Snatching their next meal often requires insects to attack upwards, as their prey is more easily visible towards a clear sky background. But this is not the case for killer flies (Coenosia attenuata), which take the risk of performing swift aerial dives to grab their prey from above. Intrigued by this behaviour, Paloma Gonzalez-Bellido from the University of Minnesota, USA, with Sergio Rossoni from the University of Cambridge, UK, and collaborators from Cambridge and the University of Lincoln, UK, set off to understand how these dives compare with upwards or sideways attacks more commonly found in insects in the wild to learn about the flies’ unorthodox hunting strategy.
The team released pairs of killer flies in a transparent rectangular arena, dangled a moving black bead in front of them to give them a target to attack, and filmed their hunting manoeuvres in 3D using high-speed cameras. The researchers then used the recordings to extract the flies’ velocity and acceleration during each dive and found that the flies that were taking off from the ceiling reached significantly higher accelerations than those taking off from the walls or the floor of the arena. The dive accelerations were also higher than those measured in free-falling unconscious killer flies, suggesting that they work together with gravity to propel themselves downwards even faster. The researchers also found that the aeronauts could use the angle of their line of sight – the imaginary line between the fly and its prey – to time their take-off to target and influence their dive angle.
Aiming to understand how these flight characteristics affect the flies’ hunting strategy, the researchers looked into the factors that determine how the killers navigate in for a kill. Do the flies head directly toward their target, or do the killers observe how their meal changes direction to predict where their victim is going to arrive and aim to intercept it at some point in the future? The researchers converted these strategies into mathematical models, used the data for the position and velocity of real flies and the position of the bead target, and simulated potential flight paths that could have resulted from each strategy. They then compared those paths with the real-life trajectories of the killers and were able to show that the second strategy, where the attackers aim to intercept their quarry at some point in the future, best predicted the killer flies’ paths, no matter where they took off from. However, they also noted that the flies that took off from the ceiling and further ahead of their target were particularly vulnerable to missing their meal. The researchers then adapted their model of the flies’ navigation strategy to include the acceleration needed for the killers to turn a corner and attempt a follow-up attack – a measure of the steering effort required – and showed an even higher compatibility of this adapted model to real-life flight trajectories.
But why perform aerial dives given the additional steering effort required, compared with a direct attack from below? Well, for starters, Gonzalez-Bellido and her colleagues discovered that dives provide the fastest route for a first encounter with a victim and that plummeting from above produces a greater impact than attacks mounted from the walls or the floor of the arena. If, in addition, we include that killer flies in the wild are not limited by the size of a laboratory flight arena and would certainly attack again after the first near-miss, there is only one conclusion to make: dives are a powerful attack strategy, but a head start isn't helpful if you are a hungry fly.