Desert ecosystems in the huge spaces between the wet places vary greatly in biological value. Some of the hyper-arid deserts support little life of any kind, except after rare rainstorms when they are briefly visited by species from surrounding, better-watered deserts. One can travel for hours in parts of southern Libya, Algeria or Peru, eastern Saudi Arabia and central Oman, northern Sudan, Chad, Niger, or northern Chile without seeing a blade of grass or a bush. A daily lizard or gazelle, or at some seasons a migrating bird, comes as a surprise. Some authorities have asked, are they ecosystems at all? Are the connections between species durable enough to class them as fullyinteractive systems? (Moore 1978). It is more debatable whether these wildernesses deserve the scarce resources available to conservationists. Perhaps conservationists should concentrate their efforts on the desert's edge, that is, on arid ecosystems interposed between the hyper-arid wildernesses and the savannahs and grasslands. It is in these areas where the harvesting of wild products often offers a valuable economic alternative (Box 5.6), and more generally, where many species have great biological value and are under maximum threat.

Hunting

Indiscriminate hunting is one of the biggest threats to desert biological sustainability, so damaging that if allowed, it would soon eliminate its own raison d'être. Hunting may have eliminated the postglacial mega-fauna of North America, and much later, North Africa and southwest Asia lost most of their large mammals to hunters in Roman times. It is said that on the day on which the Emperor Titus inaugurated the Roman Coliseum, 5 000 wild animals were slaughtered. The more distant African and Asian deserts lost their large mammals after guns became available in colonial times, when they were also infected with the European hunting ethos (Anderson and Grove 1987). In desert Central Asia, the native Przewalski's horse (Equus caballus przewalskii), was hunted to extinction in the wild by about 1870. Hunting and competition with domestic camels has reduced the population of the wild Bactrian camel (Camelus bactrianus ferus). Przewalski's gazelle (Procapra przewalskii) is now isolated to only four local populations around the Qinghai Lake in desert China (Li and Jiang 2002). Hunting continues. Income from trophy hunting in Namibia was estimated in 2003 to be 14 per cent of all tourist earnings (Humavindu and Barnes 2003). Large convoys of air conditioned caravans follow hunters across the deserts of Arabia, Sudan and Kazakhstan. The remaining populations of large mammals, such as various species of gazelle, oryx (Oryx beisa), addax (Addax nasomaculatus), Arabian tahr (Hemitragus jayakari) and the Barbary sheep (Ammotragus lervia) are on the brink of extinction. Game bird populations are declining fast, particularly that of the Houbara (Chlamydotis undulata macqueenii), which is the choice for hunters with falcons in Arabia and Kazakhstan (Tourenq and others 2005). Most of these endangered species need urgent protection: captive breeding (of oryx as in Oman); reintroduction; restrictive legislation and its enforcement; and so on. Surprisingly (to some) help is coming from the hunters, at least as concerns the Houbara; they have supported captive breeding, radio tracking and other conservation measures in Saudi Arabia and the United Arab Emirates (Bailey and others 1998).

Grazing

Another threat to dry ecosystems may come from pastoralists, but here the argument is not so simple. Pastoralists do indeed use the high nutritive value of arid grasslands in wet seasons. The Fulani in Niger take their cattle to the edge of the desert in the wet season, specifically for this (Penning de Vries and Djeteyé 1982), as do the Kabbabish camel nomads in Sudan (Wilson 1978).

The argument about the damage that grazing might be doing to these ecosystems is unfinished. Some authorities believe that overstocking has removed valuable species, and reduced grazing value. Their evidence includes the replacement of palatable by unpalatable species, particularly the woody shrubs that have invaded many arid areas, as in Arizona (Guo 2004) or in southern Africa (Wiegand and others 2005). Further evidence lies in research into recovery times after intense grazing. In one case, full grass cover only replaced woody scrubland after 20 years of protection (Valone and others 2002). Estimates of recovery times for the Mojave ecosystems range from 50 to 300 years, but in some cases could reach 3 000 years (Lovich and Bainbridge 1999).

