Journal of Comparative Physiology. A

J. Comp. Physiol. 132, 191-201 (1979)

9 by Springer-Verlag 1979

Directional Sensitivity, Invariance and Variability of Tectal T5 Neurons in Response to Moving Configurational Stimuli in the Toad Bufo bufo (L.)* J.-P. Ewert, H.-W. Botchers, and A. v. Wietersheim Neuroethologyand BiocyberneticLaboratories, University of Kassel, Heinrich-Plett-Strasse40, D-3500 Kassel, Federal Republic of Germany Accepted April 2, 1979

Summary. 1. Different configurational visual stimuli traversed, at constant angular velocity, the centers of the receptive fields of single tectal T5 neurons (n = 42) in various directions of the x - y coordinates. The neuronal responses were recorded extracellularly in paralyzed toads Bufo bufo (L.). All data were processed by a computer, in part on-line and off-line. 2. Some of the T5 neurons showed no obvious change in their discharge rate when a stimulus traversed their receptive fields in different directions. Other neurons of the T5 group exhibited directional sensitivity which could be correlated with the shape of the moving stimulus. 3. T5(2) neurons, as described previously, were activated strongly by a stripe moving along its axis (worm-like), but they responded weakly, if the stripe axis was oriented perpendicular to the direction of movement (antiworm-like). The selective responsiveness was found to be invariant for the stimulus movement direction. 4. In other types of T5 neurons the selective response to the configurational stimuli tested (worms, antiworms) was a function of the movement direction. Various subtypes with different kinds of correlation between stimulus configuration and movement direction have been identified. Some of these neurons obviously correspond to the previously recorded T5(1) neurons. 5. Longterm recordings from the same neuron indicated that the response property - at least in some T5(1) neurons can change, when identical series of stimuli were repeatedly presented over a period of several hours.

Introduction

It is known that the linkage between stimulus geometry and movement direction plays a prominent role for the discrimination of prey objects in toads Bufo bufo (L.) (Ewert 1968; summary c.f. Ewert, 1976). Recent behavioral studies have shown that this important configurational discrimination is invariant for the direction of stimulus movement in the x - y coordinates of the visual field (Ewert et al., 1979b; Beck and Ewert, 1979). The question of directional sensitivity concerning particular stimulus transformations was recently investigated in retinal class R3 ganglion cells of Bufo bufo (Ewert et al., 1979a). The present paper analyzes tectal T5 neurons. It is supposed that T5 neurons are driven both by retinal class R2 and R3 neurons. Class T5(2) neurons appear to be part of a system which recognizes prey a n d " c o m m a n d s " the orienting turn of the prey-catching sequence (Ewert, 1974). Response properties of these neurons to horizontally moving visual stimuli have already been investigated (Ewert, 1974; von Wietersheim and Ewert, 1978). Directional sensitivity of T5 neurons in frogs to horizontally moving uniform standard stimuli has been noted by Grfisser and Grtisser-Cornehls (1968), but this phenomenon was not studied quantitatively. The present study analyzes the activity of tectal T5 neurons fn response to different configurational stimuli traversing the centers of their receptive fields in various directions of the x - y coordinates.

Material and Methods

* Supported by the Deutsche Forschungsgemeinschaft Ew 7/6 Abbreviation: ERF, excitatory receptivefield

Subjects. Commontoads Bufo b. bufo (L.) werekept under standard laboratory conditions. The experiments were performed during June through September.

0340-7594/79/0132/0191/$02.20

192

J.-P. Ewert et al. : Tectal T5 Neurons in the Toad Dw~ =

-1

0

+1

DORSAL 900

135 ~

0 D ,,I NASAL

/

9

180 ~

TEMPORAL

225 ~

315 ~ 270 ~ VENTRAL

A

Stimulus Parameters. Visual stimuli were two-dimensional rectangular pieces of black cardboard moved on a h o m o g e n e o u s white background. The a m o u n t of the stimulus background contrast was C ~ 0 . 9 . The following standard stimuli were used: a square of 2 ~ edge length (S) and stripes of 2 ~ x 8~ or 2 ~ x 20~ in size. Stripes were moved in the direction of their long axis (worm-like: W1 or W2), and perpendicular to it (antiworm-like: A 1 or A2). The standard visual angular velocity was v=7.6~ The stimulus m o v e m e n t direction could be adjusted in the visual field x - y coordinates for m e = 0 , 45, 90, 135, 180, 225, 270, 315 ~ (c.f. Fig. 1A). The distance between the eye of the toad and the stimulus was held constant at d = 25 cm.

