Jan Wessnitzer

A Model of Non-elemental Associative Learning in the Mushroom Body Neuropil of the Insect Brain

We developed a computational model of the mushroom body (MB), a prominent region of multimodal integration in the insect brain, and tested the models performance for non-elemental associative learning in visual pattern avoidance tasks. We employ a realistic spiking neuron model and spike time dependent plasticity, and learning performance is investigated in closed-loop conditions. We show that the distinctive neuroarchitecture (divergence onto MB neurons and convergence from MB neurons, with an otherwise non-specific connectivity) is sufficient for solving non-elemental learning tasks and thus modulating underlying reflexes in context-dependent, heterarchical manner.

Open source tracking and analysis of adult Drosophila locomotion in Buridan's paradigm with and without visual targets

Background Insects have been among the most widely used model systems for studying the control of locomotion by nervous systems. In Drosophila, we implemented a simple test for locomotion: in Buridans paradigm, flies walk back and forth between two inaccessible visual targets [1]. Until today, the lack of easily accessible tools for tracking the fly position and analyzing its trajectory has probably contributed to the slow acceptance of Buridans paradigm. Methodology/Principal Findings We present here a package of open source software designed to track a single animal walking in a homogenous environment (Buritrack) and to analyze its trajectory. The Centroid Trajectory Analysis (CeTrAn) software is coded in the open source statistics project R. It extracts eleven metrics and includes correlation analyses and a Principal Components Analysis (PCA). It was designed to be easily customized to personal requirements. In combination with inexpensive hardware, these tools can readily be used for teaching and research purposes. We demonstrate the capabilities of our package by measuring the locomotor behavior of adult Drosophila melanogaster (whose wings were clipped), either in the presence or in the absence of visual targets, and comparing the latter to different computer-generated data. The analysis of the trajectories confirms that flies are centrophobic and shows that inaccessible visual targets can alter the orientation of the flies without changing their overall patterns of activity. Conclusions/Significance Using computer generated data, the analysis software was tested, and chance values for some metrics (as well as chance value for their correlation) were set. Our results prompt the hypothesis that fixation behavior is observed only if negative phototaxis can overcome the propensity of the flies to avoid the center of the platform. Together with our companion paper, we provide new tools to promote Open Science as well as the collection and analysis of digital behavioral data.

Evolving a Neural Model of Insect Path Integration

Path integration is an important navigation strategy in many animal species. We use a genetic algorithm to evolve a novel neural model of path integration, based on input from cells that encode the heading of the agent in a manner comparable to the polarization-sensitive interneurons found in insects. The home vector is encoded as a population code across a circular array of cells that integrate this input. This code can be used to control return to the home position. We demonstrate the capabilities of the network under noisy conditions in simulation and on a robot.

Place memory in crickets

Certain insect species are known to relocate nest or food sites using landmarks, but the generality of this capability among insects, and whether insect place memory can be used in novel task settings, is not known. We tested the ability of crickets to use surrounding visual cues to relocate an invisible target in an analogue of the Morris water maze, a standard paradigm for spatial memory tests on rodents. Adult female Gryllus bimaculatus were released into an arena with a floor heated to an aversive temperature, with one hidden cool spot. Over 10 trials, the time taken to find the cool spot decreased significantly. The best performance was obtained when a natural scene was provided on the arena walls. Animals can relocate the position from novel starting points. When the scene is rotated, they preferentially approach the fictive target position corresponding to the rotation. We note that this navigational capability does not necessarily imply the animal has an internal spatial representation.

A model of associative learning in the mushroom body

The mushroom body is a prominent invertebrate neuropil strongly associated with learning and memory. We built a high-level computational model of this structure using simplified but realistic models of neurons and synapses, and developed a learning rule based on activity dependent pre-synaptic facilitation. We show that our model, which is consistent with mushroom body Drosophila data and incorporates Aplysia learning, is able to both acquire and later recall CS–US associations. We demonstrate that a highly divergent input connectivity to the mushroom body and strong periodic inhibition both serve to improve overall learning performance. We also examine the problem of how synaptic conductance, driven by successive training events, obtains a value appropriate for the stimulus being learnt. We employ two feedback mechanisms: one stabilises strength at an initial level appropriate for an association; another prevents strength increase for established associations.

Resonant neurons and bushcricket behaviour

The resonant properties of the intrinsic dynamics of single neurons could play a direct role in behaviour. One plausible role is in the recognition of temporal patterns, such as that seen in the auditory communication systems of Orthoptera. Recent behavioural data from bushcrickets suggests that this behaviour has interesting resonance properties, but the underlying mechanism is unknown. Here we show that a very simple and general model for neural resonance could directly account for the different behavioural responses of bushcrickets to different song patterns.

Dietary Salt Levels Affect Salt Preference and Learning in Larval Drosophila

Drosophila larvae change from exhibiting attraction to aversion as the concentration of salt in a substrate is increased. However, some aversive concentrations appear to act as positive reinforcers, increasing attraction to an odour with which they have been paired. We test whether this surprising dissociation between the unconditioned and conditioned response depends on the larvaes experience of salt concentration in their food. We find that although the point at which a NaCl concentration becomes aversive shifts with different rearing experience, the dissociation remains evident. Testing larvae using a substrate 0.025M above the NaCl concentration on which the larvae were reared consistently results in aversive choice behaviour but appetitive reinforcement effects.

Elemental and non-elemental olfactory learning in Drosophila

Brain complexity varies across many orders of magnitude between animals, and it is often assumed that complexity underpins cognition. It is thus important to explore the cognitive capacity of widely used model organisms such as Drosophila. We systematically investigated the flys ability to learn discriminations involving compound olfactory stimuli associated with shock. Flies could distinguish binary mixtures (AB+ CD-), including overlapping mixtures (AB+ BC-). They could learn positive patterning (AB+A- B-) but could not learn negative patterning (A+ B+ AB-) or solve a biconditional discrimination task (AB+ CD+ AC- BD-). Learning about the elements of a compound (AB+) was not affected by prior conditioning of one of the elements (A+ AB+): flies do not exhibit blocking in this task. We compare these results with the predictions from simulation of several well-known theoretical models of learning, and find none are fully consistent with the overall pattern of observed behaviour.

A cricket-controlled robot orienting towards a sound source

We designed a closed-loop experimental setup that interfaces an insect with a robot for testing phonotaxis (sound recognition and localisation) behaviour in crickets. The experimental platform consists of a trackball mounted on a robot, so that a cricket walking on the trackball has its movements translated into corresponding movement of the robot. We describe the implementation of this system and compare the performance with previous cricket data and a neural model circuit on the same robot. Crickets are able to drive the robot towards the sound source, although they show substantially longer walking and stopping bouts than in more standard experimental setups. The potential and the current limitations of the robot setup together with alternative designs are discussed.

Perception for Action in Insects

We review the concept of ‘perception for action’, contrasting the traditional view of perception as internal representation with the idea of transformation in a closed loop system. This introduces recent approaches using active perception, dynamical systems theory, action-based agent architectures and consideration of the role of predictive loops. We then apply these ideas to insect behaviour and neurophysiology, with particular attention to higher brain centres. We propose an insect brain control architecture for robotics.