Figure 1.
Texture discrimination chambers.
A, Sandpaper discrimination chamber for paradigm 1. S1 and S2, sandpaper surfaces. LDP and RDP, left and right drink ports. Initial training was performed with rough (P150) sandpaper vs. smooth plastic (pictured). S1 and S2 position was randomly interchanged between trials. B, Smooth-grooved discrimination chamber for paradigm 2. S1 and S2, smooth or grooved surfaces. Left-right position of smooth and grooved surfaces was interchanged randomly between trials.
Figure 2.
Learning curves for sandpaper and smooth vs. grooved discrimination.
A, Sandpaper discrimination. Left, learning curve for 3 example rats. Day is is start of stimulus-reward contingency (Stage 3). Middle, Number of daily training sessions to reach criterion performance, for each successful learner. Right, Fraction of successful learners. B, Smooth-grooved discrimination. Left, learning curve for 2 rats. Rats were initially trained in dim light, and then switched to darkness. Middle, number of training sessions to reach criterion performance. Right, Fraction of successful learners.
Figure 3.
Short-distance (SD) and long-distance (LD) sampling strategies for texture discrimination.
A and B) Examples of SD sampling (A) and LD sampling (B) during smooth-grooved discrimination (Rat G5R2). Video frames were taken from underneath the rat. Dashed line, 2 mm from surface. LD trials were defined as trials in which the nose remained >2 mm from surface during all frames of sampling. C) Example of LD sampling during sandpaper discrimination. Dashed line, 2 mm from surface. D) Fraction of LD trials for rats performing smooth-grooved and sandpaper discrimination. E) Discrimination accuracy during LD and SD trials for smooth-grooved discrimination (left) and sandpaper discrimination (right, compiled for P150 vs. P1500 or P1200 discrimination). Asterisks, significant differences from chance (binomial exact test) or between grouops (two-sample equal proportion test). ns, no significant difference.
Table 1.
Maximal forward extend of whiskers relative to the nose.
Figure 4.
Dependence of SD and LD texture discrimination on microvibrissae vs. macrovibrissae.
A, Performance of rat B2R3 on P120 vs. P1500 sandpaper discrimination during sequential trim of micro- and macrovibrissae. Each point is one behavioral session. Bottom: number of SD, LD and total trials per session. B, Performance of rat G5R1 on smooth-grooved discrimination. There were an average of 23 SD and 25 LD trials per day. Gaps indicate days with too few LD trials to calculate mean accuracy. C, Mean discrimination accuracy for LD trials for all rats in whisker trim experiments. Filled symbols are smooth-grooved discrimination. Open symbols are P120 or P150 (very rough) vs. P800 or P1500 (smooth) sandpaper discrimination. Each point is mean of 6–21 sessions (83–802 LD trials). Error bars are SEM across daily sessions. * and **, p<0.003 and p<2×10−6 relative to chance (binomial exact test). #, p<0.003 difference between groups (2-sample equal proportion test). ns, not significant. D, Mean discrimination accuracy for SD trials for rats performing smooth-grooved discrimination. Each point is the mean of 7–21 sessions (116–385 SD trials). SD trials in sandpaper discrimination were too few to be analyzed.
Figure 5.
Psychometric curve for sandpaper texture discrimination.
A, Performance of Rat C02 discriminating rough P150 sandpaper (base stimulus) from varying smoother (test) sandpapers. Each dot is performance on one daily block (∼25 trials). Open circles, mean accuracy across daily sessions ± SEM. Bars, cumulative accuracy (total correct trials/total trials). B, Mean performance of 3 rats for P150 base discrimination, and for one rat using P120 base. Bars are SEM across daily blocks (n = 4–15 blocks per data point). The x-axis is the difference in mean grit size between the test and base sandpapers. Test sandpaper identity is marked above each point. C, Effect of whisker trimming on difficult (P150 vs. P280) and easy (P150 vs. P1500 or P800) discriminations, for rat H02. D, High-acuity discrimination of a smooth P1500 sandpaper (base) from varying rougher (test) sandpapers. P1500 was the rewarded S+ stimulus.
Figure 6.
Olfactory cues do not mediate texture discrimination.
A, Cleaning smooth-grooved surfaces with ethanol between trials (“EtOH wipe trials”) did not reduce discrimination performance relative to interleaved “non-wipe” trials. B, Exchanging P150 and P800 sandpaper samples that were used daily with new samples of each sandpaper did not alter discrimination accuracy. C, Reversing the sandpapers so that the paper backing, rather than the grit surface, faced the rat abolished discrimination. **, discrimination significantly greater than chance (p<0.0003). #, significantly less than the forward condition (p<0.04).