Project 4. Stochastic Resonance in the electroreceptive system of weakly electric fish: experimentation and computation (W. Saidel, Biology; Sunil Shende, Computer Science)

Background.  A sensory threshold either signals the presence of an external stimulus or its absence. Over the last 20 years, a notion of ‘stochastic resonance’ has developed in that white noise is not seen as a corruptor of threshold but as an enhancer at minimal stimulus energy (Ward et al., 2002; Moss et al., 2004).  Subthreshold (ie., nondetectable) visual, auditory and somatosensory  stimuli becomes detectable in the presence of noise. Stochastic resonance interacts also with behavior, as demonstrated in the electroreceptive system of Polydon (the paddlefish) where  weak white electric noise enhances the ability of this fish to obtain prey (Russell et al., 1999). The electroreceptive system of weakly electric fish is used to create an electric map of the environment by emitting an impulsive electric stimulus (an Electric Organ Discharge or EOD) from modified tail muscle cells (Moller, 1995). Frequencies range from a few tens up to several hundred per second (Moller, chapter 8, 1995). EOD field is distorted when two individuals meet, thus fishes use a distortion response called ‘Jamming Avoidance Reflex’ (JAR) consisting of a mutual action: one fish lowers and one increases the frequency (Bullock et al., 1972).

Research.  This project explores the influence of white noise on the threshold of JAR.  In addition, we will initiate a modelling of the electroreceptive system (whose anatomy and physiology is well known, Heiligenberg, 1991).  We are planning to use the Neuron programming system to simulate it and therefore study,  the consequences of noise added to defined stimuli in a model system.

Student Activities. The first goal will be to determine if stochastic resonance plays a role in threshold detection  of Apternotus sp. Firstly, students will learn to use a Matlab-based program to record and replay the ~600 Hz EOD of Apternotus albifrons, the black knife fish, and determine inter-pulse interval prior to hands-on experimentation with a script written by Drs. Shende and Saidel. Secondly, students will study the effect of electrical noise on the playback of a fish’s EOD by a sequential sequence of experiments, producing the JAR by playback. They will then determine the minimal amplitude necessary to induce a JAR, by obtaining the playback amplitude at which no JAR is obtained. The fourth step will add electrical white noise to the no-JAR amplitude.. The final step is to  filter in a variety of ways the white noise to determine if a  specific frequency region induces the SR (Honggi et al.,  1993).