A major task of the next century will be the restoring of overdeveloped rivers like the Mississippi and the Rhine to their former natural productivity.

On the 3969-kilometre Missouri River, subsidised engineering has maintained navigation and shipping for over 50 years. But the original prediction by engineers that the river would carry between 12 and 20 million tons annually was never realised. At its peak, 20 years ago, shipping reached 3.3 million tons. Despite its location in the nation's breadbasket, less than 0.01 percent of the region's corn and 2 percent of its wheat are transported by water. As a shipping route the Missouri is uneconomic, and there are calls to restore the waterway to its natural functions as a means to reduce flood losses and to increase recreational use. The Mississippi is an old river where species have had a long time to diversify and create complex communities. It is home to most freshwater mussels of the US and a third of its fish species. Its north-south orientation allows the yearly migration of waterbirds, shorebirds and songbirds between northern breeding grounds and winter habitats in South America. Throughout this huge river basin, forests, wetlands, oxbow lakes and backwaters have been lost to settlement, agriculture and river engineering. The Missouri, for example, has been shortened, deepened and narrowed, and 95 percent of its floodplain has been converted to row crop agriculture. But the separation of fish from their floodplain spawning grounds and upstream reaches has virtually eliminated some species and reduced many others.

In addition, flood control and navigation structures have adversely affected the integrity and productivity of the Mississippi Delta and the Gulf of Mexico. Sediment that for millennia was carried downstream to replenish the delta is now trapped, and coastal areas are actually subsiding as water inundates wetlands and threatens coastal communities and productive fisheries. By destroying the basic ecology of the great river system, we are losing valuable ecological services and foreclosing future options for adaptation to climate change. After the 1993 flood, a special Task Force recommended ending the nation's overreliance on engineering and structural flood control in favour of floodplain restoration. It proposed that the river be managed as a whole ecosystem rather than as short segments. Even so, the mistakes made on the Mississippi look like being repeated in South America, where five countries - Brazil, Argentina, Uruguay, Paraguay and Bolivia - are currently considering the Hidrovia Project which aims to create a 3,400 kilometre year-round shipping canal that would open up the interiors of these countries.

This plan calls for the dredging and straightening of the Paraguay and Parana rivers and draining a huge portion of the Gran Pantonal wetland - considered to be the world's largest and most pristine wetland. Critics believe the plan would cause the extinction of bird, fish and mammal species; increase flooding; and disrupt many Indian and non-Indian communities. Changes in the basic hydrology - the distribution and flow of water - of the region will reduce its ability to maintain water quality and to act as a sponge to prevent flooding and drought. Like the opening of the Amazon highway, this navigation route would fuel the expansion of agricultural lands, adding to environmental problems.

Over the last 50 years the construction of large dams has skyrocketed. Today they number more than 38,000, with over half in China. Besides the large dams, hundreds of thousands of smaller ones constrict the world's rivers. The reasons for building dams include hydropower production, flood control, irrigation and storing water in the wetter season for use in the dry season. Many were built before any environmental or socioeconomic studies were made. With the benefit of hindsight, many problems have been recognised.

One of the first problems created by dams is the displacement of people from the area destined to be inundated by the reservoir. Nearly 5 million people, over half in China, have been or will be uprooted by dam reservoirs alone. The forced removal of about 1.5 million people from the 600-km-long reservoir of China's Three Gorges Dam will exceed any other dam-related human dislocation.

Even when people are not directly displaced by dams, there can be severe human consequences. One is the "paradox of malnutrition" that frequently follows when irrigated (often monocrop) agriculture replaces traditional methods. Numerous studies have found that many large water engineering schemes have directly increased the incidence of disease. For example, development in the Senegal River Basin has resulted in epidemics of bilharzia (schistosomiasis) and rift valley fever in areas that had previously been free of these diseases. Malaria cases have proliferated, and cholera and other diarrhoeal diseases have increased.

By reducing the flow of fresh water, dams and diversions can lead to the intrusion of salt water into previously fresh water supplies - making them undrinkable. Reservoirs behind dams cannot sustain the same amount of fish as natural river systems. Higher siltation rates in the tropics than in cooler zones mean a relatively short operating time for many dams. Globally, 75 percent of hydropower dams exceed their estimated cost, and many operate at less than their expected capacity. The large amounts of rotting submerged forests in the reservoirs can emit more greenhouse gases (such as methane and CO2) than coal-fired power plants, thus weakening the rationale that tropical dams will help ease global warming. Too often financial irregularities and lax building standards have caused serious safety problems.

Many of the ecological problems associated with dams have become apparent in the Columbia-Snake River Basin in North America where hydropower development began in the 1930s. Today the number of wild salmon returning to the Columbia River is less than 6 percent of what it was before the dams were built. Decades of retrofitting older dams and designing new ones have failed to produce ways for dams and fish to coexist.

Adapted by Noeline Gannaway from Imperiled Waters, Improvised Futures, pp18-25.