Grassland condition is generally interpreted from the viewpoint of the status of the vegetative cover or its level of degradation, and is often defined with reference to the land as a capital asset, i.e. its capacity to sustain livestock. It is a long-term phenomenon, but subject to short-term oscillations and complicated by the fact that negative changes may or may not be reversible. A critical and realistic interpretation of the condition and trends of grazing land is fundamental to the identification of relevant development options. Indicators of condition are primarily found in soil and vegetation change (Livingstone, 1991) and, at a secondary level, through alterations in livestock such as numbers, species mixes and productivity per animal. Negative soil changes include decreased fertility, soil loss through gully and sheet erosion, reduced soil water-holding capacity and decreased infiltration. Net soil loss clearly occurs (Box 1) in some areas, although the resilience of the vegetation may allow substantial soil loss before any effect on plant growth is detected.
Box 1. Soil loss
Soil loss is the aggregate of nutrient extraction "mining," erosion by wind and water, and other factors. In many areas of the world, soil loss rates are greater than regeneration rates (Friend, 1992). Irreversible soil productivity losses from water erosion appear to be serious on a national scale in Algeria, Ethiopia, Ghana, Kenya, Lesotho, Morocco, Swaziland, Tunisia, Uganda and Zimbabwe. Many other African countries have serious erosion, but the effect on productivity is not clear (Dregne, 1990). Soil loss from the Loess Plateau in northern China is of major national concern because of on-site resource and productivity loss, and off-site flood risks.
The long-term depletion of soil nutrients by pastoral activities has been proposed as a factor contributing to degradation in Australia and New Zealand (Box 2) (Harris and O'Connor, in prep; Helden, 1991). Nutrient loss from grazed areas is associated with burning, soil erosion, volatilization and leaching at excreta sites, the transfer of nutrients to stock camps, and, to a lesser extent, the export of livestock products. The expression of nutrient depletion depends on grazing history and the inherent fertility levels of soils and base materials. Nutrient depletion may not be universal but it can be an insidious factor of degradation; in herder-based pastoral systems, the use of dung as fuel and the penning of livestock at night may contribute to nutrient depletion from grazed areas.
Box 2. Examples of nutrient depletion on grasslands
1. Preliminary investigations of the effect of 150 years of pastoralism on the nutrient status of soils in mountain lands of New Zealand indicate that depletion due to fire and livestock has occurred and is likely to have contributed to the degradation of some grasslands (Harris and O'Connor, in prep.). The figure below shows estimated P balance, relative to an 1850 benchmark, for an extensively grazed grassland in New Zealand.
2. In Australia, native pastures grazed by sheep for 80 to 100 years showed a decline in P compared to non-grazed areas and developing K deficiencies were associated with nutrient translocation by sheep (Hilder 1966)
Vegetation changes are indicated by changes in plant cover, biomass and biodiversity, and in the proportional occurrence of unpalatable plant species and weeds, along with shifts between vegetation states. Indicator species - often described as "increasers" or "decreasers" - are used to identify trends in grassland condition. Bush encroachment is often, but not exclusively, stimulated by overgrazing.
Livestock population changes are only indirect indicators of degradation and do not provide conclusive evidence. This is especially true in view of the resilience of some grasslands (Livingstone, 1991). When interpreting trends in condition, therefore, it is important to identify the contributing factors critically and not assume that livestock have been the cause. The blaming of livestock for desertification in Africa has been challenged on this basis. Tree cutting and inappropriate cropping are now thought to have greater effects (Simpson and Evangelou, 1984).
The extent of yield depression (attributed to overgrazing and other related factors) varies considerably according to the environment and characteristics of the vegetation. In Xinjiang (northwestern China) the grazing land was considered in "balance" with livestock populations until the mid-1970s. During the 1980s, productivity is reported to have decreased by about 30% as livestock populations became excessive, with the decline in grassland yield in some areas reportedly 50 to 60% (Ai Ming Wang, Grasslands Research Institute, Xinjiang Academy of Animal Science, pers. comm. 1992). Similarly, in Qinghai (northwestern China) grassland yields are claimed to have decreased by 35% to 64% over the past few decades (with decreases in the grass component and increases in weeds and toxic plants) (IFAD, 1993b). In Balochistan (Pakistan), yields under prevailing stocking levels have been estimated to be less than 30% of their potential (van Gils and Shabbir Baig, 1992). Pasture degradation ("overgrazing" and "desertification") is still a controversial issue. For example, UNEP claimed that 70% of Sahel rangeland is degraded whereas The United Nations Sudano-Sahelian Office (UNSO) found no sign of persistent dryland degradation in areas where pastoralism is the dominant land use (Vedeld, 1992).
