Département de Biologie, Faculté des Sciences et Technologies, Université de Lille, Lille, France.
IPHC UMR 7178, CNRS, Université de Strasbourg, Strasbourg, France.
Am J Primatol. 2024 Dec;86(12):e23682. doi: 10.1002/ajp.23682. Epub 2024 Sep 8.
Disease outbreaks are one of the key threats to great apes and other wildlife. Because the spread of some pathogens (e.g., respiratory viruses, sexually transmitted diseases, ectoparasites) are mediated by social interactions, there is a growing interest in understanding how social networks predict the chain of pathogen transmission. In this study, we built a party network from wild chimpanzees (Pan troglodytes), and used agent-based modeling to test: (i) whether individual attributes (sex, age) predict individual centrality (i.e., whether it is more or less socially connected); (ii) whether individual centrality affects an individual's role in the chain of pathogen transmission; and, (iii) whether the basic reproduction number (R) and infectious period modulate the influence of centrality on pathogen transmission. We show that sex and age predict individual centrality, with older males presenting many (degree centrality) and strong (strength centrality) relationships. As expected, males are more central than females within their network, and their centrality determines their probability of getting infected during simulated outbreaks. We then demonstrate that direct measures of social interaction (strength centrality), as well as eigenvector centrality, strongly predict disease dynamics in the chimpanzee community. Finally, we show that this predictive power depends on the pathogen's R and infectious period: individual centrality was most predictive in simulations with the most transmissible pathogens and long-lasting diseases. These findings highlight the importance of considering animal social networks when investigating disease outbreaks.
疾病爆发是对大型猿类和其他野生动物的主要威胁之一。由于一些病原体（例如呼吸道病毒、性传播疾病、外寄生虫）的传播是通过社交互动介导的，因此人们越来越有兴趣了解社交网络如何预测病原体的传播链。在这项研究中，我们构建了野生黑猩猩（Pan troglodytes）的聚会网络，并使用基于主体的模型来测试：（i）个体属性（性别、年龄）是否预测个体的中心度（即是否具有更多或更少的社交联系）；（ii）个体中心度是否影响个体在病原体传播链中的作用；以及，（iii）基本繁殖数（R）和感染期是否调节中心度对病原体传播的影响。我们表明，性别和年龄预测个体的中心度，年龄较大的雄性具有更多（度中心度）和更强（强度中心度）的关系。正如预期的那样，雄性在其网络中的中心度高于雌性，并且它们的中心度决定了它们在模拟爆发期间感染的概率。然后，我们证明了直接衡量社交互动的措施（强度中心度）以及特征向量中心度强烈预测了黑猩猩社区的疾病动态。最后，我们表明这种预测能力取决于病原体的 R 和感染期：在传染性最强和持续时间最长的疾病模拟中，个体中心度的预测能力最强。这些发现强调了在研究疾病爆发时考虑动物社交网络的重要性。
