J Neurophysiol 114: 125–137, 2015.
First published April 8, 2015; doi:10.1152/jn.00212.2015.
Adaptation and adaptation transfer characteristics of five different saccade
types in the monkey
Yoshiko Kojima, Albert F. Fuchs, and Robijanto Soetedjo
Department of Physiology and Biophysics and Washington National Primate Research Center, University of Washington,
Seattle, Washington
Submitted 3 March 2015; accepted in final form 3 April 2015
saccade; adaptation; motor learning; plasticity; monkey
PLASTICITY is an important property of the nervous system. In
the case of movement, it helps to overcome persistent motor
errors caused by growth, injury, and aging. Such motor plasticity or adaptation can be investigated quite handily for saccadic eye movements, which shift the direction of gaze between visual targets of interest. To cause adaptation, the target
can be displaced during the saccade so that the saccade appears
to have the wrong amplitude or direction. When faced with
such repeated apparent errors, the oculomotor system gradually
adapts the size and/or direction of its saccades so that gaze
eventually lands near the displaced target (McLaughlin 1967).
This adaptation paradigm has been used in primates to infer the
site(s) and mechanism of adaptation in the saccade system.
Primate saccades can be elicited in a variety of situations.
We consider five types here. The most natural type is the
scanning saccade, which we use to redirect our gaze from one
Address for reprint requests and other correspondence: Y. Kojima, 1959 NE
Pacific St. HSB I421, Washington National Primate Research Center, Box
357330, Univ. of Washington, Seattle, WA 98195-7330.
www.jn.org
interesting, usually stationary, object in space to another, e.g.,
during reading. The remaining types are elicited primarily in
the laboratory in response to jumping spots, whose timing
generally produces different response latencies. Express saccades, which have the shortest latencies (80 –120 ms) (Fischer
et al. 1993; Fischer and Boch 1983; Fischer and Ramsperger
1984), occur if a target that is being fixated disappears before
a distant target appears. Typical latencies for targeting saccades to sequentially illuminated spots are ⬃150 ms or greater
in highly trained monkeys. Saccades to a distant target can be
delayed if the monkey is required to maintain fixation until the
fixation target is extinguished. Memory-guided (MG) saccades
occur when the distant target appears only briefly and the
monkey again must maintain fixation until the disappearance of
the fixation target instructs it to saccade to the remembered
location of the extinguished target. Although targeting saccades in the monkey have been shown to undergo adaptation,
adaptation of the other simian saccade types has not been
studied systematically. Therefore, the first goal in this study
was to remedy that gap in our knowledge by documenting and
comparing the characteristics of adaptation for all five saccade
types.
Knowledge of the pattern of adaptation transfer or generalization between saccade types allows us to speculate whether
sites of adaptation are shared by different saccade types or
whether an adaptation site is unique for a particular type.
Adaptation transfer between several saccade types has been
studied in humans. For example, although results differ somewhat between laboratories, the adaptation of targeting saccades
transfers to express, scanning, delayed, and MG saccades
(Alahyane et al. 2007; Collins and Doré-Mazars 2006; Cotti et
al. 2007; Deubel 1995, 1999; Fujita et al. 2002; Hopp and
Fuchs 2002, 2010; Panouillères et al. 2014; Zimmermann and
Lappe 2009; for reviews see Table 1 in Hopp and Fuchs 2004;
Pélisson et al. 2010). However, the adaptation of MG saccades
transfers only to delayed saccades (Deubel 1995, 1999; Fujita
et al. 2002; Hopp and Fuchs 2002, 2010). Therefore, the
adaptations of targeting and MG saccades occur at different
neuronal loci. Although this information does not allow us to
identify specific neural adaptation sites, it suggests that the
saccade commands for all of the other four types pass through
the targeting saccade adaptation site. In contrast, only the
delayed saccade command passes through the MG saccade
adaptation site.
The only simian transfer study thus far revealed that targeting saccade adaptation also transfers robustly to the other four
types (Fuchs et al. 1996), again suggesting that their saccadic
commands pass through the targeting saccade adaptation site.
0022-3077/15 Copyright © 2015 the American Physiological Society
125
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
Kojima Y, Fuchs AF, Soetedjo R. Adaptation and adaptation
transfer characteristics of five different saccade types in the monkey.
J Neurophysiol 114: 125–137, 2015. First published April 8, 2015;
doi:10.1152/jn.00212.2015.—Shifts in the direction of gaze are accomplished by different kinds of saccades, which are elicited under
different circumstances. Saccade types include targeting saccades to
simple jumping targets, delayed saccades to visible targets after a
waiting period, memory-guided (MG) saccades to remembered target
locations, scanning saccades to stationary target arrays, and express
saccades after very short latencies. Studies of human cases and
neurophysiological experiments in monkeys suggest that separate
pathways, which converge on a common locus that provides the motor
command, generate these different types of saccade. When behavioral
manipulations in humans cause targeting saccades to have persistent
dysmetrias as might occur naturally from growth, aging, and injury,
they gradually adapt to reduce the dysmetria. Although results differ
slightly between laboratories, this adaptation generalizes or transfers
to all the other saccade types mentioned above. Also, when one of the
other types of saccade undergoes adaptation, it often transfers to
another saccade type. Similar adaptation and transfer experiments,
which allow inferences to be drawn about the site(s) of adaptation for
different saccade types, have yet to be done in monkeys. Here we
show that simian targeting and MG saccades adapt more than express,
scanning, and delayed saccades. Adaptation of targeting saccades
transfers to all the other saccade types. However, the adaptation of
MG saccades transfers only to delayed saccades. These data suggest
that adaptation of simian targeting saccades occurs on the pathway
common to all saccade types. In contrast, only the delayed saccade
command passes through the adaptation site of the MG saccade.
126
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
MATERIALS AND METHODS
Surgery and Training
We measured eye movements in two rhesus monkeys (Macaca
mulatta, male; monkey B: 7.4 kg, monkey W: 5.0 kg) with the
electromagnetic search coil method (Fuchs and Robinson 1966; Judge
et al. 1980; Robinson 1963). Briefly, in an aseptic surgery, we
implanted a three-turn coil of fine Teflon-coated stainless steel wire on
one eye of each monkey. The ends of the wire terminated in a small
connector fixed to the top of the skull with titanium screws and dental
acrylic. In the same surgery, we attached three receptacles to the
monkey’s skull to stabilize the head during eye movement recordings.
For more details see Fuchs et al. (1996).
After recovery from the surgery, each monkey was placed in a
dimly lit booth and trained to follow a small visual target with its eyes.
The monkey sat in a primate chair with its head restrained. We
surrounded the animal with a semicircular array of light-emitting
diodes (LEDs), which were spaced every 1° along the horizontal
meridian, and illuminated them sequentially to produce target jumps.
Each LED subtended 0.25°. The LEDs could be programmed to emit
either a red or a green light. We used the green LEDs only to elicit
delayed and MG saccades (see below). After the monkeys had been
trained to fixate an illuminated LED, i.e., the target, we trained them
to make saccades between sequentially stepping LED targets. If the
monkey’s gaze arrived within ⫾2° of the target within 0.8 s of the
target step and stayed there for at least 0.5 s, the monkey was
continually rewarded every 1–1.2 s. Note that 0.8 s allows the
occurrence of a corrective saccade. We applied the same reward
schedule for all saccade types during preadaptation, adaptation, and
postadaptation sessions.
Saccade Types and Their Adaptation
The data gathered in this study were obtained before we started the
cerebellar single-unit recording experiments reported in Kojima et al.
(2010b), where the procedures for eliciting the five types of saccade
are described in detail. Briefly, after the monkeys made targeting
saccades continually for at least 1 h, we attempted to generate express
saccades. For each animal, we determined which target amplitude
elicited a high incidence of saccades with express latencies (for more
detail, see Express saccades below). Monkey B made a high rate of
express saccades to a 10° horizontal target step and monkey W to a 15°
target step. For each monkey, these same step sizes were used for
adaptation and testing of all five types of saccade. Once the animal
reliably made saccades with express latencies, we trained it on the
delayed, MG, and scanning saccade paradigms.
Targeting saccades. After the monkey had fixated a red LED for at
least 0.6 s, another LED was illuminated within the next 0.7– 0.9 s and
the first went out. This sequence created a target step to either the right
or the left that remained within ⫾23° of straight-ahead.
Adaptation of targeting saccades was induced by a modified version of the intrasaccadic step (ISS) paradigm (McLaughlin 1967). In
the McLaughlin paradigm, the target jumped along the horizontal
meridian and then, during a targeting saccade, jumped backward so
the eye appeared to overshoot the target. The subject then made a
corrective saccade backward to acquire the target. This procedure,
when repeated over several hundred trials, caused a gradual decrease
in the size of the initial saccade. In this study, we did not examine the
gradual gain increases that are produced by a forward ISS. In our
modified version of the paradigm, we caused the error of the targeting
saccade to remain a constant percentage of its amplitude. We employed this adaptation strategy to provide the same percentage ISS to
all five saccade types, which had different accuracies. For example,
MG saccades were less accurate than targeting saccades (White et al.
