows and grazing animals did not fit well in such systems and therefore received scant attention. Although incentives have changed, there is still a dearth of appropriate technology to improve pasture lands in the CIS and overgrazing, resulting in increased erosion and reduced water infiltration, is still a common phenomenon. Overgrazing in the temperate, high altitude pastures of Central Asia is particularly severe after the transition to a market economy (Schillhorn van Veen, 1996). Zhong (1993) estimated that, in Mongolia, 40 percent of the country's 65 million hectares is degraded, especially on state farms.

Box 2.11 Riparian health, sheep and elk

RlPARIAN HEALTH ifs an important issue driving the monitoring and use of public grazing lands. However, it is often overlooked that any species of wildlife or livestock can overgraze these critical areas. A key example of such a situation exists today in Yellowstone National Park (YNP), the crown jewel of the US park system. It has recently been demonstrated that elk are severely overgrazing riparian areas in YNP. In a study comparing riparian areas in Yellowstone National Park and on the summer range of the US Sheep Experiment Station (approximately 30 miles from YNP) it was shown that grazing of sheep had a more benefical impact on riparian health, as measured by willow populations, a key indicator species. Furthermore, as a result of healthier willow communities on the Sheep Station, beaver populations are also in better condition. This work demonstrates that any grazing animal can cause environmental instability and/or degradation and that by using an appropriate livestock species environmental health can be maintained or increased (Kay and Walker, in press).

Average height of willow population inside and outside Yellowstone National Park

Box 2.12 The control of the leafy spurge.

LEAFY SPURGE is an aggressive perennial weed that typifies the impact of noxious weeds. It has now infested about 600,000 hectares in the western USA, greatly reducing the habitat for cattle, bison, and deer. In North Dakota alone it has cost the state US $75,000,000. However, grazing by small ruminants, especially goats, can greatly reduce this pest and improve plant and animal biodiversity, and this is becoming increasingly popular as the sole acceptable control method.

Sources: Walker et al., 1994 and Bureau of Land Management, 1994.

Many grazing areas of western Europe and the eastern USA are over fertilized. The highly subsidized and guaranteed milk price under the Common Agricultural Policy (CAP) of the EU in the seventies and eighties encouraged strong intensification, and this led to very high nitrogen fertilizer use on pastures. For example, in the eighties, the average nitrogen application on grassland in the Netherlands was about 500 Kg N per hectare per year (two-thirds inorganic and one-third organic), whereas the average N outflow in milk and meat did not amount to more than 82 kg N per hectare per year (Steenvoorden, 1989) As a result, these meadow soils are now overloaded with nutrients, especially nitrogen and phosphorus, contaminating ground and surface water.

In the western USA, there is currently strong concern about the impact of grazing along the banks of rivers and streams: the riparian areas. Such stream sides, which normally comprise less than 5 percent of the total area, receive often 20-30 percent of the use and therefore water quality (increased nitrates and phosphates), plant biodiversity, and trout yields may be impaired (Sheehy et al., 1996). It is estimated that more than 50 percent of riparian areas are seriously damaged (Armour et al., 1994). On the other hand, there is ample evidence that nutrient leakage into streams is minimal where streamside pastures remain in good condition (Sheehy, et al., 1996).

Driving forces

Policies continue to drive livestock production into degrading temperate grazing systems:

• In Central Asia, changes in fuel prices and privatization of land are currently the driving forces causing overgrazing. Previously, annual movements of animals were greatly facilitated by low energy prices but market pricing of transport has reduced animal movements and is changing the traditional cycle of seasonal grazing and rest periods into year-round grazing;

• In western Europe and the USA, the most significant driving forces to have an impact on the environment are those policies which foster intensification of livestock production. The Common Agricultural Policy (CAP) of the EU (then European Economic Community), which heavily subsidized dairy production (US$ 250 per dairy cow per year [Knudsen and Nash, 1990]) encouraged intensive production at high stocking levels and therefore excessive use of fertilizer and concentrate;

• The phasing out of feed subsidies in Central Europe is still too recent to have had a measurable effect;

