Having studied roost selection by bats for his PhD, Ireneusz Ruczyński was struck by the fact that researchers knew a great deal about how bats find prey using echolocation, but no-one knew how they actually find somewhere to roost after a night's hunting. Teaming up with sensory ecologists Elisabeth Kalko and Björn Siemers, Ruczyński took on the challenge of designing a lab-based experiment to find out how noctule bats(Nyctalus noctula) from the Białowieża forest in Poland find their ideal tree roost(p. 3607).
Bats change roosts frequently, probably to reduce predation risk and exposure to parasites. As Ruczyński explains, one of the problems of investigating roost finding is the assumption that a bat's echolocation is good enough: `Finding prey in a forest isn't easy', he says, so it is likely that roost finding is also challenging for the bats. The team suspected that the bats would probably use a combination of cues, such as echolocation,vision and smell to help them find a cavity.
First the team trained the bats to find roosting cavities drilled into larger alder logs placed upright in a lit flight room. Each log had 8 cavities carved into it, which were a similar size to the cavities that the bats use in nature. At first, the bats would land on the log and find one of the entrances by chance, but once they had crawled inside an experimenter rewarded them with a tasty mealworm. The bats quickly learned to associate the mealworm treat with finding a cavity, so as their performance improved the team progressively blocked up each of the entrances until only one cavity was left open.
Once each of the bats could find most of the entrances within 5 minutes,using just their echolocation and vision, the team started experiments to find out which cues improved the bats' ability to find an entrance. They found that it took the bats around 40 s searching to find an entrance, and that they would either find entrances in flight, landing very close to the entrance and crawling in, or they would land on the log and find the entrance by crawling around on it. The team expected that low light levels, similar to those experienced in the early evening when the bats are active, would help them find the entrance, however, their performance didn't improve over echolocation alone. Putting cloths smelling of other bats in the holes didn't improve performance either, while heating up the inside of the cavities only helped the bats find the entrances a little quicker when they were crawling around on the log. When the team played echolocating calls recorded from roosting bats out of a speaker in the cavity, though, the bats found the entrances quicker,in around 20 s. They were also more likely to find the entrance from flight,and spent less time finding the entrance by crawling.
This shows that finding a roost from a distance is difficult and that social cues are important for helping bats to find a roosting site. `It means that sensory constraints may promote sociality', Ruczyński says. He suspects that when bats find good tree holes they remember where they are, as they are known to use the same cavities again and again.