Recovery from degradation is possible, even in arid and semi-arid areas. The drought of 1982-1984 in central Sudan led to loss of vegetation, crop failures, soil erosion and livestock deaths; since then there has been a rapid recovery because of increases in rainfall and low levels of exploitation (Olsson and Rapp, 1991). Overgrazing, interpreted as causing a depression in forage-base diversity and production, is the most common description applied to grasslands. Conclusions that, following heavy grazing, degradation is taking place or has taken place are often based on observations without substantive data. This has led to doubt that the effects are real in some areas, and adds confusion to controversy over the reversibility of degradation (Skarpe, 1991).
Of all forms of degradation, desertification has received much attention in Africa, where it is claimed to threaten more than one-third of the area (FAO, 1992). The Sahel has been at the centre of the desertification debate since the 1960s, and the Sahara has been accused of irreversibly engulfing large areas of land each year; it is now acknowledged that there is little scientific evidence for this. Degradation, even to desert like conditions, is reported well away from the desert front, in areas that have high human population concentrations, leading to excessive wood cutting and crop cultivation in areas where it not sustainable. Local nuclei of severe vegetation and land degradation may, therefore, be at least as important as any gradual advance of the desert boundary (Skarpe, 1991).
Desert fronts in the Sahara move back and forth between northern and southern boundaries according to changes in rainfall pattern. In fact, the southern boundary of the Sahara exhibits a north-south oscillation, with the northern limit of grassland shifting as much as 200 km in a year. Therefore, in areas where desertification is said to be occurring, it is important to establish if it is part of the natural oscillation of the condition of pastoral lands, or is triggered by an avoidable pastoral activity. If the former, the process must be accepted as part of the nature of the pastoral resource. To reliably detect any trends in desertification, monitoring of trends over decades would be necessary (Olsson and Rapp, 1991; Sandford, 1983).
An understanding of grassland dynamics and system uncertainty is fundamental when assessing of whether development proposals are ecologically practicable. The variability of the system, its ecological status and the forecastability of events all affect the potential for grassland development and indicate whether negative trends, such as those associated with overgrazing, are reversible.
Traditional interpretation of vegetation change, including degradation, has been based on the model proposed by Clements for North American grasslands early in the twentieth century. The Clementsian model asserts that plant succession tends towards a stable state, or at least an asymptotic condition, and predicts that degraded vegetation will follow a progressive succession back to the "original" state when the factors causing the degradation, such as excessive grazing, have been removed. In recent decades this has been shown not always to be the case, particularly in grazing land culturally and ecologically dissimilar to the North American type.
Where a complete reversion of grassland condition is possible, it is likely to be because the vegetation was merely overgrazed and not degraded to the point where major changes in plant vigour and species diversity had been triggered. Even when reversion is possible, hysteresis may occur, which prevents a return to the original state because other constraints, such as bush encroachment or a grazing-induced shortage of soil moisture or nutrients, become more important once vegetation is depleted (Skarpe, 1991). This process is illustrated in Figure 3a.
Concepts of single equilibrium communities and simple successional pathways may be applicable in sub-humid conditions, but not in arid and semi-arid regions, where vegetation establishment and change are often responses to exceptional events rather than average conditions (Call and Roundy, 1991).
Under the new paradigm for grazed ecosystems, grasslands are often described as being "equilibrium" or "non-equilibrium." Ecosystems in equilibrium tend to have livestock populations more-or-less in balance with the resources, with vegetation and livestock controlling each other. The certainty or predictability of such systems is relatively high. In contrast, non-equilibrium systems have high levels of climatic variability and uncertainty because of "internal" positive feedback mechanisms and massive "external" influences, such as drought. The effect of livestock grazing on productivity may therefore be overwhelmed by climatic influences, and such systems are highly unpredictable (see Scoones, 1995).