1994). A computer detected the saccade end position (when the eye
velocity decelerated to ⱕ20°/s) and moved the target back 25% of the
target step amplitude from the saccade end position. Because the
LEDs were placed every 1°, the target position was rounded down to
the nearest integer. The schematic at the top of Fig. 1A shows the
schematic target movement and elicited saccade that occurs in the
“targeting” saccade paradigm.
Express saccades. To elicit express saccades, we extinguished the
fixated red LED briefly (gap paradigm) before illuminating a second
LED at a distance of either 10° for monkey B or 15° for monkey W
(Fischer et al. 1993; Fischer and Boch 1983; Fischer and Ramsperger
1984). We varied the duration of the extinction gap in 10-ms steps
from 80 to 200 ms to determine which duration elicited the highest
incidence of saccades with express latencies. The most effective gap
was 100 ms for monkey B and 160 ms for monkey W. With these gaps,
the distribution of latencies was bimodal, with the trough occurring at
⬃100 ms for monkey B and ⬃110 ms for monkey W (Fig. 1B). We
considered as express saccades only those with latencies from 50 to
100 ms for monkey B and from 50 to 110 ms for monkey W.
To adapt express saccades, we measured saccade latencies online
and applied an adapting step only if the latency was ⬍100 ms for
monkey B and ⬍110 ms for monkey W (Fig. 1A, Express). The target
was stepped backward 25% of the saccade size from the saccade end
position. If a saccade latency was greater than the maximum accepted
express value, there was no back step and instead the target was turned
off for 300 ms. We used this short duration here to gather the
maximum number of saccades before the monkey became fatigued.
Scanning saccades. To elicit scanning saccades, we simultaneously
illuminated three red LEDs, each separated by either 10° (monkey B)
or 15° (monkey W) along the horizontal meridian, and rewarded the
monkey whenever it made, at its own pace, more than three consecutive saccades between any of the three LEDs. Monkeys most often
made a saccade to the adjacent LED, i.e., 10° (monkey B) or 15°
(monkey W).
To adapt scanning saccades, all three illuminated LEDs were
extinguished at the end of the saccade and three other LEDs located
short of each LED target were illuminated to produce a backward step
of the entire three-LED array by 25% of the saccade amplitude from
the saccade end position (Fig. 1A, Scanning). The occasional (⬃1%)
saccades that were made between the two most eccentric LEDs were
not analyzed. However, the back step also occurred for those
saccades.
Delayed saccades. To elicit a delayed saccade, we required the
monkey to continue to fixate an illuminated red LED for 0.3 s before
a peripheral green LED was illuminated continuously to show the
destination for the future saccade. We required the monkey to delay its
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
Neurophysiological studies indicate that targeting and delayed
saccade adaptation occur downstream of the superior colliculus
(SC) (targeting saccade: Edelman and Goldberg 2002; Frens
and Van Opstal 1997; Melis and van Gisbergen 1996; cf.
Takeichi et al. 2007; delayed saccade: Quessy et al. 2010).
Neurophysiological studies during adaptation suggest that one
downstream structure that likely participates in targeting saccade plasticity is the oculomotor vermis (OMV) of the cerebellum (Barash et al. 1999; Catz et al. 2005, 2008; Kojima et
al. 2010b, 2011; Soetedjo et al. 2008; Soetedjo and Fuchs
2006; Takagi et al. 1998). In contrast, nothing is known about
the sites or the neuronal substrate for the adaptation of any
other saccade type. To locate these different sites and to
determine the mechanisms that underlie the adaptations of the
different saccade types, we require similar transfer data in an
experimental primate. Therefore, our second goal was to determine how adaptation of each simian saccade type affects the
other four types. Such data provide a behavioral platform from
which to launch the various neurobiological techniques that
can be applied to the behaving monkey.
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
B
% of Saccade
A Targeting
Express
Gap
Monkey-B
Monkey-W
20
20
10
10
0
127
0
0
100
200
0
100
200
Saccade Latency (ms)
Scanning
C
1st test
2nd test
Pre
Adapt
adapt saccade
Last15
#1 #2 #3
for Transfer test
test saccade 1
test saccade 2
Recovery
test saccade 3
Note: Number of test saccade types varied between experiment
Memory guided
Target &
Eye Position
Time
Fig. 1. Schematic illustration of the experimental protocol. A: the task used to elicit the 5 types of saccade examined in this study. Thick gray lines indicate both
target position and LED ON/OFF state; thin black lines indicate eye position. B: distributions of saccade reaction times from representative express saccade
experiments for monkeys B and W. The bimodal distributions were elicited by gaps of 100 and 160 ms, respectively. Reaction times between 50 and 100 and
between 50 and 110 ms in monkeys B and W, respectively, are express saccades (black bars). C: time course of an experiment. In this example, 1 saccade type
(gray) is adapted and 3 other types (red, blue, and green) are tested. We collect ⬃30 saccades of all 4 types during the preadaptation session (Pre). We then reduce
the amplitude of the adapting saccade (Adapt). During adaptation, we collect ⬃20 trials each of the 3 test saccades 3 times without backward steps. Finally, we
discontinue the backward step and monitor the recovery of the adapted saccade gain (Recovery).
saccade to the green target by continuing to fixate the red LED for an
additional 0.3 s. When the fixated red LED went off, it signaled the
animal to make a saccade to the peripheral, illuminated green LED.
After the monkey had fixated the green LED for 0.6 s, the red LED
was turned on in place of the green LED to serve as the initial fixation
spot for the next trial. We used short fixation durations here to
maximize the number of saccades before the monkey became
fatigued.
To elicit adaptation, the green LED was stepped backward by 25%
of the saccade amplitude from the saccade end position (Fig. 1A,
Delayed).
MG saccades. To elicit a MG saccade, we required the monkey to
fixate a red LED for 0.3 s and then we turned on a green LED briefly
for 0.2 s at the next target location. After the green LED was
extinguished, 0.3 s later the red fixation LED was also turned off,
serving as a signal for the monkey to saccade to the remembered
location of the green LED. When the saccade landed at the remembered target location, the green LED came on at the actual location.
After the monkey had fixated the green LED for 0.6 s, the red LED
was turned on in its place to serve as the initial fixation LED for the
next trial.
To produce MG saccade adaptation, the green LED was stepped
backward by 25% of the saccade amplitude from the saccade end
position (Fig. 1A, Memory guided).
Experimental Conditions
Once the monkey had demonstrated its ability to perform the
various saccade tasks and adapt all types of saccade, we began the
transfer experiments. In each experiment, we induced adaptation of
one saccade type (the adapt saccade) and tested whether that adaptation transferred to one or more (up to 4) of the other types (the test
saccades).
Figure 1C illustrates the time course of an experiment in which
one saccade type was adapted and three other types were tested.
Before adaptation (Fig. 1C, Pre), we collected ⬃30 of the saccades
to be adapted and ⬃30 of each of the three saccade types to be
tested. We then reduced the amplitude of the adapted saccade by
jumping the targets backward as the monkey made saccades toward
them. The backward step was applied for saccades in both the leftand rightward directions. After each ⬃10-min segment of adaptation,
we collected ⬃20 of one of the three test saccade types so that by the
end of the entire adaptation each test saccade had been sampled three
times. During an ⬃10-min segment of adaptation, a monkey adapted
approximately 370 targeting saccades, 180 express saccades, 280
delayed or MG saccades, and 550 scanning saccades in each direction.
The backward step was not applied during the collection of test
saccades.
After the ⬃20 test saccades, the amplitude of the adapt saccade had
shown little or no recovery. Across both monkeys, the average gain
recovery was 0.020 ⫾ 0.028, 0.014 ⫾ 0.016, 0.012 ⫾ 0.024, 0.017 ⫾
0.015, and 0.022 ⫾ 0.026 for targeting, express, scanning, delayed,
and MG saccade adaptation, respectively. Gain recovery did not differ
significantly between saccade types (unpaired t-test, P ⬎ 0.05 with
Bonferroni correction). Therefore, we ignored it in further analyses.
Finally, we discontinued the back step and monitored the recovery
of the adapted saccade gain (Fig. 1C, Recovery). For most experiments with monkey B, we were able to test four types of saccade in
each adaptation experiment. For most experiments with monkey W,
who did not work as long, we could test only two types of saccade in
each experiment.
Data Analysis
We did the initial data acquisition and the online data processing as
described in Kojima et al. (2010a). Briefly, eye and target position
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
First15
Delayed
3rd test
128
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
f 共x兲 ⫽ a ⫻ exp共⫺b ⫻ x兲 ⫹ c
(1)
Because all adaptations had at least 300 adapt trials, we determined
the gain change after 250 saccades as [mean normalized gain of first
15 saccades of adaptation ⫺ f(250) of exponential fit]. Finally, we
calculated the initial jump of the adaptation as [mean normalized gain
of last 15 saccades of preadaptation ⫺ mean normalized gain of first
15 saccades of adaptation] and the recovery jump at the onset of
recovery as [mean normalized gain of first 15 saccades of recovery ⫺
mean normalized gain of last 15 saccades of adaptation].
