Experimental procedures

We conducted two series of experiments: behavioral and pharmacological. In the behavioral experiments, all three monkeys performed the oculomotor foraging task for >60 min per session (nine, nine, and four sessions for monkeys S, O, and E, respectively; Fig. 2). We obtained less data from monkey E because she was also assigned to other projects. In the pharmacological experiments, only two monkeys (S and O) were used (Figs. 3-​-6).6). In each daily session, the animals received a single intramuscular injection of either normal saline (control), ketamine (0.7, 1.0, and 1.5 mg/kg), or medetomidine (α2 adrenoceptor agonist, 0.01 mg/kg) into the right femoral quadriceps after a block of 100 baseline (preinjection) trials. Monkeys continued to perform the behavioral task >60 min after the injection. The dosage of ketamine was determined based on previous pharmacological studies in monkeys (Taffe et al., 2002; Stoet and Snyder, 2006; Pouget et al., 2010; Ma et al., 2018) and humans (Ghoneim et al., 1985; Morgan et al., 2004), so that the animals could actively perform the task following injection. For the medetomidine experiments, ∼10% of the anesthetic dose was used. The injection volume was 1.0 ml for all experiments. Pharmacological experiments were separated by at least 3 d to avoid cumulative dosing effects. Different drug and dosage conditions were randomly ordered.

Behavioral data and model fitting. A, Actual behavioral data (gray bars) were compared with the best-fit distributions obtained from the foraging model (colored lines). For each animal, data from multiple sessions were combined (nine, nine, and four sessions for monkeys S, O, and E, respectively). The CDs for the fit of each model were 0.96, 0.97, and 0.92 for monkeys S, O, and E, respectively. Note that when we evaluated the goodness-of-fit, each of the three distributions with 25 bins was normalized so that the area under the curve equaled unity. Three optimal parameters for each animal are reported in the right panel. B, Optimal parameters in individual sessions. Note that the data from each monkey separated by color are clustered.

Effects of ketamine administration on oculomotor parameters. A, Eye position traces from single trials after saline (black) or ketamine (1.5 mg/kg, red) injection in monkey S. The horizontal bar indicates the 100-ms interval for measuring postsaccadic drift. B, Time course of eye movement parameters (drift size, ISI, and saccade number in every 10 min) following intramuscular injection of saline or drugs (0.7, 1.0, and 1.5 mg/kg ketamine and 0.01 mg/kg medetomidine). Data for the same experimental condition (three sessions) are connected with lines. Different colors indicate different drug conditions. The error bar indicates 1 SD and is only displayed for the maximal or minimal values of each time window.

Time courses of foraging model parameters (left column) and eye movement parameters (right) following ketamine administration in monkeys S (A) and O (B). Data points connected with lines indicate single experimental conditions (data from three sessions). Different colors indicate different drug conditions. The left side black square indicates the preinjection data for all conditions (12 sessions). The error bars and horizontal dashed lines indicate the 95% confidence interval of the bootstrap data. Filled circles denote statistically significant modulations compared with the preinjection data. Data points for different drug conditions are slightly jittered horizontally for presentation purposes only.