Humid, temperate grasslands tend to be in greater equilibrium than more arid types, and are more soil-fertility limited and resilient to livestock pressures. Grasses in humid environments are normally unable to compete with woody vegetation. Management systems in these areas have to maintain livestock populations at an optimum carrying capacity (economically or ecologically). Arid and semi-arid grazing lands, in contrast, tend to be non-equilibrium systems, are usually water limited, and may occur in areas of nutrient-rich soils and produce low quantity, high quality biomass (fast-response annual grasses with high nutrient content and palatability). These areas require sensitive and responsive management that must take account of temporal and spatial heterogeneity (Vedeld, 1992). In practice, there is a continuum between "equilibrium" and "non-equilibrium" systems and hence many grasslands cannot be characterized exclusively as either type; some exhibit a flux between the two states.
In some circumstances, the concept of "ecological threshold condition" provides a useful alternative to those of gradual retrogression and secondary succession, where secondary succession refers to the sequence of vegetation changes that follow the destruction of existing vegetation (Heady and Heady, 1982). According to the threshold paradigm, a deteriorating grassland may retain its capacity to recover up to a critical point, beyond which it cannot return to its former state. Once vegetation has been degraded to a lower successional state, it normally will not respond to a change in, or even removal of grazing. Instead, the former state may only be attained with significant managerial inputs such as fertilizer, burning, ploughing or herbicides. Often events such as drought, fire or heavy rains coincide with excessive grazing and cause a transition across a threshold between states. Multiple steady states appear to be present in many arid and semi-arid vegetation types (Friedel, 1991; Laycock, 1991). A conceptual model of thresholds and changes in plant community structure between grassland and shrub land is illustrated in Figure 3b.
Responses in productivity to varying grazing pressures (stocking load relative to available forage) within "intact" grasslands are not as clear-cut as often presented. A common recommendation for management is to reduce the level of utilization and allow higher herbage residuals, thus lessening animal impact on herbage production. Higher grazing pressures can lead to a reduction in palatable species and the increase of undesirable species, but this is not always true. Sometimes areas that are grazed heavily support the greatest populations of nutritious grasses. Grasses such as Themeda triandra and Eragrostis rigidor are of low feeding value unless they are closely grazed; grasses allowed to grow tall deteriorate rapidly in quality and a higher proportion of their biomass is lost to grazing (Simpson and Evangelou, 1984).
Reduction in grazing pressure, on obviously overgrazed land, generally leads to increased forage productivity through greater vigour of individual plants and increases in number of plants. At intermediate grazing pressures, however, the type of response may vary. Plants may be stimulated by intermediate levels of grazing intensity or, at the other extreme, be inhibited by any level of grazing. An intermediate response, with growth not being inhibited by low-level grazing intensities, is typical of many pastoral grasses. Contrasting responses of primary production to changes in grazing intensity are illustrated in Figure 3c.
It is notoriously difficult to formulate and implement practicable management policies for grazing land, particularly with a progression from humid ("equilibrium") to arid ("non-equilibrium") environments. Management systems must take grassland dynamics into account and incorporate seasonal and spatial patterns of forage supply, and often high inter-year variability.
Transhumant and migratory pastoralism take into account the high-level variability in grazed environments and maximize sustainable resource use while minimizing risk. The spatial and temporal flexibility of such systems supports higher livestock populations than is possible under sedentary grazing regimes. Unfortunately, socio-economic and political conditions often constrain or suspend transhumant and migratory pastoralism and enforce systems of inappropriate resource use. The interrelationships among resources, activities and external influences that constitute grazing management are illustrated in Figure 4.
Box 3. Alternative stocking strategies in coping with irregular seasonal forage supply (Behnke, 1992)
Key: Dotted areas = periods of supplementary feeding; solid line = ecological carrying capacity; dashed line = livestock numbers under stocking strategies A to D.
Strategy A represents high stocking rates and is insupportable within the confines of a single grazing area that is subject to large variations in rangeland productivity. Long-term sustainability of this strategy is dependant upon the exploitation of grassland resources with different patterns of forage availability, outside the area depicted. This represents the "migratory" or "transhumant" management strategy.
