ARecently, winter floods on the rivers of central Europe have been among the worst for 600 to 700 years, and dams and dykes (protective sea walls) have failed to solve the problem. Traditionally, river engineers have tried to get rid of the water quickly, draining it off the land and down to the sea in rivers re-engineered as high-performance drains. But however high they build the artificial riverbanks, the floods keep coming back. And when they come, they seem to be worse than ever.

BEngineers are now turning to a different plan: to sap the water's destructive strength by dispersing it into fields, forgotten lakes and flood plains. They are reviving river bends and marshes to curb the flow, and even plugging city drains to encourage floodwater to use other means to go underground. Back in the days when rivers took a winding path to the sea, floodwaters lost force and volume while meandering across flood plains and inland deltas, but today the water tends to have a direct passage to the sea. This means that, when it rains in the uplands, the water comes down all at once.

CWorse, when the flood plains are closed off, the river's flow downstream becomes more violent and uncontrollable; by turning complex river systems into the simple mechanics of a water pipe, engineers have often created danger where they promised safety. The Rhine, Europe's most engineered river, is a good example. For a long time engineers have erased its backwaters and cut it off from its plain. The aim was partly to improve navigation, and partly to speed floodwaters out of the Alps and down to the North Sea. Now, when it rains in the Alps, the peak flows from several branches of the Rhine coincide where once they arrived separately, and with four-fifths of the Lower Rhine's flood plain barricaded off, the waters rise. The result is more frequent flooding and greater damage. The same thing has happened in the US on the Mississippi river, which drains the world's second largest river catchment into the Gulf of Mexico. Despite some $7 billion spent over the last century on levees, the situation is growing worse.

DSpecialists in water control now say that a new approach is needed — one which takes the whole landscape into consideration. To help keep London's feet dry, the UK Environment Agency is reflooding 10 square kilometres of the ancient flood plain of the River Thames outside Oxford. Nearer to London, it has spent £100 million creating new wetlands and a relief channel across 16 kilometres of flood plain. Similar ideas are being tested in Austria, in one of Europe's largest river restorations to date. The engineers calculate that the restored flood plain of the Drava River can now store up to 10 million cubic metres of floodwater, and slow down storm surges coming out of the Alps by more than an hour, protecting towns not only in Austria, but as far downstream as Slovenia and Croatia.

EThe Dutch, for whom preventing floods is a matter of survival, have gone furthest. This nation, built largely on drained marshes and seabed, has had several severe shocks in the last two decades, when very large numbers of people have had to be evacuated. Since that time, the Dutch have broken one of their most enduring national stereotypes by allowing engineers to punch holes in dykes. They plan to return up to a sixth of the country to its former waterlogged state in order to better protect the rest.

FWater use in cities also needs to change. At the moment, cities seem to create floods; they are concreted and paved so that rains flow quickly into rivers. A new breed of 'soft engineers' wants cities to be porous. Berlin is one place where this is being done. Tough new rules for new developments mean that drains will be prevented from becoming overloaded after heavy rains. Architects of new urban buildings are diverting rainwater from the roofs for use in toilets and the irrigation of roof gardens, while water falling onto the ground is collected in ponds, or passes underground through porous paving. One high-tech urban development can store a sixth of its annual rainfall, and reuse most of the rest.

GCould this be expanded to protect a whole city? The test case could be Los Angeles. With non-porous surfaces covering 70% of the city, drainage is a huge challenge. Billions of dollars have been spent digging huge drains and concreting riverbeds, but many communities still flood regularly. Meanwhile this desert city ships water from hundreds of kilometres away to fill its taps and swimming pools. Los Angeles has recently launched a new scheme to utilise floodwater in the Sun Valley section of the city. The plan is to catch the rain that falls on thousands of driveways, parking lots and rooftops in the valley. Trees will soak up water from parking lots; houses and public buildings will capture roof water to irrigate gardens and parks, and road drains will empty into old gravel pits to recharge the city's underground water reserves. Result: less flooding and more water for the city. It may sound expensive, until we realise how much is spent trying to drain cities and protect areas from flooding, and how little this method achieves.

