LAUSR

Some territorial animals recognize familiar neighbours and are less aggressive to established neighbours than they are to strangers. This form of social recognition produces a ‘dear enemy’ effect, which may allow animals to reduce the costs of territory defence. The dear enemy effect is thought to reflect either the decreased threat posed by neighbours relative to strangers (the relative threat hypothesis) or the decreased need for escalation with increasing familiarity between neighbours (the familiarity hypothesis). We tested for a vocally mediated dear enemy effect in male brilliant-thighed poison frogs, Allobates femoralis. In this species, the familiarity hypothesis predicts a dear enemy effect, because males defend long-term stable territories and should be familiar with the calls of their neighbours. In contrast, the relative threat hypothesis does not predict a dear enemy effect, because neighbours and strangers both represent competitors for mates and likely pose equivalent threats. Acoustic analyses showed that males produce individually distinctive advertisement calls. Two playback experiments were conducted to determine whether territorial males respond less aggressively to neighbours' calls than to strangers' calls based on this acoustic identity information. In the first experiment, males pursued acoustically simulated neighbours and strangers to similar distances when calls were played both from the direction of the neighbour's territory and from a direction with no neighbours. In the second experiment, males responded aggressively to neighbours' and strangers' calls at similar threshold amplitudes and pursued these calls to similar distances. Hence, two different playback experiments failed to find evidence of a vocally mediated dear enemy effect. Our results support the hypothesis that territorial animals respond to the relative threat posed by neighbours and strangers regardless of their level of familiarity with neighbours.