Fig. 4: Effect of attention on color information.
From: Neuronal correlates of endogenous selective attention in the endbrain of crows
A Behavioral protocol of no-cue trials, same as in Fig. 1A, the only difference being the lack of a cue phase. Due to a lack of a cue the birds had to maintain color information about all three locations. In case none of the colors changed between the sample and comparison phases, the animal had to choose the center location. B Dot-raster plots and spike-density functions of an example neuron. The example neuron shows differentiation between colors at the top location. The presence of the cue increased the differentiation between colors. Top: neuronal activity in trials without cue (ncolor1 = 82, ncolor2 = 67). Bottom: neuronal activity of the same neuron in trials when the pre-cue had appeared at the top location (ncolor1 = 59, ncolor2 = 54). Purple shading indicates a significant difference in firing rate. C Effect of attention on color information. When birds attended a cued location, they had more information about color at that location. The attentional shift was quantified as the absolute average of cue dependent gain of color information (Δ PEV, [pre-cue minus no-cue]) and loss of color information (Δ PEV, [pre-cue minus cue-other], see also Fig. 3C). Binned in 200 ms bins, advanced in 200 ms steps along the time course of the trial. D Gain of information about color when cue was at the analyzed location, relative to when there was no cue. E Loss of information about color when cue was not at the analyzed location. Error bars in (C, D, E) indicate the standard error of the mean. Asterisks indicate significant difference from 0, based on a sign-rank test using Bonferroni correction. Each plot shows decadic log transformed datapoints (nbird 1 = 163 units, and nbird 2 = 193 units, per bin). Positive sign datapoints indicated with “x”, negative sign data points indicated with “o”; refer to right hand log axis for values and Statistics and Reproducibility for more details.