To assess the adaptation transfer to the test saccades, we applied an
unpaired t-test to the last 15 test saccades during preadaptation and the
first 15 test saccades of the third test saccade block (see Fig. 1C). If
the normalized gain of saccades in the third test block was not
significantly different from the normalized gain of preadaptation test
saccades (unpaired t-test, P ⬎ 0.05), we considered that adaptation
had not transferred to the test saccade, i.e., the percent transfer was
0% (Fuchs et al. 1996). On the other hand, if the normalized gain of
the test saccades was smaller and the normalized gain changes of the
adapted and test saccades were not significantly different (unpaired
t-test, P ⬎ 0.05), we considered that adaptation transfer was complete,
i.e., the percent transfer was 100%. If the latter t-test was significant
(unpaired t-test, P ⬍ 0.05), we then calculated the percent transfer as
Percent transfer ⫽ 关共GpreT ⫺ GadaT兲 ⁄ 共GfirstA ⫺ GlastA兲兴 ⫻ 100
(2)
In this equation, GpreT is the mean normalized gain of the last 15 test
saccades of preadaptation, GadaT is the mean normalized gain of the
first 15 test saccades of the third test, GfirstA is the mean normalized
gain of the first 15 saccades of adaptation, and GlastA is the mean
normalized gain of the last 15 adapted saccades before the third test
(Fig. 1C, box 1, 2, or 3 for Transfer test). We chose these gain
measures so that the ways the target stepped were always the same for
the test saccades (GpreT and GadaT, without ISS) and the adapted
saccades (GfirstA and GlastA, with ISS).
As shown in RESULTS, there often is an immediate drop in gain at the
beginning of adaptation. Therefore, we also calculated the percent
transfer in a second way, which took the drop into account by
changing the denominator of Eq. 2 to (GpreA ⫺ GlastA), where GpreA
is the mean normalized gain of the last 15 adapt saccades of preadaptation (use of this modification 1 of Eq. 2 produced the gray bars in
Fig. 5). In this case, the target condition differs between GpreA
(without ISS) and GlastA (with ISS).
To avoid drawing conclusions based only on nonsignificant results
(see the description of Eq. 2), we used the same equation to calculate
the actual percent transfer without applying the adjustment for no
(0%) or complete (100%) transfer (use of this modification 2 of Eq. 2
produced the striped bars in Fig. 5). These two calculations did not
change the conclusions of our results.
Because the transfer of adaptation varied from experiment to
experiment (Fuchs et al. 1996), we collected 10 data sets for each pair
of adapt and test saccade types for each monkey so that the total
permutation of the transfer data sets was 10 (data sets) ⫻ 5 (adaptation
saccade types) ⫻ 4 (other test saccade types) ⫻ 2 (monkeys) ⫽ 400.
To infer whether the adaptation transferred to a particular test saccade
on average, we tested the mean percent transfer obtained from the two
monkeys (n ⫽ 20 data sets, t-test). If it was significantly different
from 0 (unpaired t-test, P ⬍ 0.05 with Bonferroni correction), we
inferred that the adaptation had transferred to the test saccade type.
Because the number of saccade types tested in each adaptation
experiment varied, the number of adaptation data sets for each saccade
type did as well. For monkey B, the sets numbered 20, 18, 24, 10, and
10 for targeting, express, scanning, delay, and MG saccades, respectively. For monkey W, they numbered 22, 20, 20, 20, and 24,
respectively.
All experiments were performed in accordance with the Guide for
the Care and Use of Laboratory Animals (1996) and exceeded the
minimal requirements recommended by the Institute for Laboratory
Animal Research and the Association for Assessment and Accreditation of Laboratory Animal Care International. All procedures were
evaluated and approved by the local Animal Care and Use Committee
of the University of Washington (ACC no. 2602-01).
RESULTS
We induced gain decrease adaptation of targeting, express,
scanning, delayed, or MG saccades and tested the transfer of
that adaptation to one or more of the other four types. We first
describe the adaptation characteristics of each type of saccade
and then the transfer of that adaptation.
Course of Adaptation of Different Saccade Types
All types of saccade showed significant adaptation. However, the amount of gain change during adaptation varied from
one saccade type to another. Figure 2, A–E, show representative courses of adaptation and recovery of the five types of
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
signals were digitized at 1 kHz and stored on a hard disk through an
interface [Power 1401, Cambridge Electronic Design (CED)]. During
the experiment, we used a custom program running in Spike 2 (CED)
to control the data acquisition box (Power 1401, CED) and to analyze
the incoming data online. The data were saved after the experiment
and analyzed off-line on a computer using custom programs that ran
analysis software (Spike 2, CED). Saccade onset and end were
identified by an eye velocity criterion of 20°/s. Saccades elicited by
the initial (not intrasaccadic) target jump were selected for analysis.
Parameters of saccades and target jumps, e.g., latency, onset, offset,
amplitude, duration, peak velocity, etc., and the saccade traces for
every trial were exported to other programs (MATLAB, MathWorks)
for subsequent analysis. We eliminated saccades with a vertical
component that was ⱖ3 standard deviations from the mean and if the
horizontal and vertical initial eye positions differed from the initial
target positions by ⱖ1°. These restrictions eliminated ⬍5% of the
trials.
We calculated gain as [saccade size/target step size]. Before adaptation, the gain of some saccade types in both monkeys was idiosyncratically different. In monkey B, scanning saccades showed significantly higher gains than the other four types (unpaired t-test, P ⬍ 0.05
with Bonferroni correction). In monkey W, delayed saccades showed
significantly higher gains than express saccades (unpaired t-test, P ⬍
0.05 with Bonferroni correction, here and for all other t-tests). Therefore, in each monkey we normalized the gain of each saccade type as
[saccade gain/mean preadaptation gain].
The gain change of the adapted saccade was calculated as [mean
normalized gain of first 15 saccades of adaptation (Fig. 1C, First15) ⫺
mean normalized gain of last 15 saccades (Fig. 1C, Last15)]. To
assess whether the adaptation was significant, we compared the mean
normalized gain of the first and last 15 adapt saccades (unpaired t-test,
P ⬍ 0.05). Saccades in all of the data sets used in this study showed
a significant gain reduction during adaptation (unpaired t-test, P ⬍
0.05). We treated rightward and leftward saccades as independent data
sets because adaptation in one horizontal direction does not produce
adaptation in the opposite direction (humans: Deubel et al. 1986;
Miller et al. 1981; Semmlow et al. 1989; cf. Rolfs et al. 2010;
monkeys: Noto et al. 1999; Scudder et al. 1998; Straube et al. 1997),
so each adapted direction provides a separate data set. The average
number of rightward saccades during the adaptation session was
827 ⫾ 321, 593 ⫾ 292, 945 ⫾ 584, 411 ⫾ 108, and 429 ⫾ 103 for
targeting, express, scanning, delayed, and MG saccades, respectively.
For leftward saccades, it was 767 ⫾ 216, 569 ⫾ 232, 945 ⫾ 598,
435 ⫾ 104, and 482 ⫾ 105 for targeting, express, scanning, delayed,
and MG saccades, respectively.
To compare adaptation characteristics across saccade types, we
also fit the course of adaptation with an exponential function:
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
Targeting
Normalized Saccade Gain
Monkey-B
A
Pre
Monkey-W
Recovery
B
C
Express
D
Scanning
E Memory guided
Delayed
1.2
1.2
1.2
1.2
1
1
1
1
1
0.8
0.8
0.8
0.8
0.8
0.6
0.6
0.6
0.6
0
F
500 1000 1500
Saccade Number
Pre
Adapt
2000
0
500
1000
1500
0
500
1000
1500
2000
1.2
0.6
0
500
1000
0
1.2
1.2
1
1
1
1
1
0.8
0.8
0.8
0.8
0.8
0.6
0.6
500
Saccade Number
1000
Recovery
1.2
0
500
1000
0.6
0
500
1.2
1.2
0.6
0.6
0
500
1000
0
500
0
500
Fig. 2. Time course of adaptation measured as normalized saccade gain (black dots) and recovery (gray dots) of each of the 5 types of saccade for both monkeys.
Each dot indicates 1 saccade trial. Gray line on the black dots indicates the exponential fit of saccade gain during adaptation.
saccades for monkey B. Note that although other types of
saccade were inserted and tested during the adaptation (see Fig.
1C), we have plotted only the adapted saccades. Also recall
that the 20 test saccades inserted at various times during
adaptation caused very little recovery of the adapted saccade
(see MATERIALS AND METHODS).
As described in Straube et al. (1997), the course of targeting
saccade adaptation (Fig. 2A) was nicely fit with an exponential
function (r ⫽ 0.73, rate constant ⫽ 646). There was an initial
jump in gain when we stopped the target back step (vertical
dashed line between Adapt and Recovery; Fig. 2A), and the
subsequent recovery was gradual. The adaptation courses of
express, scanning, delayed and MG saccades (Fig. 2, B–E)
were qualitatively similar but exhibited some quantitative differences. For example, express, scanning, and delayed saccades adapted less than targeting saccades, whereas MG saccades adapted the same amount as targeting saccades but in
fewer trials. Also, scanning saccades showed an initial jump
decrease in gain at the onset of adaptation, whereas the other
types did not.
Figure 2F shows representative courses of adaptation for
each saccade type in monkey W. They were qualitatively
similar to their counterparts from monkey B.
