Moving house is a rite of passage that can test even the strongest relationships, requiring teams of people to cooperate to heave heavy sofas and large boxes up and down stairs. But we are not alone in our ability to pull together to tackle a sizeable challenge; dragging a hefty insect carcass to the nest is probably the closest equivalent of negotiating a tight turn in the stairs for a team of ants. ‘We have to talk to the other person constantly to figure out how to solve the problem’, says Helen McCreery from the University of Colorado, Boulder, USA, so how does a crowd of determined ants carrying a large morsel respond when they encounter an impediment to their mission? ‘They suddenly have to make a choice when it is not clear that one option is better than any other’, says McCreery; and the busy creatures appear to achieve this without discussion. Intrigued, McCreery, her advisor Michael Breed and computer scientist Radhika Nagpal from Harvard University, USA, decided to find out how hindered ant crews cooperate to negotiate unexpected obstructions.
As a band of burdened ants could use an almost unlimited range of strategies to navigate their way around an obstacle, the team designed three obstructions to test how the insects responded in practice. Constructing a Lego brick wall, a three-sided cul-de-sac and a completely enclosed corral coated in slippery Fluon® – to prevent the ants from escaping over the top – McCreery and Zachary Dix placed each one in the path of a crew of longhorn crazy ants (Paratrechina longicornis) as they attempted to heft chunks of tuna home, and then filmed the ants’ tactics.
Analysing the ants’ manoeuvres when they encountered the wall, the team saw that the obstructed ants simply moved to and fro along the wall until they reached the end and resumed their homeward course. ‘Ants are really good at knowing the direction of the nest’, says McCreery, explaining that the ants needed little information other than the direction of home to implement this strategy. However, using the same tactic when faced with a cul-de-sac would be doomed to failure. Realising that the strategy for escaping a blind alley must be more sophisticated, the team watched as the ants zig-zagged to and fro across the front wall of the obstruction – as they had when trying to navigate around the wall. However, as time passed, they began randomly moving backwards, away from the nest, until eventually they encountered the exit and were able to go on their way again. The ants’ strategy of using random movements allowed them to maintain a consensus while being flexible and robust enough to get them out of most tight corners, although the route that they took was not always the most efficient.
However, McCreery admits that she was surprised at how quickly the ants gave up and abandoned their precious piece of tuna when they found themselves trapped in the corral. ‘We really expected their behaviour in the trap to look a lot like the cul-de-sac, at least at the beginning. But as soon as we closed the door, the speed and group size started to drop’, she recalls, suggesting that the ants might have given up sooner than expected because they were cut off from incoming ants. And, having discovered how ants deal with being stuck in a blind alley, Nagpal is keen to apply these strategies to teach teams of robots how to cooperate to solve problems in environments that they are unfamiliar with.