BIRD VOCALIZATION

Vocalization is a general term encompassing all avian sounds produced by voice: song, calls, distress cries, alarm calls etc. Many bird songs and calls comprise a mixture of tones (harmonic sounds) and noises; for some species there is prevalence or an exclusive presence of tones, in other cases noises constitutes the whole repertoire. Spectrograms show the acoustic spectrum of each sound so that much information about the tone quality, pitch, timbre and general structure may be deduced. Pure tones are represented by relatively narrow lines, while harmonics are shown as multiple parallel lines. Harmonics are components of the waveform having frequency, which is a simple multiple of the fundamental (that is the lowest frequency component). Other factors which affect the quality of a tone and which may be read from a spectrogram are frequency modulation and amplitude modulation. The former is like the vibrato in a singer, but the modulation rate is much faster in birds than in humans; the latter is a waxing and waning of the amplitude (volume) of a sound; it may be coupled with frequency modulation.
Noise, i. e. those sounds lacking a sufficient concentration in energy within narrow margins of a principal frequency to give tones of identifiable pitch, appear as broad band patches of energy. The frequency range is often wide and may encompass the whole available spectrum.

The spectrogram in Fig. 1a represents a harsh call of a Magpie (Pica pica): this is a typical noisy call, whose energy is a broad undefined patch. However it is interesting to note that the dominant frequencies range between 1500 to about 4500 Hz, a closer range than that of the whole spectrogram.
Fig. 1b shows the whistling song phrase of a Greenshank (Tringa nebularia). In this case the four syllables that compose the song are all melodic tones (at 2500/3000 Hz), each having one harmonic component about twice the fundamental frequency. Here the dominant frequency is also the fundamental one (= the lowest frequency component).
In Fig. 1c is depicted a Blackbird (Turdus merula) song phrase: the first syllable is a whistle with frequency modulation (= rapid increase and decrease), then we can see a brief harsh note and another pure whistle with frequency modulation. This song ends with a long harsh sound, which is a spectrographic broad patch recalling that in Fig 1a. Thus song in Fig. 1c represents a typical combination of tones and noise. The blue numbers above the waveform in Fig. 1c help to identify the four song syllables described above. Also in spectrogram 1c many additional information can be drawn analyzing the dominant frequencies: the whistle frequencies are always dominant, while the harsh ending note is characterized by a dominant band around 4500/5000 Hz, more clearly defined than that in spectrogram 1a.

Whatever structure characterizes a bird vocalization, it may be classified in some different categories:
SONG is a relatively complex pattern of sounds in time that may be repeated exactly and is recognizable not only at the specific level but often at group and individual levels. (Bibliography 1)
FULL SONG is usually under the control of the sex hormones and its utterance is therefore largely confined to the breeding season. Often song and full song are the same. (Bibliography 1) Song is commonly referred to passerine birds only, while sexual and territorial vocalizations of non-passerines are typically simpler than those of passerines and therefore are generally defined as calls. Moreover song learning is characteristic of passerine birds, while non-passerine calls are considered to develop without a learning process.
SUBSONG denotes forms of quiet song. It refers to a quiet, extended warbling in which fragments of the full song may be heard, often with imitation of other species. A bird uttering subsong is usually perched low and in an inconspicuous place; the beak may be closed while singing. Subsong is not confined in out-of-season utterances and may readily be heard during the breeding season. (Bibliography 1)
CALL is a sound of brief duration used mainly to alert to danger and to coordinate group behavior. For passerine birds most of these calls do not vary with the season and in contrast to song, are not strongly influenced by the state of seasonal development of the sex hormones. Calls can be further classified as alarm calls, flight-calls, flock-coordination calls, advertising/territorial calls etc. (Bibliography 1). Non-passerine birds have calls with sexual and territorial functions analogous to passerine songs.
DUETTING occurs when the members of a pair sing simultaneously or antiphonally (precisely timed alternate singing) as part of the courtship display, to maintain the pair bond or to maintain contact. (Bibliography 1)

In these pages I have adopted some definitions of the components of a bird song performance as follows:
SONG SEQUENCE or SONG BOUT is an uninterrupted singing passage of different duration (from few seconds to several minutes) that is characterised by a nearly constant number of utterances per unit time. This singing passage may end for spontaneous bird behavioural changes or for external causes (predator presence, noises etc).
PHRASE is a relatively short sound passage forming part of a song sequence. Different phrases or several repetitions of the same phrase make a song sequence. However a song sequence may be also made by only one phrase.(Bibliography 18)
SYLLABLE is used to describe an elementary unit that forms part of a phrase, as shown on the spectrogram and waveform representations.(Bibliography 18)
ELEMENT is the smallest building block which is part of an entire syllable (when this is a complex one). One definition of an element is simply a continuous line on spectrogram.(Bibliography 18)
Fig. 2 and 3 below illustrate these criteria.

Fig . 2 shows a 20 second fragment over 1-minute song bout of a Cirl Bunting (Emberiza cirlus). Fig. 3 blows up one phrase of the same song and shows the ten syllables constituting it (each syllable is composed by two elements). It must be said that the boundaries between song element, syllable and phrase are sometimes not so clear cut as showed above; some species can have far more complex songs with blurred boundaries between the different components. Note that there is not a standard bioacoustic terminology universally recognized, therefore the terminology used in this website, altough widely applied by many research workers, is still subjective.
A concluding remark: every utterance, although well characterised and described, looses significance if it is not evaluated in relation to its role in behaviour, communication and location, to its relevance and reliability in specific identification, to its linkage with surrounding circumstances, physiological state, sex and age.

After this discussion about the general characteristics of bird vocalization I include a few possible topics for specific studies (see Bibliography 1) that could help to clarify some aspects of bird vocal behaviour.

1. Intraspecific variation of vocalization (a) Regional, (b) Group, (c) Individual.
2. Song learning.
3. Vocal repertoires and call functions.
4. Presence and consistence of certain species that are difficult to see (e.g. owls) in a delimited area utilizing automatic recording equipment.
5. Diurnal variation in song.
6. Study of song outside the breeding season.
7. Development of song in the individual.
8. Vocal mimicry.