We compared five measures that describe adaptation for the
five different types of saccade. The mean gain decrease at the
250th saccade of adaptation was greater for targeting and MG
saccades than for the other three types in both monkeys
(Fig. 3A). For the average from both monkeys, MG saccades
exhibited a significantly greater reduction than did scanning
saccades (unpaired t-test, P ⬍ 0.05 with Bonferroni correction;
Fig. 3A). In monkey B, the initial jump at the onset of adaptation was larger for scanning saccades than for the other four
types; however, it was the smallest among all test saccades
in monkey W (Fig. 3B). Finally, the jump at the beginning of
recovery was significantly larger for MG saccades than for
delayed, scanning, and express saccades in both monkeys
(Fig. 3C).
To analyze the time course of adaptation for the different
saccade types, we compared the rate constants (1/b in Eq. 1)
and the absolute gain changes (a in Eq. 1) of the exponential
fits of adaptation. Because we did not always collect adaptation
data until the gain change had reached a plateau, rate constants
and the absolute gain changes might be slightly inaccurate in
some experiments. Figure 3D shows the distribution of the
width of the 95% confidence interval for the rate constants of
the exponential fits of all adaptations. Because the width of the
95% confidence interval was ⬍1,000 in most data sets, we
excluded those with larger widths in our consideration of rate
constants and absolute gain changes. The number of experiments remaining after this exclusion was 32, 18, 33, 18, and 27
for targeting, express, scanning, delayed, and MG saccade
adaptation, respectively. Across saccade types, the rate constant was significantly greater for targeting saccades and
smaller for MG saccades than for the other four saccade types
(unpaired t-test, P ⬍ 0.05 with Bonferroni correction; Fig. 3E).
The absolute gain changes for targeting and MG saccades were
larger than those of the other saccade types (Fig. 3F). In
summary, targeting and MG saccades exhibited the greatest
amount of adaptation, but MG saccade adaptation was faster
and showed a larger recovery jump. Other saccade types were
not significantly different from each other.
Transfer of Adaptation from One Saccade Type to Another
Adaptation transfer from targeting to other saccade types.
Figure 4 shows data on the transfer from adapted saccades to
test saccades from representative experiments in monkey B.
Although we tested three types of saccade during every experiment (see MATERIALS AND METHODS), each panel in Fig. 4 shows
the data associated with only the adapted saccade and one type
of test saccade examined three times during the course of
adaptation. For example, in column T in Fig. 4, the plots in
rows E, D, and MG are from the same experiment, i.e., B627.
Because test saccade 1 was MG, 2 was delayed, and 3 was
express (see Fig. 1C), the number of adapt saccades is smallest
for the first-tested MG saccades and largest for the last-tested
express saccades.
As the gain of targeting saccades decreased (Fig. 4, column
T), the gains of all types of test saccades exhibited a gradual
decrease as well, albeit smaller. The mean percent gain decrease of the adapted targeting saccades calculated from data
taken just before and just after the third set of data from the test
saccade (see MATERIALS AND METHODS) ranged from 21.6 ⫾
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
Normalized Saccade Gain
Adapt
129
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
Initial Jump
B
Recovery Jump
0.2
0.2
0.1
0.1
0.1
0
0
0
0.2
0.2
0.2
0.1
0.1
0.1
0
0
0
0.3
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.1
0
T E S D MG
Saccade type
0
T E S D MG
Saccade type
0
∗
20
10
0
0
1,000
3,000
5,000 20,000
Width of 95% confidence interval of rate constant exponential fit
Monkey-B
E
∗
T
E S D MG
Saccade type
F
Monkey-W
∗
500
500
500
0
0
0
0.3
0.3
0.3
0.2
0.2
0.2
0.1
0.1
0.1
0
T E S D MG
Saccade type
0
T E S D MG
Saccade type
0
∗
T
Average for
2 monkeys
∗
∗
E S D MG
Saccade type
Fig. 3. Comparison of adaptation characteristics of the 5 different types of saccade. A: average amount of gain decrease at saccade 250. B: initial gain decrease
following the start of the backward step task. C: initial gain increase at the start of recovery. D: distribution of the width of 95% confidence interval of rate
constants of each exponential fit. E: rate constant of the exponential fit. F: absolute gain change of the exponential fit. Error bars are SD. Open bars show data
from each monkey. Filled bars are averages of the data from both monkeys. T, E, S, D, and MG indicate targeting, express, scanning, delay and memory-guided
saccades, respectively. *Unpaired t-test P ⬍ 0.05 with Bonferroni correction.
7.7% to 27.0 ⫾ 11.3% (Fig. 4, 4 data sets in column T) and was
always highly significant (P ⬍ 0.05, unpaired t-test). In contrast, the mean gain decrease for tested express, scanning,
delayed, and MG saccades was only 16.1 ⫾ 9.0%, 19.5 ⫾
10.3%, 8.4 ⫾ 6.6%, and 13.5 ⫾ 6.6%, respectively (Fig. 4);
however, each decrease was significant (P ⬍ 0.05). The percent gain decreases of all test saccade types were significantly
less than those of the adapted targeting saccade. For these
exemplar experiments, the percent gain transfer from targeting
saccades was 70.9%, 72.2%, 38.9%, and 57.9% to express,
scanning, delayed and MG saccades, respectively.
Figure 5 shows the percent transfer from adapted targeting
saccades to the various test saccades. In column T, each open
bar represents data calculated with Eq. 2 and averaged across
the 10 experiments performed on each monkey and the filled
bar represents the data averaged across both monkeys. Across
the monkeys, the average percent gain transfer from adapted
targeting to express, scanning, delayed, and MG saccades was
73.6 ⫾ 18.1%, 58.1 ⫾ 16.7%, 63.8 ⫾ 20.5%, and 34.6 ⫾
16.7%, respectively, and was significantly different from zero.
Thus adaptation of targeting saccades transferred significantly
to each of the other four types of saccade. The thin gray and
striped bars represent the percent transfer evaluated by modifications 1 and 2 of Eq. 2, respectively (see Data Analysis).
Adaptation transfer from express to other saccade types.
Figure 4, column E, shows data for the transfer of adaptation
from express saccades to the four other saccade types taken
from an exemplar experiment in monkey B. The mean gain
decreases of the adapted express saccades before the third tests
of the different types were ⬃20% and always significant (P ⬍
0.05, unpaired t-test). Although targeting saccades showed a
significant gain decrease, scanning, delayed, and MG saccades
did not. The gain changes of all the test saccades differed
significantly from that of the adapted express saccade; the gain
transfer was 25.6%, 0%, 0%, and 0% for targeting, scanning,
delayed, and MG saccades, respectively.
Across both monkeys (Fig. 5, column E), the average percent gain transfer from adapted express saccades to targeting,
scanning, delayed, and MG saccades was 71.5 ⫾ 22.3%, 22.1 ⫾
19.4%, 23.3 ⫾ 20.6%, and 18.8 ⫾ 19.9%, respectively, but the
transfer was significantly different from zero only for targeting
saccades. Thus adaptation of express saccades transfers only to
targeting saccades.
Adaptation transfer from scanning to other saccade types.
Figure 4, column S, shows data for the transfer of adaptation
from scanning saccades to the four other saccade types taken
from representative experiments in monkey B. The mean gain
decrease of the adapted scanning saccades was always significant. Targeting and express test saccades also showed significant gain decreases, but delayed and MG saccades did not.
The gain decreases of targeting and express saccades were not
significantly different from that of the adapted scanning saccade. Therefore, the percent gain transfer was 100%, 100%,
0%, and 0% for targeting, express, delayed, and MG saccades,
respectively.
Across both monkeys (Fig. 5, column S), the average percent
gain transfer from adapted scanning saccades to targeting,
express, delayed, and MG saccades was 67.1 ⫾ 21.5%, 68.6 ⫾
27.0%, 33.0 ⫾ 22.8%, and 22.8 ⫾ 20.5%, respectively; however, the transfer was significantly different from zero only for
targeting and express saccades. Thus adaptation of scanning
saccades transferred only to targeting and express saccades.
