Investigation of neurons involved in the analysis of gestalt prey features in the frogRana temporaria
- Cite this article as:
- Schürg-Pfeiffer, E. & Ewert, J.P. J. Comp. Physiol. (1981) 141: 139. doi:10.1007/BF01342660
- 37 Downloads
In frogs,Rana temporaria, the activity of 63 single neurons-retinal ganglion cells (classes R 1, R2, and R3) and tectal neurons (classes T5(1), T5(2), T5(3) and T7) — were quantitatively investigated in response to different configurational moving stimuli of various sizes: (i) squares, (ii) stripes moving in direction of their long axis (worm configuration), and (iii) stripes moving perpendicular to the direction of their long axis (antiworm configuration). All stimuli were moved at constant visual angular velocity. Neuronal activity was evaluated using several methods.
The influence of stimulus area on the neuronal activity was investigated with square stimuli (Figs. 1–4, 6–10). Units were observed to respond optimally according to the different sizes of excitatory receptive fields. Retinal class R1 neurons (ERF=3.5° diam.) were best activated by ≈3° stimuli, class R2 (ERF =5.4°) by 3–4° stimuli and class R 3 (ERF=7.3°) by ≈8° stimuli. Tectal class T5(1) neurons (ERF=21.5°) were best activated by ≈8° stimuli, class T5(2) (ERF=17.6°) by ≈ 4° and class T5(3) (ERF=24.7°) by 8–16° stimuli. Class T 7 neurons (ERF=4.0°) were optimally activated in the range of 3–4°.
Effects of stimulus configuration on the neuronal activity were studied with stripes moving either in worm- or antiworm-configuration (Figs. 1–11). According to the receptive field organization in terms of a central ERF and a surrounding IRF, retinal ganglion cells exhibited different selective configurational sensitivity in distinct size ranges. If stripe stimuli were not longer than the diameter of the ERF, the antiworm configuration activated a ganglion cell more strongly than the worm configuration. This effect was most obvious in class R3 neurons. If stripes were longer than the ERF diameter then the worm was clearly preferred against the antiworm. This effect was strongest in R1 and R2 neurons. Tectal class T5(1) neurons showed almost no configurational sensitivity in the present context (Fig. 6); T5(2) neurons were more strongly activated by worm than by antiworm stimuli (Fig. 7), and T5 (3) neurons showed the reverse configurational preference (Fig. 9). Class T7 neurons exhibited worm preference only for stimuli longer than 8° (Fig. 10).
When the results inRana temporaria are compared with those previously obtained inBufo bufo (Figs. 11 and 12), it becomes evident that the investigated neurons of the visual system have similar response properties. However, the acuity of configurational selectivity in frog T5(2) neurons is not as sharp as that observed in corresponding tectal neurons in common toads.
In connection with the ‘command system concept’ — and in agreement with our earlier results — evidence is given that the outputs of several tectal neuron types, each forming acommand element, together constitute acommand system, which itself activates a specific motor pattern of the prey-catching sequence: orienting, approaching, fixating and snapping. Each motor command requires simultaneous activation of command elements subservingrecognition (stimulus classification) andlocalization (strategy of catching). We suggest that class T 5 (2) neurons are command elements with recognition properties, which are determinedby interaction of neuronal networks. The correspondingdecision-making process (prey/ nonprey)precedes the goal oriented behavioral response.
excitatory receptive field
inhibitory receptive field
caudal dorsal thalamus