The costs of cleaning up the mistakes of the past have been rising and are likely to continue to do so. There have been many of these mistakes, some of them discussed elsewhere in this chapter. The reclamation of salinized land, and the revival of economies that once depended on that land, already consume large sums, and could consume much more. By far the best-known case has been the Aral Sea basin (Figure 5.5), which will take decades to restore (if it is ever achieved). The existing recovery programme will only save one basin of the former sea, and reduce only a proportion of the dust that damages health (Kingsford and others 2005).

The Aral is not unique. "Salinization and waterlogging have affected 8.5 million ha or 64 per cent of the total arable land in Iraq; [.] 20-30 per cent of irrigated land has been abandoned due to salinization" (UNEP 2002). In the Tarim River basin of China, more than 12 000 square kilometres of land was salinized between the 1960s and 1990s (Feng and others 2005). In both countries, degradation is expanding. Because salinization slowly and incrementally affects yield well before it precipitates complete failure, the problem may take years to become apparent to anyone but the local farmers. Collapse of the whole may then be sudden. This is what is said to have happened in Iraq, 3 000 years ago, where the legacy of salinized soils is almost as great today as it was then (Jacobsen and Adams 1958). Dealing with accumulated salt may take centuries, if not millennia.

The prospect of wars over water (Bulloch 1993) has not materialized. Countries that compete for water, like India and Pakistan, have kept strictly to their agreement about sharing water (in their case, the Indus Waters Treaty), even if they have resorted to war over other things. Perhaps water is too important to fight over (Alam 2002). This is not to say that water is not a source of local conflict, as it has been recently in Cochabamba in Bolivia, and continues to be in Israel and Palestine. Reclamation may not, then, be violent, but it will not be cheap. Communities and economies will need to be relocated; national and regional economies have to be adjusted - some of them for the better in the long-term, if we believe Reisner (1986); more will have to be spent on supplying water, and on international treaty obligations. Dust, as from the Aral Sea and the Owens Lake in California, has to be controlled. Sedimenting reservoirs have to be managed more carefully and on a longer-term basis.

Collapse

Accounts of the collapse of the Iraqi and Indus Valley (above) and other desert collapses, such as that of the irrigation system in Chaco Canyon in Arizona in the 12th century (Diamond 2005), show that collapse is a common hazard in deserts. This is largely because desert economies depend so utterly on water. In ancient water-supply systems, withdrawal or degradation usually brought collapse. Vulnerability may now be reduced, but it is not eliminated, as the Aral crisis shows.