Across the world, pastoralists are strongly associated with rangelands. (Reindeer herders are an exception to this generalization, in part because their key asset – the mobility of their animals – makes it possible for them to exploit nutritional resources in regions unavailable to arable farmers.) “Rangelands” is a broader term than “grasslands”, and includes regions where woody vegetation is dominant; moreover, it is common in texts describing land from the viewpoint of livestock production. Grasslands are just that, and the term has a more biological emphasis.2 Some of the ecological literature attempts to distinguish rangelands from natural grasslands (e.g. the Elsevier Ecosystems of the World premises different volumes on this dichotomy – see Bourlière, 1983; Coupland, 1993). However, closer examination of the descriptions suggests that either the origin of many grasslands is contentious or grasslands become “natural” if they are ancient human creations (see e.g. Gillison, 1993a, 1993b on the grasslands of New Guinea).
The literature uses several terms for the world’s main rangelands: African savannah, Eurasian steppe, South American savannah, North American prairies, Indian savannah, and Australian grasslands (Moore, 1970; Groombridge, 1992: 285; Solbrig, 1996). Estimates of their importance vary according to the regions included but, as figures given in the literature suggest, rangelands occupy between 18 and 23 percent of the world’s land area, excluding Antarctica (Table 3).
Estimated areas of the world’s rangelands
Whittaker and Likens (1975)
Atlay, Ketner, Dugvigneaud (1979)
Olson, Watts and Allison (1983)
Savannah (million km2)
Temperate grassland (million km2)
Total (million km2)
Rangeland as % of world land area
Rangeland as % of world land area (excluding Antarctica)
Source: Groombridge, 1992: 281.
Grasslands are usually divided into four major types: tropical grasslands, prairie/steppe, temperate grasslands and tundra, which are determined either by the underlying soils or by climatic conditions. Table 4 shows the main categories of grasslands and their major zones of concentration.
Classes of grasslands
Africa, South America, northern Australia, India
North America, Central Eurasia, South Africa
Europe, North America, Australia, New Zealand, Asia
All subarctic regions
The rangelands of the Tibetan Plateau are unique in that they are the highest grazing lands anywhere in the world. The greater part of the plateau is more than 4 000 m above sea level, and some camps are as high as 5 100 m (Miller and Craig, 1997: 58ff.).
The main floral component of rangelands – grass – exists to be grazed, and over time co-adapts to both the intensity and the quality of grazing. The long-term evolutionary history of a grassland ecosystem, as well as the history of the last few centuries, is therefore essential to understanding grassland’s response both to management and to new pressures.
In parts of North Africa and Southwest Asia, rangelands have been reduced in size, in part because the widespread use of irrigation technologies (in both traditional and, more recently, high-technology forms) has allowed agriculture to colonize much larger regions of the rangelands. As a result, what rangelands remain are considerably more arid than those exploited by pastoralists in sub-Saharan Africa. Indeed, drought conditions may be said to obtain most of the year. Responses to this have long been developed, and involve both the species used and the movement of resources. Pastoralism has traditionally been oriented around camels and sheep, with sheep becoming predominant in recent times owing to their greater marketability. The movement of water and feed resources to arid areas has been practised since before ethnographers began to describe pastoral nomads (notably through the carriage of large water-skins by camel). Today, pastoralists throughout North Africa and Southwest Asia have relatively sophisticated trucking systems (for water, feed resources and the animals themselves) which allow them to exploit areas that would be unavailable in sub-Saharan Africa (Blench and Marriage, 1998a).
The situation in Australia is somewhat different. Australian arid and semi-arid rangelands occupy nearly 70 percent of the continental land mass, and much of this area is used for extensive livestock production (Groves, 1981). Australia’s rangelands were transformed subsequent to European settlement by the following (James, Landsberg and Morton, 1999):
In many arid or semi-arid rangelands in Australia and North America, artificial sources of water are so widespread that lack of rainfall results in only localized feed shortages (Bennet, 1997). Large herbivorous mammals are able to continue grazing in areas that they would usually have abandoned (James, Landsberg and Morton, 1996). Native wild animal populations that used to rely on drinking-water from natural sources increase because they are able to persist in areas that were previously not habitable for most of the time. Such “artificial” increases in some species may have negative effects on others. The effects on native fauna are the displacement of ground-dwelling bird species; changes in the distribution and abundance of invertebrates (e.g. grasshoppers, ants and collembolans); possible recent extinction of some medium-sized native mammals; and indirect effects on wildlife populations through the changing activities of predators (James, Landsberg and Morton, 1999: 1). Another effect of artificial water sources is to maintain constant high levels of grazing pressure. Many native plant species are naturally not adapted to constant grazing and tend to be eliminated in favour of exotics (Austin and Williams, 1988).