Stimulus Presentation. Stimuli were presented to the immobilized toad by m e a n s of a special kind of perimeter described by Ewert et al. (1979a). Stimuli traversed the center of the excitatory receptive field (ERF) of the neuron according to an electronically controlled program. Two sets of stimuli were used: S, W1, A1 or S, W2, Az. The order of stimuli set presentation and the directions of m o v e m e n t were chosen at random. Successive stimulus traverses of the receptive field were separated by constant 90 s recovery pauses. Each neuron was investigated only with one of the stimulus sets. The entire stimulation program lasted about 50 rain; it was repeated as long as the neuron could be recorded at a constant signal/noise ratio. Recording and Preparation. Toads were paralyzed and prepared as described by Ewert and Hock (1972) and yon Wietersheim and Ewert (1978). All preparations were performed under anesthesia to avoid stress to the animaI. The actual recording h a d to be done in the awake paralyzed animal. During the experiment the animal was unrestrained and immobilized by succinylcholine; it just sat on a wet sponge. Action potentials of single neurons were recorded extracellularly with either stainless steel or tungsten electrodes from the central layers of the exposed optic tectum. Standard recording techniques were used (Ewert and Hock, 1972).

Some General Properties of T5 Neurons in Bufo bufo. N e u r o n s of this group do not show on or off responses to sudden diffuse light changes in the entire visual field. All of these neurons are monocularly driven; they have excitatory receptive fields of E R F g 2 6 ~ diameter (horizontal measurements). T5 neuron response is m o v e m e n t specific. They have no spontaneous activity under the conditions tested. The topographical arrangement

B

Fig. 1 A and B. " S t a n d a r d fields" for plotting functional neuronal properties in polar coordinates. Different configurationai stimuli traversed the receptive field center of a neuron in 8 directions of the x-y coordinates. Values for the neuronal activity R (and R +) are plotted in the diagram A (c.f. Figs. 2 and 3). Discrimination values DW,A (and DW,A) + are plotted in diagram B (c.f. Figs. 4 and 5). The receptive fields were located in different regions of the visual field. The "standard fields" shown here correspond to the left eye (electrode position in the right optic teetum). Experimental data obtained for neurons investigated in the right or the left optic tectum were transformed into this " s t a n d a r d field". F o r further explanations see text

of the E R F s corresponds roughly with the retinotectal projection diagram. Neurons show relatively strong adaptation, lasting up to 70 s, immediately after stimulus traverse of the ERF. N e u r o n s of this general response type are termed T5 neurons in the n o m e n clature by Grfisser and Griisser-Cornehls (1976). But there is evidence that these cells are not homogeneous in their response properties. (c.f. von Wietersheim and Ewert, 1978).

Calculation of the Neuronal Output. A total of n = 4 2 single T5 neurons were investigated quantitatively. For data analysis a single neuron was considered only when it could be tested during the entire stimulation program (all 8 directions; n = 3 4 ) , at least for the vertical and the horizontal 4 m o v e m e n t directions (n=42). The following output parameters - describing the neuronal activity - were measured during stimulus traverse of the E R F : (t) N u m b e r n of spikes, (2) time t between first and last spike recorded, (3) intervals i between successive spikes. There are various ways to determine the activity of a neuron. A c o m m o n measure is the average discharge frequency f = n/t (= R). Another possibility which considers also the total n u m b e r of spikes, is the product nf (= R +), which is related to the power of the output. The values for R and R + at each stimulus m o v e m e n t direction tested were plotted in a " s t a n d a r d field" polar coordinate system (Fig. 1A). Data Analysis. A quantitative measure for the sensitivity of a neuron to a stripe stimulus moving either in a worm-like (W) or an antiworm-like m a n n e r (A) is given by the " f o r m - c o n t r a s t " formula

Dw,A={Rw_RA} {Rw+RA } 1 where R is the response of the neuron to the worm or to the antiworm configuration respectively (Ewert etal., 1978). The values Dw,A can be expected to be between + 1 and - 1 (Fig. 1B). If the w o r m configuration is preferred, DW.A becomes positive and is plotted outside a "zero-circle" (c.f. F!g. 1B). If the antiworm configuration is preferred, Dw,A becomes negative and its values are plotted within the "zero-circle" ; the origin of the coordinates system in Fig. 1 B corresponds to DW,A= -- 1. If Dw.A= + 1 or -- 1, both stimulus configurations are discriminated by a yes/no decision. The discrimination values were determined both for R (Dw,a) and R + (Dw,A). + All data have been processed and the diagrams plotted by a Nicolet computer, in part on-line and off-line. The examples shown in Figs. 2-5 were drawn from computer plots. Each neuron

J.-P. Ewert et al. : Tectal T5 N e u r o n s in the Toad

W7"7"78'I

C

193

W7"7+78'4

0 , R~=7'5 R ~5oo

,

E25.8"78"1

E

W20,6.78.2

F R =7.5 R+=500

,

E25"8"78"5

was given an identification number, for example B 15.8.78.3.1, indicating the initial of experimenter's name, the date, the current n m n b e r of the neuron from the day, and the number of the stimulation program tested. N e u r o n s have been stimulated with the patterns S, Wll A1 by B and E, and with S, W2; A2 by W.

