Individual-based modelling of temperature-induced aggregation behaviour

Gerald Radspieler, Ronald Thenius, Thomas Schmickl
Proceedings MATHMOD 09 Vienna - Full Papers CD Volume, ARGESIM Report 35 (2009), 895 - 903




There are several species of animals from different phyla that are known to exhibit diverse forms of social behaviour. Certain insect species like honeybees implement an interesting form of emergent social behaviour, based on a simple set of rules for individual behaviour rather than on complex cognitive interactions between individuals or with a leader. A good example for the bees’ swarm behaviour is temperature induced aggregation. Young bees show a clear preference for temperatures around 36° C [1]. If a group of bees is allowed to move freely on a ground exhibiting a two dimensional temperature gradient, their likeliness to come to rest in an area of preferred temperature is positively correlated to the number of interacting individuals up to a certain level. We suggested four different hypotheses to explain the basic mechanisms of this behaviour. In order to test them, we developed an individual based (bottom-up) NetLogo [2] model which allows us to simulate a group of bees moving in a temperature gradient. By varying the influence of each of the four hypothetical mechanisms and comparing the result of the respective simulation to the results of real life experiments, we receive valuable clues to determine the likeliness of each hypothesis.

The model proved successful in replicating many aspects in the bees’ behaviour. As an example, the figure below demonstrates the model’s ability to reproduce certain aspects of the real bees’ behaviour. It shows a comparison of the results of a real experiment with bees in an arena exhibiting a temperature gradient on its ground and a simulation with a similar setup.


[1] H. Heran: Untersuchungen über den Temperatursinn der Honigbiene (Apis mellifica) unter besonderer Berücksichtigung der Wahrnehmung von Strahlungswärme. In: Zeitschrift für Vergleichende Physiologie, 34, 179-207
[2] NetLogo website:


Download PDF: