Neuronal Mechanisms Mediating Prey Pursuit Behaviour in Predatory Flying Insects
Lead Research Organisation:
University of Cambridge
Department Name: Physiology Development and Neuroscience
Abstract
Various neuronal specialisations have evolved in predatory flying insects to solve the task of efficiently capturing fast flying prey. In dragonflies, a system of small target movement detectors in the optic lobes function to discriminate prey items from the visual surroundings (Nordstrom, 2012). These neurons are thought to transduce information about prey position and velocity to a small population of target selective descending neurons (TSDNs) in the ventral cord to steer the dragonfly along an intercepting flight (Olberg, 1986; Gonzalez-Bellido et al., 2013). However, the input received by TSDNs in the brain and the mechanism triggering the initiation of pursuit has remained unexplored.
Analogous neuronal systems to those in dragonflies are hypothesised to exist in other predatory insects which perform the similar task of catching prey. For example the evolutionarily distant dipteran killer fly (Coenosia attenuata) employs an aerial hunting strategy that shares many parallels with that of dragonflies, albeit under different physiological and ecological constraints (Gonzalez-Bellido et al., 2011; Wardill et al., 2015). Damselflies on the other hand, a sister group to dragonflies, often snatch stationary prey from a hovering position, a behaviour termed gleaning (Yee et al., 2012). Whether these two groups employ similar sensorimotor control systems to those identified in dragonflies is unknown. By comparing the behaviour and neuronal specialisations at the level of descending neurons in these three taxa, we will investigate if convergent (killer fly) and/or conserved (damselfly) features of the dragonfly TSDN system are present in other predatory aerial invertebrates. Such findings will serve to elucidate general design principles relevant for controlling intercepting behaviour; in addition, any idiosyncratic properties may highlight features specialised for a given ecological niche.
Overall, the focus of this PhD is to extend understanding of the sensorimotor conversion at the level of TSDNs in predatory flying insects in three ways:
1. Identify generalised and/or idiosyncratic features of target selective descending neuron systems across three different species with varying physiology and behavioural niches
2. Characterise the input to TSDNs in the central brain, making use of the established knowledge for dragonfly TSDNs in addition to the neuroanatomical data available for dipteran species
3. Investigate mechanisms in which sensory information about prey is integrated with a decision to initiate an attack, making use of the ability to control the killer fly's hunting history in the laboratory as a behavioural correlate.
Analogous neuronal systems to those in dragonflies are hypothesised to exist in other predatory insects which perform the similar task of catching prey. For example the evolutionarily distant dipteran killer fly (Coenosia attenuata) employs an aerial hunting strategy that shares many parallels with that of dragonflies, albeit under different physiological and ecological constraints (Gonzalez-Bellido et al., 2011; Wardill et al., 2015). Damselflies on the other hand, a sister group to dragonflies, often snatch stationary prey from a hovering position, a behaviour termed gleaning (Yee et al., 2012). Whether these two groups employ similar sensorimotor control systems to those identified in dragonflies is unknown. By comparing the behaviour and neuronal specialisations at the level of descending neurons in these three taxa, we will investigate if convergent (killer fly) and/or conserved (damselfly) features of the dragonfly TSDN system are present in other predatory aerial invertebrates. Such findings will serve to elucidate general design principles relevant for controlling intercepting behaviour; in addition, any idiosyncratic properties may highlight features specialised for a given ecological niche.
Overall, the focus of this PhD is to extend understanding of the sensorimotor conversion at the level of TSDNs in predatory flying insects in three ways:
1. Identify generalised and/or idiosyncratic features of target selective descending neuron systems across three different species with varying physiology and behavioural niches
2. Characterise the input to TSDNs in the central brain, making use of the established knowledge for dragonfly TSDNs in addition to the neuroanatomical data available for dipteran species
3. Investigate mechanisms in which sensory information about prey is integrated with a decision to initiate an attack, making use of the ability to control the killer fly's hunting history in the laboratory as a behavioural correlate.
Organisations
People |
ORCID iD |
| Liria Masuda-Nakagawa (Primary Supervisor) | |
| Jack Supple (Student) |
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| BB/M011194/1 | 01/10/2015 | 31/03/2024 | |||
| 1644265 | Studentship | BB/M011194/1 | 01/10/2015 | 30/09/2019 | Jack Supple |
| Description | 1. I have discovered that descending neurons in the Damselfly integrate binocular information about moving prey objects. This means that the motor commands sent to control the legs and wings incorporate binocular information, indicating that the damselfly fuses the two binocular images in the brain, which is a central computational step in stereopsis in humans. 2. I have discovered that target selective descending neurons in dipterans are inhibited by background wide-field optic flow when the two are correlated. These neurons respond to small moving objects in the visual field and are thought to control the wings, legs, and neck during flight. The inhibition of target responses when the target and background movement are correlated indicates that these neurons actually represent an 'ego-centric' inertial frame of reference of the fly relative to the prey instead of a retinal frame of reference. This functions to avoid initiating flight maneuvers caused by sensor wobble (e.g. flapping, head rotations). Retinal frames of reference exist in lower visual processing areas of the fly, and we have shown how this is combined in the sensorimotor pathway. |
| Exploitation Route | The findings in 2. above are important in the study of target interception strategies. Most work is looks at algorithms such as proportional navigation, which is used by insects, birds, and man made systems used in transportation and guided missiles. To date, work has described various behavioural features of proportiona navigation such as gain constants and error, but little was described about what cues are used to generate an intertial frame of reference. We have suggested how earlier visual processing might be integrated to generate this frame of reference at the neuronal level, which will be of interest for further studies on the neuronal basis of interception. |
| Sectors | Aerospace, Defence and Marine |
| URL | http://www.jneurosci.org/content/38/50/10725 |