Results

If the E R F centers of tectal T5 neurons were traversed by the different configurational stimuli in various directions of the x - y coordinates, different response types could be obtained. The 9 reported here are based on quantitative measurements of 42 single neurons. The examples of neuronal response types appear to show some representative properties of the T5 neurons according to the stimulus parameters tested.

615"9,78-1

Fig. 2A-F. Directional sensitivities o f 6 different T5 neurons R (thin lines) and R + (thick lines) - in response to a 2 ~ x 2 ~ black square traversing the receptive field centers at 7,6~ on white background in different directions

1. Directional Sensitivities The neuronal activity was measured in response to a configurational "indifferent" 2 ~ x 2 ~ square~ M a n y of the T5 neurons (n = 14) showed no obvious response preference for any of the movement directions tested. They are functional radially symmetrical in the present context. Often (n = 13) the neuronal activity was particularly high in response to stimulus traverse in a certain plane of the x - y coordinates in both directions (Fig. 2A and D). We might call these neurons plane sensitive. Other neurons ( n = 10) showed a preference for stimulus movements mainly in one direction of the preferred plane. These neurons are directional sensitive (Fig. 2C, E, F). The preferred direction was different a m o n g various neurons of this type. Although the E R F s of neurons recorded in this study were highly variable and partly similarly localized, the data

194

J.-P. Ewert et al. : Tectal T5 Neurons in the Toad response to worm

response to antiworm

A

/ i

R=lO R+=IO00

i

W27"6-78,1

B

/ i

,i

W22.6 '78"t

C

815.8"78 '3

D

E24'8"78.2

gave n o i n f o r m a t i o n o n a possible r e l a t i o n s h i p between E R F locations in the visual field a n d their directional sensitivity properties. Some n e u r o n s c o u l d n o t be clearly " d e t e r m i n e d " as being either plane or directional sensitive (Fig. 2 B). I n t e r m e d i a t e types also a p p e a r e d to exist b e t w e e n radially a n d plane sensitive n e u r o n s . Very few n e u r o n s ( n = 4 ) showed total a s y m m e t r y in this respect.

Fig. 3A-D. Linkage between stimulus configuration and movement direction. The same black 2~ x 8~ or 2~x 20~ stripe traversed the receptive field center at 7,6~ either in a worm-like(left part) or antiworm-like manner (right part). The output of T5 neurons was measured as R (thin lines) or R + (thick lines)

2. Linkage Between Stimulus Configuration and Movement Direction A stripe traversed the E R F center either in a w o r m like or a n t i w o r m - l i k e m a n n e r . Fig. 3 A - D shows four representative examples of different response types. All of these n e u r o n s r e s p o n d e d selectively to b o t h stimulus c o n f i g u r a t i o n s . However, the degree of selec-

195

J.-P. Ewert et al. : Tectal T5 N e u r o n s in the Toad

W22"6'78"1

_._.8.78-3

L~o'o'~o'~

24-8.78.1

78-2

D

'8 "78.5

Fig. 4 A - D . Discrimination values Dw, A (thin lines) and D~, A (thick lines) for worm-like and antiworm-like stimuli traversing the receptive field center in different directions. The figure shows 4 response types (A-D) of T5 neurons. If values are outside the circle, worm configuration activated the neuron stronger than the antiworm. Values on the circle indicate no discrimination. If values are inside the circle (rastered area) the neuron was more strongly activated by the antiworm configuration. For further explanation see text

tivity appeared to vary a m o n g the neurons. Figure 3A: When tested with the w o r m stimulus, the neuron responded well to m o v e m e n t of stripe in all directions; but it was silent if the same stripe was moved as antiworm. Figure3B: The neuron responded well to the w o r m stimulus as described

above; however, it showed weak activity to the antiw o r m moved in horizontal and vertical directions. Interestingly it was silent if this stimulus traversed the E R F center in oblique directions. Figure 3C: The neuron was generally activated both by the w o r m and the antiworm configuration. However, the re-

