In Social Insects, Researchers Find Hints for Controlling Disease


Given that she infects ant colonies with deadly pathogens and then studies how they respond, one might say that Nathalie Stroeymeyt, a senior lecturer in the school of biological sciences at the University of Bristol in the U.K., specializes in miniature pandemics. The tables turned on her, however, in March: Covid-19 swept through Britain, and Stroeymeyt was shut out of her ant epidemiology lab. The high-performance computers she uses to track ant behavior sat idle, and only a lab technician — deemed an essential worker — was permitted to tend to the lab’s hundreds of black garden ant colonies, each housed in its own plastic tub.

With governments across the world now encouraging people to maintain space between one another to prevent the spread of the virus, Stroeymeyt drew parallels with her insect subjects. The current guidance on social distancing “rung familiar,” Stroeymeyt said, “because I’ve been seeing it among the ants.”

Such insights are at the heart of a burgeoning field of insect research that some scientists say could help humans imagine a more pandemic-resilient society. As with humans, fending off disease can be a tall order for social insects — a category that includes termites, ants, and many species of bees and wasps. Insect workers swap fluids and share close quarters. In most species, there is heavy traffic into and out of the nest. Some ant colonies are as populous as New York City.

The insects are “living in very confined environments where there’s a lot of microbial load,” said Rebeca Rosengaus, a behavioral ecologist who studies social insect behavior at Northeastern University in Boston. Many of those microbes, she added, are pathogens that could sweep through the colony like a plague. That rarely happens, social insect researchers say, and vast colonies of such species are somehow able to limit the spread of contagions.

Over the past three decades, researchers have begun to explore just how that might occur, mapping the myriad ways that colonies avoid succumbing to disease. Some of those methods can seem alien. Others, including simple immunization-like behavior and forms of insect social distancing, can seem eerily familiar. Put together, they form a kind of parallel epidemiology that might provide insights for human societies battling pathogens of their own – even if, so far, human epidemiologists don’t pay much attention to the field.

Still, those insights are what Rosengaus and some other researchers are now exploring. “How is it possible,” Rosengaus asks, “that an individual that gets exposed to a fungus or a bacteria or a virus, or whatever pathogen there is, comes back to the colony, and does not infect everyone in the colony?”

While social insects have been the subject of intense scientific scrutiny for more than a century, the threat of pathogens and other parasites, researchers say, was long overlooked. “The mainstream social insect research has ignored parasites for a very long time,” said Paul Schmid-Hempel, an experimental ecologist at the Swiss public research university ETH Zurich. Biologist E.O. Wilson’s classic 1971 survey of the field, “The Insect Societies,” does not even list “disease,” “pathogen,” “bacteria,” or “virus” in its index.

As a postdoctoral researcher at Oxford in the 1980s, Schmid-Hempel realized that the bees he studied were constantly infested with parasites. He began to formulate questions that would help launch a small field: What if pathogens were not an incidental nuisance to colonies, but a profound threat that shaped the very evolution of their societies? To what extent were things like ant colonies and beehives actually tiny epidemic states?

Observers of social insects have long known that the animals keep their homes meticulously clean. Workers deposit waste and dead bodies outside the nests. Social insects groom each other, and often themselves, frequently. But recent research has documented other adaptations that also fight infection. Some ants, for example, harvest antimicrobial tree resins and spread them around their nests, a process researchers have described as “collective medication.” Social insect species also secrete a pharmacopeia of microbe-killing compounds, which they apply to their bodies and surfaces.

The current guidance on social distancing “rung familiar,” Stroeymeyt said, “because I’ve been seeing it among the ants.”

Grooming, too, seems to have unexpected benefits. As some ants clean each other, they transfer small amounts of pathogens to their nestmates. Those mini-exposures, the biologist Sylvia Cremer writes in a recent paper, cause “non-lethal, low-level infections” that “trigger a protective immunization.” She compares the process to variolation, a once-common method for immunizing humans against smallpox by exposing them to a small amount of fluid or dried scab material from a sick person. Rosengaus’ research has documented similar social immunization behavior among dampwood termites.

