(174ca) Complex Collectives: Investigating Biofilm Patterning and Ant Traffic

The natural world is full of multicomponent communities that dynamically self-organise to aid their collective function and survival. Investigating the design principles and emergent intelligence that define these behaviours provides a means of describing and understanding these complex living systems. In this work, we develop and apply an agent based model of self-propelled spherocylinders alongside experimental analyses to explore two such collectives: a multispecies bacterial community and a crowded ant colony.

Bacteria play a significant role in human health, the environment, and industrial applications. In these environments bacteria are most often found in multispecies communities in which colonies containing different cell shapes interact closely. Recently, time-lapse microscopy has shown that these multispecies interactions can be cooperative or competitive (eLife 8:e47365), presenting dynamic and complex behaviours hypothesised to be a consequence of the shape-dependent physical and chemical interactions. However, the role of shape polydispersity in individual and collective dynamics is poorly understood. We are employing a biophysical model of variable size sphere- and rod-shaped individuals to explore how shape polydispersity influences interspecies interactions and to develop a predictive understanding of multispecies spatiotemporal patterning.

Avoiding traffic jams in crowded environments is a significant challenge for collective motion across scales, from migrating herds to cellular cargo transport. Ants possess a remarkable ability to avoid typical traffic jamming patterns at high densities and instead maintain a constant bidirectional flow between their nest and a food source (eLife 8:e48945). No large-scale spatiotemporal organisation such as lane formation or oscillatory flow is seen, suggesting that there is another dynamic mechanism allowing the ants to navigate in crowds without jamming. We use detailed experimental analyses of extant data combined with agent based modelling to explore individual ant trajectories and properties that make up the collective colony movement. By quantifying spatiotemporal patterns of density, interaction rates and ant speeds, we probe the role of the type of interactions and their rates on the group as a whole as it engages in crowded traffic. This leads us to propose feedback models by which individual ants respond to their local environment over time that capture the observed efficient movement in bidirectional crowds.