Abstract Detecting predators requires information, and many behavioral and environmental features are predicted to enhance or limit an animal’s ability to learn about potential danger. Animals living in groups are thought to be at an advantage for learning about predator presence, but individual safety also depends on cues spreading from detectors to nondetectors as unsuspecting individuals may still be vulnerable. In this study, we simulated predator presence among two groups of wild samango monkeys ( Cercopithecus albogularis schwarzi ) to mimic natural encounters where only some individuals within a primate social group have access to personal information about potential threats. We did this using visual models of natural predators placed in positions for the monkeys to encounter within the landscape for a limited amount of time. We measured the number of individuals that were observed to detect and respond to these models with antipredator reactions, relative to subgroup size. While initial detectors that were able to spot the model themselves always reacted with overt behaviors such as alarm calling or staring at the model, responses did not typically spread to all group members. The number of initial detectors was also only weakly associated with the number of individuals that responded at the end of a trial. Initial responses to leopards were much stronger and more likely to spread than those given in response to pythons or eagles, and the importance of behaviors assumed to have an antipredator function depended on the type of predator the samangos encountered. Significance statement Early detection is critical for prey to survive an encounter with an ambush predator. Social prey have the advantage of being able to rely on cues from conspecifics, though individuals that do not detect a potential threat can still be susceptible to attack. Here, we exposed wild samango monkeys to predator models to simulate predator presence to only part of a social group. Habitat visibility was a key predictor affecting collective detection, but the importance of other hypothesized factors—scanning and group spatial cohesion—depended on the predator species. Overall, our results indicate that the social effectiveness of purported risk-sensitive strategies varies based on the type of threat encountered.

中文翻译：

---

威胁检测的程度取决于野生萨芒戈猴群的捕食者类型和行为背景

摘要 检测捕食者需要信息，预测许多行为和环境特征会增强或限制动物了解潜在危险的能力。人们认为群体生活的动物在了解捕食者存在方面具有优势，但个人安全也取决于从探测器到非探测器传播的线索，因为毫无戒心的个体可能仍然很脆弱。在这项研究中，我们模拟了两组野生萨芒戈猴 (Cercopithecus albogularis schwarzi) 之间的捕食者存在，以模拟自然遭遇，其中灵长类社会群体中只有一些个体可以访问有关潜在威胁的个人信息。我们使用自然捕食者的视觉模型来做到这一点，它们被放置在猴子在有限的时间内在景观中遇到的位置。我们测量了观察到的个体数量，相对于亚组大小，这些模型具有抗捕食者反应。虽然能够发现模型本身的初始检测器总是对公开的行为做出反应，例如发出警报或盯着模型，但反应通常不会传播到所有组成员。初始检测器的数量也与试验结束时做出响应的个体数量仅微弱相关。与对蟒蛇或鹰的反应相比，对豹子的最初反应要强烈得多，也更容易传播，而且被认为具有抗捕食者功能的行为的重要性取决于萨曼戈斯遇到的捕食者的类型。意义声明 早期发现对于猎物在遭遇伏击捕食者时幸存下来至关重要。社会猎物的优势在于能够依赖来自同种的线索，尽管没有检测到潜在威胁的个体仍然容易受到攻击。在这里，我们将野生萨芒戈猴暴露于捕食者模型中，以模拟捕食者的存在，使其仅在一部分社会群体中出现。栖息地能见度是影响集体探测的关键预测因素，但其他假设因素——扫描和群体空间凝聚力——的重要性取决于捕食者物种。总体而言，我们的结果表明，所声称的风险敏感策略的社会有效性因遇到的威胁类型而异。我们将野生萨芒戈猴暴露于捕食者模型中，以模拟捕食者的存在，使其仅对社会群体的一部分进行。栖息地能见度是影响集体探测的关键预测因素，但其他假设因素——扫描和群体空间凝聚力——的重要性取决于捕食者物种。总体而言，我们的结果表明，所声称的风险敏感策略的社会有效性因遇到的威胁类型而异。我们将野生萨芒戈猴暴露于捕食者模型中，以模拟捕食者的存在，使其仅对社会群体的一部分进行。栖息地能见度是影响集体探测的关键预测因素，但其他假设因素——扫描和群体空间凝聚力——的重要性取决于捕食者物种。总体而言，我们的结果表明，所声称的风险敏感策略的社会有效性因遇到的威胁类型而异。