Mouse hue and wavelength-specific luminance contrast sensitivity are non-uniform across visual space

Mammalian visual behaviors, as well as responses in the neural systems thought to underlie these behaviors, are driven by luminance and hue contrast. With tools for measuring activity in cell-type specific populations in the mouse during visual behavior gaining traction, it is important to define the extent of luminance and hue information that is behaviorally-accessible to the mouse. A non-uniform distribution of cone opsins in the mouse potentially complicates both luminance and hue sensitivity: opposing gradients of short (UV-shifted) and middle (blue/green) cone opsins suggest that hue discrimination and wavelength-specific luminance contrast sensitivity may differ depending on retinotopic location. Here we ask if, and how well, mice can discriminate color and wavelength-specific luminance across visuotopic space. We found that mice were able to discriminate hue, and were able to do so more broadly across visuotopic space than expected from the cone-opsin distribution. We also found wavelength-band specific differences in luminance sensitivity.


Introduction 22
The mouse visual system is increasingly 1,2 being used as a model system for studying both 23 cortical sensory processing 3-7 and behavior [8][9][10][11][12] . While most physiological work has used achromatic 24 stimuli 3,13 , mice, like most other mammals, display physiological color-opponent signals in the retina 14-25 18 , through LGN 19 and possibly V1 20 . The mouse retina displays asymmetric and mixed expression of 26 its two opsins along the dorsal-ventral axis of the retina, creating opposing gradients of short and 27 middle opsins 21,22 and resulting in gradients of wavelength-band specific responses 16,19,23 . Therefore 28 the substrate for cone-driven color-opponent signals, and any hue sensitivity, exists only in the 29 overlapping "opsin transition zone" 15,19 . However, short and middle opsin responses broadly overlap in 30 V1 and higher visual areas 23 and rod-cone antagonism can also create color opponency in some 31 mouse retinal ganglion cells 17 , presenting the possibility that behaviorally-relevant color information 32 could be extracted more broadly across retinotopic space. 33 Whether mice can use hue information to guide visual behavior is an open question. There is 34 some evidence for hue discrimination 24 , but it remains unclear how this depends on overall luminance, 35 luminance contrast, or retinotopic position. Further, it is not known if the gradients in opsin distribution 36 lead to variations in behavioral luminance sensitivity across space. Such non-uniformity would impact 37 studies of visuotopically extended V1 populations, such as studies of population sparsity 25 , population 38 correlations 26 and other notions of population coding 10,27 . 39 Here we use a simple behavior, change detection, to determine where in visual space mice can 40 discriminate changes in hue and luminance at ethologically-relevant (i.e, mesopic) luminance levels. By 41 measuring detectability of luminance and color changes separately across elevation (spanning ~75º), 42 we are able to generate an estimate of wavelength-specific contrast sensitivity across visual space. 43 Mice were able to discriminate hue, but only at elevations above the horizon. We find both wavelength-44 specific luminance and hue contrast sensitivity to be dependent on retinotopic location, but that these 45 differences in sensitivity were less dramatic than expected from the cone opsin distribution, suggesting 46 behavioral access to differential activation of rods and cones. 47 48 Results 49

Behavioral Task 50
To examine the psychophysical and physiological basis of mouse color vision, we first trained 51 mice in a go/no-go change detection task 8 in an immersive visual stimulation environment customized 52 for delivering stimuli in the spectral bands of the mouse short and middle wavelength opsins (Fig 1A;  53 Materials and Methods). We use the system here to deliver a video stimulus driven by a green and 54 ultraviolet LED projector; for each point on the stimulus the green and ultraviolet intensity could be 55 independently modulated. Total luminance was in the mesopic range, over which mice are both 56 behaviorally active 28 and color opponent signals have been demonstrated in the retina 15,17 . Under this 57 paradigm, mice indicate that they have perceived a change in the stimulus by licking a reward spout 58 within 1 second of the change ( Figure 1B); subsequent licks allow reward consumption (Figure1C). 