I saw this short but interesting article on Digg about new data on how sharks hunt for food. It turns out that the seemingly random way they move around resembles a Levy flight. From the article:
He explained that the technique's movements "can be advantageous when searching for randomly distributed resources because they reduce 'over sampling' without the need for cognitive maps and sophisticated navigational abilities."
It seems like there are at least two awesome things here: 1) an idea developed in a pure mathematical context proves "useful" well after its invention; 2) evolution drives behaviors towards some sort of optimal solution. Very cool stuff!