When performing playback experiments with birds, is there a biologically relevant source level for song playbacks?

I have read that a standard value is 90 dB @ 1 m, however I could not find a good reason that justified this value. Is there evidence that songbirds' source level is really that high? Is it not different across species? What happens if the playback source level is too high? Is it transformed into a super-aggressive signal? And can we compare results obtained with playbacks of different amplitudes?

Sorry for the beginner's question....

  • 1
    Hi @lframond, could you please provide reference/s for the stated convention?
    – Thejasvi
    Jul 27 at 7:41

2 Answers 2


It is not all all a beginner's question, on the contrary, in underwater behavioral response studies a lot of money was spent to realize that responses are context dependent.

So, answer to OP is that the playback sound level should be adapted to the experiment you are carrying out. Varying sound levels is fine if one wanted to determine the onset of a certain behavior as function of sound intensity for the otherwise fix context.

Keep in mind, it is the received sound that counts and not the transmitted sound level, however, reaction to weak predator calls at close distance may be interpreted differently than that of somewhat louder calls at larger distances. Animals are smart enough to deduce distances from received signals and react differently.

Stacy DeRuiter's paper describes such a case for beaked whale responses to simulated and real naval sonar.

Ref: DeRuiter, Stacy L., et al. "First direct measurements of behavioural responses by Cuvier's beaked whales to mid-frequency active sonar." Biology letters 9.4 (2013): 20130223.


In playback experiments (any animals), it is crucial to monitor or estimate the sound level at the receiver's location, and not only at a random/standard distance to the speaker.

  • If you play back "natural" recordings, you should adjust the amplitude of your signal in order to have a sound level at the receiver's position that is similar to what happens in natural conditions.
  • if you play back sounds that does not exist in natural conditions, you should adjust the amplitude of your signal in order to comply with the sound levels which you want to test at the receiver's position.

Otherwise, your tested animal may not hear the sound if it is too quiet or behave unexpectedly if the sound is too loud.

Then, you should not compare results with different sound levels, except if the sound level has been included as an explanatory variable in your statistical model, provided you have enough replicates for each sound level etc.

PS: The sound level at e.g. 1 m can be reported for information about the power of the playback source, but it I don't think it should be a standard (except maybe for very particular experimental protocols), since it is not related to the receiver's perception of sound level.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.