As anyone who has ever heard a frog chorus can testify, the cacophony of trills, peeps, croaks and whistles make it difficult for us to distinguish individual frogs from the background din. Female frogs, though, have two inner ear organs that are relatively more sensitive to the range of frequencies that their males produce, which helps the females find them. Because different frog calls have different sound properties, Carl Gerhardt and colleagues wanted to know how this affected females' preferences for different calls in four closely related species of American treefrog(p. 2990).

Three of the treefrog species – two species of grey treefrog and the canyon treefrog – all produce calls with two distinct sound peaks at different frequencies. The fourth species, the bird-voiced treefrog, has a single high-frequency peak in its song. The team wanted to test two competing hypotheses for how frogs recognise sounds. The matched-filter hypothesis predicts that the nervous system is more sensitive to the specific frequencies in a male's call, so if a call has two sound peaks, a signal with both peaks is preferred over a call with just one of the peaks. The pre-existing-bias hypothesis takes into account that there is sometimes a mismatch between female preference and the calls that males produce, which could lead to sexual selection on male calls.

Using simplified electronic signals which accurately mimicked a natural call, the team tested females' responses in sound-proof rooms to calls played out of one of two speakers 2 m apart. If a female preferred a call, she would move towards the speaker producing it. Comparing the preferences of the four species to the number of sound peaks in a call, they found that grey and canyon treefrogs with double-peaked calls preferred this type of call over a call with only one of the frequency peaks. The bird-voiced treefrog preferred the calls with single high-frequency peaks, not two, just like the males'natural calls. This finding was consistent with the matched-filter hypothesis,and not the pre-existing-bias hypothesis.

Knowing that the higher-frequency peak in the call stimulated one inner ear organ, while the lower frequency peak stimulated the other, the team tested which peak the frogs preferred when given a choice between a two- or one-peaked call. The bird-voiced treefrogs, as expected, overwhelmingly chose the single high-frequency peak. They found that the other frogs responded differently, however: some populations of grey treefrogs, for example,preferred the high-frequency peak much more than grey treefrogs from another population in another part of the country. This suggests that there is probably a bias towards certain frequencies in the nervous system of different frogs, and that different populations could evolve to prefer and produce different calls.

So did female grey treefrogs prefer high- or low-frequency peaks? The team found that preference depended on sound intensity: both species of grey treefrog preferred lower frequencies, which stimulate one ear organ, at 65–75 dB. At 85–90 dB, however, one species preferred higher frequencies, which stimulate the other ear organ, while the other had no preference. This means that females' frequency preference differs depending on how intense the sound is: in other words, how close she is to a male. With this in mind, researchers will have to be very careful when designing future experiments.

Gerhardt, H. C., Martínez-Rivera, C. C., Schwartz, J. J.,Marshall, V. T. and Murphy, C. G. (
). Preferences based on spectral differences in acoustic signals in four species of treefrogs(Anura: Hylidae).
J. Exp. Biol.