196

J.-P. Ewert et al. : Tectal T5 Neurons in the Toad

sponse was optimal for the worm moved in horizontal directions and for the antiworm moved in vertical directions. In other words: These neurons were best activated if the axis of the stripe had a horizontal orientation - rather independently of whether it moved horizontally or vertically. Figure 3D." This neuron had properties similar to that described in C but the directional preference for the antiworm was more selective. When the values for the " f o r m contrast" Dw, A and DW,A + are considered in this investigation, four main response types can be distinguished among the 42 neurons tested. Three examples of each type are shown in Fig. 4 A - D : e) Neurons ( n = l l ) showed a relatively high degree of selective response to both configurational stimuli at all movement directions tested (Fig. 4A). The worm was preferred over the antiworm by

the worm was preferred varied among the neurons of this type. No obvious changes in response types c~-3 were found during stimulation with the 2 ~ x 8 ~ (c.f. initials B and E) or the 2 ~ x 20 ~ stripe (c.f. IV). The following Table 1 shows average discrimination + values DW,A (Dw,a) for worms and antiworms moving horizontally and vertically. It is evident that neurons of type c~ exhibit a strong worm preference for all 4 movement directions. Similar relations were found in type fl; however, the configurational selectivity is not so strong. Type 7 shows worm preference mainly for movements in horizontal directions. In type 3 the configurational discrimination is weak for all directions or the antiworm is mainly preferred.

3. Longterm Recordings from the Same Neuron

Dw, A = + 1.

fl) Neurons (n = 13) preferred the worm configuration at all movement directions except for one sector (Fig. 4B). In this particular range there was either no discrimination (Dw,Am0) or even the antiworm was preferred (Dw,A< 0). This kind of " d i s t o r t i o n " could occur in different directions among neurons of this type. ~) Neurons (n = 11) preferred the worm configuration only if the stimulus was moved in a certain plane of the x - y coordinates (Fig. 4 C). For movements perpendicular to this plane the two configurational stimuli were not discriminated, or the antiworm was preferred. The orientation of the plane preferable for worm-like stimuli could be different in various neurons of this type. Some of these neurons are "orientation sensitive ". 3) Few neurons (n=7) showed a preference for the worm configuration in only one movement direction. For the other directions either both configurational patterns were equally effective o r the antiworm was the stronger stimulus. The direction at which

Regarding the variety of neuronal response types a-b, it is reasonable to ask whether the properties of such a neuron are constant over time. Figure 5A and B shows two examples in which neurons have been recorded for over 4 h. The same stimulation program was repeated 4 (Fig. 5B) or 5 (Fig. 5A) times. The time separation between the trials was constant; each trial lasted about 50 min. The neuron of Fig. 5A showed in the first two trials a response characteristic of type fl, during the third it changed to 7, in the fourth again to fl, and in the fifth it corresponded almost - but not quite - to c~ (if R + is considered as output). Figure 5B shows another example in which the neuron responded according to 7 in the first three trials and changed to 6 in the fourth. It is important to note that during the entire test series in both neurons the response activity to the worm stimulus was relatively constant for all movement directions tested (similar to that shown in Fig. 3A, left). Response variability was restricted mainly to the antiworm configuration.

Table 1

"Response type" of T5 neurons (n =42)

Stimulus movement in horizontal directions

Stimulus movement in vertical directions

0o

90 ~

DW,A

a fi 7 3

(n= (n = (n= (n=

11) 13) 11) 7)

0.71 0.32 0.44 --0.24

180 ~ (Dw,A) +

DW,A

(0.91) (0.49) (0.58) (--0.31)

0.62 0.25 0.34 0.08

(D~f,A)

(0.84) (0.62) (0.57) (0.12)

Dw,A 0.59 0.20 0.03 --0.18

270 ~ (Dw,A) +

Dw ,a

+ (Dw.A)

(0.93) (0.56) (-0.01) (--0.50)

0.58 0.13 -0.13 --0.38

(0.87) (0.27) (--0.30) (--0.61)

197

J.-P. Ewert et al. : Tectal T5 Neurons in the Toad

A

I

If

W7~7,78.1-1

7,7,78-1.2

E24.8,78,2,4

Fig. 5A and B. Variability of the discrimination values Dw,A (thin lines) and D~,A (thick lines) during long-term recording from the same neuron. Two examples, A (1 to 5) and B (1 to 4). Explanation see text

Discussion Previous experiments in c o m m o n toads Bufo bufo (L.) have shown that tectal T5 neurons respond selectively or with great sensitivity to various behaviorally relevant configurational stimuli - such as worms and

antiworms - traversing horizontally the centers of their E R F s (Ewert, 1974). The present results obtained for the response of T5 neurons to different movement directions in the x - y coordinates give further insights into this kind of configurational sensitivity and discrimination.

