Social insects, which don't have very large nervous systems, are capable of remarkably sophisticated behavior, such as the direction-giving dance by bees or the lifesaving rafts formed by fire ants. In these cases, the benefits of this behavior—more food or survival, respectively—are pretty obvious. But there are also cases where the benefits are less than obvious, so how do insects collectively decide to engage in a risky activity?

Researchers are studying a species of ant, the weaver ant Oecophylla smaragdina, that can move vertically amid trees by building a ladder using its own body. The effort takes a lot of workers away from foraging for as long as the ladder is in place, making it a major investment. But in most cases, the rewards will be uncertain; there's only a payoff if the ants find a significant food source at the newly accessed level.

To make the decision, ants appear to judge the distance between their location and the destination. But not every ant makes the same judgment, and it's possible to trick the ants into building longer ladders by moving the destination.

Taking a gamble

Most people find risk/reward decisions difficult—you just have to look at things like cryptocurrency or NFTs to see that we often we get it wrong. But weaver ants have to somehow make that evaluation collectively. In their natural habitat, they regularly build ladders that allow them to move up the tree canopy they call home. Those ladders require the commitment of a significant number of workers for as long as they're left in place.

And the ants form them with an uncertain payoff. Until they explore their newly accessible destination, there's no way to tell whether there's enough food there to make the risk pay off. "This makes chain formation akin to a gamble," the authors of a new paper write. "The colony must invest a proportion of its capital (number of ants) to perform a task with unknown outcomes."

To understand the decision-making process, the researchers used video recordings to track the behavior of the ants while they built a ladder. Their system involved giving the ants access to a horizontal rod. An adjustable-height platform could be placed beneath the rod to provide a destination, and a food source was placed on the far end of the platform, providing a reward for building a bridge between the rod and platform.

Although the ladders are several ants wide, the researchers found that nearly half of the ants that joined in the process did so at the tip of the growing chain. The tip is also where about 90 percent of the ants left the chain, which helps ensure that the whole thing remains stable instead of collapsing from internal instabilities. Once an ant commits to the chain, it simply latches on to its neighbors and stops moving and stays that way unless the tip of the ladder retracts.

Once the ladder is close to the platform, those ants arriving at its tip start to reach for the platform, which eventually completes the connection.

Going the distance

Using data from video recordings, the researchers showed that the process wasn't really affected by most of the things that could potentially influence whether a ladder grew successfully. Whether the ants that reached the tip joined in the chain or not doesn't appear to be linked to its length or the distance from the tip to the platform. Instead, ants seemed to spend more time in the chain when the platform was close.

That explains the growth of successful chains but not their initiation. Here, the key determinant was the distance between the rod the ants started on and the destination platform. If the two were 11 cm or more apart, the ants never built a ladder across the distance.

But things were a bit more complex than a yes/no decision. Even when the distance was long enough that the ants didn't complete a ladder, there were consistently a handful of ants that would try to start one at the bottom of the rod. They just couldn't get many of their peers to join in. This suggests there's at least some diversity in ant behavior.

The researchers also found that they could induce the ants to build longer ladders by using a movable platform. They could start the distance within the range that ants would normally bridge and keep moving the platform down to keep the tip within that range. That suggests that the ants don't make a decision as to whether to build a ladder until they're at the tip of a growing one.

Overall, the results suggest a complex behavior can be built from several simpler ones. For example, there seem to be different behaviors involved in starting, joining, and staying within a growing ladder. And the key decision that stands in for risk/reward is how far the destination is when an ant is ready to join that ladder. That distance is likely to be a pretty reasonable proxy for how much of the colony's workers will end up giving up their productive labor in order to make the ladder succeed.

PNAS, 2023. DOI: 10.1073/pnas.2216217120  (About DOIs).