Flight phases in the song of skylarks: impact on acoustic parameters and coding strategy

Abstract

Skylarks inhabit open fields and perform an aerial song display which serves as a territorial signal. The particularly long and elaborate structure of this song flight raises questions about the impact of physical and energetic constraints acting on a communication signal. Song produced during the three distinct phases of the flight - ascending, level and descending phase could be subject to different constraints, serve different functions and encode different types of information. We compared song parameters during the ascending and the level phases. We found that the structure of the song varied with the phase of the flight. In particular, song had a higher tempo when skylarks were ascending which might be related to higher oxygen and energetic demands. We also explored which phase of the song flight might encode individuality. Earlier studies reported that skylarks reduced their territorial response to established neighbours if the neighbour song was broadcasted from the correct adjacent boundary, but reacted aggressively if the neighbour songs were broadcasted from an incorrect boundary (mimicking a displaced neighbour). Such differential response provides some evidence for individual recognition. Here, we exposed subjects to playback stimuli of neighbour song in which we had replaced either the song produced during the level or the ascending phase by the relevant song of the neighbour from the incorrect border. Singing response was higher towards stimuli in which the 'level phase song' was replaced, indicating that skylarks could be able to recognise their neighbours based on song of this phase. Thus, individuality seems to be primarily coded in the level phase of the flight song.

Figure 1. Sound spectrogram of a complete flight song of a male skylark (above) and of two excerpts of the same (below).

Dotted line squares highlight 20 s of song produced during the ascending and the level phases as used for acoustic analysis and playback; fine structure of the syllables produced in each phase is shown below (A and B) at higher scale for the song in the continuous line squares. The arrows indicate the transition between the song produced during the ascending, the level and the descending phase.

Figure 2. Sound spectrogram of a short part of a flight song of a male skylark.

Each syllable is labelled with a number and syllables belonging to the same syllable type have the same number.

Figure 3. Design of the playback experiment.

S is the subject, with its left and right neighbours established respectively in the territories at the left side (NL) and the right side (NR). The composition of each stimulus broadcasted by the loudspeaker placed at either boundary is indicated in the box. Phases of songs broadcasted from the correct side (belonging to the correct neighbour) are in bold. Dotted lines indicate the territorial borders of subjects S, NL and NR. S1 indicate stimuli in which song produced during the ascending phase was broadcasted from the correct and song produced during the level phase from the incorrect side, S2 indicate stimuli in which song produced during the level phase was broadcasted from the correct and song produced during the ascending phase from the incorrect side.

Figure 4. Skylarks respond more strongly to S2 stimuli.

Mean scores (±SE) of the first and the second principal components for responses to the two stimuli (S1: level phase of flight song from neighbour of the opposite side; S2: ascending phase of flight song from neighbour of the opposite side). * p = 0.024. Variance explained by the scores is given in parentheses. There was no effect of the order of presentation (GLM, F = 0.53 p>0.1).