She and colleagues have also found evidence that, when some members of a black carpenter ant colony encounter pathogenic bacteria, they are able to develop an immune response and share it with their nestmates, making the entire colony more resistant. The ants who have been exposed appear to be passing along immune system compounds, mouth-to-mouth, ahead of the infection, readying their nestmates’ bodies for the possibility of exposure. Rosengaus compares this adaptation to a world in which a human could French kiss someone who has received a vaccine — and then gain the benefits of that vaccine indirectly.

These kinds of findings challenge assumptions that social living, by creating ripe conditions for diseases to spread, is automatically a risk to individuals. “Both the risk and the mitigation of risk come from sociality itself,” says Nina Fefferman, a professor of ecology and evolutionary biology at the University of Tennessee, Knoxville who studies disease transmission. Other individuals may get us sick. But they can also offer the care, food, and knowledge that saves our lives. “Everything is all rolled into this very complicated set of constraints and goals,” Fefferman said.

For all of Undark’s coverage of the global Covid-19 pandemic, please visit our extensive coronavirus archive.

For social insect researchers, one elusive question is whether, like human public health departments that impose coronavirus quarantines on households and occupancy limits on restaurants, social insect societies actually change their interactions to make it harder for diseases to spread — a phenomenon sometimes called organizational immunity. Most social insect colonies have complex systems for dividing up tasks. Some workers may end up caring for the queen, or feeding larvae, or standing on guard duty, or foraging. Decades of research have analyzed that division of labor in terms of task efficiency. But, starting in the early 2000s, mathematical models suggested that those social divisions might also slow down infections. By only interacting with a few designated workers, for example, a queen may be less likely to get sick.

Testing some of those theories on real colonies, researchers say, has been difficult. But the advent of automated insect tracking systems has opened up new possibilities, allowing researchers like Stroeymeyt to construct detailed pictures of who is interacting with whom inside an ant colony, for example.

To map an ant social network, Stroeymeyt and her fellow researchers glue tiny QR code tags, some smaller than a square millimeter, to ants’ thoraxes. Once each ant in a colony has been tagged — Stroeymeyt estimates she can personally saddle 500 ants with QR codes in a 12-hour day — the colony is placed in an observation box. Cameras overhead read the QR codes and record each ant’s position two times per second, for hours on end. The process generates data about every single contact between ants in the colony — hundreds of thousands of datapoints that, with high-powered computers, can be resolved into a detailed picture of the ant colony’s social network.

In 2014, Stroeymeyt and her colleagues mapped the networks of 22 colonies, tallying the interactions in each of them over the course of a few days. Those networks, they showed, did not emerge from random interactions of ants. Their interactions were more compartmentalized. Certain ants had more contact with each other than with other members of the colony.

A Lasius niger queen ant and worker ants tagged with tiny QR codes, some smaller than a square millimeter. Visual: Timothée Brütsch / University of Lausanne
The ant colony is placed in an observation box with multiple cameras overhead to record each ant’s position twice every second. Visual: Nathalie Stroeymeyt / University of Lausanne

The automated tracking procedure used by Stroeymeyt and her colleagues detects and highlights ants according to their status — red (workers), blue (queen), or brown (dead individuals). A unique ID number for each ant is displayed in green. When two ants physically interact, they are connected with a yellow line.
Visual: Courtesy Nathalie Stroeymeyt

At least in theory, those kinds of modular networks alone could slow the spread of infection in the colony. A human virus, after all, spreads more quickly through a lively party of 100 people than it does among 20 isolated clusters of five friends each, who mostly just hang out with each other.

But the bigger breakthrough came after the team exposed individuals in 11 colonies with the deadly ant-infecting fungus Metarhizium brunneum, with the other 11 serving as controls. Once the ants sensed the pathogens, those networks changed: Their modularity increased, and different task groups in the colony interacted less than before. Foragers exposed to the fungus demonstrated fewer contacts. Even unexposed ants started interacting differently, keeping a higher proportion of their contacts to smaller circles of nestmates. This process, Stroeymeyt told me, is not unlike social distancing. “It’s a very cheap and easy way to protect the colony from an epidemic,” she said.

