In dragonflies, descending visual neurons code prey direction in population vector form

International Conference on Invertebrate Vision - 2013-08-01Gonzalez-Bellido PT, Peng H, Yang J, Georgopoulos AP, Olberg RM10.3389/conf.fphys.2013.25.00058
The population vector is the weighted vectorial sum activity of an ensemble of neurons and it was first shown to predict the direction of an upcoming arm movement in monkeys (Georgopoulos et al. 1983, 1986). In this study we show that in the dragonfly Libellula luctuosa, the population vector algorithm also decodes the target direction information relayed from the brain to the wing motor centers by a group of 16 neurons. Moreover, these 16 neurons (named Target Selective Descending Neurons or TSDNs; Olberg 1986), perform such directional information coding with high accuracy across 360°. This is significant because the monkey motor cortex requires upwards of 200 neurons to achieve the same performance (Georgopoulos et al. 1988). To obtain the TSDNs directional tuning curves, we impaled a total of 51 TSDNs from 38 animals with sharp electrodes and recorded their responses to a battery of 3497 target trajectories. The target had a constant speed and size, but random location and direction. To confirm cell ID, Lucifer yellow was injected into 32 of the recorded cells. Although the preferred direction, receptive field and morphological traits (3D tracings) for each TSDN type were consistent among animals, spike rates were not. Importantly, the TSDN spatial (receptive field) and temporal (latency) properties matched the area of the retina where the prey is focused and the reaction time, respectively, during predatory flights. The findings of this study are published in Gonzalez-Bellido et al. 2013. Figure 1. (A) Dragonfly L. luctuosa (image courtesy of Steve Thompson). (B) Example of TSDN tuning curve fitted to data points. (C) Tuning widths for each TSDN type.