From ants to algorithms
A lot of people know that ants live in the ground, build ant hills and lay down invisible chemical trails. Temnothorax rugatulus does none of that. Instead, these ants live in free-form crevices, and their communication and division of labor are there for all to see, including scientists.
Which is why Stephen Pratt and Gabriele Valentini are so interested in these tiny ants. In a new study, the Arizona State University researchers explored how each member of a T. rugatulus colony, a complex functional organization with no central control, manages to effectively communicate among themselves and to make decisions that allow them to thrive.
Understanding how each member of the colony communicates and how labor is divided can help researchers design algorithms for complex artificial systems or better understand complex organic systems, like the brain, said Valentini, a postdoctoral research associate whose background is in computer science and complex systems.
“Modeling heterogeneity informs engineering,” said Valentini, who is also the lead author of the study.
For example, modeling heterogeneity could help design algorithms for robot swarms and, if humans were to decide to colonize another planet, such swarms could be dispatched from Earth and tasked with deciding where to build structures that humans would inhabit.
Given their relatively small colony size of about 200 members, their longevity and their overt physical contact with one another, T. rugatulus serve as a near-ideal model for exploring complex systems.
“They’re very nice for getting this fine-scale detailed description of individual behavior and at the same time observing this interesting collective behavior,” said Pratt, a professor of life sciences and the study’s principal investigator. Of note, each ant lives two to three years, long enough to have the same ant participate in multiple studies.
Pratt, who studies insects’ complex social lives, explains that ant colonies acts as a kind of collective intelligence.
He and other researchers have known for a long time that ants divide labor among specialists. But previous research on collective intelligence has often assumed that all ants behave the same and follow the same rules. Not so, say Pratt and Valentini. Different ants undertake very different roles.
To find out what those roles looked like, Pratt, Valentini and their colleagues looked at how a colony of T. rugatulus behaved when they were compelled to find a new nest. Moving to new quarters means the ants have to find new crevices, reach a consensus on which one is best, and finally, move into their new home. All that activity requires clear communication and a division of labor.
“And they have to do that without anyone seeing the big picture,” Pratt said. “It’s not like there’s any one ant that gets informed about all the possible nests and figures out which one is best, and then tells everyone where to go. There’s no ant who knows everything.”
That’s why the researchers tracked and observed each ant's movements. They paint-marked each ant with a unique color code and then observed their emigration process.
What the researchers came up with was this: a huge data set, to which Valentini put order using statistical analysis. Valentini’s analysis showed a mash-up of active ants and passive ants on a continuum of behavior.
“It’s easier to model these systems if you treat all the ants as identical, but we know that’s not true,” Valentini said.
Active ants include primary workers, which make up about one-third of the colony. Primary ants play a major role in choosing a new nest site. These ants leave the nest, explore possible new nest sites and then bring each other to candidate nest sites using tandem runs.
Tandem runs are a way of recruiting other members of the colony to a new nest and sharing information about its location. In a tandem run, one ant finds a nest that she thinks is a suitable new home and leads another ant to the candidate nest to assess it.
In addition to primary ants, about one-fifth of the colony consists of secondary ants that are carried to a candidate nest by a primary ant. Eventually, these secondary ants begin to carry other nest-mates from the old nest to the new nest. Of note, secondary ants aren’t the colony’s decision-makers, but, instead, they help implement the move to the colony’s new nest.
Video courtesy Gabriele Valentini.
Once the colony achieves a quorum, the workers ramp it up: They quickly carry the remaining members, known as passive ants, to the new nest. Passive ants make up about 40% of the colony.
Intriguingly, the researchers found another segment of the colony that did not behave like the other ants. Pratt and Valentini dubbed these ants “wanderers.” The wanderers made up about 10% of the colony and did not appear to be decision-makers or implementers. However, they made several visits to the candidate nest.
Pratt and Valentini speculate that the wanderers are foraging for food or perhaps they are primary ants that are dissatisfied with their new home and are still looking for better alternatives. “Perhaps they are very picky, and never find a nest that’s good enough,” Valentini said.
No matter, the researchers now have a better idea of what these ants are about.
“As biologists, we gathered this data, but it sat around for a long time, and we weren’t sure how to handle it,” Pratt explained. “It was when we started working with engineers, like Gabri and Ted Pavlic, and physicists like Sara Walker, they just brought a whole new perspective to it.”
The study, “Division of labour promotes the spread of information in colony emigrations by the ant Temnothorax rugatulus,” was published online in the “Proceedings of the Royal Society B” on April 1. In addition to Valentini and Pratt, the paper’s authors include Naoki Masuda, Zachary Shaffer, Jake Hanson, Takao Sasaki, Sara Imari Walker and Theodore Pavlic.
Top illustration: Temnothorax rugatulus ants have a stereotypical posture that they use to carry nestmates. Credit: Alex Cabrera/ASU