Is bird song a reliable signal of aggressive intent? A reply

We advocate assessing the reliability of signals of aggressive intent by eliciting aggressive signaling from a subject, giving the subject an opportunity to attack a model, and testing whether the subject's displays predict a subsequent attack. Using this design, we found that most singing behaviors are poor predictors of attack in song sparrows (Melospiza melodia). Laidre and Vehrencamp (Behav Ecol Sociobiol, DOI 10.1007/s00265-007-0539-3, 2008) suggested altering our experimental design to make the model more realistic; it remains to be seen whether such design changes would change the association between display and attack. Laidre and Vehrencamp (Behav Ecol Sociobiol, DOI 10.1007/s00265-007-0539-3, 2008) also suggested that the reliability of soft song, the one display that predicts attack in song sparrows, can be explained by a vulnerability cost. We question the rationale for a vulnerability cost for this display and suggest instead that soft song has a competing functions cost, in that, by using soft song to counter an intruder, a male sacrifices other possible functions of vocal signaling.