198

J.-P. Ewert et al. : Tectal T5 Neurons in the Toad

Directional Symmetries and Asymmetries

Response Types and Classes

There are T5 neurons showing a relatively high degree of selective responsiveness to both kinds of configurational stimuli tested. This relationship is invariant for the direction of movement (Fig. 4A). The underlying principle might be based on symmetrical functional coupling with preconnected nerve nets (Ewert and von Seelen, 1974). Neurons of this response type correspond to the previously described class T5(2) (Ewert and von Wietersheim, 1974a; von Wietersheim and Ewert, 1978). Neurons showing such selective responsiveness appear to be relatively constant according to this characteristic. There are also T5 neurons showing various degrees of configurational sensitivities at different movement directions (Fig. 4B-D; c.f. also 3C and D). At least some of these neurons (types fi and 7) correspond to class T5(1) (yon Wietersheim and Ewert, 1978). The directional variability of configurational sensitivity in different T5(1) neurons is probably based on asymmetric functional coupling which can be reached generally in different ways (Takashi and Fujiwara, 1979; for review see Grfisser and Grfisser-Cornehls, 1973). For example, it seems quite likely that units of type 7, which have opposite planes of maximal response for worm and antiworm stimuli, simply have elongated ERFs, such that the worm is best when travelling along the longer axis. The results from n = 11 type 7 neurons indicate that the greatest discrimination between worm and antiworm configurations is seen when the stimuli move horizontally (c.f. Table 1). However, the ERFs of these neurons have mainly vertically (!) elongated shapes. Thus, the configurational response property of these neurons is not simply linked to the shapes of the ERFs (c.f. Ewert and v. Seelen, 1974). Directional sensitivity for moving configurational stimuli (worms vs. antiworms) was also observed in retinal R3 ganglion cells (Ewert e t a l , 1979a). However, these neurons preferred the antiworm over the worm figure at all movement directions, provided the length of the stimuli did not exceed the size of the ERF diameter. This configurational preference was strongest for vertical movements. The response characteristic of these neurons was relatively constant over time. - More specifically, 29% of the R3 neurons showed no preference and 71% exhibited sensitivity for vertical movements. Among those 60% showed some preference for me~270 ~ 20% for md~90 ~ and 20% had no preference in the vertical plane. 1

We might ask whether response types c~-6 (Fig. 4AD) belong to different neuronal classes or whether they are (at least types t - f ) representing different "states" of neurons belonging to the same class. Indeed long-term recordings from the same neuron indicate that fi may change to 7 (Fig. 5A), and that 7 can change to 6 (Fig. 5B). But type ~ differs from the others by it's strong selectivity and apparent resistance to change.

t Recent studies suggest a subtype among R3 neurons which is activated more strongly by the w o r m rather than by the antiworm (Ewert, unpubl, data)

Response Variability At least in some of the T5(1) neurons the directional dependence of configurational sensitivity appear to change with time (Fig. 5 A and B). There are various possibilities to explain these results: (1) Statistical variance due to random processes, such as fluctuations in membrane potential and number of quantal units of transmitter released. This in connection with network mechanisms might cause differences in nerve pulse frequency and timing. (2) The activities of two (or more) different neurons (classes), both being close to the tip of the electrode, might have been recorded, and their relative distances to the electrode varied over time. (3) The oxygen supply and thus the metabolic state of the nervous tissue underwent some alteration during the longterm experiment. (4) Certain neurons have their response characteristic altered due to variable selectivities of system components in the neural network. During our long-term recordings it was mainly the response to the antiworm-configuration that underwent changes, the response to the worm configuration remaining almost constant. This might exclude explanations (1) and (3). We recorded only those neurons showing a relatively high signal/noise ratio, and during the entire experiment the spikes had same amplitude and time course, which might exclude possibility (2). Therefore (4) may be the best explanation. Response variability and fluctuations in receptive field properties have been reported for neurons of the optic tectum in frogs (Grfisser and Grfisser-Cornehls, 1976), the thalamic-pretectal region in toads (Ewert, 1971), the pretectum in cats (Sprague et al., 1968), the visual cortex of cats (Bear et al., 1971 ; Horn and Hill, 1969; Noda and Adey, 1970; Spinelli and Barrett, 1969) and the visual cortex of macaque monkeys (Poggio, 1972). In a study by Tomko and Crapper (1974) it was shown that some of the neurons from the striate cortex of unanesthetized monkey exhibited variability in response to identical photic stimuli (stationary achromatic spatial grating pattern). However,

J.-P. Ewert et al. : Tectal T5 Neurons in the Toad response variability was not found in all neurons. Tomko and Crapper (1974) assume that the " n o n changing" neurons are complex cells and that the latter may have the potential to respond optimally to various visual features of the environment. Also neurons in the auditory cortex were found to be highly labile (for summary see Manley and Mfiller-Preuss, 1978). Whereas previous studies have used primarily pure tones and artificial stimuli, Manley and Mtiller-Preuss (1978) investigated responses of single cortical cells in monkeys to biologically relevant sounds. These authors found spontaneous changes in response selectivity to specific vocalizations, in primary as well as secondary cortical cells. This neuronal response variability can be explained by changes of the behavioral state (Beaton and Miller, 1975) or by changes in the arousal level (Newman and Symmes, 1974); but other causes may be also responsible.

