I love this chart, from a 1981 review paper on whether animals can anticipate earthquakes.
It plots instances of odd behavior, for a multitude of species, based on time before a quake and distance from the epicenter. To me, it’s like a data-based graphic equivalent to herding cats. The caption:
Fig. 1. Distribution of animal behavior incidents according to the distance from the epicenter and the time before the main shock of 36 different earthquakes in Europe, Asia, North America, and South America. Symbols indicate reports on the following animals: catfish, eels, other fish, frogs, snakes, turtles, sea birds, chickens, other birds, dogs, cats, deer, horses, cows, rats, and mice. Data are nonsystematic and collected from many sources. [Kilian, 1964; von Hentig, 1923; Simon, 1975; Lee et al, 1976; Academia Sinica, 1977a, b; Shaw, 1977; Rikitake, 1978a, b; Tributsch, 1978]
A little while back, I wrote about research into whether animals can somehow predict earthquakes. It’s a great subject!
For years people have wrestled with the question of whether a link can be found, whether animals can somehow act as a warning system that will tell us to seek shelter before an earthquake. But one of the challenges, whether in terms of first identifying the behavior or of later monitoring your warning-animal, is consistency and knowing that you’ve made a robust observation. And one phrase in that caption pretty much nails the problem for drawing broader conclusions: “data are nonsystematic and collected from many sources”.
Which leads us to ask, what would be systematic data? Hard to say. As I wrote in that Beast article, most of us can’t sit around staring at Rover, taking careful notes, just in case an earthquake strikes. And yet that might be what it takes. As one researcher wrote more than 30 years ago:
Should we find that our animals indeed do sense impending earthquakes so much the better. If we do not, however, there is danger of extrapolating our results to the statement that no animal anticipates seismic events and that further studies are not warranted. I don’t know how to deal with this paradox except to urge the USGS to lobby for sufficient funds to allow installation of a greater diversity of species in seismically active areas. An important part of the lobby would be a plea for patience and understanding of the gamble and costs incurred as one simply maintains the experiments and waits for an earthquake to happen.
That was written by a UCLA biologist who’d spent a couple years observing pocket mice and kangaroo rats near Palm Springs. The rodents seemed to run on their wheels and dart through passages more often just before earthquakes.