Il y a 4000 ans, l'humanité a vécu une collision entre la Terre et un gros corps céleste. Un véritable cataclysme dont on pensait qu'il ne s'était pas produit depuis la lointaine préhistoire ! C'est la théorie que défend aujourd'hui la Française Marie-Agnès Courty. Au terme de plus de quinze ans d'enquête minutieuse, cette géologue du CNRS est parvenue à reconstituer le scénario suivant : il y a environ 4000 ans, un bolide venu de l'espace (astéroïde ou comète) d'environ 1 km de diamètre aurait percuté l'hémisphère Sud, près des îles Kerguelen. Sous la violence de l'impact, des millions de fragments de la croûte océanique ont été catapultés dans l'espace. À peine quelques heures plus tard, ils ont retraversé l'atmosphère et sont tombés sur l'ensemble de la planète en une pluie de météores incandescents. Marie-Agnès Courty a reconstitué ce scénario à mesure qu'elle a découvert, dans des échantillons de sol du monde entier datant de 4000 ans, les indices de la catastrophe. Dont le plus troublant : des fossiles d'organismes marins en provenance des hautes latitudes australes, retrouvés sur les terres émergées du Moyen-Orient, de l'Europe ou de l'Asie centrale.

À l'époque où se situent les faits, la Terre était déjà peuplée de 30 millions d'êtres humains. En Égypte, la pyramide de Gizeh venait d'être achevée, l'empire d'Akkad régnait sur la Mésopotamie, les premières grandes dynasties se succédaient en Chine et l'Europe était organisée en de nombreuses tribus. En marquant en profondeur la mémoire collective, un tel cataclysme aux accents de fin du monde est peut-être à l'origine de mythes célèbres, comme ceux de l'Apocalypse ou de Sodome et Gomorrhe.

Days of heavy rains have burst river banks, isolated many communities and made helicopters and boats the only way of delivering relief supplies.

The government's Meteorological Department issued a warning that there will be heavy monsoon rains until Sunday.

Malaysia's southernmost state, Johor, has been the hardest hit, with 50,000 of the 60,000 evacuees, the government's website said. People are being housed in 300 flood-relief centres.

All six of the dead were in Johor. The website describes how a truck driver saw two people in a car being swept away.

The bodies were later recovered but "we could not launch a rescue operation earlier as the current was too swift," Fire and Rescue Department operations commander Hazmi Ali said.

"The evidence from 1918 is pretty impressive that in poor communities the death rate was just extraordinary. This is a remarkable increase in mortality in just a short period of time," said lead author Dr. Chris Murray, director of Harvard University's Initiative for Global Health.

The study, funded by the U.S. National Institute of Aging, will appear in Saturday's issue of The Lancet. Murray's co-authors are from Harvard, the University of Queensland, Australia and the Johns Hopkins Bloomberg School of Public Health in Baltimore.

The head of the World Health Organization's global influenza program admired the work, but said while it is an interesting exercise, it doesn't really change the task facing those working to prepare the world to weather a future pandemic.

"In terms of disease control and pandemic response and those sorts of thing, the bottom line is no matter what approach you take, the figures are pretty big and the steps needed to prepare for it are substantial," Dr. Keiji Fukuda said from Geneva.

Underestimating globalization

A U.S. infectious diseases expert challenged the paper's suggestion that the developed world would get off relatively lightly from a 1918-like pandemic, saying the study and an accompanying editorial underestimate the effect of globalization and overestimate the capacity of developed-world medical systems to cope with a crush of gravely ill people.

"The paper and the editorial have no sense at all of a modern global just-in-time economy, where the kind of drugs and medical services that they assume will be available in a modern world just won't be there," said Dr. Michael Osterholm, director of the Center for Infectious Diseases Research and Policy at the University of Minnesota.

But the author of the editorial, Dr. Neil Ferguson of Imperial College in London, supported the paper's assertion that the developing world would take a far bigger hit and contested Osterholm's prediction that supply chain disruptions would amplify the impact of a pandemic in developed countries.

"A pandemic is a relatively short-lived thing," said Ferguson, a leading mathematical modeller.

"There may be crises over a few weeks, but I would be skeptical whether we would face famine and the sort of food shortages and other problems that would lead to dramatically enhanced mortality. I just don't find it plausible."

Mortality estimates vary widely

Osterholm had earlier predicted a 1918-like pandemic could kill between 180 million and 360 million worldwide.

Those numbers, and others, are what inspired Murray and his colleagues to do this work. They had attended meetings of influenza experts where death projections were raised. The estimates varied wildly - from a low of two million to 360 million to even one billion people.

Murray said there seemed to be little scientific basis to those projections, so he and his colleagues set out to see if they could come up with an evidence-based estimate.

They hunted for death registries from the time, finding good records for 27 countries. In the case of India and the United States, they were even able to mine data from nine provinces and 24 states respectively.

In each case, they looked at nine years of deaths - the three years before 1918, the three years during the pandemic and the three after. They averaged the years before and after to come up with a standard death rate for that jurisdiction, then attributed any excess deaths that occurred in 1918-1920 to pandemic influenza.

They made exceptions for Britain and France to factor out the death toll of the First World War, and in Finland to account for a civil war that overlapped the end of the Spanish flu. Murray said similar adjustments were not made for Commonwealth countries like Canada, Australia and New Zealand that also fought in the First World War.

They then used the excess mortality rates - which varied from a low of 0.2 per cent in Denmark to a high of 7.8 per cent in parts of India - to calculate what a pandemic of similar virulence would do to the current global population.

Murray admitted he and his co-authors were startled by the projection they arrived at; they were expecting something more in the range of 10 million excess deaths.

He said the research points to a need to start looking for realistic ways cash-strapped countries can try to minimize the death toll, given that expensive antiviral drugs and scarce flu vaccine are likely to be unavailable except in the richest countries.

Consider what a significant disruption of a life pattern this is. Hibernation has evolved for the same reason most animal behaviour has evolved: as a strategy to maximise survival. Some creatures that need a lot of energy to get around have learnt to shut themselves down in the winter months, when the food to provide that energy is simply not available (or too much energy would be expended in searching for it). Zoologists have realised in recent decades that many species have an instinctive and finely tuned way of weighing up the balance between how much effort needs to be expended to acquire a certain food item, and how much energy is available, in return, in the item acquired. The general law is: if the second is less than the first, don't do it. This has been christened "optimal foraging".

Hibernation could be seen as a version of this: if the food search is going to be hopeless, it makes sense to stop foraging altogether. Instead, fatten yourself up before the hopeless time, then sleep it out. This is a strategy that has evolved - in bears, hedgehogs, bats and other species - over millions of years and it has persisted as a piece of behaviour because it has been successful.

If some European brown bears in the Cantabrian mountains are now stopping hibernation, we can draw two conclusions. First, something quite enormous is happening in the world around them, and if you want to hazard a guess that that something is global warming, you would have as good a theory as any other.

Second, they are abandoning a survival strategy - which has been successful - for the unknown. What if they give up hibernation because of rising winter temperatures, but then when they are active in winter, are unable to find enough food?

We are already witnessing what a problem such disruption of natural cycles can cause for other creatures. In Britain, insects are hatching earlier in the spring, but migratory birds that depend on the "flush" of caterpillars to feed their young are coming back at the same time as they have always done, and thus may be starting to miss out on the feast: this may be one of the reasons why many of our woodland birds are now in sharp decline.

Climate change is perceived as a terrible problem for human society, and rightly so; but we should not lose sight of the fact that, to the natural world and its inhabitants, the warming also presents a mortal predicament.