59 Following pre-training on the change detection task (see Materials and Methods), we switched to 60 change detection sessions in which the ultraviolet and green intensity, centered on the mouse short 61 and middle wavelength bands, respectively, were varied independently on each trial. Each trial 62 contained a change in intensity for a 15º test circle on a mean luminance background at one of four 63 elevations: -10º, 10º, 30º, and in some cases 50º (relative to both the horizon and the placement of the 64 rotating mouse platform). We varied position only along elevation because both rods and cones are 65 relatively uniform across the azimuthal axis of the retina 29 . Eye position did not change with stimulus 66 location (Figure 1 -figure supplement 1). To achieve sufficient trials to cover this stimulus space, we 67 presented a total of 127659 trials (n=4/5 total mice trained, 284 sessions). To control for motivation, we 68 calculated a running average of the reward rate and selected trials where this reward rate remained 69 above 4 rewards per minute; only these engaged trials (44%, 56112/127659) were used for analysis. 70 71 Figure 1. Change detection task in an immersive visual stimulation environment capable of delivering short-and middle-wavelength band stimulation. A, The visual stimulation environment, with the positions and size of stimuli shown; the colored edges were not a part of the presented stimulus, but indicate the color scheme used to denote elevation throughout the other figures. B, A schematic of the task. The background was set mean intensity for each wavelength band. At variable times, t n , the intensity of short and middle wavelength bands within a 15º diameter circle changed. If the mouse licked within 1 second of this change (indicate by the dark grey boxes), a reward was delivered. Schematic short and middle band intensities are shown on the lower plot and corresponding stimulus changes for each epoch are schematized in blocks and projected as circles onto a sphere. The schematized circles are larger than actual stimuli, for clarity. C, Example performance in a single session across 250 trials. Each lick is shown relative to stimulus change; the response window is overlaid in grey. A histogram of lick times is shown above. Error bars are S.E.M. D, The distribution of change times (t in panel B) across the trials used for analysis of change detection performance. The distribution follows the log sampling distribution, enforcing roughly equal probability of a change occurring as the mouse continues to wait. We first examined our results to estimate the relative luminance contrast sensitivities to short 77 and middle-wavelength band stimulation across the visual field. Although the green and ultraviolet 78 projector LEDs nearly isolate responses of the middle and short wavelength sensitive opsins 30 , they do 79 not necessarily isolate responses of individual cones, most of which express a combination of the two 80 opsins. Nor is it necessarily a measurement of the relative weight of the cone opsins themselves, as 81 rods also may contribute to this light sensitivity at these luminance levels. Rather, we present a 82 measure of the relative perceptual weight to stimuli of the middle and short wavelength bands covered 83 by our stimulus LEDs (Figure 2 -figure supplement 1), as combined through both cone opsins and 84 rods. 85 Total luminance change detection saturated by ~30% at all elevations (Fig 2A) and the half-saturation 86 threshold (hereafter referred to as "threshold") < 12% for each elevation ( Fig 2B). The highest 87 sensitivity was in the upper visual field, at 5.5% threshold, a threshold that is consistent with previous 88 reports for a 15º (~0.07 cyc/º) stimulus 8,31-33 . Sensitivity to increments in contrast was similar for all 89 elevations (10.5-12.1%). Consistent with previous physiological measurements in V1 of mouse 20 and 90 other species 34,35 , sensitivity to decrements in contrast was higher than sensitivity to increments (5.4% -91 10.4%). Notably, this was most pronounced in the upper visual field (difference: 5%) than the lower 92 visual field (difference: 1.3%). 93 To determine the independent contributions of short and middle wavelength bands we 94 examined change trials that contained increments or decrements of only one of the two LEDs. For the 95 short wavelength band (i.e. UV), contrast sensitivity was non-uniform, with the highest sensitivity in the 96 upper visual field (Fig 2C,D; 8% threshold). As with total luminance, mice were more sensitive to 97 decrements than increments in contrast. The non-uniformity across elevation was more pronounced for 98 short-wavelength specific sensitivity than total luminance, but was restricted to decrements. The 99 middle-wavelength (i.e. blue/green) luminance contrast sensitivity was also non-uniform, across 100 elevation, but with the opposite relationship as the short wavelength and total luminance. The middle-101 wavelength (i.e. green) luminance contrast 102 sensitivity was nearly identical across the 103 tested positions (Fig 2E-F). Middle-104 wavelength sensitivity was very similar for 105 increments and decrements, again in 106 contrast to total and short-wavelength 107 luminance contrast sensitivity. In summary, 108 we found that luminance contrast sensitivity 109 was non-uniform, with significant opposing 110 wavelength-band specific non-uniformities, 111 though less than what would be predicted 112 from the opsin expression or photoreceptor 113 response alone. were successful, we expected the axis of opposition to closely match unity. 