Collective mechanical adaptation of honeybee swarms Journal Article uri icon



  • Honeybee Apis mellifera swarms form clusters made solely of bees attached to each other, forming pendant structures on tree branches (1). These clusters can be hundreds of times the size of a single organism. How these structures are stably maintained under the influence of static gravity and dynamic stimuli (e.g. wind) is unknown. To address this, we created pendant conical clusters attached to a board that was shaken with varying amplitude, frequency and total duration. Our observations show that horizontally shaken clusters spread out to form wider, flatter cones, i.e. the cluster adapts to the dynamic loading conditions, but in a reversible manner - when the loading is removed, the cluster recovers its original shape, slowly. Measuring the response of a cluster to a sharp pendular excitation before and after it adapted shows that the flattened cones deform less and relax faster than the elongated ones, i.e. they are more stable mechanically. We use particle-based simulations of a passive assemblage to suggest a behavioral hypothesis that individual bees respond to local variations in strain. This behavioral response improves the collective stability of the cluster as a whole at the expense of increasing the average mechanical burden experienced by the individual. Simulations using this rule explain our observations of adaptation to horizontal shaking. The simulations also suggest that vertical shaking will not lead to significant differential strains and thus no adaptation. To test this, we shake the cluster vertically and find that indeed there is no response to this stimulus. Altogether, our results show how an active, functional super–organism structure can respond adaptively to dynamic mechanical loading by changing its morphology to achieve better load sharing.

publication date

  • September 14, 2017

has restriction

  • green

Date in CU Experts

  • January 28, 2021 6:45 AM

Full Author List

  • Peleg O; Peters JM; Salcedo MK; Mahadevan L

author count

  • 4

Other Profiles