Simple Brain Simulation Resources 
Home > Visual System > Amphibian Retina

DanceFloor by dynamic artist Jenny James. Copyright 2006 (used with permission)

Frog and Toad Eyes

by David D. Olmsted (Copyright - 2000, 2006. Free to use for personal and educational purposes)
Last Revised September 23, 2006

How the Frog Uses its Eye

How the frog uses it eye is best described by Lettvin, Maturana, McCulloch, and Pitts (1959):

“A Frog hunts on land by vision. He escapes enemies mainly by seeing them. His eyes do not move, as do ours, to follow prey, attend suspicious event, or search for things of interest. If his body changes its position with respect to gravity or the whole visual world is rotated about him, then he shows compensatory eye movements. These movements enter his hunting and evading habits only, e.g. as he sits on a rocking lily pad. Thus, his eyes are actively stabilized. He has no fovea, or region of greatest acuity in vision, upon which he must center a part of the image....
The frog does not seem to see or, at any rate, is not concerned with the detail of stationary parts of the world around him. He will starve to death surrounded by food if it is not moving. His choice of food is determined only by size and movement. He will leap to capture any object the size of an insect or worm, providing it moves like one. He can be fooled easily not only by a piece of dangled meat but by any moving small object. His sex life is conducted by sound and touch. His choice of paths in escaping enemies does not seem to be governed by anything more devious than leaping to where it is darker. Since he is equally at home in water and on land, why should it matter where he lights after jumping or what particular direction he takes? He does remember a moving thing provided it stays within his field of vision and he is not distracted.”

Frog Eye Characteristics

The large round lens of the frog shown as a filled black oval in figure 1 gives the animal a large field of view. The frog is naturally nearsighted (myopic) to -6 diopters giving it a focus of approximately 6 inches. Frogs and toads can change their focus by moving the lens out towards the cornea. In frogs the focus range is a few diopters and in toads the focus range is 5 diopters giving a best myopia of -1 diopters. During this accommodation the pupil also increases in size. (Grusser and Grusser-Cornehl - 1976). The advantage of nearsightedness is that it blurs the background clutter making foreground object characterization much easier. Only 75% of the light intensity entering the eye reaches the retina as shown in figure 2.

Figure 1
Top View of the Frog Eye (originally from Szent-Gyorgyi - 1914 but scanned from Grusser and Grusser-Cornehls - 1976)

Figure 2
Spectral Transmission Curve of the Eye of the Frog Rana esculenta (originally in Grusser-Cornehls and Saunders - 1975 but scanned from Grusser and Grusser-Cornehls - 1976)

The minimum separable vertical stripe period in an optokinetic test is 6 to 7 minutes (Birukow - 1937 as reported by Grusser and Grusser-Cornehls - 1976). In this kind of test a drum having vertical stripes in rotated around a frog who will normally track the stripes for a certain distance before resetting its gaze back forward. Yet tectal neurons can respond to moving dots of light as small as 0.1 to 0.2 minutes (Grusser and Grusser-Cornehls - 1976). This is in spite of the fact that the frog Rana esculanta has an average of 35 visual receptors per degree which is equivalent to approximately 1 every 2 minutes (Alexander-Schafer - 1907 as reported by Grusser and Grusser-Cornehls - 1976) indicating that neural circuitry is able to refine the native resolution of the retina.

A quiescent frog will close its eyelids synchronously every 0.5 to 5 minutes. They will also be closed by:

  1. eye irritation
  2. whole body vibration
  3. sudden bright light
  4. large approaching dark objects
  5. prey swallowing

Frog eyes also exhibit a rhythmic motion possibly due to heartbeat or muscle drift. This motion has period of 1.3 seconds and an amplitude of 28 minutes. Every 15 to 20 seconds the lungs empty causing a larger 1 degree movement of the eyes. (Grusser and Grusser-Cornehls - 1976). One would expect the frog's brain to have neural circuitry that allows it to ignore these types of false motions.

Retinal Cell Anatomy

The optic nerve of the frog consists of approximately 470,000 unmyelinated fibers and 15,000 myelinated fibers. In the toad it consists of 320,000 unmyelinated fibers and 10,000 myelinated fibers. The neurons of the retina which produces the output fibers are the ganglion cells. The frog Rana pipiens has approximately 440,000 small ganglion cells (7 to 10 microns in diameter) and 12,000 of the larger ganglion cells 14 to 20 microns in diameter. (Maturana - 1959). With the frog brain consisting of only 16 million neurons one can see that the retina makes up a significant portion of its brain's total neuron number.

