Without fail, desert ants make a beeline for their nests after a meandering foraging trip - and they manage this without spotting any landmarks or sniffing any scent trails. `It's amazing what these animals can do with a brain that's only a tenth of a milligram', says Bernhard Ronacher. To find out how desert ants measure distances on their long marches in search of food,Ronacher teamed up with Rüdiger Wehner of the University of Zürich and Gunnar Grah of Humboldt University in Berlin to test desert ants in hilly terrain (p. 4005).

Desert ants use dead reckoning (path integration) to make their way home,which means that they must be able to measure both the direction and distance they've travelled from the nest. But what distance information do ants gather on their journeys across a hilly landscape - do they project their trajectories onto the horizontal plane as they travel over undulating terrain to measure the ground distance they've covered (`as the crow flies'), or do they measure the actual distance walked, which is much longer than the ground distance if the ants travel up and down hills?

To find out, the team travelled to Wehner's field site on saltpans near Maharès, Tunisia. They trained three groups of desert ants to walk from their nest to a food site through L-shaped aluminium channels. The first group walked a straight stretch on the ground, then for the second leg of their trip they took a right-angled turn up a steep incline, walked another short straight stretch and then down a steep incline to ground level, where a bowl filled with biscuit crumbs awaited their arrival. The second group stayed at ground level, with the second leg of their L-shaped trip covering the same ground distance that the first group had travelled. Finally, the third group also walked along the ground, but the second leg of their journey was much longer, covering the same walking distance as the first group.

The angle at which the ants subsequently set off for home would reveal whether the climbing ants (the first group) had gauged the ground or the walking distance during their L-shaped journey. Would climbing ants act like ants that had travelled the ground distance or like ants that had travelled the walking distance? When each ant had secured a treat from the bowl and held the crumb in its jaws for the homeward journey, the team scooped the unsuspecting insect into a box and took it to a flat field some distance away. To determine which direction an ant considered to be homeward bound, the team placed the ant on a grid painted on the ground and traced its movements on a piece of grid paper. They found that the ants that had climbed up the steep slopes set off at the same angle as the second group of ants, which had travelled the same ground distance as the first group. But the third group of ants, which had covered the first group's walking distance, took off at a completely different angle. Ants from the first two groups also travelled similar distances before they began circling to search for the nest entrance. The team concluded that ants measure the ground distance as they walk across hills, and don't measure the total distance walked. Importantly, desert ants incorporate this information into their path integration to calculate the correct angle to take a beeline back to the nest. If these ants ever find themselves crossing desert dunes, they certainly won't get lost on the way home.

Grah, G., Wehner, R. and Ronacher, B. (
). Path integration in a three-dimensional maze: ground distance estimation keeps desert ants Cataglyphis fortis on course.
J. Exp. Biol.