Fish Schools as Self-Organizing Adaptive Information Networks

Little stills our thoughts more than gazing at schools of fish as they dart and wheel beneath the surface, providing us with impressive displays of coordination and presumed solidarity in the process.  Yet conflict exists within schools, as members have competing interests, such as balancing social information to enhance resource discovery rate with density-dependent competition.  As such, group members must be proficient at parsing out salient from specious information from the crowd; how they do so in a manner that is both rapid and adaptive is somewhat paradoxical.  Our work focuses on using fish schooling behavior to learn how the components of such self-organizing networks can complement the opposing effects of resiliency and sensitivity.  Here we use a spatially explicit particle model to explore the adaptive potential of groups when individuals interact according to a primitive signal-to-noise filter based on the psychophysical principle of Weber’s Law.  By scaling individual attention to background motion cues group members generate context dependent asymmetries in the underlying communication networks – responses that are not possible with a standard topological filter.  The result is a mechanism to explain how social animals can optimize resiliency to noise, while remaining sensitive to novel stimuli prior to undergoing any changes in their internal state, such as increasing individual vigilance level.  Our results provide a robust mechanism to optimize collective adaptability across a broad range of ecological conditions and group sizes.