Strategy B represents a possible moderate stocking rate policy in which managers attempt to compensate for fluctuations in feed supply by quickly and deliberately adjusting livestock numbers (the "tracking" strategy). As in the "migratory" strategy, livestock leave the local grassland, but in this case primarily through disposal or death, rather than relocation. Development options are:
Risks of this strategy are associated with the simultaneous collapse of livestock prices, markets and grassland carrying capacity.
Strategy C represents another moderate stocking rate policy, this time where livestock numbers are held constant and shortfalls in grassland forage are offset by the purchase or cultivation of forage. In this case, pastoralism and farming are closely integrated either through market exchanges or mixed agro-pastoral systems. Development options are:
This option is well developed in the pastoral areas of North America, Australia and New Zealand. The system permits the maintenance of high livestock numbers and may therefore have long-term deleterious effects on natural grassland. In addition, this may be compounded by the conversion of the best grazing land to cropping.
Strategy D represents the second extreme: this time of extremely light stocking to the point where forage requirements rarely exceed forage supply. Grassland forage is wasted in most years. Economically, this is not consistent with the requirements of pastoralists.
The fundamental challenge of grazing management is to optimize, simultaneously, the interception and conversion of solar energy into primary production and the efficient harvest of primary production by livestock. Grazing management involves the manipulation of kinds and classes of livestock, stocking rate, grazing season and grazing intensity to optimize these two opposing processes and maximize livestock production per unit area on a sustainable basis. The managerial task of optimizing primary production and efficient forage harvest is further complicated by climatically induced variation in plant production and the widespread occurrence of selective grazing (Briske and Heitschmidt, 1991).
The "grazing optimization" hypothesis suggests that an optimal grazing intensity can increase primary production over that of an ungrazed system ("C" in Figure 3c). Some evidence exists to support this hypothesis, but it does not appear to be a significant ecological process operating on a regular basis in grassland systems. Illustrations of the grazing optimization hypothesis tend to exaggerate the potential increase in primary production resulting from an optimal level of grazing relative to the potential decrease, which may occur in response to severe grazing (Briske and Heitschmidt, 1991).
Grazing management systems have generally been based on forms of rotational or sequential grazing with inter-grazing resting to provide an opportunity to improving grassland production and ensure a level of sustainability. However, in both semi-arid and temperate grazing lands, the advantages of rotational systems over continuous grazing are not as tenable as once believed (Lambert et al., 1983; van Gils and Shabbir Baig, 1992)
Managing livestock and forage resources to accommodate seasonality of forage supply is a major issue in livestock production. Different approaches to the "matching" of forage supply and livestock forage requirements are illustrated in Box 3. The issues supporting either a conservative or opportunistic stocking policy for grazing lands are complex. Australian studies on the long-term effects of stocking rates, or management policies, on the success of managing droughts is summarized in Box 4. Traditional pastoralists often use a combination of conservative and opportunistic stocking policies.
Options for managing inter-year variability and droughts include: maximizing grazing distribution; sale of stock as soon as drought is indicated; an efficient breeding herd or flock with minimal unproductive animals; use of special-purpose pasture; use of fodders not used in "normal" years; expansion of areas available for grazing; weaning and selling stock earlier; and purchase of supplementary feeds.
Box 4. Long-term effects of different stocking rates or management policies on the success of managing droughts - Australian experience.
The tendency for livestock populations to fluctuate with major inter-year changes in forage supply carries a high risk of damage to grasslands during droughts. Studies in Australia have examined the long-term effects of different stocking rates or management policies on the success of managing droughts: Preliminary findings for cattle systems showed that a "low-stock" strategy displayed a low grassland degradation risk and facilitated the improvement in both productivity and stability of the grassland. In contrast, a "high-stock" strategy carried higher potential income but required very high drought management skills to avoid environmental damage. An intermediate "average-stock" strategy was believed to carry the greatest risk of environmental degradation because of an associated "wait and see" attitude to drought management. (Foran and Stafford Smith, 1991) Another study, using a sheep-based system (Stafford Smith and Foran, 1992), indicated an economic advantage to de-stocking as soon as winter rains failed, but highlighted problems of climate prediction and uncertain livestock market responses.
The moving of livestock to areas less affected by drought at that particular time requires the availability of climatically heterogeneous grasslands. This is a convincing argument against the forced settlement of pastoralists. The feasibility of the options outlined above is a major issue in sustainability. If conditions are at the extreme, the system is forced into "survival" mode, where management is simply not possible; instead, the long-term resilience of the ecosystem is relied upon.
