Farm animals may be able to sense earthquakes hours before they actually occur, according to new research that followed the movements of cows, chickens, and dogs during temblors in northern Italy.
“We saw a clear relationship that the closer [the earthquake] is to a farm, the longer the animals are excited before the earthquake,” said Martin Wikelski, managing director at the Max Planck Institute of Animal Behavior and a professor at the University of Konstanz in Germany.
Indigenous Knowledge and anecdotal evidence of animals being able to sense natural disasters have been around for millennia. In his Natural History, 1st-century Roman naturalist Pliny the Elder noted that birds on sea vessels showed strange behavior before impending earthquakes, and scientific explorer Alexander von Humboldt noted in his travel narrative that locals in Venezuela “attentively observe the motions of dogs, goats, and swine,” the last of which he said can sense quakes because of “delicate olfactory nerves.”
Such evidence is not relegated to history books. Prior to the Haicheng earthquake of 1975, for instance, Chinese authorities ordered the evacuation of the city at least in part because of observations of strange animal behavior. After the Indian Ocean earthquake and tsunami of 2004, Indonesian students learned to repeat songs instructing them that when animals run uphill en masse, students should drop everything and follow them because it could mean a tsunami is approaching, Wikelski said.
A Lucky Shake
But the idea that animals could sense these disasters had never been empirically proven. The problem is, to test whether animals can predict an earthquake in a given spot, scientists would need to know when an earthquake would actually occur in that given spot—and humans cannot predict earthquakes. Wikelski and his team opted for the next best thing: aftershocks.
They had their chance in October 2016, when a quake occurred in central Italy with a magnitude of roughly 5. The scientists rushed to the area and asked farmers whether they could outfit their animals with accelerometers.
Several farmers refused the request, but the researchers found a family in Capriglia who had already noticed their animals were more active during the recent earthquake. The family had even fed the animals wine and sugar to calm them down.
“[Farmers] trust some of their animals because they know they’re very sensitive,” Wikelski said.
The farmers identified animals that were particularly sensitive to quakes, and about 20 cows, dogs, sheep, rabbits, chickens, and turkeys were fitted with devices.
Although scientists had equipped the animals to test their ability to detect aftershocks from the M5 quake, an M6.6 earthquake happened just days after the first.
“We’re incredibly lucky that we have [data collected] before, during, and after a major earthquake,” Wikelski said.
“Stock Market Crash”
When scientists initially examined the data, published recently in Ethology, they didn’t notice anything unusual. But on a closer look, the animals’ movements in some areas were like mayhem—particularly those that were kept in stables, from which they couldn’t escape.
“When [farm animals] are not in the stable and out in the field, they don’t really seem to care,” Wikelski said.
But those that were in stables showed a different response. The farm kept some cattle chained in a stable, whereas at other times cattle were allowed to roam in the pasture. Sheep were generally allowed to freely move inside a separate stable.
“We see that there is a massive excess of activity of all the animals [in a stable],” Wikelski said. “If that is sustained for more than an hour, then we know something is wrong.” Wikelski noted that examining the data was like watching the financial markets for signs of impending trouble. When all animals start moving in different directions, the data appear like a “stock market crash.”
There were a number of aftershocks and earthquakes recorded after the animals had been fitted with accelerometers, and the researchers observed that the closer the animals were to the earthquake, the sooner they started to react. In a case where the epicenter was nearly below the farm, the animals showed reactions 15–18 hours beforehand.
“That’s a pretty decent warning time,” Wikelski said.
Sifting Through the Lore
Susan Hough, a seismologist with the U.S. Geological Survey (USGS) who was not involved in the study, said she really likes the approach the researchers took. “It’s a very old idea that animals sense earthquakes—that lore has been around forever,” she said.
She had heard anecdotes about animals but noted that many of these stories come after the fact—like people recalling that their cats were acting strangely right before the earthquake.
“People always start from the time of the earthquake and then go back,” she said. “The problem is, you can’t do science that way,” she said.
Regardless, the study’s results “aren’t overwhelming” for Hough, who added that the plots showing animal activity versus seismic activity aren’t exactly intuitive. “To my eye, the correlation doesn’t really jump out at me when I look at it,” she said.
Morgan Page, a geophysicist with USGS, also can’t see much of a pattern from their data. “It’s cool that someone actually tried to [document animal behavior prior to an earthquake],” she said, but the data look more randomly distributed to her.
“Even with sophisticated instruments, we have trouble detecting earthquakes,” Page noted.
Hough said she’d like to see more data before making conclusions, and Wikelski agreed.
“It’s always really important for everyone to know that this is all preliminary. We can’t use this as a predictive system yet,” he said, but added that after his team learned to read the data from the first few quakes, they managed to successfully predict a small (magnitude 3) quake that happened right below the farm.
“It’s as good a start as we can get,” he said.
—Joshua Rapp Learn (@JoshuaLearn1), Science Writer
Learn, J. R. (2020), Some farm animals might have a sense about impending earthquakes, Eos, 101, https://doi.org/10.1029/2020EO148162. Published on 24 August 2020.
Text © 2020. The authors. CC BY-NC-ND 3.0
Except where otherwise noted, images are subject to copyright. Any reuse without express permission from the copyright owner is prohibited.