13 December 2005

New discovery on how natural complex systems work

The Department has been studying some of the complex systems found in nature for a number of years. Some of the most interesting systems are social insect communities, particularly ants. We study Pharoah's ants (Monomorium pharoensis) and have recently made a startling discovery following one that also grabbed the headlines a year ago.

Ants communicate with each other by laying chemical trails - for example they will lay such a trail to a new food source and as long as the food is still there other ants will reinforce the trail by laying more of the pheremone chemical.

Last year we discovered that the geometry of the trail was important - the angle of the branches contained information that helped the ants to get back to the nest, [Jackson et al., 2004 Nature, 432, 907-909].

More recently we have discovered a new chemical that is laid at the start of a branch when the food source has become exhausted. This tells the other ants not to bother with that part of the trail - the main trail chemical may not have fully evaporated by then, [Robinson et al., 2005 Nature, 438, pp 442].

This make their foraging system much more efficient. Ants as a community solve some very hard dynamic problems and the way they do this is of great interest to Computer Science. Some of our work in this area is funded by BTExact who are interested in solving such problems - these may be involved in the management of mobile phone networks, internet resources etc.

Our approach is to build a computer simulation of the ant community and to try to understand how it works, validating this against experiments carried out with our collaborators in Animal and Plant Sciences.

Researchers involved: Duncan Jackson (joint DCS & APS), Elva Robinson (joint APS & DCS), Francis Ratnieks (APS), Mike Holcombe (DCS)