Periodic Reporting for period 1 - CollectiveDecisions (Neural basis of zebrafish collective decision-making)
Periodo di rendicontazione: 2023-08-01 al 2026-01-31
Understanding how animals make decisions in complex environments is a central question in biology and neuroscience. Individual animals have limited sensory and cognitive capacities, yet many species overcome these constraints by forming groups that share and evaluate information collectively. Such collective decision-making enables more accurate navigation, foraging, and threat avoidance, but the behavioral algorithms and underlying neural circuits that generate these emergent group behaviors remain largely unknown. The CollectiveDecisions project addresses this fundamental gap by establishing the juvenile zebrafish as a powerful experimental model to investigate how individual and collective decisions arise from individual animals' neural computations. Zebrafish provide an ideal system for this endeavor because, at the juvenile stage, they display rich social interactions and cognitive abilities while remaining experimentally accessible to advanced imaging and manipulation techniques.
The project’s overarching objectives are:
1. To identify the behavioral algorithms that govern individual and collective decision-making in zebrafish groups under controlled visual conditions.
2. To characterize the neural representations of sensory and social cues underlying these behaviors using advanced functional imaging.
3. To causally link neural circuit activity to collective decisions through optogenetic and laser-ablation approaches.
By bridging behavior, neuroscience, and computational modeling, the project aims to uncover general principles of collective information processing that may apply across animal species and inspire new algorithms for distributed sensing and decision-making in artificial systems.
The project’s overarching objectives are:
1. To identify the behavioral algorithms that govern individual and collective decision-making in zebrafish groups under controlled visual conditions.
2. To characterize the neural representations of sensory and social cues underlying these behaviors using advanced functional imaging.
3. To causally link neural circuit activity to collective decisions through optogenetic and laser-ablation approaches.
By bridging behavior, neuroscience, and computational modeling, the project aims to uncover general principles of collective information processing that may apply across animal species and inspire new algorithms for distributed sensing and decision-making in artificial systems.
During the first reporting period (August 2023 – Jan 2026), major progress has been achieved in developing behavioral paradigms, optical technologies, and analytical frameworks for dissecting individual and collective decision-making in zebrafish.
Behavioral Algorithms of Collective Interactions:
Four major behavioral projects have been initiated. We established virtual-reality tracking systems that allow real and virtual zebrafish to interact across arenas, enabling systematic dissection of distance-keeping, alignment, and leader-follower dynamics. These experiments revealed the sensory features that most strongly drive social attraction and alignment.
Phototactic Navigation in Juvenile Zebrafish:
We discovered that juveniles navigate light gradients using temporal rather than spatial luminance cues—a strategy distinct from that of larvae. This finding provides a new foundation for studying how sensory information is integrated within groups. A manuscript reporting this work is currently under review at iScience.
Collective Visual Motion Estimation:
We demonstrated that zebrafish improve their ability to estimate visual motion drift direction when swimming in larger groups, suggesting information sharing between individuals. We developed stimulation paradigms using naturalistic motion patterns, laying the groundwork for linking these behavioral benefits to neural circuit mechanisms.
Collective Threat Evasion:
By presenting looming stimuli to individual fish, we quantified how escape reactions propagate across group members. The resulting behavioral data have been incorporated into computational models that capture the transfer of information between individuals.
Technological and Methodological Developments:
We built large-scale tracking systems capable of following up to 30 animals simultaneously and designed frameworks for real-time virtual coupling between arenas. We also developed high-speed (100 Hz) imaging systems, advanced 3D projection methods that create mathematically correct virtual environments, and light-field microscopy for volumetric brain imaging at 100 Hz.
Neurobiological and Genetic Tools:
We established procedures for brain clearing, c-fos-based activity labeling, and single-cell optogenetic targeting, enabling causal manipulations of defined neurons.
Collectively, these achievements position the project at the frontier of behavioral and systems neuroscience, providing the first comprehensive toolkit for studying social decision-making in a vertebrate model.
Behavioral Algorithms of Collective Interactions:
Four major behavioral projects have been initiated. We established virtual-reality tracking systems that allow real and virtual zebrafish to interact across arenas, enabling systematic dissection of distance-keeping, alignment, and leader-follower dynamics. These experiments revealed the sensory features that most strongly drive social attraction and alignment.
