Figure 4. The number of exogenous, movement-unrelated ‘visual’ spikes to occur intra-saccadically linearly added to the executed movement’s amplitude.

(A) For every recorded neuron from experiment 1 (Figure 3A,B) and every microsaccade to occur near the visual burst interval (Figure 2), we counted the number of spikes recorded from the neuron that occurred intra-saccadically (0–20 ms after movement onset). We did this for movements directed towards the RF location (Figure 1C; Materials and methods). We then plotted radial eye position (aligned to zero in both the x- and y-axes) relative to saccade onset after categorizing the movements by the number of intra-saccadic spikes. When no spikes were recorded during the eye movement, saccade amplitudes were small (darkest curve). Adding ‘visual’ spikes into the SC map during the ongoing movements systematically increased movement amplitudes. Error bars denote s.e.m. (B) To summarize the results in A, we plotted mean saccade amplitude against the number of intra-saccadic ‘visual’ spikes for movements directed towards the RF locations (faint red dots). There was a linear increase in amplitude with each additional spike per recorded neuron (orange line representing the best linear fit of the underlying raw data). Even intra-saccadic spikes from visual neurons (more dissociated from the motor output of the SC than visual-motor neurons) were still associated with increased amplitudes (Figure 4—figure supplement 1). For movements opposite the RF locations (faint green dots and green line), there was no impact of intra-saccadic ‘visual’ spikes on movement amplitudes. The numbers of movements contributing to each x-axis value are 1772, 383, 237, 145, 113, and 78 (towards) or 1549, 238, 104, 63, 36, 23 (opposite) for 0, 1, 2, 3, 4, and 5 spikes, respectively. (C) For the movements towards the RF locations (A), peak radial eye velocities also increased, as expected (Buonocore et al., 2017). Error bars denote one standard error of the mean (A, C) and 95% confidence intervals (B). Figure 4—figure supplement 2 shows results for intra-saccadic spikes from more eccentric neurons (>4.5 deg), and Figure 4—figure supplement 3 shows the full dependence on neuronal preferred eccentricity. Finally, Figure 4—figure supplement 4 shows the same analyses of B but for the data from experiment 2.

Figure 4—figure supplement 1. Same analysis as in Figure 4B (for movements toward RF’s), but separating visual and visual-motor neurons.

Even visual neurons (≤4.5 deg eccentricity), which are more dissociated from the SC motor output than visual-motor neurons, were still associated with an increase in microsaccade amplitude for injected intra-saccadic ‘visual’ spikes. The influence of visual-motor neurons was larger than the influence of visual neurons because the former are better connected to the SC’s motor output (Mohler and Wurtz, 1976); therefore, the correlation between any one such neuron and the global output behavior of the animal is expected to be larger (this is analogous to the concept of choice probability in other research fields Britten et al., 1996; Nienborg and Cumming, 2006). Error bars denote 95% confidence intervals.

Figure 4—figure supplement 2. Intra-saccadic ‘visual’ spikes from more eccentric neurons in experiment one still linearly increased microsaccade amplitudes, but with a much weaker effect size.

(A) Radial eye position relative to saccade onset grouped by the number of ‘visual’ spikes counted in the interval 0–20 ms, for eye movements going towards the recorded neuron’s RF location (similar to Figure 4A). In this analysis, the RF was always located at an eccentricity >4.5 deg (Figure 1C). We used the same grouping and color conventions as in Figure 4A but in gradients of blue instead of red. When no spikes were recorded during the eye movement, saccade amplitudes were relatively small (darkest blue curve). Adding visual spikes in the SC map during the ongoing movements slightly increased their amplitudes (1–5, color-coded from dark to light blue), but the effect was much milder than for neurons closer in eccentricity to the foveal movement endpoints (Figure 4A). Note, however, that with proper temporal alignment of visual spikes with microsaccade onsets, even these more eccentric neurons could have a strong impact on microsaccade amplitudes (see Figure 6—figure supplement 2). (B) Mean saccade amplitude as a function of the number of intra-saccadic visual spikes (faint blue dots), similar in formatting to Figure 4B. There was a linear increase in amplitude relative to the number of injected visual spikes (blue line), similar to Figure 4B for the more central neurons. However, the slope of the effect was significantly lower (slope: 0.0098876; t = 6.7195, p=2.024⁻¹¹). The solid lines represent the best linear fit of the underlying raw data. Error bars denote 95% confidence intervals. The numbers of movements contributing to each x-axis value are 3458, 684, 375, 244, 169, and 155 for 0, 1, 2, 3, 4, and 5 spikes, respectively.

Figure 4—figure supplement 3. Injected visual spikes always increased microsaccade amplitudes, but the effectiveness was diminished with larger neuronal eccentricities.

(A) Analysis similar to that in Figure 4B (for the towards movements) but now for different neuronal preferred eccentricities (the different colors; Materials and methods). Visual spikes in neurons at eccentricities larger than approximately 4–5 deg had much lower slopes than visual spikes in more central neurons. Importantly, the slope of the relationship between injected intra-saccadic visual spikes and microsaccade amplitude was always positive across all of the tested eccentricities, meaning that there was still a positive impact of the more eccentric neurons, albeit weaker in magnitude (also see Figure 6—figure supplement 2). (B) The slopes of the curves in A now drawn as a function of neuronal preferred eccentricity (Materials and methods). There were diminishing returns with larger distances between neuronal preferred eccentricity and microsaccade amplitudes, but the slope was always positive. That is, even the most eccentric neurons were still associated with a modest impact in terms of increasing the executed movement amplitudes (an example is seen in Figure 4—figure supplement 2 and also in Figure 6—figure supplement 2). Also note how the curve of slope dependence on neuronal preferred eccentricity justifies our choice in most analyses to focus on eccentricities ≤ 4.5 deg. Error bars denote s.e.m.

Figure 4—figure supplement 4. The analysis of Figure 4B but during the spatial frequency task (experiment 2).

This figure is identically formatted to Figure 4B, but this time showing results from experiment 2 (movements toward and away from the RF locations and for neurons ≤ 4.5 deg in preferred eccentricity). Very similar observations were made in both tasks. The numbers of movements contributing to each x-axis point are 403, 86, 36, 30, 18, and 15 microsaccades (towards) and 519, 81, 37, 21, 22, and one microsaccades (opposite) for 0, 1, 2, 3, 4, and 5 spikes, respectively.