Abstract According to the dear enemy phenomenon, territory owners decrease costs of ownership by decreasing aggression toward neighbors once territory boundaries have been established. To maintain this cooperation, when territorial neighbors are caught in the act of “cheating” by failing to respect territory boundaries, they should be “challenged” via increased aggression. Most studies examining this prediction tested neighbor recognition through playback of advertisement calls. We tested whether territorial terrestrial salamanders, Plethodon serratus , would identify cheating neighbors via chemical or visual cues. In separate experiments, focal residents were exposed to either territorial substrate markings from another salamander or to a mirror (simulating a territorial neighbor) on one side of their territory for several days. During subsequent testing, the chemical or visual (mirror) cues were presented at either the trained location (cooperating) or on the opposite (cheating) side. Either olfactory or visual cues alone were sufficient for discrimination of cheating/cooperating neighbors, with cheating neighbors receiving the most aggressive displays and chemosensory sampling behavior from focal residents. Aggressive displays were particularly high in trials with visual cues, which likely indicated a more immediate threat. Moreover, responses to the visual threat may have been more escalated because the “neighbor” did not back down following threat posturing from the focal resident. Qualitatively, focal residents attempted to physically interact with the mirror images even in cooperating neighbor treatments. Therefore, in accordance with the dear enemy hypothesis, cheating neighbors elicited a higher level of threat displays than cooperating neighbors, but even cooperating neighbors could sometimes be challenged at the territory border. Significance statement For territorial neighbors, cooperating by respecting territorial boundaries can reduce the cost of ownership by reducing aggressive interactions between neighbors. Cooperating neighbors (“dear enemies”) remain in their own territories, whereas cheating neighbors travel beyond their territory boundaries. Evolutionarily, cooperation can be maintained if neighbors that “cheat” must pay a cost. Thus, location of individuals (detected via advertisement signals) is predicted to influence aggressive behavior by adjacent neighbors. We demonstrated that territorial southern red-backed salamanders responded with a higher level of aggressive displays to cheating neighbors whose presence was detected via either chemical (substrate markings) or visual (mirror) displays. In mirror trials, focal residents paced back and forth with their mirror images, exhibiting short bursts of threat posturing and tapping the image with their snouts, which is consistent with behavior toward other salamanders. Therefore, cooperation between neighbors can be maintained via increased aggression toward cheaters.

中文翻译：

---

当邻居作弊时：对南方红背蝾螈天敌现象的测试

摘要 根据亲爱的敌人现象，领土所有者通过在领土边界确定后减少对邻居的侵略来降低拥有成本。为了维持这种合作，当领土邻国因不尊重领土边界而陷入“欺骗”行为时，他们应该通过增加侵略来“挑战”。大多数检查此预测的研究通过播放广告呼叫来测试邻居识别。我们测试了领地陆生蝾螈 Plethodon serratus 是否会通过化学或视觉线索识别作弊的邻居。在单独的实验中，焦点居民被暴露在另一只蝾螈的领土基底标记或他们领土一侧的镜子（模拟领土邻居）下数天。在随后的测试中，化学或视觉（镜子）提示出现在受训位置（合作）或相反（作弊）一侧。单独的嗅觉或视觉线索足以区分作弊/合作的邻居，作弊的邻居收到来自焦点居民的最激进的展示和化学感应采样行为。在具有视觉线索的试验中，攻击性显示特别高，这可能表明存在更直接的威胁。此外，对视觉威胁的反应可能会更加升级，因为“邻居”在焦点居民做出威胁姿态后并没有退缩。定性地说，即使在合作的邻居治疗中，焦点居民也试图与镜像进行身体互动。因此，根据亲爱的敌人假设，作弊的邻居比合作的邻居引起更高程度的威胁，但即使是合作的邻居有时也会在领土边界受到挑战。重要性声明 对于领土邻国而言，通过尊重领土边界进行合作可以通过减少邻国之间的积极互动来降低拥有成本。合作的邻居（“亲爱的敌人”）留在自己的领土上，而欺骗的邻居则越过他们的领土边界。从进化上讲，如果“欺骗”的邻居必须付出代价，则可以保持合作。因此，预测个体的位置（通过广告信号检测）会影响相邻邻居的攻击行为。我们证明了领土南部红背蝾螈以更高水平的侵略性展示回应欺骗邻居，通过化学（基质标记）或视觉（镜子）显示检测到邻居的存在。在镜子试验中，焦点居民用他们的镜像来回踱步，表现出短暂的威胁姿势并用他们的鼻子轻拍图像，这与对其他蝾螈的行为是一致的。因此，邻居之间的合作可以通过增加对骗子的攻击来维持。这与对其他蝾螈的行为一致。因此，邻居之间的合作可以通过增加对骗子的攻击来维持。这与对其他蝾螈的行为一致。因此，邻居之间的合作可以通过增加对骗子的攻击来维持。