New paper on bats using echolocation to identify each other

For echolocating bats, each bat is almost always producing biosonar pulses as it “looks” around. So when bats are flying or roosting in the same place, do they recognize the voices of all the bats nearby? Do they have an omnidirectional mental ‘image’ of who is around them at all times?

Several studies with different bat species have tested their ability to vocally recognize each other using only echolocation (1–4). Expectations about this ability are nuanced. On one hand, individual variation exists and persists in almost every biological trait (even behaviors of clonal fish raised in identical environments (5)), and recognizing this variation in echolocation calls can clearly serve a useful social function in bats (6,7). If echolocating bats are masters of sound, why would they not be able to recognize each other through their echolocation calls?

On the other hand, the design of echolocation calls is shaped by natural selection and also fine-tuned in every moment for finding prey and avoiding obstacles (e.g. 8), which should reduce the social information in these calls. There are different kinds of echolocation calls, some more variable than others.  Many bats also communicate using social calls, so echolocation calls might not need to convey anything. In other words, just because individual variation exists, doesn’t mean that animals actually use it for social recognition. For example, we recognize people by their face and voice, but we don’t recognize them by the iris of their eye, even though the human iris might provide a better individual signature than the face (and we might be able train a computer to identify people by their iris better than their face). So maybe the variation is there, but bats don’t care about it.

Some big questions are: How much social information is in the echolocation calls of different bats? How much of that information do bats perceive and use? And how does the use of social information in bat calls differ between species? For example, do bats that hunt together have more information in their echolocation calls than bats that forage alone?

Some of these questions are being addressed by Jenna Kohles, a PhD student working with Dina Dechmann at the Max Planck Institute for Animal Behaviour. Jenna is studying social information and social foraging in bats. She did her Master’s Thesis with Dina on vocal recognition using echolocation calls by velvety free-tailed bats, which we just published recently (9). Here’s a synopsis….

Since at least the early 1980s (10), it’s been well known that echolocating bats eavesdrop on others foraging nearby. When a bat catches an insect, it makes so-called “a feeding buzz” (because it sounds like a buzz on a bat detector). So it’s even possible for a person (or maybe a bat) to estimate the encounter rate of insects by listening to the rate of feeding buzzes. No doubt this information would be useful for a bat in search of prey! Hearing feeding buzzes is like being a vulture and seeing a group of other vultures circling around something in the distance—a good sign there’s food.

But bats also make other kinds of echolocation calls when searching for prey or commuting to foraging locations (search-phase calls). In some bat species, like Molossus molossus, these two types of calls are actually pretty different in their design (see call images below).

Molossus molossus or the velvety free-tailed bat is a neotropical insectivorous bat that lives in stable groups (11) and forage with group members in the same areas (12). Foraging with groupmates probably helps each group member find insect swarms faster than foraging alone. It’s also important to note that these bats are adorable.

Jenna caught M. molossus from six different social groups and recorded their search-phase calls in a standardized way. She analyzed the call structure and also conducted playback experiments to test for individual discrimination. She counted their head waggles as response (see video above). She found that she could use a relatively simple computer algorithm to assign calls to individual bats, and that the bats could also discriminate individuals using recordings of the calls. However, she found no evidence for group-level vocal discrimination, which is an ability that does exist in other species when using other (very different) kinds of calls: screech social calls in greater spear-nosed bats (13).

The lack of group-specific echolocation makes sense as this would require adult bats to learn and convergence in call type, which would decrease their ability to have an optimal call design for hunting. Also, there’s a tradeoff in call design: the greater vocal individuality, the more difficult it is to have calls be similar among groupmates. A similar tradeoff exists between calls that are stereotyped (search-phase calls) and ones that are highly flexible (like those used for avoiding obstacles). Search-phase calls are constant in structure over time (relatively constant within each bat), so it is actually easier for a computer (and presumably a bat) to discriminate individuals using these calls, as compared to the calls that bats make when approaching obstacles or prey, which are highly variable across contexts (hence relatively variable within each bat). In any case, variation in call structure cannot tell us what bats actually perceive, which is why Jenna used careful playback experiments.

Jenna did her vocal recognition tests under controlled conditions and her new work is taking her to the field using GPS loggers and microphones attached to a different bat species: Noctilio albiventris, which also conveys social information in echolocation calls and also hunts with group members (4,14,15). This bat forages over open water, so to measure prey abundance and bat feeding activity at their foraging sites, Jenna sets up camera traps and ultrasonic speakers from a kayak.

Jenna became interested in bat research as an undergrad and started working with Susan Loeb, a leading researcher on the ecology of forest-dwelling bats in the Eastern USA. She then began looking for summer research opportunities working on bat behavior in a neotropical Spanish-speaking country. She found exactly this opportunity in the form of an internship working with PI Dina Dechmann and then-postdoc Teague O’Mara (now a PI at Southeastern Louisiana University), who are doing ground-breaking work integrating ecology, physiology, social behavior, and animal tracking. Jenna worked on a project studying the behavior of young tent-making bats.

I first met Jenna when she was an intern for a second time at the Rachel Page Bat Lab at the Smithsonian Tropical Research Institute in Panama, working on a project which showed that bats can learn about novel prey cues, not only from observing  members of the same species, but also from members of other species (16).  At some point while in Panama, Jenna and I became friends, I think, because she laughs at my jokes even when they are not funny. Then we met up again when I did a short postdoc at the Max Planck Institute for Animal Behaviour, where she was doing her PhD. Anyway, it’s always been fun talking with her about bat behavior and watching her career develop. Keeping in touch and watching the development of younger scientists is perhaps one of the greatest joys in academia, which nobody tells you about as one of the rewards of academia, but it’s one of the biggest perks! I’ve also learned a lot from her and Dina about social foraging in bats, which is useful in that we are now looking at social foraging in vampire bats!

References