Can animals sense earthquakes?

Dear Cecil: Can animals sense earthquakes? I've heard stories about cats carrying their litter of kittens out of a building just prior to a quake. Is there any truth in it? We've had two quakes in the last three hours, and my goldfish didn't do a thing.
— Nick E, Manchester, England

Cecil replies:

Of course animals can sense earthquakes. Human beings can sense earthquakes. Some years ago a friend of mine in California walked out the door and saw the parking lot in front of him rippling like a flag in a stiff breeze. Without hesitation he concluded: This is an earthquake. The real feat would be sensing a quake an appreciable time in advance. Scientists haven't given up hope of finding animals capable of this, with a view to figuring out how to do it themselves. But so far it's been a slow job.

The idea that animals act funny before an earthquake hits has been out there for thousands of years. However, even if we accept that it's not out of the question, we have to ask: How would this work? What would the animals be picking up on? Some possibilities, as discussed in a 2000 article in the Bulletin of the Seismological Society of America:

• P-waves — the fast-traveling primary seismic waves that show up just before things start getting tossed around. Humans aren't good at sensing these, but animals might be. "Numerous observations exist of animals displaying panic in the few seconds prior to the onset of strong ground shaking . . . including dogs barking, nervous cats jumping out of windows, birds screaming, rats running out of their holes, bees swarming, etc.," we read. Obvious problem: all a few seconds gives you is a head start on freaking out.
• Ground tilt, which supposedly has been detected hours prior to some Asian quakes. But the tilt angles are microscopic, and thus far there's no hard evidence animals can perceive such tiny shifts.
• Humidity changes. Rising groundwater levels have been detected before some quakes, leading some to conjecture that more moisture would be forced into the air, where animals could notice it. Even in explanation this sounds tortured, and it's hard to imagine how a critter could recognize quake-driven humidity amid normal variation.
• Electromagnetic field fluctuations. A lot of people favor this angle, as it's been proven earthquakes can generate detectable electromagnetic activity right before they happen. ("Earthquake lights," the creepy bluish-white flashes that sometimes appear over fault lines before a quake, may be a related phenomenon.) While land animals seem to be fairly oblivious to electrical activity, fish can be attuned to it, and it's possible magnetosensitive animals such as birds and bees could pick up on it too.
• Low-frequency sound. This one is easy to believe, given how much better animal hearing is than ours. One happenstance observation in the Mojave Desert showed dogs could detect low-frequency "booms" from earthquake aftershocks that a seismometer registered but researchers didn't notice

— Cecil Adams

From: http://www.straightdope.com/