Title: Testing a thermodynamic approach to collective animal behavior in laboratory fish schools

Abstract: Collective behaviors displayed by groups of social animals are observed frequently in nature. Understanding and predicting the behavior of complex biological systems is dependent on developing effective descriptions and models. While collective animal systems are characteristically non-equilibrium, we can employ concepts from equilibrium statistical mechanics to motivate the measurement of material-like properties in laboratory animal aggregates. Here, we present results from a new set of experiments that utilize high speed footage of two-dimensional schooling events, particle tracking, and projected static and dynamic light fields to observe and control the behavior of negatively phototaxic fish schools ($\textit{Hemigrammus}$ $\textit{bleheri}$). First, we use static light fields consisting of dark circular regions to produce visual stimuli that confine the schools to a range of areas. We find that the school has a maximum density independent of group size, and that the swim pressure increases linearly with number density, suggesting that unperturbed schools exist on an isotherm. Next, we use dynamic light fields where the radius of the dark region shrinks linearly with time to compress the schools. We find that the effective temperature parameter depends on the compression time and our results are thus consistent with the school having a constant heat flux. These findings provide further evidence for the utility of effective thermodynamic descriptions descriptions of non-equilibrium systems in collective animal behavior.