Is the quorum threshold for emergent group response in whirligigs absolute or proportional?

Groups of animals sometimes coordinate their individual behaviours to produce an emergent group response. Examples of these quorum responses include stampedes in ungulates and orientation flights in honeybee swarms. In these groups, there may be some individuals who are knowledgeable about the threat or direction to go to, and others who are not. Few experimental studies have convincingly addressed whether the number of knowledgeable individuals to trigger an emergent group response is a fixed (absolute) number or a fixed proportion (percentage) of the group. We tested whether this threshold to produce an emergent group response was absolute or proportional in an experimental study of whirligig beetles (Gyrinidae: Dineutes). When whirligig beetles see an aerial predator, individuals make a startle response. If enough beetles startle, then the whole group makes a flash expansion. In our study, we manipulated the numbers of beetles in a group that were able to see the predator model by covering their eyes. We also manipulated group size (12, 24, 48). Our results reject the absolute hypothesis and support the proportional hypothesis for how many knowledgeable whirligigs it takes in a group to elicit an emergent flash expansion. At all three group sizes the threshold was approximately 10%. We also examined the interaction of the ratio of sighted/unsighted beetles and group size on swarm density, group area and longevity (duration of the flash expansion). Longevity was significantly leptokurtic, as would be expected for a stereotyped display. This is one of the first controlled empirical studies to differentiate between absolute and proportional thresholds in producing an emergent predator avoidance response.