Behavioural plasticity and the transition to order in jackdaw flocks

Abstract

Collective behaviour is typically thought to arise from individuals following fixed interaction rules. The possibility that interaction rules may change under different circumstances has thus only rarely been investigated. Here we show that local interactions in flocks of wild jackdaws (Corvus monedula) vary drastically in different contexts, leading to distinct group-level properties. Jackdaws interact with a fixed number of neighbours (topological interactions) when traveling to roosts, but coordinate with neighbours based on spatial distance (metric interactions) during collective anti-predator mobbing events. Consequently, mobbing flocks exhibit a dramatic transition from disordered aggregations to ordered motion as group density increases, unlike transit flocks where order is independent of density. The relationship between group density and group order during this transition agrees well with a generic self-propelled particle model. Our results demonstrate plasticity in local interaction rules and have implications for both natural and artificial collective systems.

Machine learning for active matter

“…obtaining the data necessary to analyse how their social structure is reflected in their collective dynamics 35 , an important question also for other species, such as flocking jackdaws 42 and even humans 43 .…”

Spatio-temporal reconstruction of emergent flash synchronization in firefly swarms via stereoscopic 360-degree cameras

“…Traditionally, stereoscopic set-ups have used regular ( planar) cameras for the three-dimensional tracking of flocking or swarming animals, such as mosquitoes [3], midges [4,16,17] or birds [18][19][20][21][22], which has provided considerable insights into the mechanisms involved in collective dynamics, for example the structure of interactions [17,19,23] or propagation of information [18,22]. Over the past few years, however, 360-degree cameras (or 360-cameras) have started to be used widely, notably in computer vision [24].…”

An equation of state for insect swarms

“…Extracting interaction rules by observing group behaviour is a highly nontrivial inverse problem 12 that can typically only be solved by assuming a modelling framework a priori 13,14 . Appropriate model selection is made more difficult given that interactions may change in different contexts 7,8,15 . Even less work has been done to precisely determine the tasks optimized by collective behaviour.…”