The frog retina has three cell layers as shown in figure 3: the outer, middle, and inner granular layers, and two fiber layers: the inner and outer plexiform layers. The frog has a limited color discrimination as indicated by the red and green rods. The red rods are the most numerous, closely followed by red sensitive cones. The green rods are relatively scarce. Only the ganglion cells from the inner granule layer send axons to the brain. The bipolar cells (dendrites above and below the cell body) in the middle granular layer are intermediate processing elements. The bipolars connect mostly vertically while another class of neurons called horizontals and amicrines (not shown) connect mostly horizontally. The ganglion cells themselves come in several varieties. Their dendritic arbors can be thick or thin and be widespread horizontally or quite restricted. Some of the thick arbors tend to separate into two vertical layers

Figure 3
Basic Retinal Cell Shapes (Lettvin, Maturana, Pitts, and McCulloch - 1961)

Response Classes of the Ganglion Cells

Class 1 and class 2 ganglion cells project to the brain via unmyelinated fibers having conduction velocities of 20 to 50 centimeters per second. They make up 97 percent of the optic nerve (Maturana - 1959). Typical responses are shown in figure 4.

  1. Class 1 neurons (edge detectors): These neurons have oval receptive fields from 1.5 to 4 degrees in size. These neurons detect the completeness and sharpness of an edge including that from a spot which can be either brighter or darker than the background. If a boundary is brought into its receptive field in total darkness and the light switched on a continuing response occurs after an initial delay. The response is enhanced by edge movement. If an edge stops in its visual field the neuron remains activated for several minutes. In this case if the light is turned off the response ceases only to reappear less intensely when the light is turned on again.
  2. Class 2 neurons (convexity detectors): These neurons have oval receptive fields ranging in size from 2.5 to 5 degrees. Their purpose seems to be the detection of the dark leading edge (head) of any worm or bug. If the stimulus stops before the center of the receptive field the neuron exhibits a slow decaying response (time constant between 0.3 to 3 seconds) . Unlike the class 1 neuron a brief darkness or shadow will permanently stop the response until the object moves again. These type of neurons have a strong inhibitory surround region indicating that they are prevented from working in environments with substantial background clutter.
  3. Class 3 neurons (change detectors): These neurons are also known as on-off cells and they have receptive field sizes between 6 and 10 degrees. The intensity of their response is governed by the amount of light intensity change. This includes such parameters as as stimulus size but not shape.
  4. Class 4 neurons (dimming detectors): These neurons have oval receptive fields ranging from 10 to 15 degrees. They respond to a dimming of the light.
  5. Class 5 neurons (tonic darkness sense): These neurons are very few but they seem to indicate the general light level of the frog’s surroundings. When the light level is reduces these neurons slowly increase their firing rate.

The class 3 and class 4 ganglion cells project to the brain on myelinated fibers which make up only 3% of the optic nerve. Class 3 fibers have a conduction velocity of between 1 and 5 meters per second, class 4 have a conduction velocity of 10 meters per second, and a small set of fibers having a conduction velocity of 20 meters per second is thought to arise in the brain and project to the retinal neurons.

For all classes of ganglion neurons the more rapid is the angular velocity of a stimulus the greater is the frequency of the action potentials but with the consequence of a shorter pulse length. Consequently, the total number of impulses probably remains the same (figure 5).

Figure 4
Responses of the Four Retinal Ganglion Classes. Class 1 - edge detector, Class two - convexity detector, Class 3 - Change (on-off) detector, Class 4 - dimming detector. (from Grusser and Grusser-Cornehls - 1968a but scanned from Grusser and Grusser-Cornehls - 1976)

Figure 5
Signal Intensity vs. Stimulus Angular Velocity. Black spot moving on a light background. Spot 1.2 degrees for class 1 and 2. Spot 2.7 degrees for class 3. Spot 23 degrees for class 4. (from Grusser and Grusser-Cornehls - 1969 but scanned from Grusser and Grusser-Cornehls - 1976)

The left graph in figure 6 shows how the latency of signal initiation decreases with increasing light intensity. Top line (open circles) represent the “on” burst while the bottom line (dark circles) represent the “off” burst. The right graph gives the frequency in impulses per second of the first 6 impulses for both the “on” and “off” bursts. The total number of impulses in the signal is also proportional to the light intensity

Figure 6
Responses to Light Intensity of the Class 3 (Change Detector) Neuron. (Hartline - 1938)

The more central is the stimulus in a receptive field the more intense is the neural signal. This is shown in Figure 7 which gives the average impulse rate when a black spot is moved at a constant velocity in a horizontal direction. Yet the size the receptive field does not seem to correlate to the breadth of the respective ganglion cell dendritic arbors. This is shown in figure 8 which shows the top view of some class 4 ganglion cells called A, B, and C. Their receptive fields are shown in the illustration in the upper left. Notice how small is the receptive field for C and how the receptive field for B has some inhibitory flanks as exposed by the small spots of lights used as probes.

