Algorithms underlying collective decision-making during threat perception in zebrafish

• Applicant: Dr. Margherita Zaupa
• Subject Area: Cognitive, Systems and Behavioural Neurobiology
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 574832510

How much do our peers influence our decisions? How do animals integrate their own sensory experience with information from conspecifics to choose the optimal behavioral strategy? These are long-standing questions in the field of behavioral neuroscience. Animals in a collective make faster and more accurate decisions, since individuals in a group rely not only on their own perception but also on the behavior of their neighbors. A salient example of coordinated behavioral choice is the collective response to a predator threat: animals in groups are more efficient in detecting and escaping from predators, thanks to the sharing of information between individuals. However, it is difficult to disentangle whether a single animal is responding directly to the threat or indirectly through conspecifics. The complex and reciprocal nature of interactions between individuals makes it challenging to dissect the exact algorithms of collective decision-making. To tackle this problem, it is crucial to gain control of the behavior of single individuals within the group and perform causal experiments under supervised conditions. However, so far such approaches have largely been hindered by technical challenges due to the limited accessibility of the nervous system of many of the commonly studied animal species in collective behavior research. On the contrary, zebrafish is a model organism uniquely suited for this purpose thanks to its amenability to genetic and behavioral manipulations, unprecedented in a vertebrate performing social behavior. Hence, I will use the collective escape response to a predator in young zebrafish as a paradigm to investigate how animals integrate potentially conflicting environmental and social information, disentangling the rules underlying collective decision-making. I will use virtual reality to uncouple the sensory experience of individual animals to that of their conspecifics and define the strategies they employ to resolve conflicting threat and social cues. Based on the results obtained, I will build agent-based behavioral models for collective decision-making. I will then develop novel methods to gain direct control of the behavior of individual animals within real groups through optogenetics manipulations. I will exploit this unprecedented control to perform causal, hypothesis-driven experiments to test and refine the theoretical models. The highly innovative approach and methodology employed will be critical in driving a considerable broadening of our knowledge of collective decision-making to a much-needed causal understanding.

• DFG Programme: WBP Position