Phototactic Navigation in Juvenile Zebrafish:
We discovered that juveniles navigate light gradients using temporal rather than spatial luminance cues—a strategy distinct from that of larvae. This finding provides a new foundation for studying how sensory information is integrated within groups. A manuscript reporting this work is currently under review at iScience.
Collective Visual Motion Estimation:
We demonstrated that zebrafish improve their ability to estimate visual motion drift direction when swimming in larger groups, suggesting information sharing between individuals. We developed stimulation paradigms using naturalistic motion patterns, laying the groundwork for linking these behavioral benefits to neural circuit mechanisms.
Collective Threat Evasion:
By presenting looming stimuli to individual fish, we quantified how escape reactions propagate across group members. The resulting behavioral data have been incorporated into computational models that capture the transfer of information between individuals.
Technological and Methodological Developments:
We built large-scale tracking systems capable of following up to 30 animals simultaneously and designed frameworks for real-time virtual coupling between arenas. We also developed high-speed (100 Hz) imaging systems, advanced 3D projection methods that create mathematically correct virtual environments, and light-field microscopy for volumetric brain imaging at 100 Hz.
Neurobiological and Genetic Tools:
We established procedures for brain clearing, c-fos-based activity labeling, and single-cell optogenetic targeting, enabling causal manipulations of defined neurons.
Collectively, these achievements position the project at the frontier of behavioral and systems neuroscience, providing the first comprehensive toolkit for studying social decision-making in a vertebrate model.
The project has already produced several advances that go well beyond the current state of the art:
1. Discovery of a novel phototactic algorithm in juvenile zebrafish, fundamentally different from that described in larvae, revealing a developmental shift in sensory-motor computation.
2. Implementation of real-time virtual social coupling across multiple arenas, allowing controlled manipulation of social networks and quantitative testing of collective interaction rules — a level of experimental control previously not possible in vertebrates.
3. Integration of behavioral, neural, and computational approaches at unprecedented temporal and spatial resolution, including the development of a light-field microscope for fast volumetric imaging in head-restrained behaving animals.
4. New open-source software frameworks for brain-wide atlas mapping, multi-animal tracking, and agent-based modeling, providing tools that can be widely used by the neuroscience and collective behavior communities.
5. The project bridges neurobiology, computer vision, and theoretical modeling, offering a scalable framework for understanding distributed decision-making. The insights gained are expected to influence fields ranging from swarm robotics to computational neuroscience.
By uncovering how neural circuits in individual brains give rise to coordinated collective decisions, the CollectiveDecisions project will significantly advance our understanding of group intelligence in biological systems. The combination of innovative behavioral paradigms, cutting-edge imaging technologies, and computational modeling will generate general principles of collective perception and decision-making. These results will not only deepen our knowledge of animal cognition but also provide conceptual and algorithmic inspiration for artificial systems capable of distributed, adaptive decision-making.
1. Discovery of a novel phototactic algorithm in juvenile zebrafish, fundamentally different from that described in larvae, revealing a developmental shift in sensory-motor computation.
2. Implementation of real-time virtual social coupling across multiple arenas, allowing controlled manipulation of social networks and quantitative testing of collective interaction rules — a level of experimental control previously not possible in vertebrates.
3. Integration of behavioral, neural, and computational approaches at unprecedented temporal and spatial resolution, including the development of a light-field microscope for fast volumetric imaging in head-restrained behaving animals.
4. New open-source software frameworks for brain-wide atlas mapping, multi-animal tracking, and agent-based modeling, providing tools that can be widely used by the neuroscience and collective behavior communities.
5. The project bridges neurobiology, computer vision, and theoretical modeling, offering a scalable framework for understanding distributed decision-making. The insights gained are expected to influence fields ranging from swarm robotics to computational neuroscience.
By uncovering how neural circuits in individual brains give rise to coordinated collective decisions, the CollectiveDecisions project will significantly advance our understanding of group intelligence in biological systems. The combination of innovative behavioral paradigms, cutting-edge imaging technologies, and computational modeling will generate general principles of collective perception and decision-making. These results will not only deepen our knowledge of animal cognition but also provide conceptual and algorithmic inspiration for artificial systems capable of distributed, adaptive decision-making.