New paper: how vampire bats form new food-sharing bonds.

We have a new paper out today in Current Biology and it is featured in the New York Times (pdf), National Geographic, CNN, Scientific American, Discover, Phys.org, Cosmos and I have two radio interviews tomorrow for NPR and BBC.

Doing these media interviews has been a pleasant ‘distraction’ from the coronavirus pandemic which is days away from exploding in the USA. There’s not much that most of us can do right now but brace for impact.

Talking to science journalists is nice, but some journalists/bloggers ask me strange questions. One kept asking me how studies on social behavior in vampire bats could ever help people. I kept trying different angles with no success. First, I tried to talk about the evolution or emergence of cooperation, but somehow that didn’t get me far. Then I said, “Also, some general features of social networks are common in human and nonhuman animals” and he said “But how can understanding social networks help people?” And I said “Well, for example, it could help scientists model and predict how pathogens spread?” And then he said, “Ok, but why is that important?” [??!!] Mind you, this is in the midst of the coronavirus pandemic predicted to kill 480,000 people in the USA alone. 🤨

Another frustratingly common question I get is “What is the vampire bat’s function in the ecosystem? What is their purpose?”. I used to ask, “What do you mean?” But my new preferred response is “Oh, just the same as everything else– they have the same purpose as humans, butterflies, and grass”.

Here’s a nice email interview I did for an article in Gizmodo. My responses are in bold.

Very quickly, tell us (1) about vampire bat grooming behavior and (2) the blood meal sharing (not in terms of why, but how).

They groom each other by licking, like a cat. They share food by regurgitating, like a bird feeding its nestlings.

How does this cooperative behavior fit into the context of the selfish gene. Put another way, how does a trait like this manage to evolve and appear over time? 

The selfish gene is an important idea for thinking about cooperation: genes help copies of themselves in the same body or in other bodies. Not surprisingly, most food sharing in vampire bats occurs among close genetic kin; often it is extended parental care. By donating food you are helping copies of your genes in your relatives. But food sharing is not limited to kin and it is probably an adaptive trait for several other reasons. Food donations appear to promote reciprocal donations back to the original donor. That’s a hypothesis for which we have increasing evidence. Every bat has a different food donor network, which can vary in size, and vampire bats that feed more nonkin have more donors and cope better with the loss of a donor in their social network. So bats that are more generous to nonkin might do better in the long run by having more ‘backup’ donors.

What did your research uncover, that previous research did not?

To my knowledge, nobody had yet rigorously measured how two animals go from being ‘strangers’ to having a natural cooperative relationship, where one partner would even help the other at a cost to themselves. Most cooperation occurs in families (like cooperative breeders) or it is based on fixed mechanisms (like when plants and fungi trade nutrients). People who study how animals make more complex decisions about cooperation tend to either observe established relationships in nature (like chimpanzees that groom each other and form coalitions), or they study unfamiliar animals that perform a trained instrumental helping task in the lab (like a rat or a monkey that pulls a lever to deliver food to partner in another cage). Each of these approaches has its own costs and benefits. What we did was introduce unfamiliar vampire bats in captivity, so we could monitor them, and then let them form a cooperative relationship as they would in nature. You can simulate a bat’s natural social life in captivity because they are small and live in tight spaces like hollow trees. To induce helping, we just fasted a bat which gave each stranger the opportunity to help it. We did not have to train them, and we got more observations than you could from a purely observational field study. By controlling who met who, we could watch them go from strangers to food donors (or fail to bond at all).

And what is this telling us about vampire bats that we didn’t know before? 

First it tells us that the bats are not just helping each other because of kin selection. For example, previously, you could imagine that two unrelated vampire bats that happen to smell similar might help each other in a reciprocal fashion due to simple heuristic like  ‘help individuals that smell similar to you’. This would give the false appearance of two bats returning favors and having some kind of reciprocal relationship, even if it was purely a byproduct of a trait that originally evolved to help kin. From watching these bats for hundreds of hours, I have always thought this was unlikely, but it remained a possibility.

Now, this study rejects this idea. The bats form new nonkin relationships in a manner that depends on the social environment and how the social interactions change over time. Bats that groom each other increasingly over time are the ones that start sharing food reciprocally. The grooming increases up until the food donations start. If they meet as an isolated pair, they start sharing food much faster. If the bats just helped each other based on similarity, you would not see these patterns over time.

Very quickly, tell me about the experimental set up and some of the thinking that went behind it.

We introduced bats as isolated pairs of two strangers. We also introduced bats as 1 ‘out-group’ stranger and 3 ‘in-group’ bats. Then we put two whole groups together. In each case, we repeatedly fasted all the bats in the group, but one at a time, so that everyone had to the opportunity to help and to be helped, and we looked at how both grooming and food-sharing rates changed over time. The whole experiment took around 15 months.

Who tended to instigate the bonding? The donor or the recipient? In either case, what is this telling us about the behavior? 

Each bat could be a donor and a recipient multiple times. If bat A helped B for the first time, this made it more likely that B would later help A for the first time.

I love this: “The idea of using low-cost behaviors to build up to higher-cost investments can be something of much more general importance outside just food sharing in vampire bats.” Hoping you can elaborate, and possibly speak to other species.

Yeah, it could be really important for other species that cooperate. In humans or other primates, when a new individual joins a group, she needs to socially integrate. So maybe she tries just sitting near another female. Maybe they cluster to say warm. Then maybe she grooms the other a little bit. If that goes well, maybe she grooms her more. Eventually, there’s a reciprocal grooming relationship and maybe this leads to something more risky, like sharing food or forming a coalition. That’s the basic idea.

There are these fascinating fish called cleaner fish that engage in cooperative cleaning of other larger fish, called ‘clients’, even client fish that could eat them. The cleaners eat the dead skin or ectoparasites from their clients, which is mutually beneficial, but they can also exploit the client by biting their live tissue or mucous. The cooperation is enforced by both parties. Interestingly, the cleaner fish often do a little light massage of the client fish before they clean them. It’s almost like when you build-up trust by letting an unfamiliar dog sniff your hand and you give it a little pet on the head. You’re saying “See, I touched you and nothing bad happened’.

This behavior of raising the stakes‘ or ‘testing the waters’ could also be important in behaviors like courtship and mating. In humans, before marriage there is dating, and before dating, there is that period where you just figure out: is this stranger even a safe person to be around?

Even when you think about dominance interactions, it’s obvious that you want to escalate gradually. There are many other social situations where animals might want to ‘test the waters’.

The most important idea behind the original theory is to recognize that cooperative behavior is not usually a binary act like “cooperate” or “defect”. It’s often a continuously variable investment.

---