Agriculture has been important to people in deserts since domestication of crops began. Rain-fed agriculture is less important in deserts than in higher rainfall areas because of the scarcity and unpredictability of rain; alternative systems began as attempts to reduce risks imposed by rainfall variation. Irrigated agriculture has evolved in different ways in different places based on different situations and crops available (Figure 2.6).

Large perennial rivers running through deserts, for example, the Nile, Tigris-Euphrates, Rio Grande and Colorado rivers, have supported desert people and their irrigation for a long time. Of particular importance on the Nile was the annual flooding and deposit of nutrient-rich silts on alluvial soils as well as prevention of salt accumulations (Dregne 1999a). All these benefits stopped once the Awsan High Dam was built in 1970 and alternative fertilisers and salt removal techniques are being tested. The Nile Delta is shrinking for lack of silt deposits and the final solution has not been identified. The Colorado River supports some agriculture but is better known for supplying water and electricity to major urban centres in Arizona and California (Channell 1999b). Its delta has lost most of its water and hence its productivity. The Tigris and Euphrates basin has been occupied by people for millennia and is part of the Fertile Crescent with its early irrigation and urban developments (Dregne 1999b). The lower marsh has been recently drained and its contribution to agriculture eliminated. Silt in the river water and salinization of the irrigated land have been ongoing problems. Clogging by silt can be addressed by individual farmers for smaller canals, but peace and political stability are required for maintenance of large river systems of 300 km or longer.

Small-scale rainfall harvesting in deserts has been adapted to specific types of terrain, climate conditions and choice of crop (Lövenstein 1994). None of these approaches lend themselves to large scale, mechanised farming, but have provided abundant agricultural products for desert people. The terrace system was probably one of the earliest irrigation systems involving a series of stone walls across a water course. With rain, the terraced fields would fill up and excess water cascaded onto fields below. Perhaps as a next step, the hillside conduit channel system was established based on narrow channels from neighbouring hill slopes. Micro-catchments, still in use today, have for thousands of years enhanced water application to small scale catchment areas. On a larger scale, diversions have been used on small and large, perennial and ephemeral rivers to channel water onto terraced fields on adjacent plains or even at a distance. Oasis agriculture has developed wherever water is available and has taken on many forms. A unique water harvesting system for oases is the foggara, which augment the water supply of isolated oases and small villages in places where it was not sufficient to encourage extensive settlement (Box 2.4 and Box 2.5).

With much ingenuity and hard work, different people in different areas supplied by an assortment of water sources have developed systems for crop production. Original savanna crops such as millet and sorghum have been adopted for desert use. Palms, tamarisk, acacia, and Zizyphus as well as tubers, fruits and cereals were all part of the Sahara array of crops when rains were more plentiful several millennia ago (Reader 1997). Palms, however, are the only species known to have been domesticated directly in deserts (Box 2.6).

Hunting and gathering, pastoralism and irrigated agriculture all advanced, at different rates and different times, with different innovations. Baskets and pottery for transporting food and other goods allowed people to gather and carry more food and other materials. Deserts have always been a part of the global environment with desert peoples trading within deserts and with neighbouring cultures (see Chapter 3).

Population constraints within deserts, imposed by changing climates and agricultural developments, have caused wide fluctuations in the numbers of desert inhabitants (Reader 1997). With the expansion of technologies supporting people to live in deserts, the degree of fluctuation may be reduced. Nevertheless, institutions focused on resource management, as required for successful hunting and gathering, nomadism, transhumance and oasis agriculture, will undoubtedly play a large, although altered, role in future desert development.