Theta Modulation of MUA Arises from RF Center-Surround Interactions
(A) Receptive field (RF) composition with an excitatory center (blue) and a suppressive surround (gray). Small and large stimuli (black disks) are shown to illustrate their relationship with RF structures. Lower: example size tuning curve from one representative MUA channel from monkey K (suppression index [SI] 0.78, based on ∼100 trials per condition).
(B) Upper: passive viewing task; the dotted circles indicate the excitatory MUA RF center. Left: example MUA response from the same recording electrode as in (A) to three different stimulus configurations: small 2° disk (blue), large 6° disk (gray), and 2° disk with 4°–6° annulus (orange). The vertical dashed line highlights the window for spectral analyses (0.3–1 s). Right: corresponding MUA powerspectra (calculated during 0.3–1 s after stimulus onset). Note the presence of rhythmic activity in the disk-annulus condition only. Shading around lines depicts SEM. VM/HM, vertical/horizontal meridian.
(C) Population powerspectra for the same stimuli as in (B), averaged across channels for monkey K (left; n = 40) and monkey H (right; n = 57). Shading around lines depicts SEM.
See also .

Anti-Phasic Theta Oscillations from Nearby Electrodes with Adjacent RF Coverage
(A) Hypothesized mechanism underlying the neural theta oscillation. If two separate stimuli (e.g., disk and bar, simultaneously presented) each drive one population while being in the suppressive surround of the other, the anti-phase oscillation is triggered.
(B) Upper: task design; orange and purple dashed circles indicate the RFs of two representative MUA channels. Lower: MUA responses from two example MUA channels from monkey K, each driven either by the disk (orange) or the bar (purple). Inset: theta phases of both channels. Left: the channels’ RFs with disk and bar stimuli. Scale bars signify 1°. Shading around lines depicts SEM.
(C) Population distribution of phase differences between disk- and bar-selective channels for both monkeys (monkey K: mean 77° ± 1.6°, n = 385 channel combinations; monkey H: mean 134° ± 5.7°, n = 20 channel combinations).
See also .

The MUA Rhythm Does Not Depend on Microsaccades
(A) Raster plot shows microsaccade (MS) occurrences over trials and time for passive fixation task (as in ) of monkey K during presentation of the disk-annulus stimulus. The histogram above shows average MS occurrences across all trials; the red line depicts the smoothed distribution (see ). Lower: example MUA response (example channel, averaged across trials). Right: powerspectra of MS occurrences and MUA responses (population average across trials and channels, same as in ). In 37% (40/108) of the trials, no MS occurred in the time period used for the spectral analysis (0.3–1 s after stimulus onset). CI, confidence interval.
(B) Same as (A), but for monkey H. 52% (54/104) of the trials exhibited no MSs in the spectral analysis window.
(C) MS-triggered MUA from monkey K (upper) and H (lower). Zero represents the time of MS occurrence; the red and gray traces show the actual data and shuffle control, respectively. Stars on top highlight significant differences between both conditions (p < 0.05, Bonferroni-Holm corrected for multiple comparisons).
(D) Upper: single-trial example of eye movements of monkey K (x signal, blue; y signal, green). MSs are highlighted in red (n = 2). Lower: a simultaneous single-trial MUA response to a disk-annulus stimulus (disk, 2°; annulus, 6° outer diameter; 4°, inner diameter) averaged across channels (n = 40). The solid vertical line highlights the stimulus onset.
(E) Upper: the average MUA from one example channel from monkey K averaged across trials without MSs after stimulus onset. Lower: comparison of the theta power per channel based on trials with (w/) and without (w/o) MSs for monkey K (black) and H (blue) showing no significant difference (p = 0.27, n = 40 and p = 0.83, n = 57 for monkey K and H, respectively, Wilcoxon signed-rank test).
See also .

Same Theta Rhythm in MUA and Reaction Times during a Distributed Attention Task
(A) Experimental design. Monkeys had to report a luminance change in either the center (orange) or upper flanker (purple; distributed attention task) by making a saccade toward the respective stimulus change. One-third of the trials were constituted of catch trials where no target appeared.
(B) Upper: the reaction time (RT; normalized as deviation from the mean) from monkey K for distributed attention conditions. RT to the center target position is shown in orange, and to the flanker target position in purple. Non-smoothed RT data are shown as thin dashed lines. Lower: MUA response recorded during catch trials from an example channel (no target appearing; during distributed attention). Vertical gray shaded areas highlight the peak-to-peak locking between the behavioral and neural oscillation (RT to center location [orange] and MUA with corresponding RF). Right: powerspectra of the respective rhythms and the RT phase difference between center and flanker conditions (based on non-smoothed RT data). Shaded areas of RT powerspectra highlight 95% confidence intervals based on shuffled surrogate data (see ). Shading around other lines depicts SEM.
See also .

Rhythmic Distributed Attentional Sampling Is Correlated with Cortical MUA Rhythm
(A) Example cross-correlogram between RT and MUA from monkey K during distributed attention (left). Right: the distributions of correlation coefficients of RT-MUA channel combinations for monkey K (left; n = 40) and H (right; n = 57). Electrodes with significant cross-correlations to RT are shown in black (non-significant channels are in gray). The mean MUA-RT cross-correlations across significant electrodes are shown in purple for each monkey.
(B) Powerspectra of cross-correlations for monkey K (left) and H (right) averaged across significantly cross-correlated channels. The shaded area below the dashed line depicts the 95% confidence interval; shading around lines depicts SEM. Note the significant modulation of cross-correlations at theta frequencies.

Strength of MUA Response to Target in Relation to Periodic Fast and Slow RTs
(A) MUA from one representative electrode (monkey K) during trials in which the target appeared over the disk stimulus (center position; 1.21 s after stimulus onset). The MUA response to the target’s luminance change is highlighted in orange on the background of the stimulus-induced oscillation (black).
(B) Average MUA target response across channels comparing response strength during target delays that corresponded to peaks (red) and troughs (black) of the RT time course (see ).
(C) Population distributions of d′ of MUA target responses across channels for RT peaks and troughs for monkey K (left) and H (right) during distributed attention. Asterisks denote significant differences. Shading and error bars depict SEM.
See also .