System Properties There is evidence that the conflgurational sensitivity or selectivity of T5 neurons may be determined by inputs from thalamic-pretectal (TP)2 neurons, which themselves are more strongly activated by the antiworm rather than the worm configuration. Following lesions to the TP region configurational selectivity is remarkably decreased, as seen in the response of T5 neurons (Ewert and von Wietersheim, 1974b) as well as in the prey-catching behavior (Ewert, 1968)3. There is some evidence for a cholinergic inhibitory system (for results in frogs c.f. Stevens, 1973): After local application of curare to the optic tectum the activity of T5 neurons and the prey-catching behavior are "disinhibited" in response to moving visual objects - quite similarly - as after TP lesions (for results in toads see Ewert et al., 1974). In future experiments we have to investigate, whether the directional sensitivity property of T5 neurons remains after TP lesions. The present results support the idea (Ewert, 1979) that not only T5(2) but also T5(1) neurons may be influenced by inhibitory inputs from the TP region. We believe that these inputs are stronger for 2 TP includes the caudal dorsal thalamus and the pretectal region 3 In mammals the directional selectivity of Certain neurons of the superior colliculusis determined by cortical inputs. Colliculus neurons loose this particular property after ablation (Sterling and Wickelgreen, 1970) or deprivation of the visual cortex (K.-P. Hofmann, pers. comm. Mainz, 1975). This is another example, which shows that certain response properties of neurons are determined by inputs of other neurons located in distinct brain regions. New results suggestthat evenresponsecharacteristicsof cellsin the visual cortex of mammals are partly determined by thalamic inputs (Creutzfeldt, DZG Regensburg 1979)

199 T5(2) than for T5(1) neurons and that they can be modulated, a phenomenon occurring predominantly in some of the latter. What might cause such modulations? (1) The TP region itself could be a source of modulatory influence. Response variability of caudal thalamic neurons was described in Bufo americanus by Ewert (1971). Furthermore it was found that thalamic neurons have changing levels of activity. However, nothing is known at present whether their (spontaneous) effects on tectum change periodically, or over time, or if their elicited activity changes. (2) Variability in pattern discrimination can also be due to stimulus specific habituation. The TP region appears to be a source involved in these processes: Following TP lesions stimulus specific habituation in prey-catching (Ewert, 1967) as well as in T5 neurons is decreased (Ingle, 1973). (3) Thalamic inputs to the optic tectum could also be modulated by telencephalic structures. From training experiments it is known that olfactory cues associated during feeding can modify configurational prey selection (Ewert, 1968). - Recordings from tectal and thalamic-pretectal neurons to configurational visual stimuli in presence and absence of known and unknown olfactory prey stimuli will shed light in the intrinsic central mechanisms. These experiments are in progress in our laboratory. Neuroanatomical evidence of neurons projecting from pretectum to the optic tectum in Rana pipiens was given by Trachtenberg and Ingle (1973) using Fink-Heimer degeneration techniques. Recently these results have been confirmed by Wilczynski and Northcutt (1977) using H R P as marker for retrograde tracing of nerve fibers. Most recent anatomical investigations gave a detailed picture of pathways by which telencephalic structures might influence the optic tecturn directly, or indirectly via thalamic neurons (Northcutt and Wilczynski, personal communication to J.-P.E., St. Louis, 1978).

Conclusions Measurements of receptive field sizes of neurons in the central visual system may depend largely upon the stimuli used for this purpose (c.f. comments in Ewert et al., 1979 a). The best stimuli were supposed to be thin vertical bars - provided, however, that the neurons respond to them. F r o m the present study we learn that this must not necessarily be the case As far as we have been able to determine, the selective response of tectal T5(2) neurons (type c0 to behaviorally important configurational stimuli (worm vs. antiworm) is invariant with regard to movement direction in the x - y coordinates. The sensitivity of T5(1) neurons to configurational stimuli is in part

