Collective animal motion: The case of midge swarms

Abstract

Abstract

Abstract Among collective states of animal groups, midge swarms pose several puzzling questions. Their polarization order parameter is very small, and the insects interact only weakly with one another while remaining strongly bound to the swarm. In laboratory studies performed without external perturbations, the correlation length is small. By contrast, field studies reveal strong correlations, scale-free behavior, and power-law scaling of the correlation length, susceptibility, and correlation time. The dynamic correlation function exhibits data collapse only for short times when time is rescaled by the correlation time. Is there a theory that explains these disparate observations? We review several models proposed in the literature and extend our harmonically confined Vicsek model to anisotropic confinement. Within the extended criticality region of the phase diagram, numerical simulations predict elongated swarms and static critical exponents that lie between those of the two-dimensional and isotropic three-dimensional models. These values agree better with those measured in natural swarms. We also calculate probability density functions for both natural and laboratory swarms. Numerical simulations indicate that the laboratory distributions correspond to a different region of parameter space, although some open questions remain.