Search

In monkeys, saccades that repeatedly overshoot their targets adapt to become smaller by the time the monkey has made 1,000-2,000 saccades. In life, adaptation must keep movements accurate for long periods of time. Previous work describes only saccade adaptation that occurs within a few hours. Here we describe long-term saccade adaptation elicited in three monkeys by 19 days of training. Each day a monkey made saccades to track 16 degrees leftward and rightward target movements. During saccades, the target stepped back toward its starting position 6.4 degrees (40%) in two monkeys or 8 degrees (50%) in the third. After each day's adaptation, we blindfolded the monkey with goggles and returned it to its cage overnight. We found that adapting saccades for 19 days elicited significantly larger, long-lasting reduction in saccade size than did adapting for only 1 day. Further, after 19 days of adaptation we could elicit additional, apparently normal, short-term reduction in saccade size by increasing the size of the intra-saccade target movement. In contrast, we could elicit only small additional size reduction after only 1 day of adaptation. A simple model using separate short- and long-term adaptation mechanisms can reproduce many of the features of saccade gain exhibited by monkeys during a 19-day adaptation. We conclude that there is a long-term saccade-adaptation mechanism that is distinct from the well-characterized short-term system and that this newly recognized system is responsible for long-term maintenance of saccade accuracy.

Recent experiments have characterized the dependence of saccade gain adaptation on the characteristics of the visual error following inaccurate saccades. We currently know little about the potential role of non-visual information in driving saccade adaptation. The brain could use non-visual signals from the saccade burst generator or extraocular muscle (EOM) proprioceptors to determine if the eye had rotated the appropriate distance to aim at a target. Both saccade-related burst signals and EOM proprioceptive information reach the posterior vermis of the cerebellum, a brain area strongly implicated in saccade adaptation. In the experiment described here we determined if non-visual information has a significant affect on saccade adaptation. We made monkey saccades hypometric with intra-saccade target movements and then tested the recovery of saccade gain toward normal under three conditions: (1) when the target was continuously visible, (2) when the target extinguished for 1000 ms beginning during the saccade, and (3) when the monkey remained in the dark. In the first condition both visual and non-visual indications of hypometria were available. In the second, only non-visual information was available. In the third, the monkey made no visually guided saccades and very few spontaneous saccades in the dark so neither visual nor non-visual information could drive adaptation. We found that, though it was hypometric, saccade size during recovery changed the same small amount when monkeys made saccades to extinguishing targets or remained in the dark. Saccade size changed significantly (approximately 5x) more during recovery when the monkey tracked continuously visible targets. Thus non-visual information has no influence on adaptation and visual post-saccade error is the only known driver of saccade adaptation., (Copyright 2002 Elsevier Science B.V.)

Inaccurate saccades adapt to become more accurate. In this experiment the role of cerebellar output to the oculomotor system in adapting saccade size was investigated. We measured saccade adaptation after temporary inactivation of saccade-related neurons in the caudal part of the fastigial nucleus which projects to the oculomotor brain stem. We located caudal fastigial nucleus neurons with single unit recording and injected 0.1% muscimol among them. Two monkeys received bilateral injections and two monkeys unilateral injections. Unilateral injections made ipsiversive saccades hypermetric (gains >1.5) and contraversive saccades hypometric (gains approximately 0.6). Bilateral injections made both leftward and rightward saccades hypermetric (gains >1.5). During unilateral inactivation neither ipsiversive nor contraversive saccade size adapted after approximately 1,000 saccades. During bilateral inactivation, adaptation was either small or very slow. Most intact monkeys completely adapt after approximately 1,000 saccades to similar dysmetrias produced by intrasaccadic target displacement. After the monkeys receiving bilateral injections made >1,000 saccades in each horizontal direction, we placed them in the dark so that the muscimol dissipated without the monkeys receiving visual feedback about its saccade gain. After the dark period, 20-degree saccades were adapted to be 12% smaller, and 4-degree saccades to be 7% smaller. We expect this difference in adaptation because during caudal fastigial nucleus inactivation, monkeys made many large overshooting saccades and few small overshooting saccades. We conclude from these results that: (1) caudal fastigial nucleus activity is important in adapting dysmetric saccades; and (2) bilateral caudal fastigial nucleus inactivation impairs the relay of adapted signals to the oculomotor system, but it does not stop all adaptation from occurring.

Saccade accuracy is fundamental to clear vision. The brain maintains saccade accuracy by altering commands for saccades that are consistently inaccurate. For example, saccades that consistently overshoot their targets gradually become smaller. The signal that drives the adaptation of saccade size is not well understood. Previous reports propose that corrective movements and visual errors, both generated after inaccurate saccades, could be responsible for a change in saccade size. Here we show that we can elicit normal reductions in saccade size while eliciting few or no correction saccades. These normal reductions in saccade size indicate that visual errors, not correction saccades, drive the adaptation of saccades.

