Anytime the discussion of animal's supposed ability to sense earthquake precursory phenomena comes up, I have to wonder how such instincts or behaviors would have evolved or been learned. Since non-human animals virtually never die as a result of earthquakes (tsunami are another issue, however see below), there would be no survival value in such behaviors. Due to the lack of threat posed by earthquakes, as well as the rarity of large ones, many hundreds of generations of animals would be born, reproduce, mature and die between events which would kill a vanishingly small portion of their population. It just doesn't make sense that evolution would waste time on a behavior that doesn't further the survivability of a species' genes.

The case for tsuanami is just as strong. Though animals do die as a result of extraordinarily large seismic sea waves, their extreme rarity suggests the same argument as the one regarding earthquakes.

Mike Williams in Arroyo Grande, CA USA