Song Categories and Their Functions in the Field Sparrow (Spizella pusilla)

Abstract

-TWO song categories, distinguished by their acoustic structure, contextual usage, and potency in provoking responses from territorial males, are found in repertoires of individual male Field Sparrows (Spizella pusilla). Simple songs are sung spontaneously throughout the day and in long-distance countersinging by territorial males. Complex songs are sung most frequently during the dawn chorus and during territorial boundary chases. Most males sing one simple and one complex song type. Males responded more strongly to playback of complex songs than to simple songs. Our experiment supports the hypothesis that male songbirds respond differently to songs that are sung in different contexts. Received 13 November 1989, accepted 11 July 1990. IN MANY species of passerine birds, males sing two or more acoustically distinct song types (reviews in Krebs and Kroodsma 1980, Kroodsma 1982). Many hypotheses have been proposed to explain the proliferation of song types within repertoires. One class of hypotheses assumes that all songs within a species' repertoire contain the same information. Repertoires of redundant song types may be favored by both intraand intersexual selection. Large repertoires could function as effective keep-out signals by decreasing habituation in listening males (Krebs 1976), or by appearing to represent the presence of several singers (Krebs 1977). Large repertoires might also be favored by females during mate choice (Howard 1974, Catchpole 1980). Accordingly, songs should be used interchangeably and should provoke indistinguishable responses from listeners. A second class of hypotheses assumes that different song types within a repertoire provide different information to listeners. Different song types are sung in different behavioral contexts, and thus appear to provide different information to listeners (Morse 1970, Ficken and Ficken 1970, Lein 1978, Smith et al. 1978, Payne 1979, Jarvi et al. 1980, Sossinka and Bohner 1980, Catchpole 1983, Gaddis 1983, Nelson 1985, Derrickson 1987, Trainer 1987). In some warblers, males share different temporal patterns of song delivery in different contexts (MacNally and Lemon 1985, Staicer 1989). In species where song usage varies with behavioral context, different I Present address: Department of Zoology, University of California, Davis, California 95616 USA. songs may provide information on the singer's intentions or motivation (Lein 1978) and fighting ability (Lemon et al. 1987). In some species it has been suggested that some songs are directed primarily at females, while others within the repertoire are used primarily in male-male interactions (Morse 1966, 1967; Kroodsma et al. 1989). According to these hypotheses, which are not mutually exclusive, different song types should provoke different responses from recipients. To distinguish between these two forms of repertoire organization, we refer to song types that have different behavioral or contextual correlates as different song categories. In a similar way, Schroeder and Wiley (1983) grouped song types of the Tufted Titmouse (Parus bicolor) into classes based on usage and structure, and others have referred to song groups (Nolan 1978) or modes (MacNally and Lemon 1985). We suggest that three criteria must be met for song types to be separated into distinct categories: the song types must differ acoustically, they must be used in different behavioral contexts or reliably predict different subsequent behaviors, and they must elicit different responses from listeners in playback tests. If these criteria are met, a strong case is made that the songs serve different functions and convey different messages. Although we are concerned here with song-type variation, other aspects of song, such as the pattern of song delivery, could vary across contexts and so cause songs to be grouped into different categories. We documented the occurrence of two song types in male Field Sparrows (Spizella pusilla), a species in which males were previously re42 The Auk 108: 42-52. January 1991 This content downloaded from 207.46.13.21 on Tue, 27 Sep 2016 04:21:14 UTC All use subject to http://about.jstor.org/terms January 1991] Song Categories and Functions 43 ported to sing a single song type (Goldman 1973, Heckenlively 1976, Morrison-Parker 1977, Searcy 1983). Most studies of song repertoires have identified song types by subjective visual examination of sonograms. In Part 1 we present a quantitative acoustic analysis of song variation that supports the classification of songs into discrete acoustic categories, and we also estimate the size of individuals' song repertoires. In Part 2 we present observations of the contexts of singing. In Part 3 we describe an experiment to support the hypothesis that the two song types convey different messages, thus constituting distinct song categories. STUDY LOCATION AND OBSERVATIONAL