Mapping of the receptive fields in the optic tectum of chicken (Gallus gallus) using sparse noise

Abstract

The optic tectum plays a key role in visual processing in birds. While the input from the retina is topographic in the superficial layers, the deep layers project to the thalamic nucleus rotundus in a functional topographical manner. Although the receptive fields of tectal neurons in birds have been mapped before, a high resolution description of the white and black subfields of the receptive fields of tectal neurons is not available. We measured the receptive fields of neurons in the different layers of the tectum of anesthetized chickens with black and white stimuli that were flashed on a grey background in fast progression. Our results show that neurons in the deep layers of the optic tectum tend to respond stronger to black stimuli compared to white stimuli. In addition, the receptive field sizes are larger when measured using black stimuli than with white stimuli. While the black subfield was significantly larger than the white subfield for the intermediate and deep layers, no significant effects were found for the superficial layers. Finally, we investigated the optimal stimulus size in a subset of the neurons and found that these cells respond best to small white stimuli and to large black stimuli. In the majority of the cases the response was stronger to a large black bar than to a small white bar. We propose that such a stronger response to black stimuli might be advantageous for the detection of darker objects against the brighter sky.

Figure 1. Method.

a) Extracellular recordings were made from the superficial, intermediate and deep layers of the optic tectum in chickens while the contralateral eye was stimulated with white and black stimuli presented on a CRT monitor. b) Black and white stimuli on a grey background were shown in fast progression (see methods for details). c) Reverse correlation method. For each spike we looked which stimulus was presented at that time (for delay = 0 ms) and we increased the spike count for the stimulus presented at this position. The result is an array showing the number of spikes evoked by the stimulus presented at the moment that the spikes occurred. d) Next we looked which stimulus was present a short delay before the occurrence of the spikes (delay = 60 ms). For simplicity the stimulus grid is reduced to 3×3 positions and data arrays are only shown for delay = 0 and delay = 60 ms. However, analyses were done for stimuli occurring 0–200 ms before in 5 ms steps (for stimulus duration of 50 ms).

Figure 2. Receptive field of a neuron using PSTH and reverse correlation.

a) PSTH of a single unit response recorded at a depth of 197 µm (superficial layers) in response to white and black stimuli on a grey background for each grid position on a 20×20 axis. In this specific case each grid position covered 0.3×0.3 degrees of visual field. For plotting purpose the axes of the PSTH were discarded. b) Receptive field of the same neuron where every spike is reversely correlated to the white (upper row) and black (lower row) stimulus shown a short period before (delay = 0–190 ms). The arrows show the optimal latency for which the highest autocorrelation was found for black and white. The white subfield is 1.34 degrees (at delay = 100 ms) and black subfield is 1.53 degrees (at delay = 40 ms). The maximum response evoked to stimulation at one grid position in this particular example was 99 spikes.

Figure 3. Receptive fields of two neurons in the optic tectum.

a) Receptive field of a neuron recorded at 1210 µm presented as the response to white and black stimuli shown 0–190 ms before. The neuron responded with a short delay to black (40 ms) and with a longer delay to white (130 ms) and therefore could be responsive to decrease in luminance. In this particular case, the stimuli were shown on 15×15 positions which each covered 3.4×3.0 degrees of the visual field. The white subfield (at delay = 130 ms) was 10.8 degrees in size while the black subfield was 14.9 degrees (at delay = 40 ms). b) Receptive field of a neuron recorded at 1090 µm. In this particular case, the stimuli were shown on 15×15 positions which each covered 4.2×3.5 degrees of visual field. Please notice that the black subfield is larger than the white subfield (for this neuron the white subfield was 11.1 degrees in size while the black subfield was 19.4 degrees).

Figure 4. White and black subfields of the receptive fields.

Examples of the white and black subfields of the receptive fields of twelve neurons at their optimal latency recorded in the a) superficial layers (neuron 1–4), b) intermediate layers (neuron 5–8) and c) deep layers (neuron 9–12). Below the picture is the information regarding the number of grid units (15×15 or 20×20) as well as the size of each grid unit in degrees (width×height). The calculated size of the subfields is depicted in the upper right corner of each picture.

Figure 5. Receptive field properties.

a) Maximum number of spikes in response to a black or white stimulus for the superficial layers, intermediate layers and deep layers. Responses to white stimuli are shown in white and responses to black stimuli in black. b) Average size of the black subfield and the white subfield for the superficial layers, intermediate layers and deep layers. Responses to white stimuli are shown in white and responses to black stimuli in black. c) Sizes of the black subfield and the white subfield in relation to the recording depth in the optic tectum. Open circles represent the white subfields while closed circles represent the black subfields. d) Average contrast index for the superficial (shown in light grey), intermediate (shown in dark grey) and deep layers (shown in black). Negative values represent neurons whose black subfield was larger, while positive values represent neurons whose white subfield was larger. e) Distribution of the contrast index values for the superficial layers (shown in light grey), intermediate layers (shown in dark grey) and deep layers (shown in black). Please notice that the majority of the neurons has a contrast index >0 indicating a larger subfield for black than for white. Error bars represent standard errors. *p<0.05.

Figure 6. Example of the response to a white and a black bar.

PSTH and raster plot of a single unit (recorded at 1700 µm depth) in response to white and black bars swept through the receptive field in the preferred direction. The upper row shows the responses to a 0.5, 1, 2 or 4 degrees wide white bar while the lower row shows the responses to a similar-sized black bar. This neuron responds almost as strong to a small white stimulus as to a large black stimulus.