Extinction rates overestimated: study

Ecologists routinely overestimate the rate at which animal and plant species go extinct as their habitat is destroyed because a widely used calculation method is flawed, a Canadian researcher says.

Fangliang He, an ecology researcher at the University of Alberta, Edmonton, led a study that shows that the most common method used to estimate extinction rates is based on the wrong kind of data.

"It's flawed, really," He said at a press briefing organized by Nature, which published his research Wednesday.

In their paper, He, who is on sabbatical leave at Sun Yat-sen University in Guangzhou, China, and co-author Stephen Hubbell of the University of California in Los Angeles, demonstrated at least one example where the extinction rate calculated using the common method was 2.5 times larger than an estimate using the right kind of data.

Hubbell believes other estimates frequently cited by scientists and conservationists may be even bigger overestimates.

"Clearly, this is welcome news in the sense [that] we have bought a little time to save species," he said at the press briefing. "But it's unwelcome in one sense because we have a huge amount of research that was done incorrectly as a result of the incorrect method."

The paper emphasized that extinction due to habitat loss is still a real and growing threat, even if the rate is lower than previous estimates.

The "flawed" estimation method uses data on the number of species found in areas of different sizes to create a relationship called a "species-area relationship." Then it extrapolates backwards to smaller areas to figure out how many species will be lost as their habitat shrinks due to logging, farming and other types of habitat destruction.

Known problem, unknown cause

Ecologists had long realized that the resulting number was an overestimate, but for years nobody could figure out why, He said. Instead, starting about 25 years ago, they assumed that the difference between the estimate and the actual extinctions was an "extinction debt" that meant some species were on their way to extinction, but not there yet.

The species-area relationships are constructed by adding a species to the data each time a scientist encounters the first individual of a species within an area of a given size. The process is repeated for areas of different sizes to figure out how the number of species grows as the area increases.

But for the species to go extinct, the last individual of a species must be eliminated — something that can only be confirmed by sampling a much bigger area.

It took He and Hubbell eight years to figure out what the problem was.

"What this shows is that many scientists can be led away from the right answer by thinking of the problem in the wrong way," Hubbell said.

The researchers found publicly available species-area data, and used it to calculate extinction rate estimates using the traditional method. Then they recalculated, taking into account the probability of encountering the last member of a species if individuals are randomly and independently distributed. That generated much smaller extinction rates, suggesting the traditional method produces overestimates on the order of 40 to 160 per cent.

"Those numbers are very questionable," He said. "They deserve very careful scrutiny."

The researchers said that for species that aren't evenly distributed, information about their distribution and specifics about how the landscape is destroyed is needed to create true estimates.

'New' method already used: biologist

He and Hubbell specifically questioned the numbers in the 2005 Millennium Ecosystem Assessment, an international study by 1,000 scientists to analyse the state of the Earth's ecosystems and in a 2004 paper in Nature estimating species loss due to climate change.

Biologist Chris D. Thomas, who led the latter study when he was at the University of Leeds in the U.K., said in an email that he welcomes further development of analyses to assess extinction risks.

But Thomas, who is now at the University of York, said his paper actually uses the method He and Hubbell propose, estimating that five to 16 per cent of species would lose 100 per cent of their "climactically suitable" habitat by 2050. The number was published alongside the estimate, based on the species-area method, that 18 to 35 per cent of species could go extinct by 2050.

"We published the result seven years before He and Hubbell guessed this might be the case," he said. "It is a pity the authors did not realize this."

He and Hubbell's research was supported by Sun Yat-sen University, Canada's Natural Sciences and Engineering Research Council, NASA and the U.S. National Science Foundation.