200

variant with the movement direction. We cannot exclude the possibility that T5 neurons at least some of T5(1)) may change their response properties. The conditions for those changes observed here are not known at present, but we have planned to analyze them in future experiments. We assume that part of the configurational prey recognition system in the common toad is based on "hardwire" forming the fundamentals of the innate releasing mechanisms (Lorenz, 1954; Ewert, 1974). Other part may be based on "softwire" opening the possibility of modification (Schleidt, 1962) and to classify stimulus distributions from the environment into learned (acquired) classes of functional significance (Ewert, 1968; Brzoska and Schneider, 1978). Whether those systems exist as "parallel structures" or whether there is one basic structure which can be modified according to individual experience are important questions which will be investigated in our current research. We learn from the present investigation that some general visual properties, such as movement specificity, directional sensitivity, orientation sensitivity, and configurational selectivity are shared by cells in the cat visual cortex (Hubel and Wiesel, 1962, 1965) and T5 neurons in the optic tectum of the common toad. Some of the experimental measurements were assisted by Dr. B. Payne (Department of Zoology, University of Durham, U.K.) during a visit which was supported by the Deutsche Forschungsgemeinschaft. We wish to thank Dr. Ananda Weerasuriya (Department of Health, National Institute of Health, Bethesda, Maryland 20014, USA) for his comments on the manuscript.

References Bear, P.M., Sasaki, H., Ervin, F.C. : Sequential changes in receptive fields of striate neurons in dark adapted cats. Exp. Brain Res. 13, 256-272 (1971) Beaton, R., Miller, J.M. : Single cell activity in the auditory cortex of the unanesthetized behaving monkey: correlation with stimulus controlled behavior. Brain Res. 100, 543-562 (1975) Beck, A., Ewert, J.-P.: Prey selection by toads (Bufo bufo L.) in response to configurational stimuli moved in the visuai field z-y coordinates. J. Comp. Physiol. 129, 207-209 (1979) Brzoska, J., Schneider, H. : Modification of prey-catching behavior by learning in the common toad (Bufo bufo L., Anura Amphibia): Changes in responses to visual objects and effects of auditory stimuli. Behav. Processes 3, 125-136 (1978) Ewert, J.-P. : Untersuchungen fiber die Anteile zentralnerv6ser Aktionen an der taxisspezifischen Ermtidung beim Beutefang der Erdkr6te (Bufo bufo L.). Z. Vergl. Physiol. 57, 263 298 (1967) Ewert, J.-P. : Der EinfluB yon Zwischenhirndefekten auf die Visuomotorik im Beute- und Fluchtverhalten der Erdkr6te (Bufo bufo L.). Z. Vergl. Physiol. 61, 41-70 (1968) Ewert, J.-P. : Single unit response of the toad's (Bufo americanus) caudal thalamus to visual objects. Z. Vergl. Physiol. 74, 81-102 (1971) Ewert, J.-P.: The neural basis of visually guided behavior. Sci. Am. 230, 34 42 (1974)

