Can Animals Detect Earthquakes?

Over at the always-excellent Discover magazine blog 80 Beats, a fascinating post from a couple of days ago on a topic that has fascinated natural historians for several thousand years now: can animals’ strange behavior provide forewarning that an earthquake is about to strike?

I’ve been reporting a story about earthquake prediction for Parade magazine lately, and spent a day talking with some of the nation’s leading seismologists at Caltech. I came away with a renewed appreciation for the difficulty of the scientists working toward that holy grail of earthquake research, figuring out how to predict the behavior of faults that lie invisible deep beneath the surface of the earth (and which may be fundamentally chaotic in their nature anyway).

All the more intriguing, then, to hear that biologist Rachel Grant, while studying the mating behavior of toads in Italy, may have stumbled upon the first scientific evidence of animals being able to foresee what seismologists could not:

Her team was studying common toads in Italy in April 2009 when the amphibians began to disappear from the study site. This didn’t make much sense to her, the toads abandoning a breeding site in the midst of breeding season. So the researchers tracked them. They found that 96 percent of males — who vastly outnumber females at breeding spots — abandoned the site, 46 miles (74 kilometers) from the quake’s epicenter, five days before it struck on April 6, 2009. The number of toads at the site fell to zero three days before the quake. Grant says her initial reaction to the mass toad dispersal was annoyance—their flight was holding up her research. However, when they began to return the day after the earthquake, things began to make more sense.

It’s an amazing story, a seemingly irrefutable case of cause and effect. But I’m still not buying it. Here’s why.

As geologist Susan Hough explains in her delightful book Explaining the Unexplainable, scientists have been trying to chase down the link between earthquakes and animal behavior for years now. The most intriguing potential link turned up during the Maoist era in China, after the Haicheng region was struck with a series of earthquakes. Concerned that a really big one might be on the way, the authorities enlisted the peasantry in a mass effort to look for clues. Sure enough, reports began trickling in of strange behavior: “Records indicate that snakes and frogs reportedly came out of the ground, something snakes and frogs are not supposed to do during the winter hibernation season,” Hough writes.

The big quake did come, a magnitude 7.3 tremblor that struck Haicheng on February 4, 1975. Because many had been warned that a quake was on its way, fatalities were low.

In China, the successful forecast was taken as a great victory for the People’s Seismology, but from a modern perspective it doesn’t quite hold up. For one thing, along with the reports of strange reptilian behavior there were also numerous small quakes that are a more traditional indicator that a bigger quake might happen.  Hough writes:

The ground was not exactly crawling with snakes, but there were almost one hundred documented snake sightings in the Haicheng region in the month prior to the earthquake. How does one account for this? The simplest explanation is that they were disturbed by foreshock activity, including shocks too small for humans to feel. Or perhaps the reported sightings reflected only a sort of mass hysteria on the part of those doing the sightings. There is no question that the mind is capable of extraordinary invention, if suitably primed. But neither explanation sits entirely well. The number of reports… did not track either the earthquake activity or the sense of urgency among local people and officials. Either explanation could be right; the truth is, we just don’t know.

Personally, I think that Hough is being excessively even-handed in her reluctance to discount both theories. Both seem to be clearly incorrect. A snake that crawls out of its hole a full month before an earthquake, in the dead of winter, is simply doing a terrible job. And while reptiles may be able to sense vibrations in the earth that humans cannot, I doubt that they are more sensitive than seismometers.

I don’t think it was  a case of mass hysteria, either. Not exactly. But if you ask a broad population of people to look for something — especially if you are the Communist Chinese government at the height of the Cultural Revolution — a certain number of them will find it. Back in the mid-’90s, I traveled to Vietnam to do a story about field biologists working on the Laotian border. They wanted to know what birds were indigenous to the area, so they gave one of the local Vietnamese biologists a list of potential species and asked him to check off the ones he saw. A month later, when he turned in his list, they found that he had checked off every single species. In his mind, the best possible accounting of the bird population was one which showed the maximum number of species. I imagine something similar might have happened in Haicheng: asked to report anomalous behavior, the peasants reported anomalous behavior.

Follow me on Twitter: @extremefear.

3 thoughts on “Can Animals Detect Earthquakes?”

