Figure 4
Acquisition performances in a color discrimination learning task under four different background conditions. Each panel shows on the left the acquisition curves in terms of the percentage of bees responding to the CS+ (red), to the CS− (black) or not making any choice (NC; gray) during the ten conditioning trials. The pink, light gray and gray areas around the curves represent the 95% confidence interval of CS+, CS− choices and NC, respectively. On the right of each panel, a heat map shows the cumulative coordinates occupied by the bees trained under each background condition during the ten training trials. Coordinates were binned into 1 cm2. Warmer colors depict locations more frequently occupied (see color bar). The highest frequency is cut down to 10% of the maximum on the color bar. While the side of the rewarded stimulus was randomized along conditioning trials, it was placed arbitrarily on the right in the heat maps. (A) In the ‘Transparent Condition’ (N = 24), the blue and green cuboids were displayed against a dark background. (B) In the ‘Vertical Grating—No Optic Flow Condition’ (N = 17), the cuboids were displayed against the same red-and-black grating but motion cues from the background were suppressed by keeping it constantly fixed to the bee’s gaze. (C) In the ‘Vertical Grating –Optic Flow Condition’ (N = 17), the cuboids were presented against a red-and-black vertical grating, which was coupled to the bee’s movements (closed-loop conditions). (D) In the ‘Rotating Vertical Grating Condition’ (N = 17), the cuboids were shown against the same red-and-black grating, which was rotated counterclockwise around the virtual arena at a constant speed, thus generating a constant optic flow even when the bee did not move.
