Collective dynamics and model verification: Connecting kinetic modeling to data

Whirligig Beetles vs. Swarm Models: Perturb and Measure Emergent Properties

William Romey

SUNY Potsdam


Swarm simulation models that have “lifelike” properties can be readily produced with a few attraction and repulsion rules. The challenge is to obtain accurate data about how real individuals move in a group, incorporate that data into a model, then compare the emergent properties of the model with the empirical data. For the last decade we have been in an exploratory phase of group simulation models. Now, we should push for a more rigorous “strong inference” approach in which alternative hypotheses (simulations) are put forth, and rejected or verified according to rigid criteria. We will present several studies in which we compare empirical data on insect swarms with alternative simulation models. The empirical system we work with, whirligig beetle swarms, are well suited to these studies because the beetles swim in two-dimensions at the surface of water, are easily marked, filmed in the laboratory, and their trajectories determined. I will present some of the following questions my lab has studied: 1) how do individual differences (hunger, sex, size) influence position within a group? 2) how do manipulation of long vs. short range sensors influence group escape responses? 3) how many knowledgeable alarm signalers does it take to trigger group level responses? 4) do group fright responses radiate from the first responder or the center of a group? 5) what is the effect of leadership on the direction of a group as whole?