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
C
0.2
30
D
Average for
2 monkeys
Monkey-W
n. experiment
Gain Decrease
@250
Monkey-B
Rate constant
A
Absolute gain change
130
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
131
Adapt saccade
T
E
B629 40
S
∗
B608
D
ns
40
B720 40
MG
∗
B626
40
20
T
1.0
0
0.8
1.0
∗∗
0.8
20
1.0
0
∗ ∗ 0.8
1.0
20
20
0
∗∗
0.8
0
∗
Gain change
E
0
E T
1000
0.6
0
S T
2000
B619
40
0.6
ns
0
D T
500
B720 40
0.8
40
1.0
20
0
M T
500
B723
∗∗
0.8
20
1.0
0
∗ ∗ 0.8
1.0
20
20
0
∗∗
0.6
0.8
0
∗
0.6
0
0.6
T E
∗
1000
B522
B629
40
0.8
0
∗∗
0
S E
2000
0.6
0
∗
D E
500
40
40
B720
0
M E
500
B726
40
1.0
20
1.0
20
1.0
20
20
0
0.8
∗
0 ∗∗
0.8
0.8
0
∗
0.6
0.6
0
∗
T S
1000
B627
0.6
0
0.6
E S
1000
B629
40
1.0
1.0
0.8
0
∗∗
D S
500
0
B628
0 ∗
40
1.0
20
0 ∗
0.8
∗
M S
500
B723
20
1.0
20
0.8
0
40
40
20
D
0
ns
40
1.0
1.0
S
Test saccade
∗
B627
0.6
0.8
0
∗∗
0.6
0.6
0
∗
T D
1000
B627
0.6
0
E D
1000
B629
0.6
0
40
S D
2000
B626
40
B720
40
ns
0
500
M D
Saccade Number
40
1.0
MG
20
0.8
0
∗∗
1.0
20
0.8
0
∗
1.0
20
0.8
0
1.0
20
∗
0.8
0
S M
0.6
∗∗
0.6
0
1000
Saccade Number
T M
0.6
0
1000
Saccade Number
E M
0.6
0
2000
Saccade Number
0
500
Saccade Number
D M
Fig. 4. A representative transfer matrix from adapted saccade to test saccades. Gray dots and lines from the adapted saccade; black dots and lines from the tested
saccade. Label B### indicates the experiment number. Vertical bars indicate mean saccade gain ⫾ SD of the preadaptation data and 1 of the 3 tests during
adaptation. Bar graphs compare the gain change of adapted (gray) and test (black) saccades (see MATERIALS AND METHODS for details). *Unpaired t-test with
P ⬍ 0.05. ns, Unpaired t-test with P ⬎ 0.05.
Adaptation transfer from delayed to other saccade types.
Figure 4, column D, shows data for the transfer of adaptation
from delayed saccades to the four other saccade types taken
from a representative experiment in monkey B. The mean gain
decrease of the adapted delayed saccades was always significant, as was the mean gain decrease of all types of test
saccades. The gain decreases of targeting and scanning saccades were significantly less than those of the adapted delayed
saccade. Therefore, the percent gain transfer was 48%, 100%,
72%, and 100% for targeting, express, scanning, and MG
saccades, respectively.
The average percent gain transfers from adapted delayed
saccades to other test saccades across both monkeys were 68.8 ⫾
20.7%, 80.2 ⫾ 17.5%, 43.6 ⫾ 23.1%, and 38.9 ⫾ 24.1% for
targeting, express, scanning, and MG saccades, respectively
(Fig. 5, column D). The transfer to each type of test saccade
was significantly different from 0. Thus adaptation of delayed
saccades transferred to all types of saccades.
Adaptation transfer from MG to other saccade types.
Figure 4, column MG, shows data for the transfer of adaptation from MG saccades to the four other saccade types
taken from representative experiments in monkey B. The
mean gain decrease of the adapted MG saccades was always
significant. The tested targeting, express, and scanning saccades showed no significant gain decrease, but delayed
saccades did. The gain decrease for each type of test saccade
was significantly different from that of the adapted saccade,
and the percent gain transfer was 0%, 0%, 0%, and 80.1%
for targeting, express, scanning, and delayed saccades,
respectively.
In summary (Fig. 5, column MG), the average percent gain
transfer from adapted MG to targeting, express, scanning, and
delayed saccades was 14.0 ⫾ 15.8%, 15.5 ⫾ 15.9%, 18.8 ⫾
16.4%, and 40.3 ⫾ 24.3%, respectively, but was significantly
different from 0 only for delayed saccades. Thus adaptation of
MG saccades transfers only to delayed saccades.
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
Normalized Saccade Gain
0.6
132
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
Adapt saccade
T
E
S
D
MG
100
T
50
0
∗
∗
∗
∗
∗
% transfer
50
∗
S
0
100
50
D
50
∗
∗
0
100
MG
∗
∗
0
100
W
B W Ba &
vg.
50
0
∗
B W B&W
a vg.
DISCUSSION
B W B&W
a vg.
B W B&W
a vg.
∗
B W B&W
a vg.
The objective of this study was to test the adaptation characteristics of different types of saccade in the monkey and to
determine the degree to which the adaptation of one type
generalized to other types. First, we compare the adaptation
characteristics such as gain decrease, initial and recovery
jumps, and the rate constant for the five types of saccade we
examined. Then we consider the adaptation transfers between
the different types and draw inferences as to the existence of
common and distributed adaptation sites. Finally, we conclude
by comparing our data with similar data collected in nonhuman
primates and humans by our lab and others.
changes represent the slow component, MG saccades might
have adaptation mechanisms with especially robust rapid and
slow components because they exhibited a greater recovery
jump and faster gradual gain change.
The differences in adaptation characteristics suggest that the
neuronal mechanisms that produce adaptation of the various
saccade types could be different. Moreover, the mechanisms
might, at least in part, lie in different pathways within the
saccade circuitry. Whether an adaptation pathway might be
specific to a single saccade type or be shared by several types
can be inferred from the transfer of adaptation among different
saccade types.
Adaptation Characteristics of Different Saccade Types
Adaptation Transfer Between Different Saccade Types
The ISS adaptation paradigm, which has been used to reduce
the amplitude of simian targeting saccades, also produced a
significant adaptation of MG, delayed, scanning, and express
saccades (Fig. 2). However, the adaptation characteristics of
the different saccade types often were not the same. In particular, MG saccades showed the fastest and greatest amount of
adaptation and the greatest recovery jump (Fig. 3). These
results suggest that MG saccades are able to undergo the
greatest adaptation and to switch rapidly back to the unadapted
state when the adaptation paradigm is discontinued.
It has been suggested that saccade adaptation consists of
both a rapid and a slow process (Ethier et al. 2008; Smith et al.
2006; for review see Tian et al. 2009). Smith et al. (2006)
showed that it is the jump in gain after error clamps or a break
that is a marker of the fast process. Unfortunately, our experimental protocol does not allow a direct evaluation of these two
processes. If, however, the recovery jumps in gain (Fig. 4)
reflect the rapid components and the subsequent gradual gain
Adaptation transfer depended strongly on which saccade
type was adapted. In agreement with a previous study on
monkeys (Fuchs et al. 1996), our study shows that the adaptation of targeting saccades transfers to express, scanning,
delayed, and MG saccades (Fig. 6A, column T). In addition, the
present study showed that adapted delayed saccades also transferred to all other types of saccades. Adaptation of express
saccades transferred only to targeting saccades, whereas scanning saccade adaptation transferred to both targeting and express saccades. Adaptation of MG saccades transferred only to
delayed saccades.
Our transfer data suggest that the adaptation sites of some types
of saccade are shared with other types, whereas some sites are not.
Figure 6B shows how these sites might be configured based on
our data. Each saccade type can be adapted, and therefore each
has an adaptation site through which its saccade signal flows.
If adaptation of one kind of saccade transfers to another, the
tested saccade pathway traverses the adaptation site of the
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
Fig. 5. Summary of adaptation transfer among T, E, S,
D, and MG saccades for monkeys B and W. Open bars,
% transfer for each monkey with Eq. 2; filled bars,
average % transfers for both monkeys; gray bars, average % transfers for both monkeys with modification 1
of Eq. 2; striped bars are generated from modification 2
of Eq. 2 (see MATERIALS AND METHODS). Error bars are
SD. *Unpaired t-test with P ⬍ 0.05.
Test saccade
E
100
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
A
T ested Saccade
Human
Adapted Saccade
T
E
S
D
DH
H
AC
DPZ
D
DO
D
FH
H
D
Transferred
Weak transfer
No transfer
Not consistent
Not tested
DH
D
FH
C
B
Saccade
Signal
MG
D
T
E
S
D
FH
O
Monkey
Adaptation
Site
Human
Adapted Saccade
T
E
S
D MG
T
E
S
D
MG
Fig. 6. Summary and interpretation of adaptation transfer. A: summary of
transfer experiments. Area above the diagonal in each square indicates the
monkey result obtained in this study and Fuchs et al. (1996). Area below the
diagonal indicates the human result from Alahyane et al. (2007) (A), Cotti et
al. (2007) (C), Collins and Doré-Mazars (2006) (O), Deubel (1995, 1999) (D),
Erkelens and Hulleman (1993) (E), Fujita et al. (2002) (F), Hopp and Fuchs
(2002, 2010) (H), Panouillères et al. (2014) (P), and Zimmermann and Lappe
(2009) (Z). B: circuit implications based on adaptation transfer data. Arrows
indicate the route for a hypothetical saccade signal that drives each saccade
type. Circles indicate the adaptation site for each saccade type. C: distilled
version of A (see text).
adapted saccade. For example, the adaptation of MG saccades
transferred to delayed saccades, so the delayed saccade pathway traverses the MG adaptation site In Fig 6B. Similar logic
can be applied for the remaining transfers of saccade adaptation. Because adaptation of both targeting and delayed saccades transferred to all other saccade types (Fig. 6A), the
saccade pathway for the other types must traverse the adaptation site for both targeting and delayed saccades, suggesting
that they have a common adaptation site (Fig. 6B). Indeed,
when either targeting (Frens and Van Opstal 1997; cf. Takeichi
et al. 2007) or delayed (Quessy et al. 2010) saccades undergo
adaptation, the neuronal activity in the SC does not change,
indicating that both targeting and delayed saccades use an
adaptation site that is downstream of the SC.
