Harpaz Roy, Phillips Morgan, Goel Ronan, Fishman Mark C, Engert Florian
Department of Molecular and Cellular Biology, Harvard University, Cambridge 02138, USA.
Center for Brain Science, Harvard University, Cambridge 02138, USA.
bioRxiv. 2024 Aug 5:2024.08.02.606403. doi: 10.1101/2024.08.02.606403.
Complex group behavior can emerge from simple inter-individual interactions. Commonly, these interactions are considered static and hardwired and little is known about how experience and learning affect collective group behavior. Young larvae use well described visuomotor transformations to guide interindividual interactions and collective group structure. Here, we use naturalistic and virtual-reality (VR) experiments to impose persistent changes in population density and measure their effects on future visually evoked turning behavior and the resulting changes in group structure. We find that neighbor distances decrease after exposure to higher population densities, and increase after the experience of lower densities. These adaptations develop slowly and gradually, over tens of minutes and remain stable over many hours. Mechanistically, we find that larvae estimate their current group density by tracking the frequency of neighbor-evoked looming events on the retina and couple the strength of their future interactions to that estimate. A time-varying state-space model that modulates agents' social interactions based on their previous visual-social experiences, accurately describes our behavioral observations and predicts novel aspects of behavior. These findings provide concrete evidence that inter-individual interactions are not static, but rather continuously evolve based on past experience and current environmental demands. The underlying neurobiological mechanisms of experience dependent modulation can now be explored in this small and transparent model organism.
复杂的群体行为可以从简单的个体间相互作用中产生。通常，这些相互作用被认为是静态的且固定不变的，而对于经验和学习如何影响群体的集体行为却知之甚少。幼虫利用已被充分描述的视觉运动转换来指导个体间的相互作用和集体群体结构。在这里，我们使用自然主义和虚拟现实（VR）实验来对种群密度施加持续变化，并测量其对未来视觉诱发转向行为以及由此导致的群体结构变化的影响。我们发现，在暴露于较高种群密度后，相邻个体之间的距离会减小，而在经历较低密度后则会增加。这些适应性变化在数十分钟内缓慢而逐渐地发展，并在数小时内保持稳定。从机制上讲，我们发现幼虫通过追踪视网膜上相邻个体诱发的逼近事件的频率来估计其当前的群体密度，并将其未来相互作用的强度与该估计值相耦合。一个基于个体先前视觉社会经验来调节其社会相互作用的时变状态空间模型，准确地描述了我们的行为观察结果，并预测了行为的新方面。这些发现提供了确凿的证据，表明个体间的相互作用不是静态的，而是基于过去的经验和当前的环境需求不断演变的。现在可以在这种小型透明的模式生物中探索经验依赖性调节的潜在神经生物学机制。
