du 5 mai 2023 au 6 mai 2023
Publié le 25 août 2023 Mis à jour le 25 août 2023

Séminaire Luis Alberto Gómez Nava

seminaire
seminaire

Collective and synchronized motion in animal groups: from sheep to fish

Vendredi 05 mai, 14h00, salle E 4.13 hybride (jour exceptionnel)

Luis Alberto Gómez Nava

Institute for Theoretical Biology (ITB), Humboldt University of Berlin, Berlin, Germany

Collective and synchronized motion in animal groups: from sheep to fish

In this talk I will present two studies of collective behavior observed in animal groups. In the first one, we studied the spontaneous and intermittent collective displacements observed in small groups of sheep (Ovis aries Linnaeus). We performed experiments and noticed that these animals form files while moving as a group, with a clearly-defined leader at the front (i.e.: a strongly hierarchical structure). We refer to this events as Collective Motion Phases (CMPs). A second observation we made from the experimental data was that the leader of a CMP was not always the same individual, but there was a change of leaders, where each individual of the group had the same probability of playing the leading role (i.e.: a "democratic" structure). We investigated the benefits of such seemingly opposite mechanisms (hierarchical vs. democratic) and found scenarios where individuals of the group might benefit from them, like the optimal guidance while navigating a complex landscape or the optimal visit of multiple targets in space.

In the second study I will present experimental data acquired in Teapa, Mexico, where the sulphur mollies (Poecilia sulphuraria) evolved to live in sulphidic springs with high concentration of hydrogen sulphide. In this system, we studied the highly conspicuous, repetitive, and rhythmic collective dive cascades produced by many thousands of these fish that resemble neuronal avalanches observed in brain tissue at criticality. Together with the results of an agent-based model of the system, we explored in how far these fish shoals indeed operate at a critical point between a state of high individual diving activity and low overall diving activity. We explored the adaptive significance (aka 'functionality') of acting at this critical point for the fish by using machine learning algorithms and showed that the best-fitting model, which indeed is located at a critical point, allows information about external perturbations – such as predator attacks – to propagate most effectively through the shoal. Our results suggest that criticality may represent a fundamental principle of distributed information processing in biological systems including large animal collectives.