The short-latency express saccade is often considered to be
a reactive saccade like a targeting saccade (for reviews see
Hopp and Fuchs 2004; Pélisson et al. 2010). Therefore, it
might be expected to be one of the last types to join the
common final saccadic pathway. However, because we show
here that scanning saccades transfer to both targeting and
express saccades, targeting and express saccade commands
must traverse the site for scanning saccade adaptation (Fig.
6B). Similarly, delayed saccades are considered to be voluntary
saccades that require cortical processing (for reviews see Hopp
and Fuchs 2004; Pélisson et al. 2010). However, because
delayed saccades transferred to all other saccade types, the site
for delayed saccade adaptation must lie on the saccade command pathway for all other saccade types. Just because delayed
saccades apparently involve cortical processing, their adaptation site does not have to lie in the cortex. On the basis of our
transfer data, it seems plausible that the adaptation sites for
both the reactive targeting saccade and the voluntary delayed
saccade could lie downstream of the adaptation site for the
reactive express saccade.
From this simple schematic, we suggest that the adaptation
sites of targeting and delayed saccades are shared and likely are
located the farthest downstream along the saccade command
pathway. In contrast, the adaptation sites for express, scanning,
and MG saccades are likely to be separate and located more
upstream. Please note that the adaptation sites depicted in the
signal flow schematic of Fig. 6B do not necessarily refer to
specific structures in the brain; they could represent different
groups of cells in the same neural structure or different synapses on the same cell. Moreover, adaptation transfer between
any of the saccade types was never 100%, suggesting that, in
addition to the site indicated in Fig. 6B, each type also may
have a site specific to its own adaptation.
Could Experimental Conditions Account for Any of the
Differences in Transfer Between Saccade Types?
We consider three possible factors here. Although there may
be others, these have been explicitly suggested in the review
process.
In humans, the amount of forward but not backward adaptation transfer between saccade types depends on the preview
duration of the saccade target (Zimmermann 2013). In our
study of backward adaptation, the target preview duration for
scanning saccades was the intersaccadic interval (⬃1 s), for
delayed saccades the delay ⫹ the saccade latency (300 ⫹
⬃150 ⫽ ⬃450 ms), for MG saccades 300 ms, and for targeting
and express saccades the saccade latency (⬃150 and ⬃80 –120
ms, respectively). Although the preview duration for targeting
saccades was twice that for express saccades, they transferred
to each other. Also, although the preview duration for targeting
saccades was less than half that for the delayed saccade, they
still transferred to each other. Thus preview duration also is not
a factor in the transfer of backward adaptation in the monkey.
Does the amount of transfer depend on the velocities of the
different saccade types? Across both monkeys, the average
preadaptation peak velocity was fastest for scanning saccades
(⬃370°/s) and slowest for MG saccades (⬃280°/s). Targeting,
express, and delayed saccades had similar, intermediate peak
velocities (⬃340°/s). If the transfer were dictated by the
similarity of saccade peak velocity, express saccade adaptation
would have transferred to delayed saccades, but it did not (Fig.
6A). Also, if transfer preferentially occurs from faster to slower
saccades (or vice versa), scanning saccade adaptation would
have transferred to MG saccades (or MG saccade adaptation
would have transferred to scanning saccades); again, this was
not the case.
The characteristics of the adaptation process itself also did
not appear to be a factor in the adaptation transfer between
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
MG
D MG
A
CD
DH EPZ
T
E
S
T ested Saccade
Monkey
133
134
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
saccade types. For example, transfer did not always occur
between saccade types with similar adaptation speeds, e.g.,
express to scanning (Fig. 3, A and E). Nor did transfer always
occur from faster to slower adapting saccades, e.g., MG to
express, or from slower to faster adapting saccades, e.g.,
express to MG (Fig. 3).
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
Comparison with Previous Studies
Nonhuman primates. Our lab’s earlier study on the transfer
of targeting saccade adaptation in the monkey also demonstrated substantial, but somewhat greater, transfer to all the
other saccade types tested here (Fuchs et al. 1996). One
possible explanation for the modest quantitative difference
between the studies is that Fuchs et al. (1996) used an adapt
step that was a fixed percentage of the target step, whereas this
study employed an adapt step that was a fixed percentage of the
actual saccade (see MATERIALS AND METHODS). A study in humans suggested that the adaptation process indeed might be
different when elicited by different back step paradigms (Zimmermann and Lappe 2010). A direct comparison of the results
of these adaptation paradigms in the monkey might shed some
light on the adaptation process itself.
Humans. In Fig. 6A, we compare the transfer between pairs
of saccade types in the monkey with the transfer between the
same two pairs in humans. As Fig. 6A shows, there are mostly
similarities but some differences.
Targeting saccade adaptation. In both monkeys and humans, adapted targeting saccades transferred to express saccades (Fig. 6A, column T). Adapted targeting saccades also
transferred to scanning saccades in the monkey, but not consistently in humans. A possible extenuating factor might have
been the different numbers of fixed target spots used to elicit
scanning saccades in the two species. When seven (Alahyane
et al. 2007), six (Deubel 1995, 1999), or four (Zimmermann
and Lappe 2009) spots were used with humans, there was no
transfer; however, when two spots were used with either
humans (Cotti et al. 2007) or monkeys (Fuchs et al. 1996) or
three spots were used with monkeys (this study), transfer
occurred. However, one study on humans (Panouillères et al.
2014) produced no transfer when three spots were used (only
supplementary data).
In the monkey, adaptation of targeting saccades transferred
to both MG and delayed saccades, whereas transfer was inconsistent in the human. For human MG saccades, transfer depended on the delay before the saccade “go” signal. If MG
saccades were executed after delays ⬎1 s (Deubel 1995, 1999),
no transfer occurred; however, after shorter delays transfer did
occur to MG saccades (Fujita et al. 2002; Hopp and Fuchs
2010). Our delay (0.3 s) was similar to that of these last two
human studies, so the transfer to MG saccades was consistent
between the two primate species when similar conditions were
used. For human delayed saccades, targeting saccade adaptation produced no transfer whether the delay was long (⬎1 s;
Deubel 1995, 1999) or short (0.3 ms; Collins and Doré-Mazars
2006).
Express saccade adaptation. Express saccade adaptation
transferred to targeting saccades in both monkeys (Fig. 6A,
column E) and humans (Deubel 1995, 1999; Hopp and Fuchs
2002, 2010). To the best of our knowledge, transfers of express
saccade adaptation to scanning or delayed saccades have not
been investigated in humans. Express saccade adaptation did
not transfer to MG saccades in this study, but there was
significant transfer in humans (Hopp and Fuchs 2010).
There are several differences between this monkey study and
the human studies. In our study, we used shorter gap durations
to elicit express saccades. Second, the distribution of latencies
was bimodal in our study but unimodal in the human study
(Hopp and Fuchs 2010). Therefore, in our study fewer total
saccades were in the latency range of express saccades, so the
average gain change was smaller. However, when we selected
the subset of our experiments with the largest gain changes
[16 ⫾ 3.6% (n ⫽ 3) compared with 14 ⫾ 0.7% in Hopp and
Fuchs 2010], there still was no significant transfer to MG
saccades; therefore, the lower average gain change in monkeys
did not explain the lack of transfer.
Scanning saccade adaptation. Scanning saccade adaptation
transferred to targeting saccades in both monkeys (this study)
and humans (Alahyane et al. 2007; Cotti et al. 2007; Deubel
1995, 1999; Erkelens and Hulleman 1993; Panouillères et al.
2014; Zimmermann and Lappe 2009) (Fig. 6A, column S).
Scanning saccade adaptation also transferred to monkey express saccades, but transfer to human express saccades has not
been tested. Scanning saccade adaptation transferred to neither
delayed nor MG saccades in monkeys and showed only a weak
transfer to those saccade types in humans (Deubel 1995, 1999).
To provide a data set with gain changes more comparable with
those of Deubel (1999), we selected the subset of our experiments with the largest gain changes [12.3 ⫾ 1.3% (n ⫽ 4)
compared with 20.5% in Deubel 1999]; however, there still
was no transfer (0 ⫾ 0%) from scanning to MG saccades.
Delayed saccade adaptation. Delayed saccade adaptation
transferred to all other types in the monkey (Fig. 6A, column
D). In humans, no data are available for the transfer of delayed
saccades to express or scanning saccades. Adapted delayed
saccades transferred to targeting and MG saccades in both
monkeys (this study) and humans (targeting: Collins and DoréMazars 2006; MG: Fujita et al. 2002; Hopp and Fuchs 2010).
MG saccade adaptation. Adaptation of simian MG saccades
transferred only to delayed saccades (Fig. 6A, column MG). In
humans, the transfer of MG saccade adaptation to other types
was rather modest, if present at all. There was either weak
transfer to targeting, express, and delayed saccades (Fujita et
al. 2002; Hopp and Fuchs 2010) or none at all (Deubel 1995,
1999). There was no transfer to scanning saccades in either
monkeys (this study) or humans (Deubel 1995, 1999).
