Fig. 2
From: Regulation of mouse exploratory behaviour by irradiance and cone-opponent signals
Cone-opponent colour signals modulate light avoidance behaviour in mice. (A) Spectral power distributions of test stimuli matched in irradiance for melanopsin and rods but differed in irradiance for S- and/or L-cone opsin. B, C Mean ± SEM preference of Opn1mwR (B; n = 7), and Cnga3−/− mice (C, n = 13) to L- and/or S-cone ‘bright’ (○) vs. ‘dim’ (□) lighting conditions. Data analysed by one-way RM ANOVA (B: F2, 18 = 9.87, P = 0.001; C: F2, 36 = 0.53, P = 0.595) with one sample t-tests vs. null hypothesis of 0 preference as appropriate. (D) Mean ± SEM preference of Opn1mwR mice (n = 8) to stimuli represented by (○) as a function of irradiance difference relative to opposing side of chamber (□), where irradiance differed only for L- (left) or S-cone opsin (right). Data analysed by two-way RM ANOVA (opsin—F1, 14 = 0.947, P = 0.347; irradiance—F2, 28 = 0.043, P = 0.959; interaction—F2, 28 = 5.69, P = 0.009). Linear fits tested for non-zero slope (Left- F1,22 = 4.96, P = 0.037; right- F1,22 = 5.51, P = 0.028) and for difference in slopes (left vs. right—F2, 44 = 5.84, P = 0.006) via F-test. E Mean ± SEM preference of Opn1mw.R mice (n = 8) to stimuli represented by (○), which always presented a 1 log unit difference in irradiance for L-cones compared to the opposing side of chamber (□) but a variable difference in irradiance for S-cones. Data analysed by one-way RM ANOVA (F1.393, 9.751 = 5.28, P = 0.037). Linear fit tested for non-zero slope via F-test (F1, 22 = 13.33, P = 0.001)