Physical crowding truncates intrinsic Lévy-like motility into caste-specific movement regimes in social termites
Abstract
Lévy-like movement patterns, characterized by heavy-tailed step-length distributions, are often considered a highly efficient strategy for solitary searching. While recent studies have established that isolated termite workers possess intrinsic Lévy-like motility templates, how these scale-free patterns differ across functionally distinct castes—and how they adapt to the extreme physical crowding typical of natural subterranean galleries remains unresolved. Here, using high-resolution tracking of the subterranean termite Reticulitermes labralis across group sizes from solitary individuals (N = 1) to high jammed crowds (N = 1,000), we reveal a fundamental motility asymmetry between castes. Building upon established worker baseline, we demonstrate that these intrinsic behavioral templates are highly heterogeneous: although solitary individuals of both castes intrinsically exhibit Lévy-like motion, workers maintain a significantly stronger super-diffusive baseline compared to queens, reflecting an innate predisposition for extensive exploration. As social density increases, physical blocking (steric repulsion) universally truncates trajectories toward multi-phasic Brownian dynamics. However, this truncation is functionally divergent: workers retain greater mobility resilience, preserving exploration efficiency even under crowding, whereas queens rapidly succumb to physical entrapment. Integrating empirical data with an Agent-Based Model, we show that this transition emerges naturally from pervasive physical constraints. Our findings suggest that the interplay between caste-specific intrinsic templates and uniform physical blocking spontaneously organizes the colony’s division of labor: workers are tuned for efficient exploration, while queens are physically constrained to become static interaction hubs.
Fields of Study & Subjects
Diffusion and Search Dynamics
Insect and Arachnid Ecology and Behavior
Distributed Control Multi-Agent Systems
Publication details
- Journal / source
- Movement Ecology
- Publisher
- Springer Science and Business Media LLC
- Volume
- 14
- Issue
- 1
- Publication date
- 3 April 2026
- Language
- EN