Rangelands are strongly characterized by patchiness of resources and resilience in the face of climatic extremes. Especially where water resources are short, as in the semi-arid rangelands of Africa, the Near East, the New World and Australia, vegetation has adapted to patchy and variable rainfall. Reserves of seeds of particular species accumulate in the soil where they germinate when specific precipitation regimes occur. As a result, not only pasture availability, but also pasture quality may vary substantially; it is often difficult to predict which species will be abundant in a given year.
The situation in cold-weather rangelands is somewhat different since water availability is not usually a limiting factor, while the severe cold and the short flowering season bracketed by snow are. Pastures are therefore rather different in structure, with a very large number of flowering species competing to seed in a brief season. Abundance and quality are less often issues for pastoralists than is access, whether the barriers are snow or administrative. For this reason, cold-weather pastoralism tends towards systems that are based on transhumance and haymaking.
Traditional pastoralists broadly accept pasture and rainfall as givens and adapt their social and herding systems to take best advantage of them. The one exception to this appears to occur in the Andes, where herders create irrigation channels to encourage the growth of bofedales, bunch grasses that are particularly important for llama nutrition (Orlove, 1982: 100). The stone-lined leets of the United Kingdom’s Dartmoor, which were constructed in the Neolithic period, may also have the same purpose. The economic importance of extensive livestock production in Australia and North America and the greater integration with the market have helped to even out the unpredictability of rangeland productivity. Once mobile pastoralism in these regions had been eliminated through enclosure, many ranches had become large enough to allow the use of movement within the property, especially as more water sources were installed, and this blunted the impact of overall precipitation. Increasing the numbers of boreholes has often been associated with pasture seeding, thereby encouraging animal numbers to increase well beyond the long-term capacity of the land to support them. The consequence has been, as the droughts of 1996-1998 in the mid-west of the United States and Australia demonstrate all too clearly, that even extensive livestock producers with good access to infrastructural support and management tools can mistake short-term gains for long-term equilibrium and end up going out of business.
The vagaries of the weather have produced another technological response: increasingly sophisticated software programs and databases intended to assist producers both to be aware of climatic data and to make real-time use of it in managing their stock. This is particularly the case in Australia, where a series of blows to the pastoral industries have had a bad affect on the economies of States that are dependent on them and where there is intense commercial competition to find better management tools. As a result, there is both an abundance of information available to pastoral producers and a surfeit of commercial promotion, which is often larded with exaggerated claims masquerading as science. While there is no doubt that real-time and historical climatic data are valuable, it remains to be seen whether practical producers can make use of them in such a way as to gain real advantages in stock survival rates.
Control of stock numbers is another important tool; pastoralists in open access systems find it practically impossible to restrict overall numbers, especially when pasture productivity is poor. As a consequence of this, the calculation of carrying capacity has evolved into a minor industry. Despite its scientific superstructure, there is virtually no evidence that this has any scientific validity, as is demonstrated by the fact that different experts come up with widely varying figures when asked to estimate the carrying capacity of the same piece of rangeland. Even without hand-held computers and innovative software, livestock producers in enclosed systems who have experience of the landscape should be able to detect when their herds are putting too much pressure on resources. However, this is not necessarily the case; recent evaluations of Australian properties suggest that overstocking is common, with predictable consequences. One response has been the evolution of the land care movement/strategy which encourages producers to take a more holistic approach to the landscape and its management, rather than simply treating it as a more or less depleted resource.
Especially in enclosed systems, another strategy has been to seed rangelands with exotic species that are believed to have greater nutritional properties. In recent times, this has been extended from small planes, using such legumes as Stylosanthes. The consequences of introducing exotic species into vegetational systems where there is a high degree of endemism, such as that of Australia, have been well documented elsewhere; Acacia albida, an important browse plant in Sahelian Africa, is now characterized as an “aggressive weed” in Australia, where it has out-competed local species and become the subject of an expensive elimination campaign.
Obviously, these strategies are relevant mainly to developed economies, although the insertion of too many boreholes has afflicted semi-arid rangelands everywhere. Control of stock numbers in open access rangelands is basically only possible in totalitarian regimes, and thus occurred in the former Soviet Union and in the pastoral regions of the Negev, at least for the Palestinians. Unfortunately, the numerous projects, policies and strategy papers that have proposed the opposite strategy have now joined consultancy documents on destocking in some limbo where such idealistic entities are finally laid to rest.
Although pastoralists are primarily associated with rangelands, almost everywhere they make some use of forest vegetation in their annual grazing cycle; the most extreme cases are Eurasian arctic reindeer systems, which are confined to forests. Reindeer are also a special case as they depend primarily on mosses rather than browsing the trees themselves. The ability of pastoral species to digest woody vegetation is highly variable; camels, donkeys and goats can live almost exclusively on such a diet, whereas cattle, yaks, buffaloes and sheep can only consume very limited quantities, unless they have been fed on browse from an early period. Livestock species also have differing capacities to gain access to browse; camels have an advantage in that they can browse on thorny species with leaves that other species cannot reach. Unlike goats, which uproot or strip shorter plants, camels rarely damage the biodiversity of environments in which they graze.