The "Tragedy of the Commons" doctrine (Stryker, 1984) is based on the assumption that private ownership of grazing lands leads to better management. Evidence indicates that this is not true (Sandford, 1983). Nevertheless the trend towards systems of exclusive grazing rights (including privatization) continues (IFAD, 1993b). The real effects of such trends should be carefully and critically monitored.
The establishment of ranching-type livestock production systems is another major issue affecting the sustainability of use of grazing land. The worldwide attempt to replace (semi-)arid land extensive pastoralism by ranching based on North American, Australian or South African models has generally failed. The underlying assumption - that ranching is sustainable and extensive pastoralism is not - is false. There are numerous overgrazed ranches around the world and many well managed extensive pastoral lands (van Gils and Shabbir Baig, 1992). Examples of failed attempts to establish controlled communal grazing in sub-Saharan Africa are presented in Box 5.
In many grazing areas, livestock managers are losing access to key, high production resources that are occupied and exploited on a continuous basis by non-pastoralists (e.g. state farms) (IFAD, 1993b). Losing such resources reduces the capacity of pastoralists to exploit lower potential extensive land that is suitable only for grazing. Legal pastoral tenure is necessary for areas such as: water points, arable lands, transhumance routes, trees, and various types of grazing land. Factors in the design of pastoral tenure schemes and sources of tenure conflicts are described in Box 6.
Box 5. Alternative stocking strategies in coping with irregular seasonal forage supply (from Behnke, 1992)
Box 6. Factors in the design of pastoral tenure schemes and sources of conflict (Behnke, 1984)
The design of pastoral tenure schemes should include:
Land tenure conflicts are likely to occur whenever:
Sustainability of livestock production on grasslands and "straightforward" development opportunities is not as assured as was once presumed. Sustainability depends on the balance of a wide range of ecological, political, economic and social influences (Box 7). Often factors such as high human population pressure and the associated excessive livestock loadings suppress the productive potential of a grazing area, as indicated by soil, vegetation and climatic factors. Under these conditions, technical solutions cannot be expected to restore system balance and assure sustainability.
Grasslands are a varying resource; shifts in climate, whether short or long term, affect their potential productivity and must be accepted as a feature of that system. There is also a possibility that soil nutrient depletion is occurring, accelerated by rapid increases in livestock populations this century; this will have long-term effects on forage productivity.
There have been many attempts to improve the traditional extensive grazing lands of Africa and the Near East, but few have been successful, and even fewer have had a positive effect on vegetative cover (FAO, 1992). The disappointing outcome of grassland development programmes forced the recent "paradigm shift" in the approach to development, based on more accurate knowledge of grazing production systems and a greater participatory role of local people, incorporating traditional practices and customs. Projects must emphasize social reforms that maintain a system where low intensity production, usufruct, kinship dynamics, cooperation, environmentally friendly practices and nomadism are maintained.
The question of sustainability is not confined to the grazing lands of so-called "non-industrialized countries;" those of Australia and New Zealand have undergone major crises of sustainability despite high levels of technical skills and resources for development, including development programmes subsidized by government (Hughes, 1991; Joss et al., 1986; O'Connor, 1987; Pressland and Graham, 1989). The experiences of these countries highlight the caution required when considering management regimes for grazing lands in countries with less resources and greater population pressures.
Grassland use and development are linked, directly or indirectly, to the status, trends and opportunities of livestock and livestock product markets. In some countries over recent decades, the economics of extensive livestock production have become marginal. This seriously, and negatively, affects pastoralists' attitudes towards management and development.
Despite the reality that political, economic or population factors overwhelm the possibility of sustainable livestock production on some grasslands, there is still a clear opportunity for projects to contribute to the long-term potential of such areas. The approach to, and expectation of, such development, however, may require further change. Small-scale interventions are often the only practicable options; these in themselves are unlikely to assure sustainability but may act as catalysts for development towards greater technical opportunities. This "system component modification" approach gives the local people experience and reveals pathways for development that may be more fully exploited if predisposing factors, such as excessive population, are resolved.
Box 7. Factors affecting grassland sustainability and development potential. (Dickie and O'Rourke, 1984)