No reproductive fitness benefits of dear enemy behaviour in a territorial songbird

Territorial animals often respond less aggressively to neighbours than strangers. This ‘dear enemy’ effect is hypothesized to be adaptive by reducing unnecessary aggressive interactions with non-threatening individuals. A key prediction of this hypothesis, that individual fitness will be affected by variation in the speed and the extent to which individuals reduce their aggression towards neighbours relative to strangers, has never been tested. We used a series of song playbacks to measure the change in response of male great tits to a simulated establishment of a neighbour on an adjacent territory during early stages of breeding, as an assay of individuals’ tendencies to form dear enemy relationships. Males reduced their approach to the speaker and sang fewer songs on later playback repetitions. However, only some males exhibited dear enemy behaviour by responding more strongly to a subsequent stranger playback, and when the playback procedure was repeated on a subset of males, there was some indication for consistent differences among individuals in the expression of dear enemy behaviour. We monitored nests and analysed offspring paternity to determine male reproductive success. Individuals that exhibited dear enemy behaviour towards the simulated neighbour did not suffer any costs associated with loss of paternity, but there was also no evidence of reproductive benefits, and no net effect on reproductive fitness. The general ability to discriminate between neighbours and strangers is likely adaptive, but benefits are probably difficult to detect because of the indirect link between individual variation in dear enemy behaviour and reproductive fitness and because of the complex range of mechanisms affecting relations with territorial neighbours. Significance statement The dear enemy effect, in which animals respond less aggressively to familiar neighbours compared to strangers, is probably beneficial because it reduces aggressive interactions with non-threatening individuals. However, no study has ever tested whether there actually are fitness benefits for individuals with a greater tendency to form dear enemy relationships. Our study used experimental playbacks to simulate neighbours and strangers, and we found no relationship between dear enemy behaviour and reproductive success in a songbird. However, our approach to test adaptive hypotheses of this widespread territorial behaviour and our longitudinal playback design to examine the development of familiarity towards a neighbour and discrimination of neighbours and strangers are likely to be important tools to advance our understanding of territorial behaviour and individual recognition.