Animal signals are often complex, raising the possibility that different aspects of a signal may convey different types of information. Birdsong, for example, may simultaneously advertise the singer's condition during early life (through song complexity), location of song learning (through local song structure) and current condition (through song output). Each of these aspects of song is behaviourally salient: females generally prefer complex over simple song, local over nonlocal song, and high over low rates of song output. We examined the degree to which these three components of song convey complementary versus redundant information, in eastern song sparrows, Melospiza melodia melodia. One measure of song complexity, syllable repertoire size, was positively related to syllable sharing with other birds in the local population. This finding suggests that these aspects convey somewhat redundant information and that local-sounding song may in part advertise song-learning ability. A second measure of song complexity, song repertoire size, was positively related to song output, despite the widely differing developmental timescales over which song complexity and output are thought to vary. Condition during early development may have carryover effects on adult condition; alternatively, complex singers may have most to gain from advertising this attractive signal. Finally, despite our finding that both syllable sharing and song output were related to song complexity, syllable sharing and song output were not themselves significantly correlated. Overall, the three ostensibly independent aspects of song investigated here appear to overlap substantially in the information they convey.
We studied song complexity, syllable sharing and song output in 47 male song sparrows breeding near Newboro, Ontario, Canada in 2006 and 2007. Song sparrows are closed-ended learners: males do not add to or otherwise alter their song repertoires during adulthood (Nordby et al. 2002). As in other populations of this eastern subspecies (Hughes et al. 1998), males at our study site rarely share complete song types, but often share individual syllables. In a Pennsylvania population of M. m. melodia, partial song matching in response to song playback occurs more frequently than expected by chance, and is associated with more aggressive territorial defence (Anderson et al. 2005). Thus, even in populations where complete song sharing is rare, alternative forms of song matching appear to exist and to be behaviourally salient (Burt et al. 2002). At our Ontario study site, males with repertoires containing highly shared syllables also have larger cloacal protuberances, lower indexes of stress and fewer blood-borne parasites than males with repertoires consisting of less common syllables (Stewart & MacDougall-Shackleton 2008).
We investigated the relationship between song complexity and syllable sharing in 29 males breeding in 2006. We investigated the within-season repeatability of song output for 26 males breeding in 2007, the relationship between song complexity and song output for 22 of these males, and the relationship between syllable sharing and song output for 17 of these males. This sample includes eight males that bred in both 2006 and 2007 and were also included in the 2006 song complexity/syllable sharing analysis. We captured all subjects in mist nets or seed-baited Potter traps during April or May, and provided them with unique combinations of coloured leg bands to allow individual identification in the field. All subjects defended territories, attracted mates and initiated breeding at the study site.