Bats 'recognise others' voices'

As if flying around in the dark swooping and diving to catch insects was not tricky enough, bats also listen for their fellow hunters.

A study has revealed how these winged mammals recognise other bats' voices.

They are able to differentiate the ultrasonic "echolocation" calls that other bats make as they navigate.

In the journal PLoS Computational Biology, the scientists report that the bats have an internal "reference" call to which they compare others.

Yossi Yovel from the Weizmann Institute of Science, Israel, and his colleagues from the University of Tuebingen in Germany recorded the echolocation calls of five greater mouse-eared bats.

The bats use these brief bursts of sound in sonar navigation - bouncing sound waves off their surroundings to find their way and locate prey.

Dr Yovel's team tested the bats' ability to identify the others by playing the recorded sounds to them.

"Each bat was assigned two others it had to distinguish between," Dr Yovel explained. "So we trained bat A on a platform, playing a sound from bat B on one side and from bat C on the other. He had crawl to where the 'correct' sound was coming from."

Each of the subjects was taught that a call from just one of the other bats was correct.

So during this training exercise, if the bat A made the right choice, and crawled towards the sound from bat B, it was rewarded with its favourite food - a mealworm.

"Then, in the next stage - the test - we rewarded them no matter what choice they made, and they still chose correctly more than 80% of the time," said Dr Yovel.

"So we knew the bats were able to distinguish individuals. But it wasn't clear what they're using to discriminate one from the other.

"If you think of this in comparison with humans, it's like being able to recognise a person just by listening to the same one-syllable yell in different voices.

"The bats learned the voice by listening to hundreds of very short 'yells', but they then were able to recognise an individual based on one single yell."

Modelling sound

In the second part of the study, Dr Yovel's team designed a computer model to mimic the way in which the bats compared the different sounds.

"The model takes all the calls the bat thought were A, and all the calls it thought were B, and tries to understand what differences it is using to match them up," said Dr Yovel.

"Our analysis showed that each bat has a typical distribution in the frequencies it emits, probably a result of the differences in each animal's vocal chords."

He thinks the bats may have an internal "prototype" - a sort of reference sound against which they can compare these subtle differences.

This could explain how bats remain in a group when flying at high speeds in darkness, and how they avoid interference between each others' echolocation calls.

RELATED INTERNET LINKS
PLoS Computational Biology
University of University of Tuebingen
Weizmann Institute of Science
The BBC is not responsible for the content of external internet sites