Saccadic gain (saccade amplitude/target amplitude) can be reduced gradually by repeatedly stepping the target backward during the saccade. The gain reduction produced by this paradigm is thought to be driven by an error signal created by the backstep. We investigated the effects of varying the timing of this error signal relative to the end of the saccade by using two different paradigms in macaques. In the brief backstep paradigm, the target was stepped backward 30% during the saccade but extinguished after different durations. For very short backstep durations (32 ms), little gain reduction occurred. As backstep duration increased, the amount of gain reduction also increased. When backstep duration reached 80 ms, the amount of gain reduction was just under that achieved during the conventional adaptation paradigm in which the backstep remained visible for 1000-1200 ms. In the delayed backstep paradigm, as the saccade occurred, we extinguished the target and then, after a delay, illuminated it for 1 s at the backstep location. In most experiments with short delay times of 16-64 ms, the saccadic gain reduction reached that achieved during conventional adaptation. At delays of 112-208 ms, the amount of gain reduction decreased to approximately 75% of that reached during conventional adaptation. With still longer delays, the amount of gain reduction decreased more gradually. At delays of 750 ms, average gain reduction was 10%. By delays of 1.5 s, gain reduction had fallen essentially to zero. Taken together, these data suggest that the error signal must be present for a limited time ( approximately 80-100 ms) after the saccade to produce the most robust saccadic gain adaptation. However, errors present as long as 750 ms after the saccade still can produce a significant gain reduction.

The brain maintains saccade accuracy by modifying saccades that are consistently inaccurate (e.g. hypermetric). To determine whether this adaptation is influenced by the visual background we used several different target and background movements to elicit changes in saccade gain. In almost all cases, the target spot drove gain changes. The background had no effect on, or slightly reduced, adaptation. We conclude that the saccade adaptation mechanism is driven almost entirely by stimuli on or near the fovea and is affected very little by visual stimuli falling more peripherally.

Saccadic accuracy, measured as the ratio of the size of a saccade to the size of the target step that elicits it, i.e., saccade gain, can be altered by jumping the target surreptitiously during the targeting saccade. The gain change produced by this paradigm does not generalize or transfer to saccades of all sizes. Instead, the amount of transfer decreases the more the tested saccade differs in amplitude and direction from that adapted. Here, we tested the limits of this saccade-size specificity by attempting to impose quite different gain states on saccades in the same direction. We altered the saccadic gain by intrasaccadic target jumps of 30% of the initial target step, either forward to produce a gain increase or backward to produce a gain decrease. Three different conditions were studied: (1) saccades to target steps of 20 degrees or 7 degrees were adapted in individual sessions with backward and forward jumps, respectively; (2) saccades to target steps of 20 degrees caused backward target jumps during the same session in which saccades to 7 degrees target steps caused forward steps; (3) the target jumps accompanying 20 and 7 degrees saccades were the same as in (2), but in addition, there were intermediate-sized saccades to 13.5 degrees target steps with no intrasaccadic target jumps. Saccadic gain adaptation was quite flexible. In condition 2, we could simultaneously increase the gain of saccades to 7 degrees target steps while decreasing the gain of saccades to 20 degrees steps in the same direction. Intermediate horizontal saccades to 13.5 degrees target steps experienced gain reductions (average: 6.9%), which were not the sum of gain changes expected from separate 20 degrees gain decreases and 7 degrees gain increases alone, as predicted from condition 1. If adaptation at 20 degrees and 7 degrees occurred while an animal also tracked a non-adapting 13.5 degrees target step (paradigm 3), the gain reduction of saccades to the 13.5 degrees step was reduced considerably (3.4%). Thus, the mechanism that adapts saccade size can support a robust gain increase for saccades of one size while simultaneously supporting a robust gain decrease for saccades only 13 degrees larger. Furthermore, the presence during adaptation of a non-adapted target step with a size intermediate to the two adapting steps reestablishes a nearly normal gain within only 6.5 degrees of a robust gain increase and decrease. These data indicate that saccadic gain adaptation can set very different gain states for saccades with rather similar vectors.

The gain of saccadic eye movements can be altered gradually by moving targets either forward or backward during targeting saccades. If the gain of saccades to targets of only one size is adapted, the gain change generalizes or transfers only to saccades with similar vectors. In this study, we examined the spatial extent of such saccadic size adaptation, i.e., the gain adaptation field. We also attempted to adapt saccade direction by moving the target orthogonally during the targeting saccade to document the extent of a direction or cross-axis adaptation field. After adaptive gain decreases of horizontal saccades to 15 degrees target steps, >82% of the gain reduction transferred to saccades to 25 degrees horizontal target steps but only approximately 30% transferred to saccades to 5 degrees steps. For the horizontal component of oblique saccades to target steps with 15 degrees horizontal components and 10 degrees upward or downward vertical components, the transfer was similar at 51 and 60%, respectively. Thus the gain decrease adaptation field was quite asymmetric in the horizontal dimension but symmetric in the vertical dimension. Although gain increase adaptation produced a smaller gain change (13% increase for a 30% forward adapting target step) than did gain decrease adaptation (20% decrease for a 30% backward adapting target step), the spatial extent of gain transfer was quite similar. In particular, the gain increase adaptation field displayed asymmetry in the horizontal dimension (58% transfer to 25 degrees saccades but only 32% transfer to 5 degrees saccades) and symmetry in the vertical direction (50% transfer to the horizontal component of 10 degrees upward and 40% transfer to 10 degrees downward oblique saccades). When a 5 degrees vertical target movement was made to occur during a saccade to a horizontal 10 degrees target step, a vertical component gradually appeared in saccades to horizontal targets. More than 88% of the cross-axis change in the vertical component produced in 10 degrees saccades transferred to 20 degrees saccades but only 12% transferred to 4 degrees saccades. The transfer was similar to the vertical component of oblique saccades to target steps with either 10 degrees upward (46%) or 10 degrees downward (46%) vertical components. Therefore both gain and cross-axis adaptation fields have similar spatial profiles. These profiles resemble those of movement fields of neurons in the frontal eye fields and superior colliculus. How those structures might participate in the adaptation process is considered in the DISCUSSION.