• Grazing on public lands in the western USA is being driven by a larger concern over how public lands should be utilized (ecotourism vs agriculture, mining and/or timber). Currently riparian areas are the focus of the debate. In the past, the importance of riparian areas was under estimated and they were largely viewed as sacrifice areas, while upland grazing (>95% of the area) was the focus of range improvements. Also contributing to the stress on riparian areas has been the inappropriate conversion of sheep permits to cattle permits. One particularly important area of this debate has been the cost charged for grazing public lands. Some have argued that the grazing land is under-valued and that fees are therefore too low. However, in a recent CAST (1996) report it was demonstrated that grazing on public land is more expensive than with private leases. CAST reports on an animal unit month basis, a cost of $19.59 for public grazing and $17.18 for private leases. Use of public grazing lands are now the focus of conflicts between opposing national environmental and livestock interest groups, with active involvement of the federal Government because both groups claim that some form of public administration is required. There are many examples of effective wildlife preservation and effective livestock-wildlife integration by these respective groups (Sheehy et al., 1996). Even though there are conflicts on land use the public administration of these lands has been largely successful, as demonstrated by the continuous trend in rangeland improvement.

Response: Technology and policy options

The first priority in reducing the environmental impact of the over-fertilized grazing systems in the EU and USA, and enhancing their environmental condition, is to phase out the current system of milk and meat subsidies. The current shift in EU policies to replace the subsidies on these products by subsidies on farmers' income, and tie this to landscape maintenance, is a very positive development. In addition, much can be done to achieve a better nutrient management. Work in the Netherlands (Box 2.13) shows clearly the possibility of reducing input levels by introducing a ley farming system with more farm grown fodders, careful application techniques and lower nitrogen inputs, without substantial reduction of the output. Experience from Europe and the USA shows that farmer education is a key factor in arriving at better nutrient management. The introduction of a mineral balance at farm level has helped farmers to recognize the excess nitrates and ammonia which are emitted (Aarts, 1996). Experience with farmer study groups in the Netherlands, clearly shows that it is possible to decrease the nutrient loading by 25 to 50 percent with an improved financial result on many farms (Aarts, ibid). Similar results are emerging from the USA, where, under the motto "it is better that we do it now voluntarily, than have it imposed by the Government", farmers, for example, in the Chesapeake Bay area, are drastically reducing their fertilizer input levels (O'Connell, personal communication).

A greater reliance on market mechanisms, combined with a less restricted use of wildlife, are essential elements in the management of riparian areas in the western USA. Once such land is available to the highest bidder (be it private conservation organizations or livestock ranchers) and exploitation of wildlife is attractive, more land would go into wildlife or combined livestock-wildlife production, and grazing pressure would be more evenly spread. Some regulation of the riparian areas might be necessary and community-based landscape objectives and values (rather than interference by federal agencies and national environmental pressure groups) would probably be more effective.

Research needs. For the Central Asian and Russian grasslands, there is a need to increase research on pasture management. The changed incentive framework fosters the use of farm-grown grass and fodders, but there is a large gap in skills and tradition in applied grassland research. More research is also needed on better land use policies for the temperate summer pastures of the Central Asian highlands.

For the highly intensive grazing systems, further research in nutrient management is clearly essential. Experience in the Netherlands clearly shows the potential for such research. It should be multidisciplinary and involve the different animal science disciplines (breeding, nutrition and pasture agronomy), environmentalists and socio-economists. It should look critically at input-output management and the relationship between cause and effect.

Box 2.13 Reducing inputs, without losing production; The "Marke" experience.

THE EXPERIMENTAL farm "the Marke", located on sandy soils in the eastern part of the Netherlands, aims at testing how, through very careful nutrient management, the emissions in nitrates, ammonia and phosphates can be brought within the future Dutch policy requirements, without reducing output. Below are the results:



Milk production (kg/ha)



Purch. Concentrate (kg/ha)



Area grassland (%)



Area fodder crops (%)



Fertilizer application (kg N/ha)



id (kg P/ha)



N balance (accum. in soil and losses to air and (ground) water







Average nitrogen content in upper ground water



This work clearly shows that it is possible, with better information, to reduce inputs to about one-third, with only a 10 percent decrease in output. Key technological inputs are: changes in the farming systems to allow for crops with can use manure more efficiently, much lower N percentage in the feed, and a higher production per cow (with low fat percentage). This case highlights the need, and current move, towards more knowledge intensive agriculture, whereby knowledge replaces physical inputs.

Source: Aarts, 1996

Regardless of the environmental setting, there is a clear need for better monitoring tools which can effectively document the current status of grazing lands and project possible changes in these lands as weather or usage patterns change.

Contents - Previous - Next