  1. Hi Jeff,

    I feel it necessary to critique your post – please don’t take it personally.

    You give a favorable comment to the Discover article, which opens with this:

    “The wave of high-profile seismic activity so far in 2010 has been another reminder that we humans could use all the help we can get in predicting earthquakes.”

    We do need all the help that we can get, and you are actually doing the opposite.

    You also stated this, about the peer reviewed paper by a zoologist:

    “It’s an amazing story, a seemingly irrefutable case of cause and effect. But I’m still not buying it. Here’s why”

    The problem is, that you then went off to examine other stories, rather than this controlled experiment, and the scientific publishing of its results.

    Susan Hough is a fine person, and respected seismologist. You also question her:

    “Personally, I think that Hough is being excessively even-handed in her reluctance to discount both theories. Both seem to be clearly incorrect.”

    Yet, what have you examined? An excerpt from a pop-science book (meant for lay-people? Did you read the actual accounts of these events, by qualified scientists? It does not appear so, because you left out some of the most critical details, like how the animal behavior, just like this frog study, increased as the quake date approached, peaking just before. It is not as you summarized –

    “A snake that crawls out of its hole a full month before an earthquake, in the dead of winter, is simply doing a terrible job. And while reptiles may be able to sense vibrations in the earth that humans cannot, I doubt that they are more sensitive than seismometers.”

    What does a seismometer have to do with this? Do you have some evidence that physical vibration was the key factor? The authors of the paper did not make this assumption.

    (just FYI, sensitivity in seismometers is a tricky trade off between measuring the smallest vibrations – called Earth Hum, which is non-stop – and large magnitude events, like the 8.8 that happened last month in my back yard.)

    There is more to this than meets the eye, and I don’t think you’ve given it your adequate attention. You will have to, as this recent zoologist shows, go out of the area of seismology to get more information.

    I welcome you to a project thread, where there is a post about this. There are links to the papers that you can read further into this.

    http://saposjoint.net/Forum/viewtopic.php?p=25037#p25037

    I do find your cite interesting, and wish you the best!

    T.Roc

  2. Thanks, Thomas, I appreciate your reading my post and critiquing it so thoughtfully. Your points are well taken, but I remain a skeptic. The most telling part of Hough’s coverage, for me, was her statement that “The number of reports… did not track either the earthquake activity or the sense of urgency among local people and officials.” Obviously, I haven’t seen the original data, but the implication is that the animals’ strange behavior did not increase gradually and peak just before the event. Also, bear in mind that the more recent incident in Italy was no a controlled experiment, but a single anecdotal observation, which as we know cannot be used in science to infer correlation, let alone causation.

  3. Hi again, Jeff –

    I’m glad that you’ve kept your skeptic title, I’m not suggesting that you change.

    My main reasons for responding here were that you mentioned that you plan to write a story about prediction, and that you had already interviewed a seismologist.

    As you mentioned above, we shouldn’t just take one report, and then make our decision. In fact, I would use that to counter-weigh the report that you mentioned – it is only one among hundreds, that extend back well over 2000 years.

    I would also contest that you state that this study by Rachel Grant was not a controlled experiment – what was it that they were doing there? Everything was done according to standards, and this is a peer reviewed publication, not a newspaper article. A group of toads was studied before, during, and after the period in question, and among other things, at no time, not during this group or in other reported studies, have these frogs stopped there mating process once it started.

    Is this correlation a causation? No, we agree on that. Was the study expressly done to study the response to an EQ? No, but that does not effect the results. There are countless other credible stories out there if you look – recorded in daily logs and even video cameras. Check some zoos or wildlife reserve records.

    However, really, this is a side issue. If you want to write about EQ prediction as a possibility, you have to understand the various precursors. There is overwhelming evidence that supports them, of which only a minor one is animal response.

    I have suggested in my thread, that an apparent connection is animal navigation via the Earths magnetic field, which is now well established for several species, and several well noted relations (as precursors) to variations in the Earths magnetic field prior to EQs.

    Keep in mind that over a million people have died in the last 50 years due to EQs. Any clues as to how to possibly prevent more like these is a very worthy cause. You will have to decide the slant of your story. If you want it to be balanced, you will have to seek information outside seismology.

    Happy Hunting!

    T.Roc

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.