147 We measured the axis along which opposite changes in short 148 and middle wavelength stimulation effectively canceled each 149 other at several elevations. The performance across pairwise 150 combinations of changes in short and middle band luminance 151 ( Figure 3B) was fit with an ellipse ( Figure 3C, top), and the 152 major axis of this ellipse was taken to be the equiluminance 153 line ( Figure 3C, bottom). We found the mouse to be more 154 sensitive to middle band stimulation than expected at all 155 elevations tested, including at 30º where given the eye 156 positioning (Figure 1 -figure supplement 1) short-opsin 157 expression dominates. In fact, surprisingly, the mice were 158 more sensitive to middle than short-band changes at all 159 elevations, with the following middle/short ratios: 3.4, 3.6, and 160 2.25 at -10º, 10º, and 30º, respectively. 10º, and 30º, respectively; both predict far less middle-band 173 sensitivity that we observed. This result suggests that rod 174 opsin sensitivity contributes significantly to mouse perceptual 175 sensitivity at these light levels, at least as much as 60% 176 ( We continued to ask if mice could report a change in 180 hue independent of any luminance change. Instead of explicitly 181 creating a device that normalized total luminance during hue 182 changes 38-40 , we presented sufficient combinations of 183 wavelength-band specific luminance changes to experimentally 184 determine when hue changes occurred independent of 185 luminance changes. We start with the assumption that no 186 change in luminance or hue contrast is not discriminable to the 187 observer. All luminance and hue contrast changes that are 188 behaviorally indistinguishable from this "no change" condition form an ellipse of non-discriminability in 189 hue-luminance space analogous to a MacAdam ellipse of non-discriminability in color space 38 . As noted 190 above, the axes of this ellipse that more closely matches the stimulus "luminance-balanced" line 191 ( Figure 3A, unity line) specifies this equiluminance line for the wavelength band sensitivities at that 192 elevation; because mice are more sensitive to luminance that hue contrast, this was always the major 193 axis of the ellipse ( Figure 3A, purple ellipses). In an HSL color model, this line is equivalent to that of 194  . Relative short and middle-wavelength band weights and hue discrimination across elevation. A, Schematic representation of two possible relative short and middle band weight scenarios, with putative hue contrast sensitivity. The left scenario shows equal weight of short and middle bands, leading in balanced sensitivity and chance performance to equal and opposite changes in short and middle band contrast, along the unity line. The right scenario shows higher middle band sensitivity, resulting in a positive shift of the slope, where the larger changes in short band contrast are required to balance changes in middle band contrast. For both scenarios low luminance and hue contrast should yield to low change detection performance (purple ellipses). If hue discrimination occurs along the major axis of this purple ellipse, and should be high near the edges (orange ellipses). B, Observed relative weights of short and middle band weights, as in panel A., at elevations of -10º (green outline, bottom), 10º (light blue outline, middle), and 30º (dark blue outline, top). C, Fit of the relative weights in panel B with two-dimensional Gaussians, including a plot of the major axis of the fit. These major axes are isolated in the plot below. Color corresponds to stimulus elevation (green lines: -10º, light blue lines: 10º, dark blue lines: 30º). D, schematic of sensitivity testing after attempted compensation for relative short and middle band weight. The central region should be along the line of equiluminance, so subsequent testing was focused there (arrow to panel E). E., Performance of change detection at four elevations (green lines: -10º, light blue lines: 10º, dark blue lines: 30º, pink lines: 50º) after attempted balancing of short and middle weights. Because short and middle weight were not exactly balanced, performance was fit with a two-dimensional Gaussian and hue sensitivity measured along the major axis. F, Hue sensitivity at each elevation. Fit with hyperbolic ration function is shown overlaid on mean performance; mean performance line thickness shows S.E.M. across mice. The stimulus is schematized above the performance, showing the corresponding equal and opposite relative change in each wavelength band for each condition. G, Contrast sensitivity at each elevation, from fits in panel F. across space, this equiluminance line changes with elevation (Fig 3C), so we attempted to create 197 conditions of uniform luminance contrast across elevation by adjusting for the relative short and middle-198 band sensitivities ( Figure 3D). By examining change detection performance along this experimentally 199 defined axis of hue change, we can ask if mice can discriminate hue independent of luminance ( Figure  200 3A, orange ellipses). 