English

Exploring locust hopper bands emergent patterns using parallel computing

Neurons and Cognition 2023-03-01 v1 Adaptation and Self-Organizing Systems Physics and Society

Abstract

To date, the mechanisms underlying the diversity of the emergent patterns of collective motion in locust hopper bands remain to be unveiled. This study investigates the role of speed heterogeneity in the emergence of the most common patterns (frontal and columnar), following the Self-Organization framework. To address whether marching activity intermittency and density-dependant hopping individual behaviours could underlie the formation of such patterns, a three-zone Self-Propelled Particles model variant was formulated. In this model, individuals alternated between marching and resting periods, and were more likely to hop when crowded. The model successfully predicted the emergence of both patterns of interest, with the presence of a density-dependent hopping probability being a necessary condition. Short to absent pause periods mostly resulted in columnar shapes, similar to the ones observed in the brown locust (Locustana pardalina) and long pause periods rather resulted in frontal shapes, such as exhibited by the Australian plague locust (Chortoicetes terminifera). Furthermore, the density profiles of simulated frontal formations displayed the same shape as empirical profiles of Australian plague locust hopper bands. Both simulated and experimental paint marking experiments showed that locusts initially located at different positions in the band were find together at its front after a few hours marching; an expected global behavior in hopper bands undergoing activity intermittency. These results represent an important first step towards a cross-species comparison of locust mass migration patterns.

Keywords

Cite

@article{arxiv.2302.14064,
  title  = {Exploring locust hopper bands emergent patterns using parallel computing},
  author = {Adrian Bach},
  journal= {arXiv preprint arXiv:2302.14064},
  year   = {2023}
}

Comments

Masters project supervised by Jerome Buhl (The University of Adelaide)

R2 v1 2026-06-28T08:50:59.160Z