After every seismic disaster, reports pour in of animals behaving strangely before the event.
Dogs seem to bark at nothing, cats appear unusually anxious and fearful, vermin leave their underground homes en mass, and bird colonies suddenly vanish.
But what do these behaviours mean? Are the animals reacting to other stimuli that just happen to coincide with earthquakes and volcanic eruptions? Or could they be sensing and reacting to changes in the Earth that are imperceptible to humans?
Dr. Rupert Sheldrake, a London-based biologist and author of the book Dogs That Know When Their Owners Are Coming Home says whatever the underlying reason for it, animal behaviour before seismic events is worth studying.
“I’ve been tracking most of the major earthquakes over the last 20 years, and the general pattern seemed to be unusual [animal] behavior,” he says. “It seems to me almost negligent that it hasn't been studied more by scientists so far.”
Sheldrake says China had modest success predicting earthquakes in the 1970s by collecting eyewitness reports of odd animal behaviour.
A team of British researchers also documented the strange case of toads in Italy that vanished from their traditional mating ground days before a large quake.
But Dr. Andrew Michael, a seismologist with the United States Geological Survey, says any connection between animal behaviour and seismic events is purely coincidental.
“If dogs can really sense a coming earthquake, why is it just a couple of dogs out of the millions of dogs in an area?” he says. “I think it tells us that dogs periodically act strange, and occasionally there's an earthquake. It doesn't tell me [the strange behaviours] were actually related to the earthquake.”
Michael believes developing more sensitive seismic instrumentation would be a better investment than studying animals, which respond to other stimuli in their environment.
“Our sensors are not also going to be sensitive to other animals in the area, food, prey, predators or the weather,” he says.
“We'll get a much more stable signal from building a better sensor than studying a bunch of animals that really care more about where their next meal is coming from.”
Sheldrake says it's not clear how or why animals seem to know when disaster is imminent, but he believes collecting and publicizing anecdotal reports of strange behaviour could be another tool to increase preparedness in earthquake and tsunami-prone areas.
“At the very least, I think we should explore it,” he says. “You'd have to do research to see what else is going on [that could explain the strange behaviour] ... but I think this could be a low-cost research program that could actually provide effective warnings.”