201 We found hue discrimination to depend on elevation. Examining the luminance-adjusted data, 202 we found hue discrimination was negligible at -10º, but mice were capable of varying levels of hue 203 sensitivity at all other elevations tested (Fig 3F). The performance along the equiluminant line at -10º 204 was not well fit by a hyperbolic ratio function, and the performance in catch trials (0% contrast change) 205 was not significantly different from any point along the line (p > .05, student's t-test). We were able to fit 206 the performance at each of the other elevations tested, up to 50º above the horizon. Hue contrast 207 sensitivity was highest at for decrements in short-middle opponency at 10º elevation (13.4%); hue 208 sensitivity was nearly identical for decrements in short-middle opponent contrast at 30 and 50º and for 209 increments at all elevations above the horizon (29.3 -32.1%).  Figure 2 -figure supplement 1. B, normalized difference between the predicted (cone opsin expression, cone opsin functional response, V1 response) and observed behavioral weights, for -10º, 10, and 30º. The difference between the behavior middle/short ratio at each elevation and the middle/short ratios at those elevation predicted by several measures of cone opsin weight in the literature was divided by the observed ratio to estimate the proportion of the middle band weight provided by rods under our behavioral luminance conditions

Discussion 212
The finding that both wavelength-specific luminance ( Figure 2) and hue contrast ( Figure 3E-F) is 213 not uniform is in accordance with the distributions of both retinal and primary visual cortical 23 214 responses; however, we found that middle-band sensitivity was both higher and more uniform than 215 expected (Figure 2). This suggests that rod sensitivity contributes significantly to perceptual sensitivity 216 at these light levels (Figure 3 -figure supplement 1). 217 This finding may be important for studies of the mouse visual system that use visuotopically 218 extended stimulation 10,25,42-44 , especially those that measure the underlying population representation of 219 the stimuli. Because the spatial scale of luminance and contrast adaptation can be large 45 , the 220 adaptation to large single-band stimuli (such as those produced by LCD or other sRGB displays) in 221 these studies may underestimate the contrast sensitivity for cells in upper visual field. This spatial scale 222 is especially relevant because of the scale of mouse vision -50% differences can be seen across a 223 small number (~5) receptive field diameters. 224 Our results also demonstrate that hue sensitivity depends on retinotopy, and that some 225 retinotopic locations appear to not support hue discrimination. A goal of many large-scale data 226 collection efforts, both completed 18 and underway (brain-map.org/visualcoding) as well as smaller-227 scale surveys 3,13,46,47 from retina to V1 is the classification or clustering of response properties in order 228 to define functional channels. Because color-opponent cells, both single and double 48 , are thought to 229 underlie such behavior, our findings indicate that mice may have at least one, likely at least two, color-230 opponent cell types; the presence of such functional cell types may depend strongly on retinotopy. 231 Notably, our animals are housed in an environment with fluorescent lighting that does not provide UV-B 232 for reflection (Figure 2 -figure supplement 1), suggesting that the behavior we observed is 233 developmentally specified, not learned, and not lost through lack of use. with the hypothesis 15 that non-uniform wavelength-specific sensitivity is matched to the luminance 239 statistics of natural scenes (UV in the upper visual field, green in the lower), and high sensitivity to 240 decrements in the short wavelengths may be particularly helpful during the shift towards UV in the 241 spectral radiance distribution the during twilight hours 50 . Another hypothesis, that short-middle 242 opponency is useful for identifying mouse urine posts 17 , is inconsistent with our demonstration of a lack 243 of hue discrimination at -10º, at least in absence of significant excursions in eye position or head 244 movements. Our results suggest that, under the mesopic condition during which mice are often active, 245 hue signals may be mediated by rod-cone opponency, and this may facilitate the specialization of cone 246 opsin distributions for sampling natural luminance statistics 15,51 . 247 248

Methods 249
All procedures are approved by the Allen Institute for Brain Science Institutional Animal Care and Use 250 Committee. 251

Animals and Surgical Preparation 252
