THE ECOLOGY OF BEHAVIOURAL CONTAGION IN NATURAL SYSTEMS

在整个动物王国，社会信息的传递几乎塑造了生活的方方面面。信息传递发生在许多情况下，可能涉及直接沟通或通过无意的社会线索传播信息，例如当个人通过观察他人学习信息时。当个人获得信息并采取新的行为时，这可以传递给其他人，并且这种行为可以在社会中传播。例如，在20世纪40年代，人们观察到英国山雀（Paridae）打开牛奶瓶盖，这种行为似乎在整个种群中传播。通过个体之间的传播，行为的社会传播有可能导致短期变化（如觅食行为的改变）以及长期的多代过程（如文化的出现）。根据定义，社会传播依赖于个体之间精细的社会联系，而众所周知，许多动物社会的“社会网络”与行为的传播密切相关。然而，人们对行为传播的模式以及形成这种模式的因素仍然知之甚少。特别是，存在两种截然相反的假设。20世纪70年代形成的一个长期存在的假说在不同的系统（从计算网络到动物社会）中具有很大的影响力，它认为，与通常在小集群或小集团内连接的个人之间的强联系相比，弱的社会联系经常架起网络的桥梁，因此能够增加扩散潜力。然而，最近一种革命性的行为社会学假说认为，如果采取一种行为需要社会强化，那么个人群体内部的紧密联系就变得最重要。然而，要验证这些假设并研究影响动物系统内行为传播的社会和生态因素，需要(1)在同一社会系统内大规模跟踪野生个体，(2)具有高分辨率监测行为的能力，最重要的是，(3)对潜在控制因素进行实验操作。在这里，我们将对牛津附近威瑟姆的野生鸟类（大山雀）进行非常详细的研究，涉及数千只个体，追踪它们的整个一生，以了解行为是如何传播的。我们之前的工作已经证明，行为是通过这些鸟类之间的社会关系传播的，并且有可能使用新型的自动选择喂食器来操纵种群水平的组成部分和个体水平的社会性，以精确控制哪些个体可以在林地的每个区域喂食。我们将结合我们过去的协议，允许监控和操纵个人的社会关系和行为，以测试行为传播的方式，以及生态和个人因素如何影响这一点。首先，通过控制基本的生态因素（密度和混合），我们将确定这些因素如何影响网络结构和行为传播的后果。其次，我们将测试个体社会特征的差异如何通过操纵行为的社会起点和特定个体的存在来影响新行为的流动。此外，我们将研究行为变化如何反馈到社会结构本身。我们的总体目标是发展对生态学、社会性和行为传播之间相互作用的综合理解。这些见解可以用来理解生态因素的社会后果，个体特征在形成传播过程中的作用，并允许预测行为传播在不同情景下，在不断变化的条件下以及受到扰动时如何变化。

Across the animal kingdom, the transfer of social information shapes almost all aspects of life. Information transmission takes place in many contexts, and may involve direct communication or information spreading via inadvertent social cues, for instance when individuals learn information through observing others. As individuals gain information and adopt new behaviours, this can then be passed onto others, and the behaviour can spread through the society. For instance, in the 1940's, British tits (Paridae) were observed opening milk bottle tops, and this behaviour appeared to transmit across the population. By transmitting from individual to individual, the social spread of behaviour has the potential to cause both short-term changes (such as modifications to foraging behaviour) as well as long-term multi-generational processes (such as the emergence of culture). By definition, social transmission depends upon the fine-scale social connections between individuals, and the 'social networks' of many animal societies are known to hold much relevance to the spread of behaviour. Nevertheless, the mode by which behaviour spreads, and the factors that shape this, remain less well understood. In particular, two contrasting hypotheses exist. A long-held hypothesis formed in the 1970's which has been highly influential across diverse systems (ranging from computational networks to animal societies) argues that weak social links often bridge the network and therefore enable increased diffusion potential in contrast to strong links that are usually between individuals who are connected within small clusters - or cliques - anyway. However, a more recent hypothesis that is currently revolutionizing the sociology of behaviour spread states that if adopting a behaviour requires social reinforcement, then strong ties within clustered cliques of individuals become most important. Yet, testing these hypotheses and examining the social, and ecological factors shaping behaviour spread within animal systems requires (i) large-scale tracking of wild individuals within the same social system, (ii) the ability to monitor behaviour at high resolution and, crucially, (iii) experimental manipulations of the potential governing factors.Here, we will use an exceptionally detailed study of wild birds (great tits) at Wytham, near Oxford, involving thousands of individuals tracked over their entire lifetimes, to understand how behaviour spreads. Our previous work has demonstrated that behaviour is transmitted across social ties between these birds, and that it is possible to manipulate population-level components and individual-level sociality using novel automated selective feeders to control precisely which individuals can feed in each area of the woodland. We will combine our past protocols to allow monitoring and manipulation of individuals' social ties and behaviour to test the way behaviour spreads and how ecological and individual factors influence this. First, by manipulating fundamental ecological factors (density and mixing) we will determine how these affect network structure and the consequences for behaviour spread. Second, we will test how differences in individuals' social characteristics affect the flow of new behaviours by manipulating the social start points of behaviour and the presence of particular individuals. In addition, we will examine how behavioural changes may feedback onto social structure itself. Our overall goal is to develop an integrated understanding of the interplay between ecology, sociality, and the spread of behaviour. These insights can then be used to understand the social consequences of ecological factors, the role of individual characteristics in shaping transmission processes, and allow the prediction how behavioural spread may alter under different scenarios, in changing conditions and when subject to perturbations.