Fig. 9. Rod pathway is involved in the nonlinearities of chromatic integration.
a Chromatic-integration curves of sample linear, nonlinear Off, nonlinear On, and nonlinear On–Off cells for full-field stimulation during mesopic and photopic light levels. b Distributions of chromatic nonlinearity indices during mesopic and photopic brightness levels. Indices of linear cells remained unchanged but were reduced for all nonlinear cell classes (linear cells: p = 0.02, n = 75; nonlinear Off cells: p = 4.9 × 10−3, n = 18; nonlinear On cells: p = 3.9 × 10−3, n = 10; nonlinear On–Off cells: p = 1.1 × 10−4, n = 29). c Same as a, but (instead of the increased light level) for application of the pH buffer HEPES, which blocks the influence of the horizontal cells in the retina. Shaded regions around the curves show mean ± SEM. d Same as b, but before and during application of HEPES at the same (mesopic) light level. Nonlinear Off and On–Off cells displayed a significant effect (nonlinear Off cells: p = 1.2 × 10−3, n = 31; nonlinear On–Off cells: p = 1.7 × 10−6, n = 50) while nonlinear On and linear cells did not change their chromatic-integration properties (nonlinear On cells: p = 0.15, n = 8; linear cells: p = 0.21, n = 156). All statistical comparisons are based on the two-sided Wilcoxon signed-rank test (statistics summary: ***p < 0.001, *p < 0.05, n.s. not significant). For all boxplots, the central line and the box mark the median and the interquartile range (IQR) from the first to the third quartile, respectively, and whiskers extend to the maximum and minimum within the central range of 1.5 × IQR. The original data points are displayed next to each boxplot.