All animals used in this study (n=5) were C57Bl/6J male mice aged 30-300 days obtained from 253 The Jackson Laboratories. To fix the animal's head within the behavioral apparatus, a single surgery to 254 permanently attach a headpost was performed. During this surgery, the animal was deeply 255 anesthetized with 5% isoflurane and anesthesia maintained throughout the surgery with 1.5-2% 256 continuous inhaled isoflurane. The mouse was secured in a strereotax with ear bars; hair was removed 257 and the exposed skin sterilized with three rounds of betadine. An anterior-posterior incision was made 258 in the skin from anterior of the eyes to posterior of the ears. The skin was removed in a tear drop shape 259 exposing the skull. The skull was leveled and the headpost was placed using a custom stereotaxic 260 headpost placement jig. A custom 11 mm diameter metal headpost with mounting wings was affixed to 261 the skull using dental cement. The exposed skull inside the headpost was covered with a thin protective 262 layer of clear dental cement and further covered with Kwikcast. The animal was allowed to recover for 263 at least 5 days prior to the initiation of behavioral training. 264 After headpost implantation animals were kept on a reverse light cycle (lights OFF from 9AM to 265 9PM) and behavioral testing was done between 9AM and 1PM. Mice were habituated to handling 266 gradually, through sessions of increasing duration. Mice were also habituated to the behavioral 267 apparatus, first by allowing periods of free exploration and subsequently with head fixation sessions 268 increasing from 10 minutes to 1 hour over the course of 1 week. Water restriction began with 269 habituation; all mice were maintained at 85% of the original body weight for the duration of training and 270 testing. 271

Stimulus Environment and Stimuli 273
Ultraviolet and human-visible stimuli were provided across a range of retinotopic locations using 274 a custom spherical stimulus enclosure 19 ( Figure 1A, Fig 2-figure supplement 1). A custom DLP-275 projector designed for the mouse visual system provided independent spatiotemporal modulation of 276 ultraviolet (peak 380nm, Figure 1B) and green (peak 532nm, Fig 1B) light. The projection system 277 operated at 1024x768 pixel resolution and a refresh rate of 60Hz, achieving a maximum intensity of ~3 278 cd/m2. Planar stimuli were spatially warped according to a custom fisheye warp for presentation on a 279 curved screen; the fisheye warp was created through an iterative mapping protocol using the 280 meshmapper utility (http://paulbourke.net/dome/meshmapper/) calibrated on the behavioral 281 environment to achieve maximal accuracy. 282 Stimuli were presented in the right visual field and consisted of 15º diameter circles of varying 283 color on a mean intensity background using custom written software extensions of the PsychoPy 284 package (http://www.psychopy.org). The background intensity was 1.52 cd/m2. For some testing 285 sessions, the color of the display was adjusted to match the mouse's spectral sensitivity in order to 286 create uniform and balanced sensitivity to the projector's LED sources across the visual stimulus 287 enclosure. To do so, a custom warp was applied that included a spatially dependent adjustment of the 288 intensity of each LED (near-UV and 'green'), according to the results shown in Figure 3C. 289 Animals were head-fixed on a freely rotating disc in the center of the spherical enclosure and 290 allowed to run freely during the course of training and testing. A lick spout was positioned 291 approximately 0.5 cm in front of the mouse within range of tongue extension. 292 In some experiments, infrared short-pass dichroic mirrors (750nm short-pass filter, Edmund 293 Optics) were placed in front of each eye to allow for video tracking of the pupil. Cameras (Mako and 294 Manta, AVT technologies) placed behind the animal were aligned to record a reflected image of the 295 pupil; infrared illumination and a reference corneal reflection was provided via an LED positioned near 296 the camera. Movies of the eye position during presentation of the stimuli used in the task was acquired 297 at >=60Hz, with the eye occupying >60% of the image at 300 x 300. Data from these sessions were not 298 included in the performance analysis to avoid any potential artifact caused by the infrared dichroic. 299 300

Behavioral Task 301
Animals were first shaped to associate changes in luminance with a reward. After each change 302 in luminance a water reward was automatically delivered, regardless of mouse licking behavior. During 303 these sessions, the reward was constant at 10µL. Incorrect licks were punished by resetting the trial, 304 such that the mouse had to wait longer for the next change. This "shaping" phase lasted a minimum of 305 two days, but for most mice extended to several weeks. For some animals (2/5), subsequent epochs of 306 this automatic reward shaping served as task reinforcement when performance in testing blocks 307 dropped. 308 During each testing session a circle was presented at a single visuotopic location and remained 309 at this location for the duration of the session. At non-regular intervals, again selected from an 310 exponential distribution, the color and/or luminance of the stimulus was changed, and the mouse had to 311 report detection of change by licking the reward spout within 1 second of the change in order to receive 312 reward ( Figure 1C). Licks were detected through a capacitive sensor connected to the reward spout.

