Schooling increases risk exposure for fish navigating past artificial barriers

Artificial barriers have become ubiquitous features in freshwater ecosystems and they can significantly impact a region's biodiversity. Assessing the risk faced by fish forced to navigate their way around artificial barriers is largely based on assays of individual swimming behavior. However, social interactions can significantly influence fish movement patterns and alter their risk exposure. Using an experimental flume, we assessed the effects of social interactions on the amount of time required for juvenile palmetto bass (Morone chrysops × M. saxatilis) to navigate downstream past an artificial barrier. Fish were released either individually or in groups into the flume using flow conditions that approached the limit of their expected swimming stamina. We compared fish swimming behaviors under solitary and schooling conditions and measured risk as the time individuals spent exposed to the barrier. Solitary fish generally turned with the current and moved quickly downstream past the barrier, while fish in groups swam against the current and displayed a 23-fold increase in exposure time. Solitary individuals also showed greater signs of skittish behavior than those released in groups, which was reflected by larger changes in their accelerations and turning profiles. While groups displayed fission-fusion dynamics, inter-individual positions were highly structured and remained steady over time. These spatial patterns align with theoretical positions necessary to reduce swimming exertion through either wake capturing or velocity sheltering, but diverge from any potential gains from channeling effects between adjacent neighbors. We conclude that isolated performance trials and projections based on individual behaviors can lead to erroneous predictions of risk exposure along engineered structures. Our results also suggest that risk perception and behavior may be more important than a fish's swimming stamina in artificially modified systems.