Figure 7
Class 2 Neuron has Greatest Response from the Center of its Receptive Field. Each division represents 0.5 degrees. (from Grusser and Grusser-Cornehls - 1968a but scanned from Grusser and Grusser-Cornehls - 1976)

Figure 8
Size of the Ganglion Cell’s Dendritic Arbor does not Correlate with Receptive Field Size for the Class 4 Cells (Dimming Detectors). (Stirling and Merrill - 1987)

The rare class 5 neurons do not seem to be the only tonic darkness detectors for such neurons have also been found in the frontal organ of the frog. The frontal organ is analogous to the reptilian pineal eye and is located on the dorsal surface of the head where light diffuses through the skin. The response of the frontal organ nerve is shown in figure 9.  The 100% firing rate represents the mean of 76.5 spikes per second obtained after 15 minutes of dark adaptation. As light levels increase the firing rate declines. The data was taken from the frog Rana pipiens collected during August and September.

Figure 9
Response of the Frontal Organ Nerve to Steady White Light. (Hamasaki and Esserman - 1976)

Significantly, the class 3 and 4 neurons are inhibited by eye blinking as shown in figure 10. This is probably accomplished by the fibers projecting to the retina. In figure 10: A = normal class 4 response, B = normal class 3 response, C1 = class 4 response to eye blink, C2 = class 3 response to eye blink, D = class 4 response to a slow increase in light intensity. Using a photocell the eyelid was found to transmit 50% to 60% of the incident light.

Figure 10
Retinal Class 3 (Change Detectors) and Class 4 (Dimming Detectors) are Inhibited by Blinking in the toad Bufo bufo. (Ewert and Borchers - 1974)

Responses of the Toad in Low Light Levels

Compared to frogs, toads have a more complex and versatile visual system due to their more general eating behavior. Toads not only eat moving insects like frogs but will also hunt for white grubs crawling on the ground. Insects are normally dark when highlighted against the sky while grubs are normally light when highlighted against the dark ground. Consequently, the convexity detector (class 2) neurons of toads respond to small light objects in addition to small dark objects. As shown in figure 11 toads can respond to light objects even when in a dark environment (-2 is just less than full moonlight). This is equivelent to responding  at a low 0.01 Rh per second where Rh is a measure of the number of photo-isomerizations per rod (molecular transformation in response to a photon of light). In figure 11 the proportion of test sessions in which snaps occurred is given by the vertical ordinate while the light illumination level is given by the horizontal abscissa in log units. 

Figure 11
Snapping in the Toad Bufo bufo Requires That Light Levels Exceed a Threshold. (Aho, et al - 1993)

The class 2 retinal neurons sum light energy over time. The rods and cones wait until they have collected enough light energy before producing a signal. The greater is the light energy the shorter is this latency as shown in figure 12. In figure 12 the response latencies are given in seconds vs light stimulus intensity. The light intensity of 100 is slightly less than the light level during full moonlight which according to figure 11 is the snapping threshold.

Figure 12
Class 2 Ganglion Cell Response Latencies Decrease with Increasing Light Levels in the Toad Bufo bufo. (Aho, et al - 1993)

When a toad is very hungry so that it is motivated to snap in low light levels the tongue strike location onto a moving worm like target moves progressively backwards as shown in figure 13. The light rectangle is moving towards the right . Light reduction in the figure progresses from the upper left (200), lower left(2), upper right (0.1), to lower right (0.2). Despite the increase latency from 1.5 seconds to 3 seconds from light levels 0.1 to 0.02 the targeting does not get significantly worse suggesting some sort of compensation is made for the darkness.

Figure 13
The Response Latencies Cause Progressively Less Accurate Tongue Targeting in the toad Bufo bufo. (Aho, et al - 1993)

Variations in Toad Retinal Responses

A variation in the dark object or light object toad acquisition response is found if a less then ideal prey stimulus is used such as a square. Ewert and Siefert say this (1974):

“To be sure, square patterns were not so effective for prey-capturing behavior as horizontal stripes (“worm” like patterns). Nevertheless, the squares proved to be especially advantageous in these experiments: In the case of an optimal target shape, the toad seldom betrays through behavioral changes a recognition of limited variations in stimulus pattern contrast. Horizontal stripes, whose shape fit the category of “worm” prey, attracted equally in both black and white. On the other hand, when the form of a stimulus less accurately represents prey from nature (square objects), the importance to prey evaluation of the other stimulus parameters - such as the direction of the background contrast - increases.”

