Temporal Relationships in White-Throated Sparrow Song

Abstract

Bird song has been likened to human music for centuries (e.g., Anonymous 1717, Hartshorne 1973). The analogy between bird song and music is perpetrated to some degree by the use of musical terms to describe avian vocalizations: individual sound elements are called and strings of notes are called songs. In most modern studies of song this analogy has fallen out of favor (e.g., Dobson and Lemon 1977). However, recent studies have shown that fixed relationships between the frequencies of adjacent notes, a defining feature of human music, occur in the songs of several species, and the birds seem to use this feature to recognize their songs (e.g., see Weary et al. 1991). The aim of the present study was to determine if there are other predictable relationships between notes. In particular we investigated whether the duration of song notes sung by White-throated Sparrows (Zonotrichia albicollis) are correlated. When an individual produces a relatively long first note, do the subsequent notes tend to be longer? These temporal relationships between notes are termed rhythm when describing music. The familiar tune is still recognized when it is sped up or slowed down (changes in tempo), because the temporal relationships between the notes (and internote intervals) are maintained. The only formal study of rhythm perception in nonhuman animals was done in the laboratory on European Starlings (Sturnus vulgaris) (Hulse et al. 1984). This study showed that birds can learn to discriminate rhythmic from arrhythmic sound patterns, and can generalize this discrimination over changes in tempo. How important are the temporal relationships among sounds in nature? For birds, there is some evidence that these relationshipsare important. Becker's (1982) review pointed to several temporal features of song that are important in species recognition. For example, the manipulation of the intervals between notes in a song affects recognition in several species, including White-throated Sparrows (Falls 1963). Also, interspecific comparisons have demonstrated that birds which sing longer songs also tend to use greater intervals to separate them (Dobson and Lemon 1975, Weary and Lemon 1988). Thus, certain aspects of timing do seem important in bird song. Other evidence is harder to interpret. For example, Lambrechts and Dhondt (1987) found that during the course ofa song, Great Tits (Parus major) tend to sing with progressively longer intervals between notes. The rhythmic aspect of this behavior is that when Great Tits produce longer inter-note intervals, they also produce longer inter-phrase intervals (phrases consist of groups of notes). The arrhythmic aspect is that there is no corresponding increase in the duration of the notes themselves. Thus, in one respect Great Tits maintain rhythm over a change in tempo, but in another they do not.