Drought hit the Amazon last year, with devastating results. Rivers fell, fish rotted and routes to schools and hospitals were cut off. Now studies of the dry spell suggest that such conditions could become increasingly common.

The drought first caught scientists' interest because its cause was unusual. Dry spells in the Amazon usually occur in El Niño years, when warm water off the Pacific coast of South America sets up a pattern of circulating air that inhibits rainfall in the Amazon. But last year was not an El Niño year.

Instead, the drought was caused by a circulation pattern powered by warm seas in the Atlantic — the same phenomenon responsible for last year's unusually intense Atlantic hurricane season. The result was a dry spell that hit particularly hard in the western Amazon, a region that normally has more rainfall than other parts of the forest.

“It caught people by surprise,” says William Laurance, an ecologist with the Smithsonian Tropical Research Institute, based in Balboa, Panama. “We hadn't seen a drought like that before.”

How will the rainforest respond to drought? Credit: A. LIMA/EMPICS/AP

The media quickly blamed global warming, but climate researchers warn that it may not be possible to make such a link — in part because vast areas of Amazonia have no weather stations.

But early results from studies suggest that the Amazon could have more such events in the future. In a paper to be finalized this month, researchers from the United Kingdom and Brazil use a climate model developed at the Hadley Centre for Climate Prediction and Research in Exeter, UK, to study whether the 2005 event was a one-off or a taste of things to come.

The team says the model predicts the latter. “The 2005 situation will be more frequent by 2050,” says co-author Jose Marengo of the Center for Weather and Climate Prediction in Cachoeira Paulista, Brazil.

Last year's drought also bolsters a controversial 2004 finding about the future of the Amazon rainforest. Peter Cox of the Centre for Ecology and Hydrology in Winfrith, UK, who is also an author of the new paper, had predicted that more frequent droughts would wipe out around 65% of the Amazon forest cover by 2090 (P. M. Cox et al. Theor. Appl. Climatol. 78, 137–156; 2004). Many questioned the result when it was published, in part because some of the droughts predicted by the model were caused by warm waters in the Atlantic — a phenomenon that hadn't been observed at the time. Now that such an event has been seen, says Laurance, researchers are looking at Cox's results with a little less scepticism.

Other studies disagree, predicting less drought and forest destruction than Cox's model. But more and more evidence suggests that drought will take a serious toll on the Amazonian forest.

Daniel Nepstad, an ecologist at the Woods Hole Research Center in Massachusetts, has been simulating drought since 1998 with a plastic canopy covering 10,000 square metres of forest in the eastern Amazon. His results show that during drought, more large trees die off, all trees produce less wood, and there are more fires. In a paper under review at Ecology, Nepstad calculates that the combination of dead trees and reduced wood formation in Amazonia as a whole led to an extra half-a-billion tonnes of atmospheric carbon in 2005. In contrast, the Kyoto Protocol on Climate Change aims to reduce global carbon emissions by a billion tonnes annually by 2012.

Nepstad is now lighting test fires to assess whether one fire makes the forest more vulnerable to others. “Our biggest fear is that the forest will be invaded by highly flammable grasses,” he says.