Over the past 50 millennia, Australia has witnessed the extinction of many species of large animals, including a rhinoceros-sized wombat and goannas the size of crocodiles. Debate about the possible cause of these extinctions has continued for more than 150 years and one of the crucial questions raised is how long humans and megafauna coexisted in Australia. We need to know the overlap of time to make an informed choice between the two main theories regarding the causes of these extinctions. If humans and megafauna coexisted for a protracted period then climate change is the more likely cause. However, if the megafauna became extinct shortly after the arrival of humans, then humans are the likely culprits.

The archaeological site at Cuddie Springs in eastern Australia appears to be well preserved. This dusty claypan holds within its sediments a rich cache of flaked stone and seed-grinding tools, and side by side with these clear signals of human culture are the bones of a dozen or more species of megafauna. Drs Judith Field and Stephen Wroe of the University of Sydney, who excavated the site, claim that it provides unequivocal evidence of a long overlap of humans and megafauna, and conclude that aridity leading up to the last Ice Age brought about their eventual demise. In the long-standing explanation of this site, artefacts such as stone tools and extinct animal remains were deposited over many thousands of years in an ephemeral lake — a body of water existing for a relatively short time — and remained in place and undisturbed until the present day.

There is no disputing the close association of bones and stones at Cuddie Springs, as both are found 1 to 1.7 metres below the modern surface. The dating of these layers is accurate: ages for the sediments were obtained through radiocarbon dating of charcoal fragments and luminescence dating of sand grains from the same levels. Intriguingly, some of the stone show surface features indicating their use for processing plants, and a few even have well-preserved blood and hair residues suggesting they were used in butchering animals.

But is the case proposed by Field and Wroe clear-cut? We carried out a reanalysis of the scientific data from Cuddie Springs that brings into question their conclusions. The amount of anthropological evidence found at the site is remarkable: we estimate there are more than 3 tonnes of charcoal and more than 300 tonnes of stone buried there. Field and Wroe estimate that there are approximately 20 million artefacts. This plethora of tools is hard to reconcile with a site that was only available for occupation when the lake was dry. Furthermore, no cultural features such as oven pits have been discovered. If the sediment layers have remained undisturbed since being laid down, as Field and Wroe contend, then the ages of those sediments should increase with depth. However, our analysis revealed a number of inconsistencies.

First, the charcoal samples are all roughly 36,000 years old. Second, sand in the two upper levels is considerably younger than charcoal from the same levels. Third, Field and Wroe say that the tools and seed-grinding stones used for plant and animal processing are ancient, yet they are very similar to implements found elsewhere that were in use only a few thousand years ago. Also interesting is the fact that a deep drill core made a mere 60 metres from the site recovered no stone artefact or fossil bones whatsoever. These points suggest strongly that the sediments have been moved about and some of the old charcoal has been re-deposited in younger layers. Indeed, one sample of cow bone found 1 metre below the surface came from sediments where charcoal dated at 6,000 and 23,000 years old is mixed with 17,000-year-old sand. The megafauna bones themselves have not yet been dated, although new technological developments make this a possibility in the near future.

We propose that the archaeologists have actually been sampling the debris carried by ancient flood channels beneath the site, including charcoal transported from bushfires that intermittently occurred within the catchment. Flood events more likely explain the accumulation of megafauna remains, and could have mixed old bones with fresh deposits. European graziers also disturbed the site in 1876 by constructing a well to provide water for their cattle. Given the expense of well-digging, we speculate that the graziers made sure it was protected from the damage caused by cattle hooves by lining the surface with small stones collected from further afield, including prehistoric quarries. This idea is consistent with the thin layer of stones spread over a large area, with cattle occasionally breaking through the gravel surface and forcing the stone and even cattle bones deeper into the waterlogged soil.

The lack of conclusive evidence that humans and megafauna coexisted for a lengthy period casts doubt on Field and Wroe's assertion that climate change was responsible for the extinction of Australia's megafauna. However, we do not suggest that newly arrived, well-armed hunters systematically slaughtered all the large beasts they encountered. Recent studies based on the biology of modern-day large mammals, combined with observations of people who still practise a traditional hunter-gatherer lifestyle, reveal an unexpected paradox and suggest a further possible explanation as to what happened. Using a mathematical model, it was found that a group of 10 people killing only one juvenile Diprotodon each year would be sufficient to bring about the extinction of that species within 1,000 years. This suggests that here, as in other parts of the world, the arrival of humans in lands previously inhabited only by animals created a volatile combination in which large animals fared badly.