This study investigates the role of the Drosophila giant fibre (GF) descending neurons in coordinating escape responses away from attacking predators. This work builds on previous research demonstrating the role of GFs in initiating a short-duration escape take-offs (von Reyn et al., 2014). This new study shows that flies with genetically silenced GFs are captured more frequently by damselfly predators than control flies. Damselflies were successful in capturing GF-silenced flies at slower attack speeds that were necessary to capture control flies. Furthermore, damselflies were able to approach closer to GF-silenced flies during attacks. Interestingly, the subtended size of the damselfly head for GF-silenced captures often exceeded the 40o size threshold previously shown to initiate an action potential in GFs. This suggests that GF responses are important for preventing predators from approaching close enough for successful prey capture. However, by analysing escape responses to artificial looming stimuli, GF-silenced flies were found to initiate take-off with the same latency as control flies but took longer to execute the take-off.
This work makes the argument that giant escape fibres have large axon diameters (which confers higher conduction velocities) to short-circuit other parallel descending pathways that control longer behavioural programs, in favour of shorter programs. This contrasts with the prevailing view that large axon diameters in these circuits are specialised simply to reduce reaction times of the escape response.
Overall, the article provides useful information about the function of Drosophila GFs in an ecologically relevant context. The argument that there is more to large axon escape neurons than simply reducing reaction times is valuable to those broadly interested in the neuronal control of behaviour.
A few minor points:
• Figure 3A: I was initially confused by the steps in the trajectory averages in Figure 3A. I realise these are due to averaging across trajectories of different lengths, as indicated by the bottom row ‘n’ plot. Perhaps include a note in the figure legend along the lines of ‘Note that step changes in trajectory averages are due to averaging trajectories of different durations.’
• Figure 3D: Why is % escape larger for GF silenced flight for fast speeds (>0.13m/s, ~40%) compared to slower speeds (<0.13m/s, ~20%)? This indicates that a higher percentage of GF silenced flies escape faster peak speeds compared to slower speeds?
• Line 224-226: ‘Our analysis of damselfly attack kinematics further indicates that the GF generates timely escape responses by integrating multiple predatory visual features including attack speed and angular size of the predator on the fly’s retina.’. Not clear how the results demonstrate an integration of multiple predatory visual features, please elaborate.
• It would be useful in the discussion to elaborate more on the known physiological/anatomical properties of the Drosophila GF. E.g., what is the response latency when measuring action potentials in the cervical connective and how does this compare with other descending neurons? Has anything been previously reported about the output synapses of GFs? Which neurons do they output onto, and what neurotransmitter is used?
• One useful article missing from the reference list: Perge et al., (2012) Why Do Axons Differ in Caliber? https://doi.org/10.1523/JNEUROSCI.4254-11.2012 This article discusses various advantages of having large axon diameters other than conduction velocities. For example, larger axons increase spike timing precision, and can supply more synaptic terminals. Perhaps these advantages could contribute the mechanism underlying the proposed short-circuiting of parallel pathways.