In summary, there is substantial agreement regarding the
transfer of adapted saccades to test saccades across nonhuman
primates and humans (Fig. 6C). 1) Adapted targeting saccades
transferred to express saccades, to MG saccades elicited with
short delays, and to scanning saccades elicited with only a few
target spots in both species and to delayed saccades in monkeys
but not humans. 2) Adapted express saccades transferred to
targeting saccades in both species and to MG saccades in
humans but not monkeys; other saccade types were not tested
in humans. 3) Adapted scanning saccades transferred to targeting saccades in both species and showed little if any transfer to
delayed or MG saccades in either species. In monkeys, there
also was a transfer to express saccades, but this transfer has not
been tested in humans. 4) Adapted delayed saccades transferred to all saccade types in the monkey; transfer to express
and scanning saccades has not been tested in humans. 5)
Adapted MG saccades showed little if any transfer to targeting,
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
express, or scanning saccades but transferred to delayed saccades in both primate species.
Speculation on Possible Neuronal Sites of Adaptation
Conclusions
Before this report, the detailed characteristics of adaptation
in the monkey had been extensively documented only for
targeting saccades. With the help of those behavioral data,
significant progress has been made during the last 20 years in
understanding the neuronal basis of targeting saccade adaptation. In this report we provide comprehensive adaptation data
for four other types of simian saccade. Equipped with this
foundation, we can expect similar progress to be made in
understanding the neuronal basis of adaptation of express,
scanning, delayed, and MG saccades. A logical place to begin
such studies would be the OMV, whose P cells discharge for
all five saccade types.
Our second contribution documents for the first time the
transfer of adaptation between five types of saccade in the
monkey. These data allow us to create a simple hierarchical
model of how adaptation sites might be roughly configured for
the different saccade types. For some of the reasons mentioned
above, it currently is premature to equate the adaptation sites in
Fig. 6B with actual neuronal structures, with the possible
exception of the most downstream site that probably involves
the cerebellum. The rough correspondence of the adaptation
transfer data in monkeys and humans provides some optimism
that the neuronal mechanisms revealed in the monkey may
have relevance when extrapolated to humans.
GRANTS
This study was supported by National Institutes of Health (NIH) Grants
EY-019258 (R. Soetedjo), EY-023277 (Y. Kojima), and EY-00745 (A. F.
Fuchs) and RR-00166 from the National Center for Research Resources
(NCRR), NIH. The contents of this article are solely the responsibility of the
authors and do not necessarily represent the official views of NCRR or NIH.
DISCLOSURES
No conflicts of interest, financial or otherwise, are declared by the author(s).
AUTHOR CONTRIBUTIONS
Author contributions: Y.K. and A.F.F. conception and design of research;
Y.K. performed experiments; Y.K. analyzed data; Y.K. interpreted results of
experiments; Y.K. prepared figures; Y.K. drafted manuscript; Y.K., A.F.F.,
and R.S. edited and revised manuscript; Y.K., A.F.F., and R.S. approved final
version of manuscript.
REFERENCES
Alahyane N, Fonteille V, Urquizar C, Salemme R, Nighoghossian N,
Pelisson D, Tilikete C. Separate neural substrates in the human cerebellum
for sensory-motor adaptation of reactive and of scanning voluntary saccades.
Cerebellum 7: 595– 601, 2008.
Alahyane N, Salemme R, Urquizar C, Cotti J, Guillaume A, Vercher JL,
Pélisson D. Oculomotor plasticity: are mechanisms of adaptation for reactive and voluntary saccades separate? Brain Res 1135: 107–121, 2007.
Barash S, Melikyan A, Sivakov A, Zhang M, Glickstein M, Thier P.
Saccadic dysmetria and adaptation after lesions of the cerebellar cortex. J
Neurosci 19: 10931–10939, 1999.
Catz N, Dicke PW, Thier P. Cerebellar complex spike firing is suitable to
induce as well as to stabilize motor learning. Curr Biol 15: 2179 –2189,
2005.
Catz N, Dicke PW, Thier P. Cerebellar-dependent motor learning is based on
pruning a Purkinje cell population response. Proc Natl Acad Sci USA 105:
7309 –7314, 2008.
Collins T, Doré-Mazars K. Eye movement signals influence perception:
evidence from the adaptation of reactive and volitional saccades. Vision Res
46: 3659 –3673, 2006.
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
Because the generation of different saccade types is thought
to involve different brain regions (for reviews see Gaymard et
al. 1998; Hopp and Fuchs 2004; Lynch and Tian 2006; Pélisson et al. 2010; Pierrot-Deseilligny et al. 2004), it seems
logical to expect that the sites where they undergo adaptation
may also be distributed.
Considerable progress has been made in understanding the
adaptation mechanisms for the extensively documented simian
targeting saccade. Studies using a variety of techniques have
placed the site of adaptation of targeting saccades downstream
of the SC. First, SC saccade-related activity does not change
during adaptation of targeting saccades (Frens and Van Opstal
1997; Takeichi et al. 2007). Second, there is significant adaptation transfer from targeting saccades to saccades evoked by
low-current electrical stimulation of the SC (Edelman and
Goldberg 2002; Melis and van Gisbergen 1996). Finally, stimulation of the SC to act as a surrogate error signal induces
saccade adaptation artificially (Kaku et al. 2009; Soetedjo et al.
2009).
Unit recording together with inactivation, lesion, and anatomical evidence have identified the OMV as an important
component of the downstream adaptation pathway. The gradual change in amplitude that occurs during saccade adaptation
is associated with gradual changes in the discharge patterns of
neurons of the OMV (Catz et al. 2008; Kojima et al. 2010b), its
output, the caudal fastigial nucleus (cFN) (Inaba et al. 2003;
Scudder and McGee 2003), and the brain stem burst generator
to which cFN neurons project (Kojima et al. 2008). Moreover,
the error signal for targeting saccade adaptation reaches the
OMV over climbing fiber inputs to OMV Purkinje (P) cells that
originate in the SC (Catz et al. 2005; Kaku et al. 2009; Soetedjo
et al. 2008, 2009; Soetedjo and Fuchs 2006). However, even
targeting saccade adaptation apparently has a cortical component because it can be influenced or driven by cognitive
processes, such as reinforcement (Madelain et al. 2011), attention (Khan et al. 2014), or the nature of a perceptual task
(Schutz et al. 2014).
There are a few scattered studies concerning possible sites of
adaptation for delayed, MG, and scanning saccades. When
delayed saccades are adapted, SC saccade-related activity also
does not change (Quessy et al. 2010), indicating that the
adaptation site for delayed saccades also is downstream of the
SC. Because OMV P cells discharge for all the saccade types
examined in the present study (Kojima et al. 2010a), it is
possible that the OMV could constitute at least part of the final
common adaptation site of targeting and delayed saccades
hypothesized in Fig. 6B. Neurons in the lateral intraparietal
(LIP) of the simian cortex show no change in activity during
adaptation of either MG or targeting saccades in the monkey
(Steenrod et al. 2013). Therefore, the site of MG saccade
adaptation is likely downstream of the LIP. Finally, human
studies have suggested that the adaptation of scanning saccades
involves some part of the parietal cortex (Cotti et al. 2007,
2009; Gerardin et al. 2012; Panouillères et al. 2014) and the
lateral cerebellum (Alahyane et al. 2008; Panouillères et al.
2012, 2013).
135
136
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
Lynch JC, Tian JR. Cortico-cortical networks and cortico-subcortical loops
for the higher control of eye movements. Prog Brain Res 151: 461–501,
2006.
Madelain L, Paeye C, Wallman J. Modification of saccadic gain by reinforcement. J Neurophysiol 106: 219 –232, 2011.
McLaughlin S. Parametric adjustment in saccadic eye movements. Percept
Psychophys 2: 359 –362, 1967.
Melis BJ, van Gisbergen JA. Short-term adaptation of electrically induced
saccades in monkey superior colliculus. J Neurophysiol 76: 1744 –1758,
1996.
Miller JM, Anstis T, Templeton WB. Saccadic plasticity: parametric adaptive control by retinal feedback. J Exp Psychol Hum Percept Perform 7:
356 –366, 1981.
Noto CT, Watanabe S, Fuchs AF. Characteristics of simian adaptation fields
produced by behavioral changes in saccade size and direction. J Neurophysiol 81: 2798 –2813, 1999.
Panouillères M, Alahyane N, Urquizar C, Salemme R, Nighoghossian N,
Gaymard B, Tilikete C, Pélisson D. Effects of structural and functional
cerebellar lesions on sensorimotor adaptation of saccades. Exp Brain Res
231: 1–11, 2013.
Panouillères M, Habchi O, Gerardin P, Salemme R, Urquizar C, Farne A,
Pélisson D. A role for the parietal cortex in sensorimotor adaptation of
saccades. Cereb Cortex 24: 304 –314, 2014.
Panouillères M, Neggers SF, Gutteling TP, Salemme R, van der Stigchel S,
van der Geest JN, Frens MA, Pélisson D. Transcranial magnetic stimulation and motor plasticity in human lateral cerebellum: dual effect on
saccadic adaptation. Hum Brain Mapp 33: 1512–1525, 2012.
Pélisson D, Alahyane N, Panouillères M, Tilikete C. Sensorimotor adaptation of saccadic eye movements. Neurosci Biobehav Rev 34: 1103–1120,
2010.