313
No water was present on the reward spout before the first lick; if the animal correctly detected a 314 change, a water reward (3-10µL, depending on animal and stage of training) was delivered through this 315 spout ( Figure 1D). Sessions were 50-60 minutes and typically included ~300 trials. 316 Mice were first trained to associate changes in a 15º stimulus at 100% luminance contrast with 317 a reward. In these sessions (total of 3 to 25 sessions), the contrast of a stimulus (10º elevation, relative 318 to the horizon) changed at exponentially distributed intervals from 50% positive relative to the 319 background to 50% negative (from white to black), or vice versa (black to white). If a lick occurred 320 within 1 second of an actual stimulus change ( Figure 1B), a reward was delivered to the spout and 321 liquid reward was consumed subsequent licks (Figure1C). If a lick occurred outside of this window the 322 trial was aborted, extending the time the mouse must wait and effectively creating a 'time-out' period. 323 Mice advanced from this protocol after performance exceeded 75% for consecutive sessions. 324 In subsequent testing sessions, the intensity of the ultraviolet and green intensities were varied 325 independently on each trial. Each trial contained a change in LED intensity for a 15º test circle on a 326 mean luminance background at one of four elevations: -10º, 10º, 30º, and in some cases 50º. The first 327 8-20 trials of each session were 100% contrast changes, as described for the training blocks, with 328 rewards automatically delivered. The number of these daily "free" rewards was reduced to 8 for as long 329 as the mouse received >1.0mL of reward during training or performed well enough to reach satiety and 330 disengage from the task. We attempted to correct for sessions with poor performance by increasing 331 these "free" rewards on subsequent days before gradually reducing them again. To control for 332 motivation in the results, we calculated a running average of the reward rate and selected trials where 333 this reward rate remained above 4 rewards per minute; only these engaged trials (44%, 56112/127659) 334 were used for analysis. 335 336 Analysis 337 All analyses were done using Python and common scientific packages (numpy, scipy, 338 matplotllib, and pandas). Code is publicly available from github.com/danieljdenman/mouse_chromatic 339 and includes a Jupyter notebook that contains code for generation of our figures from the data. Data 340 from each training session was saved and combined into a common data structure that was used for all 341 analysis. Individual sessions were analyzed to drive adjustments in the training parameters such as the 342 number of automatic "free" rewards. Following data collection for all animals, all sessions were loaded 343 into a single object for analysis. This data structure can be recreated from the NWB 52 files made 344 available from <github.com/danieljdenman/mouse_chromatic.>. 345 To quantify performance, from each trial the following parameters were extracted: change times 346 (the time of stimulus change), lick times (the time of each lick, as detected through the capacitive 347 sensor connected to the reward spout), and the stimulus conditions. A trial was scored "correct" if the 348 first lick after a change time occurred with one second, and if there was actually change in intensity of 349 either green or ultraviolet at that change time. 350 For each mouse, the percent correct was computed for each pair of LED state transitions, i.e., 351 each pairwise combination of change in short-band luminance and change in middle-band luminance 352 (e.g., Fig 3C). For each mouse, performance was ignored if 3 trials were not presented for those 353 conditions. For fitting, missing data were replaced via a nearest neighbors approach, with the mean of 354 the surrounding data. Our sampling strategies focused on the areas of changing performance, ensuring 355 that cases of missing data were limited to the areas where performance had saturated at or near the 356 lapse rate. Psychophysical curves for wavelength band-specific and hue sensitivity were taken from the 357 appropriate slices of this color space. Sensitivity was taken from the c50 parameter of fit a hyperbolic 358 ratio fit 53 . 359 A total of 5 mice entered training on the task; one mouse failed to reach consecutive sessions of 360 75% performance during the initial high luminance contrast change detection phase, and so did not 361 continue to testing in the hue contrast discrimination phase. We did not use any statistical methods to 362 determine mouse or trial sample size prior to the study, determining based on stability and consistency 363 of results when sufficient samples had been collected. Statistical tests were student's t-test unless 364 otherwise specified. 365 Eyetracking analysis was done via a semi-automated algorithm; full details are available from 366 <http://help.brain-map.org/display/observatory/Documentation>. Briefly, the algorithm fits an ellipse to 367 the pupil or corneal reflection (CR) area, respectively. A seed point is identified by convolution with a 368 black square (for the pupil) and white square (for the corneal reflection). An ellipse was fit to candidate 369 boundary points identified using ray tracing using a random sample consensus algorithm. The fit 370 parameters were first reported in coordinate centered on the mouse eye, and subsequently converted 371 to visual degrees by projection based on the position of the dichroic mirror and the relative position of 372 pupil and corneal reflection. Coordinates for eye position were extracted independently for each frame 373 of the eye position movie. 374