Changing courses continues to disrupt the natural ebb and flow

Except for global warming, there has been no more drastic human alteration of the landscape in the past 50 years than the damming, regulation and diversion of the world's rivers," Daniel Beard, former head of the US Bureau of Reclamation, the US's premier dam builders, once told me.

He was not boasting, he was apologising. "Those who promoted dam projects were not honest about costs and benefits," he added. "The truth is that dam proponents would say just about anything to get a project approved." The decisions to build were "political, benefiting particular politicians or their benefactors rather than solving a problem". Like Beard, modern environmentalists have come to see large dams as engines of environmental destruction and a burden rather than a boon to economic development.

But it was not always so. For many years, greens in North America and Europe supported dam builders. Rather like wind power today, hydroelectricity was seen as a new, clean, cheap source of electricity. And hydroelectric dams do generate a huge amount of electricity, about a fifth of the global total. More than 60 countries depend on them for more than half their power. One dam, at Itaipu on the River Parana between Brazil and Paraguay, has a generating capacity equivalent to a dozen conventional power stations. China's Three Gorges on the Yangtze will be bigger.

The world's large dams hold approaching 7,000km3 of water. They barricade more than half the world's rivers, and on rivers such as the Colorado and the Nile they can hold two or three times the annual flow. Yet they remain an essentially experimental technology. Their hydrological, ecological and social effects have been huge. But for many years their status as symbols of modernism insulated them from serious appraisal. Even in the 1990s, less than half of all proposed dams had an environmental impact appraisal before construction. Even less had the consent of the people they displaced. Only recently have serious steps been taken internationally to establish if their benefits outweigh the environmental, social and economic downsides.

The focal point for this re-evaluation has been the World Bank. In the second half of the 20th century, it spent an estimated $75 billion (£40 billion) on building large dams in 92 countries. But by the late 1980s its own cost-benefit analyses questioned the value for money of this investment. Amid rising global opposition, the bank pulled out of funding a high-profile dam on the River Narmada in India. And in a quandary over how to proceed, it appointed a World Commission on Dams to assess the successes and failures of large dams, and to come up with ground rules for what a successful dam project might look like. The commission's final report, launched in a blaze of publicity in London in 2000, was even more scathing about large dams than the bank could have feared at the outset.

The commissioners endorsed many of the environmentalists' most trenchant criticisms. Most dams just do not deliver as advertised, they said. Average cost overruns were 56 per cent. Half of hydroelectric dams produced much less power than promised; two thirds of those built to supply cities delivered less water than promised; a quarter deliver less than half of the projected volume. Dams built to irrigate fields were no better. A quarter irrigated less than 35 per cent of the land intended. Even dams that promised to protect against floods "have increased the vulnerability of river communities to floods", often because their reservoirs have been kept full to maximise hydroelectric production.

Sanjeev Khagram has been at the heart of this revolution in thinking. He is a Harvard University academic of Ugandan Asian descent and has been a leading analyst of the growth of grass-roots opposition to dams in India.

His insight, in Dams and Development , into the campaign against the Narmada and Tehri dams in India is spot-on. He then considers how the Narmada campaign in India energised the major actors in the global movement against dams in Europe, North America, China and around the globe. Finally, he puts all this in the context of wider debates about the nature and purpose of economic development strategies and transnational governance.

But Khagram is not simply an observer. From 1998 to 2000 he was the senior policy adviser to the World Commission on Dams and author of much of its final report. Nobody was closer to the action. It is probably fair to say that no one else could have written this book. Certainly no one with such insight. So it is a shame that he does not offer his views on whether the World Bank and others have learnt the lessons he laid out so skilfully in the commission's report. The bank is today back in the dam-building business, putting its money into large dams from Uganda to Laos. It claims to be following a new agenda based on seeking consent from locals. But opponents beg to differ. Where does Khagram stand?

Managing Water Resources spreads its net wider than dams. It is the product of a series of lectures at the Centre for Water Research at Oxford University, and covers among other topics the history of underground water abstraction, the privatisation of the English water industry, the prospects and mythology of water wars and the development of American water resources.

There is worrying evidence that underground water reserves in arid areas - mostly formed during wetter eras thousands and sometimes tens of thousands of years ago - are being rapidly depleted. "Groundwater, emerging as springs, touches deep roots in human consciousness," says Mike Edmunds of the Oxford Centre in the opening chapter. Virtually all ancient civilisations write about the life-giving properties of springs. Their loss "still stirs passions", he says.

And virtually all the world's oases are running dry. The sands of the Gulf region of Saudi Arabia, Kuwait, Oman and their neighbours receive around 6km3 of new water a year in rainfall; but pumping, mostly for uneconomic agriculture, removes more than 20km3 annually.

A brief but thorough study by Peregrine Horden of the history of water use in the Mediterranean - from the irrigation that underpinned ancient Egyptian society on the Nile to Roman aqueducts and the infiltration galleries and agricultural terraces of medieval Spain - is testimony to the central place of water in most successful societies.

From these perspectives, the era of big dams from the 1930s to the 1980s may well, as Martin Reuss puts it in his look at American water resources, come to "be seen as a blip on the screen. Practically, only so many dams can be built; the reservoirs behind US dams store about 60 per cent of the entire average annual river flow of the country". As Beard puts it: "In America, and soon round the world, the era of big dam building is over."

Fred Pearce is an environmental writer and the author of The Dammed .