Pierrot-Deseilligny C, Milea D, Muri RM. Eye movement control by the
cerebral cortex. Curr Opin Neurol 17: 17–25, 2004.
Quessy S, Quinet J, Freedman EG. The locus of motor activity in the
superior colliculus of the rhesus monkey is unaltered during saccadic
adaptation. J Neurosci 30: 14235–14244, 2010.
Robinson DA. A method of measuring eye movement using a scleral search
coil in a magnetic field. IEEE Trans Biomed Eng 10: 137–145, 1963.
Rolfs M, Knapen T, Cavanagh P. Global saccadic adaptation. Vision Res 50:
1882–1890, 2010.
Schutz AC, Kerzel D, Souto D. Saccadic adaptation induced by a perceptual
task. J Vis 14: 4, 2014.
Scudder CA, Batourina EY, Tunder GS. Comparison of two methods of
producing adaptation of saccade size and implications for the site of
plasticity. J Neurophysiol 79: 704 –715, 1998.
Scudder CA, McGee DM. Adaptive modification of saccade size produces
correlated changes in the discharges of fastigial nucleus neurons. J Neurophysiol 90: 1011–1026, 2003.
Semmlow JL, Gauthier GM, Vercher JL. Mechanisms of short-term saccadic adaptation. J Exp Psychol Hum Percept Perform 15: 249 –258, 1989.
Smith MA, Ghazizadeh A, Shadmehr R. Interacting adaptive processes with
different timescales underlie short-term motor learning. PLoS Biol 4: e179,
2006.
Soetedjo R, Fuchs AF. Complex spike activity of purkinje cells in the
oculomotor vermis during behavioral adaptation of monkey saccades. J
Neurosci 26: 7741–7755, 2006.
Soetedjo R, Fuchs AF, Kojima Y. Subthreshold activation of the superior
colliculus drives saccade motor learning. J Neurosci 29: 15213–15222,
2009.
Soetedjo R, Kojima Y, Fuchs AF. Complex spike activity in the oculomotor
vermis of the cerebellum: a vectorial error signal for saccade motor learning? J Neurophysiol 100: 1949 –1966, 2008.
Steenrod SC, Phillips MH, Goldberg ME. The lateral intraparietal area codes
the location of saccade targets and not the dimension of the saccades that
will be made to acquire them. J Neurophysiol 109: 2596 –2605, 2013.
Straube A, Fuchs AF, Usher S, Robinson FR. Characteristics of saccadic
gain adaptation in rhesus macaques. J Neurophysiol 77: 874 – 895, 1997.
Takagi M, Zee DS, Tamargo RJ. Effects of lesions of the oculomotor vermis
on eye movements in primate: saccades. J Neurophysiol 80: 1911–1931,
1998.
Takeichi N, Kaneko CR, Fuchs AF. Activity changes in monkey superior
colliculus during saccade adaptation. J Neurophysiol 97: 4096 – 4107, 2007.
Tian J, Ethier V, Shadmehr R, Fujita M, Zee DS. Some perspectives on
saccade adaptation. Ann NY Acad Sci 1164: 166 –172, 2009.
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
Cotti J, Guillaume A, Alahyane N, Pelisson D, Vercher JL. Adaptation of
voluntary saccades, but not of reactive saccades, transfers to hand pointing
movements. J Neurophysiol 98: 602– 612, 2007.
Cotti J, Panouilleres M, Munoz DP, Vercher JL, Pélisson D, Guillaume A.
Adaptation of reactive and voluntary saccades: different patterns of adaptation revealed in the antisaccade task. J Physiol 587: 127–138, 2009.
Deubel H. Separate adaptive mechanisms for the control of reactive and
volitional saccadic eye movements. Vision Res 35: 3529 –3540, 1995.
Deubel H. Separate mechanisms for the adaptive control of reactive, volitional, and memory-guided saccadic eye movements. In: Attention and
Performance. XVII. Cognitive Regulation of Performance: Interaction of
Theory and Application, edited by Gopher D, Koriat A. Cambridge, MA:
MIT Press, 1999, p. 697–721.
Deubel H, Wolf W, Hauske G. Adaptive gain control of saccadic eye
movements. Hum Neurobiol 5: 245–253, 1986.
Edelman JA, Goldberg ME. Effect of short-term saccadic adaptation on
saccades evoked by electrical stimulation in the primate superior colliculus.
J Neurophysiol 87: 1915–1923, 2002.
Erkelens CJ, Hulleman J. Selective adaptation of internally triggered saccades made to visual targets. Exp Brain Res 93: 157–164, 1993.
Ethier V, Zee DS, Shadmehr R. Spontaneous recovery of motor memory
during saccade adaptation. J Neurophysiol 99: 2577–2583, 2008.
Fischer B, Boch R. Saccadic eye movements after extremely short reaction
times in the monkey. Brain Res 260: 21–26, 1983.
Fischer B, Ramsperger E. Human express saccades: extremely short reaction
times of goal directed eye movements. Exp Brain Res 57: 191–195, 1984.
Fischer B, Weber H, Biscaldi M, Aiple F, Otto P, Stuhr V. Separate
populations of visually guided saccades in humans: reaction times and
amplitudes. Exp Brain Res 92: 528 –541, 1993.
Frens MA, Van Opstal AJ. Monkey superior colliculus activity during
short-term saccadic adaptation. Brain Res Bull 43: 473– 483, 1997.
Fuchs AF, Reiner D, Pong M. Transfer of gain changes from targeting to
other types of saccade in the monkey: constraints on possible sites of
saccadic gain adaptation. J Neurophysiol 76: 2522–2535, 1996.
Fuchs AF, Robinson DA. A method for measuring horizontal and vertical eye
movement chronically in the monkey. J Appl Physiol 21: 1068 –1070, 1966.
Fujita M, Amagai A, Minakawa F, Aoki M. Selective and delay adaptation
of human saccades. Brain Res Cogn Brain Res 13: 41–52, 2002.
Gaymard B, Ploner CJ, Rivaud S, Vermersch AI, Pierrot-Deseilligny C.
Cortical control of saccades. Exp Brain Res 123: 159 –163, 1998.
Gerardin P, Miquée A, Urquizar C, Pélisson D. Functional activation of the
cerebral cortex related to sensorimotor adaptation of reactive and voluntary
saccades. Neuroimage 61: 1100 –1112, 2012.
Hopp JJ, Fuchs AF. Investigating the site of human saccadic adaptation with
express and targeting saccades. Exp Brain Res 144: 538 –548, 2002.
Hopp JJ, Fuchs AF. The characteristics and neuronal substrate of saccadic
eye movement plasticity. Prog Neurobiol 72: 27–53, 2004.
Hopp JJ, Fuchs AF. Identifying sites of saccade amplitude plasticity in
humans: transfer of adaptation between different types of saccade. Exp
Brain Res 202: 129 –145, 2010.
Inaba N, Iwamoto Y, Yoshida K. Changes in cerebellar fastigial burst
activity related to saccadic gain adaptation in the monkey. Neurosci Res 46:
359 –368, 2003.
Judge SJ, Richmond BJ, Chu FC. Implantation of magnetic search coils for
measurement of eye position: an improved method. Vision Res 20: 535–538,
1980.
Kaku Y, Yoshida K, Iwamoto Y. Learning signals from the superior colliculus for adaptation of saccadic eye movements in the monkey. J Neurosci 29:
5266 –5275, 2009.
Khan A, McFadden SA, Harwood M, Wallman J. Salient distractors can
induce saccade adaptation. J Ophthalmol 2014: 585792, 2014.
Kojima Y, Iwamoto Y, Robinson FR, Noto CT, Yoshida K. Premotor
inhibitory neurons carry signals related to saccade adaptation in the monkey.
J Neurophysiol 99: 220 –230, 2008.
Kojima Y, Soetedjo R, Fuchs AF. Behavior of the oculomotor vermis for five
different types of saccade. J Neurophysiol 104: 3667–3676, 2010a.
Kojima Y, Soetedjo R, Fuchs AF. Changes in simple spike activity of some
Purkinje cells in the oculomotor vermis during saccade adaptation are
appropriate to participate in motor learning. J Neurosci 30: 3715–3727,
2010b.
Kojima Y, Soetedjo R, Fuchs AF. Effect of inactivation and disinhibition of
the oculomotor vermis on saccade adaptation. Brain Res 1401: 30 –39, 2011.
SACCADE ADAPTATION CHARACTERISTICS IN MONKEY
White JM, Sparks DL, Stanford TR. Saccades to remembered target locations: an analysis of systematic and variable errors. Vision Res 34: 79 –92,
1994.
Zimmermann E. The reference frames in saccade adaptation. J Neurophysiol
109: 1815–1823, 2013.
137
Zimmermann E, Lappe M. Mislocalization of flashed and stationary visual
stimuli after adaptation of reactive and scanning saccades. J Neurosci 29:
11055–11064, 2009.
Zimmermann E, Lappe M. Motor signals in visual localization. J Vis 10: 2,
2010.
Downloaded from http://jn.physiology.org/ by 10.220.32.247 on November 2, 2016
J Neurophysiol • doi:10.1152/jn.00212.2015 • www.jn.org