A Curio story

Why do army ants march in three lanes?

Returning ants take the middle, outbound ants take the sides, and nothing anywhere in the column is in charge.

While an army ant raid is running, the traffic sorts itself into three lanes. Ants coming home, many of them carrying prey, take the middle lane. Ants heading out to the raid take the two outer lanes. Nobody assigns any of this. There is no lead ant, no controller, nothing anywhere in the column directing traffic.

That is the part worth sitting with. Human traffic needs painted lines, signs, and traffic law to hold a lane. The ants get a working three-lane highway out of nothing but the way individuals walk.

Where the lanes come from

The study behind this is Couzin and Franks, "Self-organized lane formation and optimized traffic flow in army ants," published in Proceedings of the Royal Society B in 2003. They worked on Eciton burchellii, and they paired a model with field observation — the lanes are not something that only exists in a simulation; the structure is there in real raiding columns.

Their answer for where it comes from is almost disappointingly small: the lanes emerge from simple differences in individual turning behaviour. No signals, no ranks, no lane assignments. Tiny differences in how individual ants turn are enough to sort two-way traffic into a returning lane flanked by two outbound ones — and the other half of the paper's title, "optimized traffic flow," is what the sorting buys.

The road is made of smell

The individual ant makes this more impressive, not less. Workers in these raiding columns are nearly blind. They are not following the road by sight; they follow pheromone trails laid by their nestmates. The route exists as chemistry on the ground rather than as knowledge in any ant's head.

So the highway has no engineer and no map. A single army ant on the trail knows almost nothing about the raid it is part of, and the raid works anyway. The route lives in the world, held in scent, and no individual ant has it.

They build the infrastructure too

Lanes are not the only civil engineering these colonies manage without a plan. Army ant colonies build physical structures out of their own linked bodies, including bridges, and the bridges are not fixed: Reid and colleagues showed in PNAS in 2015 that army ants dynamically adjust living bridges in response to a cost-benefit trade-off. The structure answers the traffic running over it. Garnier's 2016 name for the whole category is hard to beat: "living architectures."

It is worth being blunt about what this is not. It is not intelligence in the way we usually mean the word. No ant understands lanes, bridges, or trade-offs. Each one is a nearly blind animal doing something simple, and the competence sits one level up, in the pattern.

Disclosure of bias before the sign-off: I build Curio, an iOS app that serves five short curiosity stories a day — things like this one, cut to the length they actually need. If that sounds like a feed you'd want, it's Curio on the App Store. If not, no harm done; the ants keep their lanes either way.

One strange true thing, five times a day

Curio is a full-screen feed of short written curiosity stories — like this one, but they end before your coffee cools. Free on iOS, no account needed.

Download on the App Store