J.-P. Ewert et al. : Tectal T5 Neurons in the Toad Ewert, J.-P. : The visual system of the toad : behavioral and physiological studies on a pattern recognition system. In: The amphibian visual system: a multidisciplinary approach. Fite, K.V. (ed.). New York, San Francisco, London: Academic Press 1976 Ewert, J.-P. i Neuroethology. Introduction to the neurophysiological fundamentals of behavior (English edition). New York: Springer 1979 Ewert, J.-P., Hock, F.J. : Movement-sensitive neurones in the toad's retina. Exp. Brain Res. 16, 41 59 (1972) Ewert, J.-P., Seelen, W. yon: Neurobiologie und System-Theorie eines visuellen Mustererkennungsmechanismus bei Kr6ten. Kybernetik 14, 167 183 (1974) Ewert, J.-P., Wietersheim, A. von: Musterauswertung durch Tecrum- und Thalamus/Praetectum-Neurone im visuellen System der Kr6te(Bufo bufo L.). J. Comp. Physiol. 92, 131-148 (1974a) Ewert, J.-P., Wietersheim, A. von: Einfhil3 yon Thalamus/ Praetectum-Defekten auf die Antwort yon Tectum-Neuronen gegent~ber bewegten visuellenMustern bei der Kr6te Bufo bufo (L.). J. Comp. Physiol. 92, 149-160 (1974b) Ewert, J.-P., Hock, F.J., Wietersheim, A. yon: Thalamus/Praetectum/Tectum: Retinale Topographie und physiologische Interaktionen bei der Kr6te Bufo bufo (L.). J. Comp. Physiol. 92, 343-356 (1974) Ewert, J.-P., Botchers, H.-W., Wietersheim, A. von: Question of prey feature detectors in the toad's Bufo bufo (L.) visual system : a correlation analysis. J. Comp. Physiol. 126, 4 3 4 7 (1978) Ewert, J.-P., Krug, H., Sch6nitz, G. : Activity of retinal R3 ganglion cells in the toad Bufo bufo (L.) in response to moving configurational stimuli: Influence of the movement direction. J. Comp. Physiol. 129, 211-215 (1979a) Ewert, J.-P., Arend, B., Becker, V., Borchers, H.-W.: Invariants in configurational prey selection by Bufo bufo (L.). Brain, Behay. Evol. 16, 38-51 (1979b) Fite, K.V., Scalia, F.: Central visual pathways in the frog. In: The amphibian visual system: a multidisciplinary approach. Fite, K.V. (ed.). New York, San Francisco, London: Academic Press 1976 Grfisser, O.-J., Grfisser-Cornehls, U. : Die Informationsverarbeitung imvisuellen SystemdesFrosches. Kybernetik 1968, 331-360(1968) Grfisser, O.-J., Grfisser-Cornehls, U.: Neuronal mechanisms of visual movement perception and some psychophysical and behavioral correlations. In: Handbook of sensory physiology, Vol. VII/3A. Jung, R. (ed.), pp. 333429. Berlin, Heidelberg, New York: Springer 1973 Grtisser, O.-J., Gr/isser-Coruehls, U. : Neurophysiology of the anuran visual system. In: Frog neurobiology. Llin~ts, R., Precht, W. (eds.), pp. 297-385. Berlin, Heidelberg, NewYork : Springer 1976 Horn, G., Hill, M.: Modifications of receptive fields of cells in the visual cortex occurring spontaneously and associated with bodyly tilt. Nature London 221, 186-188 (1969) Hubel, D.H., Wiesel, T.N. : Receptive fields, binocular interaction and functional architecture in the cat's visual cortex. J. Physiol. London 160, 106-154 (1962) Hubel, D.H., Wiesel, T.N.: Receptive fields and functional architecture in two nonstriate visual areas (18 and 19) of the cat. J. Neurophysiol. 28, 229-289 (1965) Ingle, D. : Disinhibition of tectal neurons by pretectal lesions in the frog. Science 180, 42~424 (1973) Ingle, D., Sprague, J. : Sensorimotor function of the midbrain tectum. Neurosci. Res. Program Bull. 13, 169-288 (1975) Lorenz, K.: Das angeborene Erkennen. Natur und Volk 84, 285-295 (1954) Manley, J.A., Mfiller-Preul3, P. : Response variability of auditory cortex cells in the Squirrel Monkey to constant acoustic stimuli. Exp. Brain Res. 32, 171-180 (1978) Newman, J.D., Symmes, D. : Arousal effects on unit responsiveness to vocalizations in squirrel monkey auditory cortex. Brain Res. 78, 125 138 (1974)

J.-P. Ewert et al. : Tectal T5 Neurons in the Toad Noda, H., Adey, W.R. : Changes in the neuronal activity in association cortex of the cat in relation to sleep and wakefulness. Brain Res. 19, 263-275 (1970) Poggio, G.F.: Spatial properties of neurons in striate cortex of unanesthetized macaque monkey. Invest. Ophthalmol. Visual Sci. 11, 368 377 (1972) Schleidt, W.: Die historische Entwicklung der Begriffe "Angeborenes ausl6sendes Schema" und ~ Ausl6semechanismus" in der Ethologie. Z. Tierpsychol. 19, 697-722 (1962) Spinelli, D.N., Barrett, T.W. : Visual receptive field organization of single units in the cat's cortex. Exp. Neurol. 24, 76-98 (1969) Sprague, J.M., Marchiafava, P.L., Rizzolatti, G.: Unit response to visual stimuli in the superior colliculus of the unanesthetized mid-pontine cat. Arch. Ital. Biol. 106, 169-193 (1968) Sterling, P., Wickelgreen, B.G.: Function of the projection from the visual cortex to the superior colliculus. Brain, Behav. Evol. 3, 210-218 (1970)

201 Stevens, R.J. : A cholinergic inhibitory system in the frog optic tectum: its role in visual electrical responses and feeding behavior. Brain Res. 49, 309-321 (1973) Takashi, N., Fujiwara, M.: A neural model for the development of direction selectivity in the visual cortex. Biol. Cybernetics 32, 1 8 (1979) Tomko, G.J., Crapper, D.R. : Neuronal variability: non-stationary responses to identical visual stimuli. Brain Res. 79, 405-418 (1974) Trachtenberg, M.C., Ingle, D. : Thalamus-tectal projections in the frog. Brain Res. 79, 419-430 (1974) Wietersheim, A.v., Ewert, J.-P. : Neurons of the toad's (Bufo bufo L.) visual system sensitive to moving configurational stimuli: a statistical analysis. J. Comp. Physiol. 126, 35-42 (1978) Wilczynski, W., Northcutt, R.G.: Afferents to the optic tectum of the leopard frog: an HRP study. J. Comp. Neurol. 173, 219-229 (1977)