A while back this video was making the rounds:

It’s a self-reinforcing circular mill composed of- now here’s a change of pace! – army ants. I thought I’d reintroduce it in honor of this week’s festivities.

Labidus praedator

The species is Labidus praedator, a swarm-raiding army ant from Central and South America, and these circular mills are a common byproduct of army ant navigation. Labidus is completely blind, so ants in this genus get about by following the insect in front and laying down a chemical trail. The system works well enough in a straight line.

The trouble begins when the ants loop around and intersect their own path. The poor insects end up on a mobius strip of their own making, circling around and around until some either chance to leave the mill and the circle is broken, or they run of steam and perish. Thus, the Ant Death Spiral.

The phenomenon was first noted by pioneering army ant biologist Schneirla, who published a paper on it in 1944 under the less sensational term “circular milling”.  Schneirla’s detailed analysis is worth a read, not just for the natural history and a surprising amount of physics, but for a remarkable concluding sentence in which he asserts that people are clearly better than ants.*

In any case, youtube user l314kimo recently created a clever computer simulation showing how the mill arises. A simple set of behavioral rules given to digital ants is enough to recreate the phenomenon:

I used to see ant spirals all the time when I lived in Paraguay, and not just in the field. Labidus has no qualms about raiding through rural houses, and I’d come home to find circles of ants whirling about on top of my plates in the kitchen, or sometimes an intimate ring of 5-6 ants on a coffee mug. Unnaturally round objects, mostly.

You’d think spiral-induced mortality would be selected against, that ants would have evolved a counter-measure to such obviously maladaptive behavior. But army ant colonies are huge, their daily intake immense, their fecundity explosive. A few hundred spinning workers lost around the margins may not make all that much difference.


*”It may be observed that while army ants are constitutionally susceptible to the predominance of circular-column behavior and can be freed from it only by the incidental fact of environmental variation, man is by no means susceptible in the same sense, with his cortical basis for versatile corrective patterns which under encouragement may reduce milling to the minor role of an occasional subway rush.”