In a recent podcast you talked about spiders and ants counting steps etc. How does anyone know/find out whether these creatures count? And if they do, how do they do it? What would they use that's equivalent to our number system ?
Chris - Well, not just spiders but probably many insects can do this. The example I gave were spiders counting their steps in order to know how to build their webs but also, there was a very elegant paper that’s published in Science a year or so ago by Matthias Wittlinger who’s at Ulm University in Germany.
He and his colleagues were looking at how ants navigate. Ants use the sun and they also use a compass – they have in their brain a body clock and they're able to use this mental compass of the time of the sun moving across the sky to navigate by. But they also count their steps and the reason that the scientists know that for sure is that they performed the very delicate task of putting ants on stilts.
They cut hairs off the back of a pig and then glued those hairs onto the ants’ legs to lengthen the ants legs, and when they did this, of course, the ants’ strides became twice as long as they would normally be. When they followed the ants around, all of the ants overshot their nest because they walked twice as far as they should’ve done.
When they actually, paradoxically, cut the ant’s legs off so they had much shorter legs, they didn’t walk far enough. They're all circling around, thinking, “Where’s my nest? I should be there.” It was clear that they're actually counting the number of steps they've taken in order to find their way around.
In their brain and their nervous system, they must have some kind of neural integrator circuit that, every time they take a step, notches up another 1, 2, 3, and so on, so they can find their way around by counting steps. So there’s the evidence: ants can do it, it’s likely that other insects probably do the same.
Ben - So ants on stilts will overshoot their home. Poor ants.
I had a look on google and it doesnt look like they know much about it.
The experiment, at least in the form explained in your Answer, is not a good enough proof that ants can count. As long as they are capable of measuring time duration _and_ remembering how long it would take for them to walk between two points, ants with longer legs that walk faster than usual will overshoot their nest by simply timing their walk. This would equally be a valid explanation of what the experimenters saw. Or they could be measuring the energy they expend to walk between two points, and lengthened legs may be giving better efficiency for them, which would be another valid explanation. An interesting tangent is if scientists can determine the optimum length of legs for these creatures. For example, if ants with legs lengthened 2x can walk 2x as fast as usual but expends only 1.3x more energy to do so, it would mean that the length of their legs have still _not_ been optimized for walking by the process of the evolution. Of course, with overlong legs, their body would be farther from the ground, which makes it harder to feed, and that could be a counter-balancing factor that prevents longer-leg mutants from being selected evolutionary. Thanks for an interesting story, anyway. Gitster, Wed, 28th Jul 2010