The effect of contrast is shown in figure 14 where the impulses per second are compared at different contrasts with black on white and white on black during the summer. The square size is 4 degrees and it was moved at a constant angular velocity. The greater is the contrast the more intense is the neural signal especially for white on black.

Figure 14
Reduced Contrast Reduces the Intensity of the Neural Signal for Class 2 (Convexity Detector) Neurons in the Toad Bufo bufo. (Ewert and Siefert - 1974)

In the summer the neural response to to the squares changes with the location of the square in the eye's visual field as shown in figure 15. The sensitivity to white on black squares increases towards the lower visual field where the ground is usually in view. In contrast the sensitivity to black on white increases towards the upper visual field where the sky is usually in view. Summer responses are shown on the left of the figure  while winter responses shown on the right. The top row gives the responses for the upper visual field, the middle row gives the responses for the middle visual field, and the bottom row gives the responses for the lower visual field. The square was moved at a rate of 7.6 degrees per second.  During the winter when the toads normally hibernate the responses do not show any significant variation indicating that this selective sensitivity is probably modulated by hormonal changes.

Figure 16 shows that this retinal effect is mirrored in the toad’s prey orientating behavior as well by showing the turning responses per minute vs. square size. The summer response shown on the left and the winter response shown on the right. White open circles represent white on black stimuli while back filled circles represent back on while stimuli. The open circle line is below the closed circle line in the upper right graph. Squares moved at a rate of 20 degrees per second. About half of the 20 frogs tested showed greater overall responsiveness and these are shown in the top row. In contrast the other half showed little responsiveness to the square and these are shown in the bottom row (perhaps due to their motivational or mental health state).

Figure 15
Responses of Toad Class 2 (Convexity Detector) Neurons in Summer and Winter for Different Regions of the Visual Field in the Toad Bufo bufo. (Ewert and Siefert - 1974)

Figure 16
Behavioral Response of the Toad Reflects the Retinal Signal in the Toad Bufo bufo. (Ewert and Siefert - 1974)

References

Aho, A. -C, Donner, K., Helenius, S., Oleson-Larson, L., and Reuter, T. (1993). Visual Performance of the Toad (Bufo bufo) at Low Light Levels: Retinal Ganglion Cell Responses and Prey-Catching Accuracy. Journal of Comparative Physiology A 172:671-682)

Ewert, J.-P. and Borchers, H.-W. (1974). Inhibition of Toad (Bufo bufo) Retinal On-Off and Off Ganglion Cells Via Active Eye Closing. Vision Research 14:1275-1276

Ewert, J.-P, and Siefert, G. (1974) Seasonal Change of Contrast Detection in the Toad’s (Bufo bufo) Visual System. Journal of Comparative Physiology 94:177-186

Grusser, O.-J, and Grusser-Cornehls, U. (1976). Neurophysiology of the Anuran Visual System. In Frog Neurobiology - A Handbook, edited by R Llinas and W. Precht. Springer-Verlag: New York

Hartline, H.K. (1938). The Response of Single Optic Nerve Fibers of the Vertebrate Eye to the Illumination of the Retina. American Journal of Physiology 121:400-415

Hamasaki, D.I. and Esserman, L. (1976). Neural Activity of the Frog’s Frontal Organ during Steady Illumination. Journal of Comparative Physiology 109:279-285)

Lettvin, J.Y., Maturana, H.R., McCulloch, W.S., and Pitts, W.H. (1959) What the Frog’s Eye Tells the Frog’s Brain. Proc. Inst. Radio Engr. 47:1940-1951. Reprinted in Embodiments of Mind (1965) edited by Warren S. McCulloch, MIT Press

Lettvin, J.Y., Maturana, H.R., Pitts, W.H., and McCulloch, W.S. (1961). Two Remarks on the Visual System of the Frog. In Sensory Communication edited by Walter Rosenblith, MIT Press and John Wiley and Sons: New York

Maturana, H.R. (1959) Number of fibres in the optic nerve and the number of ganglion cells in the retina of Anurans. Nature, Lond. 183:1406

Stirling, R.V. and Merrill, E.G. (1987) Functional Morphology of Frog Retinal Ganglion Cells and their Central Projections: The Dimming Detectors. Journal of Comparative Neurology 258:477-495



Web site by David D. Olmsted. He can be contacted at brainsim1-contact at yahoo dot com (this is an anti-spam tactic. Type the address as normal). Original site established August 21, 1998 by David D. Olmsted. New home page published August 25, 2006

Information compiled by David D. Olmsted © 1998 to 2006 (Free to use for personal and educational use)