Extinction 'Tipping Points' Possibly Predictable

Thousands of plants and animals worldwide are listed as threatened or endangered, but the point of no return for these diminishing populations has been impossible to predict. A new study suggests a way to determine when extinction becomes inevitable.

If the findings from a laboratory experiment prove applicable in nature, they could help ecologists step in to save species before it's too late, researchers say. For now, the study is the first step in moving a mathematical theory into the real world, where endangered species are vanishing at a rate that may range from 10 to 100 times the so-called background extinction rate. [Read "Mass Extinction Threat: Earth on Verge of Huge Reset Button?"]

Statistical harbingers

When ecologists model the decline of species (a computer simulation of sorts), they see tipping points — sets of circumstances that make extinction all but certain.

To date, mathematical modeling has revealed a few statistical harbingers of tipping points. Right before a system reaches the point of no return, it goes through a phase called "critical slowing down." That phase is the statistical equivalent of the gut feeling you may experience right before a canoe tips over or a rollercoaster makes a plunge: that the system can no longer recover from perturbations in the environment (like your last-minute attempt to balance the canoe), and a dramatic change is imminent.

In nature, those perturbations might be small changes in temperature or precipitation, or simple normal fluctuations in how many offspring a species produces.

"The ability of the system to respond to perturbations, to these little nudges, is diminished," study researcher John Drake, an ecologist at the University of Georgia, told LiveScience. "So lots of little nudges accumulate, and that's what we call critical slowing down."

To find out if critical slowing down can predict extinction in real-world ecosystems, Drake and Blaine Griffen of the University of South Carolina used millimeters-long crustaceans called water fleas. The tiny algae-eaters were split into two groups and fed until their population stabilized. After about 150 days, the researchers stopped feeding one of the groups.

Unsurprisingly, the starving water fleas struggled to survive. By day 416, all populations in their group were extinct. By analyzing the population fluctuations as the water fleas slid toward extinction, the researchers found that critical slowing down did occur. In fact, the statistical warning signs of extinction showed up eight generations, or 110 days, before the last water fleas perished.

From laboratory to field

Translating the laboratory results to the field is likely to be difficult. Natural systems are much more complex than a limited number of water fleas in a controlled lab setting. And careful monitoring and analysis will be necessary to get the data that might be used to predict extinction.

Even if extinction can be predicted, ecologists would need to figure out how to reverse the problem in lots of different ecosystems.

"A great deal of system-specific knowledge is going to be needed to apply these things in any sort of real-world setting," said ecologist Stephen Carpenter, the director of the Center for Limnology at the University of Wisconsin. "That's not a criticism, it just says we have more work to do."

The fact that Drake and Griffen were able to demonstrate the statistical precursors to extinction in living organisms "adds momentum" to the idea of replicating the results in the field, said Carpenter, who was not involved in the study.

Drake agrees.

"Our contribution was to experimentally demonstrate critical slowing down in a biological population," he said. "Now it remains to see whether we can scale that up to applications in nature."