Such research, of course, has only just recently been made possible. As Stroeymeyt points out, it’s not clear whether, in the absence of pathogens, the ants’ modular social networks have evolved in order to respond to the threat of infection, or whether pathogen suppression is just a useful side effect of patterns that have evolved for other reasons. And while the particular mechanism documented in the research was successful in slowing the pathogen’s spread, it may be just one of a number available to the colony. In addition, one recent paper raised questions about whether lab conditions, using pathogens like M. brunneum, necessarily do much to reflect the disease conditions that colonies battle in the wild.

Still, Stroeymeyt and her colleagues’ findings have been widely discussed among insect researchers. And, as she points out, ant distancing would suggest that humans aren’t alone in reordering our societies in the face of epidemics.

If anything, Stroeymeyt said the ants’ success may offer some validation, and inspiration, to humans struggling through a pandemic. Human public health departments are only a couple of centuries old, while ant societies have been evolving for millions of years. “It’s very rare to find a colony collapsing under the weight of a pathogen,” Stroeymeyt said. “We know that their mechanisms are extremely effective.”

While insect epidemiologists study the work of human epidemiologists, the reverse appears to be less common. In theory, researchers say, social insects could be an ideal model system: a kind of miniature society, with few ethical constraints, in which to explore how disease travels through networks. But, Schmid-Hempel points out, collecting detailed information about insect health is difficult. “In humans, you have a lot of really great data, compared to what we have in social insects,” he said. One day researchers might find it useful to test out epidemiological principles in insect societies. “I’m sure it’ll come,” Schmid-Hempel said. “But it’s not yet at that point.”

One of the few researchers to bridge the divide is Fefferman, the University of Tennessee researcher. Trained in applied mathematics, Fefferman studies how infections move through networks — insect networks, human networks, computer networks, and even networks in online games. Her research has been published in both entomology and epidemiology journals. A paper she co-wrote in 2007 about a virtual epidemic in World of Warcraft gained extensive attention from public health experts.

Fefferman’s research on human epidemiology, she said, draws from her study of insects. “You can look at social insect colonies very much as successful cities,” she said. “And then you can say, well, what are the strategies that social insects use, both behaviorally and how they evolve them, that we can then borrow from?”

As an example, she brought up termite cannibalism. When exposed to a bad outbreak, some termites immediately eat the colony’s young. Doing so, Fefferman argues, helps them eliminate a pool of “highly susceptible” individuals who are likely to serve as a reservoir of infection, allowing the epidemic to linger in the nest.

“It’s very rare to find a colony collapsing under the weight of a pathogen. We know that their mechanisms are extremely effective.”

Human societies are unlikely to adopt cannibalism as a public health strategy. But the basic principle, Fefferman argues, may be relevant during the coronavirus pandemic. “If we think about abstracting that,” she said, “that’s school closures.” The lesson from the termites could be “separate the kids. The kids are going to be a massive puddle of transmission that’s going to infect everybody. Don’t do that.”

This kind of thinking has led Fefferman to build models that aim to find the most effective way to distribute medicines in the midst of a flu epidemic. A new paper she’s working on, about how companies can structure their workforces to prepare for pandemics and other disasters, is inspired by the cohort-based model that many insect colonies use to distribute tasks — though that’s not likely something she would readily advertise when the final paper is published.

Indeed, Fefferman said she doesn’t typically cite the influence of entomology on her work, at least when she’s talking with public health experts.

“I’d never run into a public health meeting and be like, ‘Guys, BUGS!’” she said. “But maybe if I did, it would be fantastic.”


Michael Schulson is a contributing editor for Undark. His work has also been published by Aeon, NPR, Pacific Standard, Scientific American, Slate, and Wired, among other publications.