This is the field station of Ruediger Wehner of the University of Zurich. For over 40 years, Wehner, his students, and visitors from other universities have studied the ants here. The station is some 5 km outside of the little town of Mahares, Tunisia, on the Mediterranean coast. As the photo above shows (it is a good number of years ago now), I have been there, and I actually did more than hold poles for sighting. Rüdiger Wehner provides research infrastructure as well as intellectual collaboration. In the rich history of the Mahares experience, much about the fabulous ants has been discovered.
This is where the ants live, and the work is field experimentation. We set up feeding stations for ants, using watermelon as bait. We trap them at the feeding station in small bottles, and offer them biscuit crumbs. After an ant gets a hold of a piece of biscuit, she is motivated to head home, and we can release her and study her ability to find her way home.
One of the study species, Cataglyphis bicolor. The long legs and the raised abdomen help to lift the ant off the hot ground. It also moves fast.
On the open salt pan, where most of the nests are, C. fortis must keep track of the distance and direction from its starting point, its nest, as it travels in foraging. This strategy, known as path integration, is the only means to navigate home in an environment devoid of visual landmarks and devoid of odours until very near the nest.
We recently investigated the memory properties of one component of path integration, the odometer, or memory for the outbound distance travelled. A nest was surrounded by an enclosure (where Ajay Narendra is in the picture) Ants were forced to come out foraging down one of the channels. They arrived at a feeder, at which they found crumbs of cookies. We let them run back with a piece of cookie in a different channel of the same construction.
This picture was taken in Alice Springs, but the channels used in Mahares were of the same construction.
Because path integration must be good enough to get the ant home the first time (or there will not be a second time), we predicted that repeated training down a channel at the same outbound distance will not improve performance. Because a foraging run is typically less than an hour, we predicted that memory for the outbound distance travelled will show decay after a day's delay. And because the function of path integration is to keep track of the current journey, we predicted that earlier journeys (which we experimentally manipulated to have a different outbound distance from the current journey) will not influence the odometric estimate of the current journey.
All these predictions were confirmed. Details are in a publication of 2006: Cheng, K., Narendra, A, & Wehner, R. (2006). Behavioral ecology of odometric memories in desert ants: Acquisition, retention, and integration. Behavioral Ecology, 17, 227-235.
The same predictions were also confirmed in our Central Australian desert ants Melophorus bagoti, in a publication of 2007.