What makes a ‘local song’ in a population of ortolan buntings without a common dialect?

Sing softly to evoke a response only from a recent intruder

Low-amplitude soft songs have been described in many birds, but in most species, research has addressed only broadcast songs. Soft songs may have a similar or distinctive structure in comparison to broadcast songs produced in order to defend a territory and attract females. In some species, such soft songs were found to be produced in an aggressive context and were the best predictors of conflict escalation and later physical attack. However, such observations are not consistent across all species studied. Several hypotheses have been put forward to explain the function(s) of soft songs and why they are so quiet. Studies on the ortolan bunting (Emberiza hortulana) show that males produce soft songs similar in structure to broadcast songs during close interactions with conspecific intruders. However, experiments with the playback of loud and soft songs as well as taxidermic mount presentation revealed that soft song in this species does not fulfil aggressive signal criteria. Here we performed an experiment with two loudspeakers simulating movements of the intruder in order to test whether the soft songs are used to evoking a response from a nearby but not localised rival. We found that males responded with similar strength to songs played back from single and double speaker systems. Simultaneously, males produced more soft songs during and just after the phase of the experiment in which we simulated short flights of the intruder. Our results indicate that soft songs might be used during close interactions with rivals without being an aggressive signal and are used as short-range signals to check if the rival is still around. Our results also provide an alternative explanation of soft song behaviour in comparison to the hypotheses of eavesdropping avoidance and readiness to fight.

Trade-offs in the production of animal vocal sequences: insights from the structure of wild chimpanzee pant hoots

Background

Vocal sequences - utterances consisting of calls produced in close succession - are common phenomena in animal communication. While many studies have explored the adaptive benefits of producing such sequences, very little is known about how the costs and constraints involved in their production affect their form. Here, we investigated this issue in the chimpanzee (Pan troglodytes schweinfurthii) pant hoot, a long and structurally complex vocal sequence comprising four acoustically distinct phases – introduction, build-up, climax and let-down.

Results

We found that in each of these phases, and for the sequence as a whole, there was a negative relationship between the number of calls produced and their average duration. There was also a negative relationship between the total duration of some adjacent phases. Significant relationships between the fundamental frequency of calls and their number or duration were found for some phases of the sequence, but the direction of these relationships differed between particular phases.

Conclusions

These results indicate that there are trade-offs in terms of signal duration at two levels in pant-hoot production: between call number and duration, and between the relative durations of successive phases. These trade-offs are likely to reflect biomechanical constraints on vocal sequence production. Phase-specific trade-offs also appear to occur between fundamental frequency and call number or duration, potentially reflecting that different phases of the sequence are associated with distinct types of information, linked in different ways to call pitch. Overall, this study highlights the important role of costs and constraints in shaping the temporal and acoustic structure of animal vocal sequences.

Electronic supplementary material

The online version of this article (10.1186/s12983-017-0235-8) contains supplementary material, which is available to authorized users.

Song duration mediates responses of territory owner in a songbird species with a small song repertoire

Song is a sexually selected trait that is involved in mate attraction and territory defence in birds. Songs may convey information about different male quality components. They are flexible in terms of frequency, amplitude, and duration. Although changes in song duration are common, the function of this behaviour has been studied less strongly. It is known that song duration changes within a singing bout and may provide information about aggressive motivation. We tested whether the elongation and shortening of songs affects the responses of territorial ortolan bunting males to neighbour song playback. If changing song duration signals level of aggressiveness, then songs differing solely in duration may elicit behavioural responses of different strength. We performed two tests with different males assigned to two experimental groups and measured approaching and vocal response. In Experiment 1, we tested 18 males, which responded to the playbacks of elongated and normal (control) neighbour songs. In Experiment 2, we tested 17 males, which responded to the playbacks of shortened and normal (control) neighbour songs. Males responded significantly stronger to longer songs in both experiments as measured by the approach variable. Vocal response was not affected by treatment, but it was affected by the order of playback presentation. Our results indicate that song duration might be used for signalling current aggressive motivation during close interactions with rivals.

Sequential information in a great ape utterance

Birdsong is a prime example of acoustically sophisticated vocal behaviour, but its complexity has evolved mainly through sexual selection to attract mates and repel sexual rivals. In contrast, non-human primate calls often mediate complex social interactions, but are generally regarded as acoustically simple. Here, we examine arguably the most complex call in great ape vocal communication, the chimpanzee (Pan troglodytes schweinfurthii) 'pant hoot'. This signal consists of four acoustically distinct phases: introduction, build-up, climax and let-down. We applied state-of-the-art Support Vector Machines (SVM) methodology to pant hoots produced by wild male chimpanzees of Budongo Forest, Uganda. We found that caller identity was apparent in all four phases, but most strongly in the low-amplitude introduction and high-amplitude climax phases. Age was mainly correlated with the low-amplitude introduction and build-up phases, dominance rank (i.e. social status) with the high-amplitude climax phase, and context (reflecting activity of the caller) with the low-amplitude let-down phase. We conclude that the complex acoustic structure of chimpanzee pant hoots is linked to a range of socially relevant information in the different phases of the call, reflecting the complex nature of chimpanzee social lives.

Phonological-dependent territorial responses in yellowhammers (Emberiza citrinella)

In songbirds, males sing to attract potential mate and to defend their territory. Information about the sender such as its sex, its motivational state, its strength or identity can be encoded through subtle modifications in songs. In this study, we investigated whether territorial responses of yellowhammer males may be affected by modifications of song syntax and phonology. Yellowhammer song is mainly composed of three elements among which the last one is a long and low frequency note that is supposed to be the main component used by males to assess their potential rival. We carried out field experiments and played back either fully natural songs or partly natural and partly artificial songs to test this hypothesis. We identified that phonology rather than syntax modifications altered territorial responses. Our results also suggest that the phonology of the first song element plays a critical role in driving territorial responses.

Acoustic model adaptation for ortolan bunting (Emberiza hortulana L.) song-type classification

Automatic systems for vocalization classification often require fairly large amounts of data on which to train models. However, animal vocalization data collection and transcription is a difficult and time-consuming task, so that it is expensive to create large data sets. One natural solution to this problem is the use of acoustic adaptation methods. Such methods, common in human speech recognition systems, create initial models trained on speaker independent data, then use small amounts of adaptation data to build individual-specific models. Since, as in human speech, individual vocal variability is a significant source of variation in bioacoustic data, acoustic model adaptation is naturally suited to classification in this domain as well. To demonstrate and evaluate the effectiveness of this approach, this paper presents the application of maximum likelihood linear regression adaptation to ortolan bunting (Emberiza hortulana L.) song-type classification. Classification accuracies for the adapted system are computed as a function of the amount of adaptation data and compared to caller-independent and caller-dependent systems. The experimental results indicate that given the same amount of data, supervised adaptation significantly outperforms both caller-independent and caller-dependent systems.