John K. Douglass



My general research interests are in neuroethology and neuronal mechanisms of information processing. I received my undergraduate education at Oberlin College (B.A. in Biology, 1980), where my interests in animal behavior led to a field research project concerning whale behavior in Newfoundland, Canada. As a graduate student at Duke University (Ph.D. in Zoology, 1986), I studied sensory mechanisms that underlie animal behavior. For my dissertation, I investigated the ontogeny of light and dark adaptation in an estuarine shrimp (Palaemonetes pugio). The development of the visual system in this animal is of particular interest because the compound eyes undergo a major transformation, from apposition to facultative superposition optics, around the time of metamorphosis from the planktonic larva to the more benthic postlarva. My subsequent research has included studies of visual pigments and spectral sensitivity in birds, amphibians and butterflies, and an investigation of the directional sensitivity of water motion-sensitive hairs on the crayfish tailfan (Procambarus spp.). The latter project led me to the first demonstration in single neurons of stochastic resonance , a nonlinear dynamical effect in which the detectability of weak, nonrandom signals is enhanced by the presence of random noise. Stochastic resonance helps explain the exquisite efficiency of biological sensory systems at detecting very weak signals in a noisy world. Subsequent research has shown that stochastic resonance can operate in central as well as sensory neurons, and can be exploited behaviorally. A provocative implication for the general role of noise in information processing is that animals not only can benefit passively from environmental noise; they also may have evolved ways to exploit noise by optimizing internal sources within their nervous systems.

Since joining the ARL Division of Neurobiology at the University of Arizona, I have been investigating the neural bases for visual motion detection and processing in blow flies (Phaenicia sericata). The detection and analysis of motion is among the most fundamental sensory tasks facing animals that rely upon vision for their survival. Information about visual motion is often essential for finding food & mates and avoiding predators, as well as for navigation and visual equilibrium.

Most physiological investigations of vision in flies have been limited to large neurons at peripheral levels (photoreceptors and primary visual interneurons) or large, wide-receptive-field neurons at deeper levels (premotor neurons and their inputs from large tangential cells). I have focused on very small neurons at intermediate processing levels, and have obtained the first intracellular recordings and stainings from several identified cell types that had long been expected to play crucial roles in early motion processing. My methods of investigating motion processing in the fly brain include using computer-generated visual stimuli during intracellular recordings (Fig.1 and Fig.2) from visual interneurons, intracellular staining and anatomical reconstructions of these neurons, and computational modeling of motion processing networks. This research is part of collaborations with Drs. Nicholas Strausfeld , Chuck Higgins, Irina Sinakevitch, and my student, Jennifer Talley. Two major goals are to understand basic principles of visual motion processing, and to use this knowledge to develop new biologically inspired analog VLSI motion processing chips. I have discovered that directional sensitivity to motion arises quite early in the visual processing pathways of flies, just as in lower vertebrates. My recordings provide the most direct information to date about the nature of elementary motion detecting circuits in insects, and have identified small-field retinotopic neurons only two to three synapses removed from the primary photoreceptors already exhibit some forms of motion selectivity, including orientation-selective and direction-selective responses (Douglass and Strausfeld 1995, 1996, 1998). These findings provide the basis for new models of insect motion detection that are based on known properties of identified neurons (e.g. Douglass and Strausfeld 2000a,b). Other major conclusions involve the manner in which separate, parallel processing streams are segregated among pathways that specialize in distinct aspects of motion information. I am currently focusing on elucidating the properties of these segregated pathways, including recordings from peripheral neurons that may participate in the very earliest stages of motion detection.


Please see Biographical Sketch for references

jkd@neurobio.arizona.edu

Home