|
In biological systems visual motion plays an important role in the control of locomotion and navigation. The coordination of flight movements presents an exceptionally challenging task for visual guidance. One major problem is that the large range of stimulus intensities and dynamics resulting from the variety of different flight manoeuvres in a heterogeneous environment might exceed the limited instantaneous operating range of sensory and nerve cells. Neuronal adaptation has been proposed to self-calibrate the system to the currently prevailing range of stimulus characteristics. In this talk it is examined whether adaptation helps visual motion- sensitive neurons to extract from their complex input those stimulus features that are relevant in a momentary behavioural situation, in particular novel and unexpected stimuli. The blowfly (Calliphora vicina) is used as an animal model for the study of neuronal processing of self-motion induced visual input, so-called optic flow. In the fly visual system optic flow is processed by a small number of individually identifiable neurons, which respond fairly specifically to certain types of self-motion or to the motion of objects in the environment. We tested how motion adaptation affects the sensitivity of these neurons for different types of stimuli. Adaptation to random velocity modulations of a grating pattern led to an overall decrease of the response amplitude and to a degradation of the linear representation of motion velocity. Remarkably, this degradation was not caused by a decrease in reliability, but by an increase in non-linear processing, leading to strong responses to abrupt changes in motion velocity, but attenuated responses to more gradual velocity changes. In a second series of experiments a sustained uniform motion stimulus was interspersed with sudden changes in velocity or in other stimulus parameters (contrast, grating orientation or spatial frequency). Adaptation consistently led to an accentuation of the responses to such stimulus discontinuities. Finally, when using naturalistic stimuli, which were reconstructed from trajectories during semi-free flight, adaptation enhanced neuronal responses to objects approached by the fly. We therefore propose that in optic flow processing neuronal adaptation plays an important role in feature extraction and novelty detection. |
|