Category Archives: Theoretical Neuroscience – P. Dayan and L. F. Abbott

Cercal System

There seems to be ample evidence that the evolution of the cercal system is a result of predator demands on various species, a trade-off between touch sensitive types of hairs and wind sensitive types of hairs. This prey driven evolution is to be expected but it is important that this has been proved. With prey drive evolution, understanding the nature of the attack is important to understanding the way the system responses to such an attack. Spiders have two types of attack behavior, one is a running rush and the others is a pounce. A blunt object moving through a viscous medium creates a small shockwave in front of it that precedes the object. For a spherical object the shockwave can be modeled by a 3d Gaussian. Thus, there is warning, but not much, the system must be very sensitive. In addition, the system responds to high frequency sound, as in cricket singing. This represents another evolutionary parameter in addition to predator pressure. I’m not sure just how this is encoded, by what types of hairs.

The wind sensitive system has enough information floating around for the start of a complete model. First there is the stimulus, shockwave of a predator and the structure of a day’s normal airflow, not at all like white noise. The sensory complex is made-up of four cells, the hair cell, the cell that forms the constraining slot. A model of the flow of air around the hair cells on the cerci and models of air flows around a cylindrical array exist. How the hair cell vibrations are encoded by the sensory neuron and any computational action by the slot cell besides acting as a constraint I haven’t found yet but perhaps the literature on touch receptors might help as the wind receptors are believed to be evolved from them. Next there is a sensory neuron with a glial companion. Studies of olfactory sensory complexes show that the dendrite of the sensory neuron and the glial companion are electronically connected. Axons arbors of most of the various types of sensory neurons have been published. The neural specific map has been found (a helix!) and information on how this map connects to at least some interneurons exists. Thus at least a rough outline of the computational pipeline between a stimulus and a response exists although I don’t know of how the interneurons connect to the leg motor neurons or even if they are in the same ganglia.

I looked briefly into the jump response, much more about the locust than anything else. One thing is that the locust takes about 1/2 sec to power-up their catapult to jump. This seems to be much too long for a system as sensitive as the cercal so their must be a locking mechanism to store the jump potential and still allow movement. I did find something for a certain cricket that has more muscles and more different types of neurons to work them. It would be interesting to see if cercal structure that varies between touch reception and wind reception depending on predator pressure makes for different jump structure.

I’d like to have a review paper with maybe a research proposal by the end of the year. Probably an ambitious goal but why not?