Vocal performance and the usage of song types in Pale-legged leaf-warbler Phylloscopus tenellipes: a contradictory study

Performance of vocal signals, such as birdsong, may be a subject to trade-offs among acoustic traits that limit the signal outcomes. Trilled songs in which syllables are repeated in rapid succession present males with a performance challenge resulting in a trade-off between trill rate (rate of repetition) and frequency bandwidth (range of frequency). Individuals of a species might differ in ability to produce high-performance songs that are close to a performance limit. It was suggested that high-performance songs might honestly reveal a high-quality singer, and both males and females might use those songs to detect high-quality singers. To test further this performance hypothesis, I asked whether the usage of song types depended on their performance in the Pale-legged leaf-warbler (Phylloscopus tenellipes), a species with individual repertoires of up to 11 trilled song types. I analyzed 125 song types (n = 25 males) taken from both spontaneous singing and singing elicited by conspecific playback. Songs showed the same performance trade-off between trill rate and frequency bandwidth as found in other passerines. Individuals did not differ relative to one another in their average vocal performance. Males did not preferentially use either high- or low-performance song types in response to playback-simulated territorial intrusion. They also did not modulate the performance of the same song type to signal aggression. Besides, song performance did not differ between predominant song types and all other song types. Since the predominant song type of a given male usually the only one that used in spontaneous singing, a receiver presumably cannot obtain information about the male’s performance ability from his spontaneous singing. Overall, my results did not confirm the role of vocal performance in male–male interaction.