The counter-argument depends on the model of these ecosystems as "pulse-response" (see Chapter 1). In seasonal or multi-annual dry periods, the ecosystem is said to be incapable of supporting enough stock to do it any damage in the brief, wet pulses. Only if stock is fed supplements brought in from wetter areas, or is moved back and forth from wet to dry areas (which, it is true, are common practices), is "overgrazing" even possible. Nomadism, this school also argues, is an ideal system for using the patchy effect of rainfall on grazing. Further, especially in the Old World, the character of many arid ecosystems depends on grazing, and because some of the native grazing animals have become domesticated, grazing by them could be vital in maintaining the original ecosystem dynamics. Conservation may then depend more on the choice of grazing intensity and grazing species, than their total exclusion. Finally, some authorities believe that overgrazing is much less of threat in arid climates where forage dynamics are primarily driven by climatic cycles, than in semiarid or sub-humid climates where grazing is more likely to be the chief control on the availability of forage (Briske and others 2003).

Conservation priorities

Another issue in the conservation of these systems is the choice of priority. If a rare species has priority, like the long-lived desert tortoise in the arid Mojave (Gopherus agassizii, a federally-listed, threatened species), complete protection from hunters and graziers may be imperative. If the objective is the conservation of an ecosystem as a whole, there is more room for manoeuvre, but there are problems there too. If there are functioning patches of different size, all the way from micro-ecosystems sheltered by bushes or trees, to patches of many hundreds of square kilometres, what is the optimum size of protected areas to ensure conservation? The failsafe rule is: as large as possible.

Community-based conservation

A more practical question is also exercising conservationists: should conservation be "expertled" or "community-based"? (Berkes 2004). Community-based conservation does not always succeed, mainly because development and conservation goals usually conflict, and because the phrase has become too loosely used. But with care, the idea of community conservation should resonate in deserts where people and their grazing animals have been part of the ecosystem for thousands of years, and where there is huge wealth of indigenous knowledge, as Triulzi (2001) describes in Syria.

Beauty

The beauty of deserts, so essential to tourists and residents, is a further priority for conservation. There are at least three conflicting desert aesthetics: desert as wilderness, as challenge, or as scenery. The wilderness-seekers - stereotypically, backpackers, and the well-educated - use deserts to find their "intellectual humility" (Leopold 1949). To them, development and the numbers of visitors need stringent control. The challenge-seekers need technology to achieve their aims (they might be SUV drivers). They want to drive over exciting terrain and may compromise with, even welcome, development. Given the openness of many deserts and their accessibility to motor transport, their activity can be intrusive (Sax 1980). The tourist aesthetic is subdivided many more times: the Wissa Wassef romanticism of the image on the title page of this chapter; the desert fathers (Box 5.2) and/or romantic decay; Marlboro County; walkabout; T.E. Lawrence; Rudolf Valentino; and so on. And tastes change.

These various demands need to be balanced with the needs of people who live and work in the desert, let alone with the demands of people beyond the desert for water or energy. The task is not impossible. Most visitors to the Grand Canyon or Uluru are impressed by the sensitivity with which they are managed under the onslaught of tens of thousands of people in a season, even if a minority complains one way or another. An advantage of some deserts over many other global environments is that they are remarkably resilient. Sand dunes loose the tracks of people or vehicles after a gentle breeze. Bare rock may not register a moderate number of footfalls. But the desert is not everywhere so unforgiving; there are many places where the demands clash, or where trampling, littering, and disturbance do lasting damage. Vehicle tracks in many deserts are visible on satellite images; they may persist for very many years (Belnap and Warren 2002).

All forms of conservation (particularly of biological value and beauty) are best met with some degree of statutory protection, backed by adequate funds. The trajectory of the total desert area under protection is upward, which is encouraging, but it would need to be accompanied by less easily monitored trajectories of community involvement, visitor satisfaction, biological change and funding, to confirm it as a wholly good story.