Eingangs-ausgangsbeziehungen und modell eines drehreflexes der haustaubeColumbia livia

摘要

Pigeons sitting on a turntable are exposed to horizontal$$dot phi $$-ramp-and-hold stimuli (=$$dot phi $$-RP-stimuli;$$dot phi $$angular velocity) (Fig. 1). At the same time, the mean rateE(t)of impulses/0.1 s or 0.2 s ofM. abductor indicis, M. abductor pollicisorM. flexor pollicisis registered. If the rotation axis lies in or in front of the pigeon's head the muscles studied respond phasic-tonicly to an ipsilaterad$$dot phi $$-RP-stimulus; and purely tonicly or also phasic-tonicly to a contralaterad$$dot phi $$-RP-stimulus (Figs. 2C and 4). The heightEPof the tonic response component is independent from the direction of rotation and is only determined by the height$$dot phi _k $$of the$$dot phi $$-plateau (Fig. 2A). The heightEDof the phasic response component is dependent upon the rotation direction and is determined, at a given rotation direction, by$$dot phi _k $$(Fig. 2A) and by the ramp slope$$ddot phi _0 $$($$ddot phi $$angular acceleration) (Fig. 3A). This and the dependence of the tonic component and, with some pigeons, also of the phasic component uponr(distance between pigeon and rotation axis) lead to the conclusion that the tonic component is caused by the centripetal accelerationbp, and the phasic component is caused by the angular acceleration$$ddot phi $$and, with some pigeons, by the tangential accelerationbt. The influence of the acceleration components uponEDandEPis dependent upon the pigeon's position angle Θ relativ to the radiusvectorr(Figs. 4 and 5). The tonic component is the response to a quasi-pitch or a quasi-roll of the pigeon during steady rotation. This results from a comparison of the$$dot phi $$-RP-responses of the pigeon in a radial or tangential position (Fig. 6A) to the pitch and roll responses (Fig. 6B). A model has been developed on the basis of the measured stimulus-response relations (Fig. 7). It contains three input channels (x-,ut- andup-channel) with low-pass filters for the angular, tangential and centripetal acceleration. The channel gain depends upon Θ. The sumŵ(t)of all channel outputs is sent through a pure-time-delay element and finally through a half-wave rectifier with a threshold and an upper limitation. The transfer characteristics of the model agree rather well with those of the pigeon. The model can be interpreted in terms of neurophysiolo