The icon indicates free access to the linked research on JSTOR.

The stories are widespread across cultures: a few hours before an earthquake, all the animals start acting agitated. Dogs howl for no reason, horses buck, fish swim in circles. Some stories are so deeply embedded that they become adages—think rats abandoning a sinking ship. But can animals really sense oncoming disasters?

JSTOR Daily Membership AdJSTOR Daily Membership Ad

Stories of animals behaving erratically before earthquakes have circulated for thousands of years, dating as far back as ancient Greece. A notable example was the 1975 evacuation of Haicheng, China purely on the basis of animals’ strange behavior, an act believed to have saved thousands of lives when a magnitude 7.3 quake struck shortly afterward. In the wake of the devastating 2004 Indian Ocean tsunami, widespread reports had animals fleeing inland prior to the wave’s impact.

Such widespread reports led researchers from the University of California system to study the possible use of animals as quake predictors. In hundreds of interviews with animal owners, most reported odd behavior prior to a tremor. But most of the positive reports of predictive behavior were recorded after the earthquakes hit. When only those instances recorded before a quake were included, the odd behavior was not statistically associated with earthquakes. People were likely viewing normal behavior as odd after a quake occurred.

Selective memory also seems to be in play when the tsunami reports are examined. Despite peoples’ reports, radio-collared elephants in Sri Lanka showed no such movements, sometimes even moving closer to the coast. They only moved inland around the time the wave hit, suggesting that the elephants reacted to the impact rather than anticipated it. It is likely that the fleeing animals were observed after the tsunami hit, and then placed earlier by selective memory.

Additionally, researchers found no plausible mechanism by which the animals might detect earthquakes. Earthquake waves travel faster than the sound, so there’s no real way animals could hear them. Perhaps animals are detecting weak foreshocks, but these should be detected by seismographs. Shifts in magnetic fields have also been occasionally detected in association with earthquakes but there is no evidence yet that animals react to these.

Still, the idea persists. At least one Chinese city installed 24-hour surveillance in a snake farm to detect strange behavior. Japanese authorities continue to conduct experiments with catfish. The anecdotes of animals’ uncanny predictions have clearly captured people’s imaginations, even if the research doesn’t quite bear out the claim.


JSTOR is a digital library for scholars, researchers, and students. JSTOR Daily readers can access the original research behind our articles for free on JSTOR.

ASEE Prism, Vol. 16, No. 7 (MARCH 2007), p. 15
American Society for Engineering Education
Biotropica , Vol. 38, No. 6 (Nov., 2006), pp. 775-777
Association for Tropical Biology and Conservation