Skip to main content
. 2019 Jan 25;12:279. doi: 10.3389/fnbeh.2018.00279

Figure 3.

Figure 3

Honeybees transferred learned information from free flight experiments to the virtual environment. (A) T-maze experimental set-up for free-flying honeybees. The T-maze consisted of a 70 cm long tunnel, 5 cm in width and 4 cm high with a 55 cm long head side. At the end of the legs, one of the colors (yellow or blue) was rewarded with sucrose whereas the other was punished with potassium chloride solution. Bees flew toward the entrance of the T-maze, walked until the point of decision and decided for one side. Afterwards they flew back to their hive or in case of a wrong decision could enter again. Sides were switched in a pseudo-random fashion to avoid side preferences. After acquisition they were transferred to the VE and were tested in a scenario similar to the T-maze without reward (see also Figure 2A). (B) Left side: Acquisition curve of correct choices (N = 6 animals) in the T-maze with five trials depicted together from trial 1 to 25. 50% of correct choice equals chance. Right side: extinction curve of choices in the VE with the first five tests in the VE. In the VE, the same bees as in A (N = 6 animals) were faced with a similar situation to the T-maze and could choose one or the other color by walking toward it. Arriving at one color (10 cm distance) was counted as a decision. Afterwards the scenario was set back to CC and they could choose again. No reward was presented during the test. Sides were changed to avoid side preference dependent choices. In the first trial, all bees chose the previously rewarded color over the unrewarded. The first test trial differed significantly from chance (binomial test p = 0.015). (C): Walking trajectories in the VE during representative test situations in a bee trained to yellow in the T-maze. Reaching the rewarded color was counted as correct choice in (B). After 30 s, the colors switched sides. (D) Percentage of correct and incorrect choices in the VE with trained colors (yellow or blue) after acquisition in the T-maze pooled for all bees (N = 6 animals, n = 150 choices). Significantly more choices were delivered to the previous C+ over the C– (binomial test p = 0.0001). (E) Walking trajectories of a bee trained in the T-maze in a VE scenario without colored stripes but with skyline and checkerboard. Comparison of percentage choices for left and right in the VE shows no significant side differences (N = 6, n = 114 choices, Two one-sided t-test, p = 0.03). (F) Walking trajectories of naïve bees in a scenario with colored stripes, skyline, and checkerboard. No preferences for one of the colors was found (Two one-sided t-test, p = 0.018, N = 12 animals, n = 12 choices). VE, virtual environment. Asterisks indiacte statistical significance.