The differing capacities of individual species have had a long-term impact on both ethnic specializations and the balance of environments. In the Horn of Africa, peoples that depend on browsing species, such as the Somali camel herders, border on arid-zone cattle producers such as the Boran and Turkana in Ethiopia and Kenya. Years of intensive grazing of grasslands cause gradual invasions of woody vegetation and drive away grazing species, which are unable to digest the lignin. It seems likely that pressure from biting flies also increases, as the shade provides a greater range of habitats, to judge from work in the Cameroon grasslands (Boutrais, 1995). Peoples who specialize in browsing livestock, such as Somali and Rendille, then move in and take over. However, over a long time span, grasslands reinvade, making the land again suitable for cattle and sheep.
Throughout much of Africa, trees are the characteristic haunt of tsetse and other biting flies, and this makes anything more than seasonal use problematic. In vertical transhumance in the Himalayas, the winters are spent in forested areas, the animals only moving to meadows during the summers (Chakravarty-Kaul, 1997). This has more to do with the cycles of snow than habituation to diet, but animals in these systems have become more adaptable than those in Sahelian Africa by virtue of the major diet changes to which they are exposed within the course of a year.
Increasing the amount of browse that cattle can consume would certainly improve conditions for pastoralists in parts of Sahelian Africa, although resources for browsing species would thereby probably be reduced. Some experiments with camels in Australia may be promising in this respect. Recent research on watering camels and cattle at the same trough, when both are feeding in wooded grasslands, suggests that the enzymes that allow camels to digest browse may transfer across to cattle and increase their abilities to digest browse. If this were expanded to encourage semi-arid cattle producers to include some camels in their herds and adapt their watering strategies it might have a major impact on survival in some regions.
It is generally considered that
Bedu herds in Jordan in the pre-modern era were limited to some 150 to 200
animals, since this was all that could be managed by the labour available
in a family unit. However, two changes have occurred to change this
situation radically: the introduction of water trucks and the widespread
availability of subsidized feeds. The buying of feeds has become the
single most important household expenditure. Every individual livestock
producer interviewed in a survey of 400 households spent some money on
feeds. Feeds are purchased in bulk by the government and sold according to
allocations denoted by the 1991 livestock census. It is government policy
not to allow the purchase of subsidized feed for camels. Despite this,
many of the camels, especially those kept for milk on the western edge of
the rangelands, are fed on purchased feed. The use of these feeds has
spiralled in the last few years with the ever-declining rangeland
resources. The system of allocating subsidized feeds on a per-head basis
has created a major incentive to increase herd sizes, and in the Badia –
the rangelands covering most of eastern Jordan – herds of 1 000 to 2 000
sheep are common. The forage resources cannot support herds of this size,
and the desert is increasingly used to store animals while sacks of feed
are trucked in.
Supplementary feeding seems to have had little place in traditional pastoralism anywhere in the world, in part because herds were very small and pasture resources vast in the pre-modern era. However, stocking winter hay was practised all across the temperate world, where snow or other climatic conditions made it impossible to provide the herds with adequate feed. Grass is usually cut in autumn, bundled in stores and rationed to the herds during the winter months. Mechanized grass-cutters and transport have increased the efficiency of this process, and the helicopter has made it possible to drop hay on herds isolated by snow. However, this is expensive and, although it was often subsidized from military budgets in Central Asia in the past, its use is now only sporadic.
In this century, changes in supplementary feeding have been extremely significant following the increasing availability of agro-industrial by-products, and transport systems to deliver them to remote areas. In semi-arid Africa, products such as cottonseed and groundnut cakes and molasses are now regularly sold to pastoralists, together with mineral licks. Throughout semi-arid West-Central Africa, cotton production was introduced in the colonial era as a cash crop, and its cultivation has remained an integral part of the economy in some countries. The main by-product of ginning cotton locally is cottonseed cake, an oily compressed cake that acts as a nutritious livestock feed. In the 1960s, cottonseed cake was introduced as an experimental diet supplement in Nigeria, and had to be given away to herders (Otchere, 1986). After some time, pastoralists gradually realized the value of such supplements; however, so did more intensive producers, especially those with stall-fed animals based around cut-and-carry businesses. As a consequence, oilseed cakes have become so highly valued that supplies are regularly bought up by wealthy urban entrepreneurs and rarely reach the markets or become available to ordinary cattle producers (Kaufmann and Blench, 1989).
The increased globalization of markets has also led to a highly significant international trade in animal feeds. Where herds have expanded far beyond carrying capacity, as in most of the semi-arid steppeland of the Near East, diets need to be supplemented with purchased feeds. The political significance of pastoralism in many countries in the region has had the consequence that national governments are tempted to subsidize feeds, thereby helping to swell herds to wholly unrealistic levels (Box 1).
2 There are two parallel series of international congresses, the International Rangelands Congress and the International Grasslands Society, whose meetings alternate but which are attended by largely the same constituency. So similar are these meetings that it has recently been proposed to merge the two societies, although this proposal remains controversial.