Summary
Many territorial animals are less aggressive towards neighbours than they are towards strangers. This phenomenon is known as the ‘dear enemy’ effect and it occurs because strangers represent a considerably higher threat to territory take-over compared to neighbours. Some evidence has suggested that large repertoires may constrain neighbour–stranger discrimination. We tested whether songbirds with large repertoires exhibit neighbour–stranger discrimination, conducting a playback study on a songbird with a large vocal repertoire, and a comparative analysis of the dear enemy effect across all published studies of songbirds. In our playback study, we broadcast neighbour and stranger songs within the breeding territories of red-eyed vireos, Vireo olivaceus, a songbird species with a large song repertoire (ca. 50 songs per individual). Vireos responded significantly more aggressively to playback of stranger versus neighbour songs; subjects approached closer to the loudspeaker, had a lower latency to approach the loudspeaker, spent more time near the loudspeaker and sang more soft songs during stranger trials than during neighbour trials. We examined song sharing between red-eyed vireos and found low levels of song sharing between neighbours, suggesting that red-eyed vireos may discriminate among conspecifics based on individually distinctive song types. We then conducted a comparative analysis of neighbour–stranger discrimination across the published literature on songbirds, using a phylogenetically controlled analysis to explore whether species with large repertoires are less likely to discriminate between neighbours and strangers. Across 34 species, we found no evidence that songbirds with large repertoires are constrained in their ability to distinguish between neighbours and strangers. We conclude that large song repertoires do not inhibit neighbour–stranger discrimination in red-eyed vireos specifically, or songbirds generally.
Methodology
We conducted a playback study with red-eyed vireos at the Queen's University Biological Station (44°34′N, 76°19′W) north of Kingston, Ontario, Canada. We conducted playback experiments from 24 May to 4 July 2015, a time when all red-eyed vireos at our site had established their breeding territories and when most subjects were incubating eggs or in the early stages of chick rearing. We studied 28 males occupying breeding territories in eight different woodlots at our study site (average ± SE distance between woodlots: 593.5 ± 92.9 m). The birds were not banded, and instead we relied on location information and features of acoustic recordings to distinguish between different males (sex was identified by song, because only males sing in this species; Cimprich, Moore, & Guilfoyle, 2000). We identified individuals by following birds on their breeding territory, paying careful attention to the movement patterns of each of our subjects, monitoring the song posts and perches they used and the parts of the forest they occupied. We verified the identities of individuals by comparing recordings of the songs they sang during playback trials to songs we collected in previous focal recordings collected during observation sessions. We based our analyses of repertoire size and song sharing on 21 males where we had recorded at least 250 songs from each bird. We based our analysis of playback responses on 21 males (14 of these males were the same males used for repertoire and song-sharing analyses) after excluding three males due to uncertainty regarding identity (no shared songs were detected during playback trials when compared to previous recording sessions), two males due to neighbour interference during playback and two additional males due to a lack of response to our playback stimuli.