UNEP Website GEO Home Page
Rivers, Wetlands and Lakes


Wet desert environments are biologically the richest places in the desert. The richest of all are the perennial rivers. Ephemeral rivers have value, and although they profoundly affect the lives of the communities and wildlife that live near them (Jacobson and others 1995), their value and vulnerability are small compared to those of the perennial rivers, the good management and conservation of which is of utmost importance.

The most difficult issue in conserving the biological value of perennial rivers is the assurance of flow. Some, like the Tarim River, were desiccated by irrigation decades ago; most of the others have been progressively depleted. But the mere suggestion that water should be left for wildlife is risible to most managers of irrigation schemes. The second, closely linked issue is the quality of the water, and in this, the interests of conservationists and farmers come closer together. The main contaminant is "return flow": water returning to the river from irrigated fields. Return flow is always more saline than the water taken from the river in the first place. Saline return flow increases problems for downstream farmers, as when it flows from the Punjab and Sindh in Pakistan. In the Colorado River in the U.S., salinity was a minor problem in the 1960s; by the 1980s, nearly 40 per cent of its salinity came from return flow (Law and Hornsby 1982). The issue is international on the Colorado River (and on the Rio Grande that flows between Texas and Mexico), because the U.S. and Mexico agreed, in 1974, that the salinity of crossboundary rivers should be kept below a threshold, but the salinity of water that flows to Mexico is, even so, above that threshold. The U.S. Congress, long ago, allocated US$1 billion to the problem (Reisner 1986); a desalination plant at Yuma (cost US$256 million) opened in 1992, but closed eight months later, partly because of high running costs. Hopes of a reopening are now being raised. In China, the maximum annual salinity in the upper reaches of the Tarim River rose from 1.3 grams per litre in 1960, to 4.0 in 1981-1984, and to 7.8 in 1998, as it was increasingly supplemented by return flow (Feng and others 2005). Salinity damages the ecology of perennial rivers and their floodplains. In the southwestern United States, the salt cedar, an alien, salt-tolerant tree, has invaded hundreds of thousands of hectares of alluvial plains, altering their avian and invertebrate ecology. Attempts at control have met with mixed results (Shafroth and others 2005). Return flow may also carry residues of agricultural chemicals and toxic trace elements.

Dams themselves severely interfere with the ecology of the perennial rivers. They deprive rivers of sediment, and thus depleted, most rivers excavate their channels, and this isolates and desiccates their former floodplains. When siltdeprivation is added to the replacement of cyclical floods and low flows by the more constant flow from a dam, other problems appear. After the closure of the Glen Canyon dam on the Colorado River, riffles (where the river flows shallowly over stones) expanded at the expense of pools and this favoured fish that spawned in gravel over those that did not (Magirl and others 2005). The combined effect also discourages lateral migration of the channel (as when meanders move), and this interferes with the ecology of the early-successional native cottonwood tree (Tiegs and Pohl 2005). Deep, cold water is released from dams, and this, together with all the other changes, is endangering the humpback chub in the Colorado (Petersen and Paukert 2005). Smoothing out the flow alone has altered the habitat of clams in the Colorado Delta (Cintra-Buenrostro and others 2005). Siltfree rivers create further troubles when they reach the ocean. Silt from the Nile once protected the Mediterranean delta coast from erosion, but after inauguration of the Aswan High Dam, no longer does so (Fanos 1995). Fewer nutrients now enter the Mediterranean, leaving it even more of a marine "desert" (Azov 1991).


Wetlands have less economic value than perennial rivers, but because their economic development value vastly increases when they are drained, they are under even more pressure. Because wetlands are the resting or roosting places for huge numbers of migratory birds, the most important have been given supposedly strong protection in international agreements, notably the Ramsar Convention of 1971, and the Convention on Biological Diversity (CBD) of 1992; yet this has done little to stem their loss (Lemly and others 2000). Desert wetlands vary hugely in size and vulnerability, and may also change substantially with management and climatic oscillations (Figure 5.11).


Most desert lakes are dry, seasonally or for years at a stretch. The smooth surfaces of some are best-known as speed tracks, but not all are smooth. Some of their features, such as the flamingo pools in the huge Salar de Atacama in Chile, or the remarkable wind-propelled rocks at Racetrack Playa in Death Valley (Bacon and others 1996) deserve careful management, but many, like the remoter parts of the vast Salar de Atacama, and Umm-as-Samim in Oman, are protected simply by remoteness and the dangers of travel.

Lakes fed by perennial rivers, like the Aral Sea, are more biologically rich, and more vulnerable. Many have suffered severely. Owens Lake in the upper Mojave dried up in 1926 (except for a few shallow wetlands), drained by an aqueduct to Los Angeles, opened in 1913. The dry, salty lake bed now releases an estimated 900 000 to eight million tonnes of dust a year, the most prolific single source of dust in the U.S. The plume is obvious 40 km downwind (Reheis 1997). Before they were drained, lakes like this seasonally supported millions of birds. Even though saline, Lake Eyre in South Australia - the world's largest ephemeral lake - occasionally supports thousands of waterfowl (Kingsford and Porter 1993). A less well-known set of lakes is probably even more biologically valuable. These are the remote, groundwater-fed lakes, some of them in extreme deserts. A few in Libya and many in China collect in hollows between huge dunes. Others are strange anomalies, like the stairway of fresh and saline lakes at Wanyanga (Ouanyanga) in extremely arid, northernmost Chad (Figure 5.12). Species endemic to isolated water bodies like these, as the desert pupfish of the Sonoran Desert, are vulnerable to extinction and need special attention (Fagan and others 2002).

Rehabilitating wet desert habitats

Given the massive size of the engineering structures that have ruined the ecology of rivers, wetlands and lakes, and the millions of people who now depend on them for water, their complete ecological restoration is unthinkable, at least in the short-term. In this time frame, better management of flow regimes, rerouting of saline return flows to special canals (as has been proposed in several irrigation schemes), and the preservation of a few remnant wetland or lacustrine ecosystems (as is also happening in many places), would alleviate some of the more urgent conservation problems. Desalinization, as may again happen on the Colorado, is too much to hope for on a major scale. In the longer-term, the financial, human, and environmental costs of maintaining huge waterdelivery systems may foreclose them. Rivers, wetlands and lakes might again return to their prelapsarian glory, but at huge human cost.

Two further wet ecosystems, both of high biological value, need rather different forms of conservation. The first is in damp desert "hollows", places to which water gravitates and feeds a shallow watertable or at best a few springs. The second are ecosystems that have been isolated by post-glacial climatic change: the "sky-islands" (see Chapter 1). In some of both, isolation has allowed the evolution of unique species, or sub-species. Unfortunately for conservationists, people (indigenous or exploitative, malevolent or innocent) also gravitate to these places, which then become the sites of intense conflict. Their conservation needs strict restrictions on interference.

© UNEP 2006