MODELING OF EMERGENT BEHAVIORS IN BIOLOGICAL NONEQUILIBRIUM COLLECTIVES

 November 19th, 2026, at 10:30 am
In Person Location: Howey N201
Zoom Link TBA

Thesis Advisor:
Daniel I Goldman
Department of Physics
Georgia Tech

Thesis Committee:
Jennifer Curtis
Department of Physics
Georgia Tech

J.C. Gumbart
Department of Physics
Georgia Tech

Peter Yunker
Department of Physics
Georgia Tech

Mingyuan Zhu
Department of Biochemistry and Biophysics
Texas A&M University
Abstract:
Many aspects of physics often rely on conservation laws such as the conservation of mass and energy. However, many biological systems exert internal forces, breaking these laws on a local scale. Active matter physics is a burgeoning field of study that seeks to analyze systems which do not follow these conservation laws. In this thesis defense, we present an analysis of 2 such systems: fire ant clustering and plant roots, using modeling to link individual behaviors to global phenomena.

First, we show how the strategies of individual fire ants can affect their collective structure. Many social insects cluster, form collective structures to conserve resources, prevent predation, or avoid separation. These systems act as "smart" active matter systems: individuals must adapt local strategies to create global structural stability under perturbation. S invicta fire ants are known to raft when placed in water to improve collective survival. To investigate the effects of such perturbations on these rafts, we subjected 100–1000 ants to Faraday waves at 20 Hz in a water-filled container and mechanical oscillations at 30 Hz in a dry container, recording CO2 emission as measure of energy use. We observed that ants used emitted more CO2 in perturbations with higher peak forces. To explain these group-level phenomena as emergent from individual behaviors, we propose a model where ants enter a passive state (using less energy and exerting less cohesion force) when in contact with enough neighbors. To test this strategy, we developed a Discrete Element Method (DEM) simulation that shows that changing the sensitivity to neighbors creates a negative correlation between clustering efficiency and energy expenditure.  This hypothesis implies that ants use history dependent decision making: to test this, we alternated periods of perturbations and stillness. In these trials, ants on perturbed water spiked CO2 emissions initially, then returned to baseline, even on secondary perturbation. 
Next, we describe plant navigation strategies as an emergent behavior from cell growth. Plants navigate their environments through directional growth to attain resources. This process is achieved through the coordinated behavior of large numbers of cells. How these organ-level behaviors emerge from cellular growth is a major question in plants and developmental biology more broadly. Here, we describe one such behavior: circumnutation, the continuous helical tip movement of roots, in O sativa rice roots. To analyze the emergence of circumnutation from coordinated cell behavior, we combine a Discrete Element Method (DEM)  simulation, where each particle represents an individual cell on the root epidermis with a putative cell-signaling process into the model. This model successfully recapitulated the circumnutation dynamics quantified experimentally across a range of gel stiffnesses by modulating the threshold required for cell-to-cell signal propagation. Roots grown in stiffer gels developed longer mature cells, suggesting that increased gel stiffness lowers the signaling threshold.

Together, these studies show how simulation can be used to test how emergent phenomena emerge from local interactions in smart active matter systems.