Vampire bats can starve to death if they don’t feed for a mere three days, so strong social ties can be key to survival. For instance, a bat will sometimes share food with a hungry member of the same roost, regurgitating any blood it has consumed into the mouth of the hungry bat—a bit like a bloody French kiss. That’s a true “friend.” Evolutionary biologists have dubbed this behavior “reciprocal altruism.” But vampire bats can also form bonds with strange bats from outside the roost, building up trust with mutual grooming first before moving on to food sharing, according to a new paper in Current Biology.
What’s being tested here is a game theory model first proposed in 1998 colloquially known as “raise the stakes.” It’s similar to the famous prisoner’s dilemma, in which two criminal suspects are arrested and separately offered a deal. If one of them confesses and the other doesn’t, the defector will go free and the other suspect will get 20 years in jail. If both suspects confess, they will each get ten years in jail. The correct strategy, therefore, is to always confess, since one should assume the other party will act solely in his or her self-interest. Both players will reap the most benefit by cooperating with each other.
But cooperative behavior isn’t always quite so simple as a binary choice between cooperating or defecting; it’s more like a continuously variably investment. The “raise the stakes” model of relationships holds that two strangers can make low-risk, incremental investments to see if there is potential for further cooperation. If the other party reciprocates in kind, it builds trust and a relationship can form. If not, no relationship will develop, and nobody has spent too much time and energy on a worthless (from a survival standpoint) connection.
“I always thought this model is the most obvious and intuitive way to think about how these relationships form in nature,” lead author Gerald Carter, an evolutionary biologist at Ohio State University, told Ars. “You test the waters of the relationship by making small cooperative investments, and if things go well and you get a return—meaning that the relationship is mutually beneficial and not exploitive—then you increase your investments more and more.”
But it’s proven challenging to test that theory within the animal kingdom. Granted, there have been human studies, in which college students played economic games in the lab where they were encouraged to cooperatively raise the stakes. For animals, however, “You’d have to introduce them as strangers and then watch them start helping each other,” said Carter. While there have been prior studies looking at cooperative grooming behavior in primates, they didn’t shed any light on how those bonds formed in the first place.
Vampire bats turned out to be the perfect choice for such an experiment. According to Carter, the creatures are highly social and also quite small, making it easier to keep them in a laboratory. “You can simulate their natural social life, but in captivity,” he said. “They spend almost all of their time inside of a space the size of a small closet, like inside a hollow tree. That’s very difficult to do with baboons or chimpanzees, or other animals that range over much larger areas. To put them all together would disturb their natural social behavior.”
Carter and his colleagues collected 39 vampire bats from different roosts in Panama (Las Pavas and Tolé), separated by hundreds of miles, and kept them in captivity for 15 months. They created pairs (one from each location) or small mixed groups of bats for two to four months. Next they withheld food from one bat and observed the subsequent interactions with its partner or cage mates. Then they put all the bats together in a single roost for a year, repeatedly fasting unfamiliar bats to see if others in the roost would sacrifice some of their food to feed the hungry bat.
Carter et al. found that while there was a great deal of grooming connections between the bats, sharing of food was relatively rare, although it happened more frequently than would otherwise happen by chance. They observed a marked increase in grooming frequency before a bat opted to share food with another, leveling off after, and found that bonds between bats who were strangers were more likely to form when familiar bats were not present.
Most intriguing, according to Carter, was that the vampire bats formed relationships more quickly when they were introduced in isolated pairs. Put two large groups of vampire bats together and in-groups and out-groups tend to form, preventing bonds from forming with unfamiliar bats. That’s significant because scientists have pondered whether the observed cooperation among vampire bats, including food sharing, might not be based on their social history but rather on proximity—say, if a fellow bat smells familiar and thus is deemed worthy of help.
“The isolated pairs [result] rejects that because it shows that it actually does depend on the availability of alternative partners,” said Carter. “If bats are just matching individuals that are the most similar to them, then when you have the two groups together, there should be even more new food sharing, because there are more opportunities for matches. That’s the opposite of what we saw.”
That said, the in-group/out-group dynamics in bats are a bit different from a similar dynamic in humans, according to Carter. “Animals are just wary of unfamiliar anything, including unfamiliar individuals of the same species,” he said. “But humans also have a tendency to look for a badge of group membership,” even something as simple as a red T-shirt versus a blue T-shirt. “You can create an in-group very easily and quickly, and people start aligning with those groups and making irrational decisions based on completely arbitrary group cues,” he said.
Carter hopes to build on this work further by exploring how the vampire bats might choose their bonded partners “The key experiment is to make good and bad partners available to see if the bats choose good ones over bad ones,” he said. “That’s the crucial contingency test that tells you that the bars are engaging reciprocal, conditional cooperation—a cheat is punished while a more cooperative bat is rewarded.” Another experiment might explore lower-cost investments, since even before bats start grooming each other, they are learning to tolerate a strange bat’s close proximity.
Carter’s lab also just published another, especially timely paper in the Journal of Animal Ecology, specifically looking at how behavioral changes among vampire bats can indicate the spread of a pathogen or parasite through the social network. It’s well-known that those who are more socially connected are more likely to infect others, should they contract a virus. And people who are obviously sick tend to become less socially connected, helping to slow the infection rate. “We also know that it’s not just who’s connected to who that matters, but the strength of those connections—actual time together, or the association rate,” said Carter. This latest study demonstrates that the type of social relationship also matters.
The lab injected vampire bats with lipopolysaccharide, which triggered the immune system such that the injected bats showed all the usual signs of being ill (e.g. lethargy). But the injection did not contain a live pathogen and wore off after a few hours, so the bats weren’t actually harmed by the experiment. Carter and his team then analyzed how bats showing symptoms of an illness changed social behavior within the colony.
They found that a vampire bat that shows signs of sickness will have some types of weakened social connections, marked by reduced grooming from other bats in the roost. But a mother vampire bat will not reduce grooming if it is her own offspring showing symptoms. Notably, the presence of sick vampire bats did not result in a decrease in food-sharing behavior, probably because it is so critical for survival.
Bats are bats and people are people, but there are still some useful parallels that can be drawn, according to Carter. “COVID-19 in China was spreading and clustering mainly within families, because these social connections would not be reduced by sickness behaviors, or the closure of schools and public events,” he wrote on the laboratory blog. “In other words, a sick individual might be less socially connected in general, but they might spend the same time, or even more, with their family.”