Can animals forecast danger?

We put John's question to zoologist Max Gray.

Max - So, John is in good company here. People have been asking this question apparently since the 4th century BC. It's the first recorded documentation of dogs being able to sense earthquakes in ancient Greece.

Chris - When you say sense earthquakes, do you mean predict, almost presage their arrival?

Max - So, this is the question that people have been asking and it's not really very clear. People think that dogs can sense earthquakes way before we can and then if we could somehow tap into that, we might be able to predict earthquakes which are notoriously difficult to predict. So, people have looked at this. People have looked at studies of dogs and other animals going missing. I think it was in San Francisco before an earthquake in the 80's and there were some anecdotal evidence that more animals go missing in the week before an earthquake. Then people have gone away and done a research into this and realised that actually, the robustness of that claim is very weak indeed. But in the early 2000's, somebody did actually get some concrete data about this, a man called Stanley Coren who was doing a completely separate study at that time about hearing in dogs, had a set of kennels with 200 plus dogs in and then as an earthquake happens to happen. He managed to get data on how many of those dogs responded in the certain timeframe beforehand and it was a vast majority. And most of the dogs.

Chris - Responded how? Barking or.

Max - Yeah, they get very stressed. They get nervous, mostly barking, whining, hiding in the corner, that kind of thing. but some of the dogs didn't and it was mostly the dogs that were in some way impaired in their hearing or dogs with big floppy ears that kind of obstruct their ears that way. So, it seems to me that dogs may have the ability to kind of hear the deeper rumblings that precede an earthquake. However, that's not really going to be very useful diagnostically. We have this one piece of evidence. There is again, more anecdotal evidence about it. Maybe cows appear to reduce their milk production slightly before an earthquake. Sharks migrate away to deeper water before tropical storms which seems to be based on a prediction due to the lowering of atmospheric pressure. But we have much better ways at measuring that through meteorology and so, all of this kind of stuff is very, very anecdotal so we don't really have the evidence. Even if we did, is it in any way useful? You can't really use a kennel of 200 dogs as an earthquake predictor. It's a little bit inhumane and you can't tell for sure. I mean, most animals are aversive to things they're not familiar with. They don't like new sensations, new sounds, loud noises, unfamiliar smells, anything. It's called novaphobia [sic], the fear of the new. This is present in most animals. So, there's no way we would know whether it's a prediction of a natural disaster or whether you know something unfamiliar and pungent happens to be nearby that was entirely harmless to us.

Chris - I've heard Novas are quite good cars but there we are.