Fig. 3

A subset of RGC inputs to tectum encode gradual changes in display luminance. A Retinal input layers of the tectal neuropil visualized in the left tectum of a 7dpf atoh7:gcamp6s larva. For clarity non-retinal layers are not labeled. B Threshold image of GCaMP6s signal in A. C Image in A with colored overlay denoting active ROIs detected during presentation of 30s ramp stimulus. D Layer distribution of ramp-responsive RGC ROIs. Data from 367 ROIs detected in 7 larvae. One-way ANOVA with Tukey?s multiple comparisons test was performed, p value <0.0001 is for each pairwise comparison between SFGS5/6 and every other group. E Average responses of ON, OFF, and DUAL response classes detected by linear SVM trained on PyrN 30s ramp data. Data from 545 ROIs detected in 6 larvae. F Comparison of response class distribution between PyrNs and RGC inputs. Note greater proportion of ON responsive units among RGC inputs compared to PyrNs. N=30 neurons from 5 larvae, two-way ANOVA, interaction effect p=0.002976, with posthoc unpaired t test. G Comparison of response onset (LUX at half-max) in PyrNs and RGC ROIs during 10s ramp stimulus presentation. Two-way ANOVA, interaction effect p=0.1657, with posthoc unpaired t test. H Comparison of response duration (width at half-max) in PyrNs and RGC ROIs during 10s ramp stimulus presentations. Two-way ANOVA, interaction effect p<0.0001, with posthoc unpaired t test. Note: for t test comparisons in F?H only p values that reached significance are shown. Scale bar: 50?m in A?C