From milling to herding: aggression-driven transition toward low-energy ordered states
Year: 2026
Type: preprint
Abstract: Collective behavior studies how large-scale coordination emerges in groups of individuals, such as animals, that only perceive their immediate neighbors. Analytic models of collective motion predominantly describe homogeneous swarms; in such swarms, large coherent patterns like mills are known to emerge. Among animals, however, inter- and intra-species heterogeneous behavior is ubiquitous. Importantly, the effect of animal-like heterogeneous aggression on the collective states of swarms is not well understood. Here, we introduce novel heteroge- neous rules in a Vicsek-like swarming model to simulate unsteady, animal-like aggressive behavior among self-propelled agents. We find that milling behavior becomes unstable to aggression, and that an aggression-driven transition exists between newly identified steady collective states. Mill destruction and herding are new transition mechanisms propelling the transition from milling to flocking and flocking to banding. Despite aggression being present, the new ordered phases in the swarm are relatively energetically inexpensive. Additionally, we find that swarm welfare deteriorates with density. This is the first time that the effect of aggressive behavior on large swarms has been modeled analytically. The newly defined behavioral rules are inspired by aggressive behavior in fish, but are broadly applicable to agonistic behavior in animal groups.
Source: Research Square
Authors: Flora Grilli, Grgur Tokić, Dick K.P. Yue
Institution: Massachusetts Institute of Technology
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
OpenAlex ID: W7235091745
