Files in this Data Supplement:

• Supplemental Figure 1 -

  Fig. S1. Bees returning to the hive that was maintained in this hut had to choose between two patterns to gain access to the hive.

• Supplemental Figure 2 -

  Fig. S2. A Y-maze was set up in this hut.

• Supplemental Figure 3 -

  Fig. S3. Experimental apparatus: (A) A Y-shaped, dual-choice apparatus at the feeder; (B) An entrance chamber, which split the access to the hive into two paths, was added to the entrance of a normal hive. Bees had to make a choice between two patterns for accessing the hive. Only the entrance behind the positive pattern was open during training.

• Supplemental Figure 4 -

  Fig. S4. Results of the learning tests in Series 3 experiments carried out at the Bee Station at Wuerzburg University. The visual patterns were blue/white horizontal versus vertical gratings at the feeder, and black/white horizontal versus vertical gratings at the hive. The horizontal grating at the feeder and the vertical grating at the hive were the positive patterns in the morning, and the vertical grating at the feeder and the horizontal grating at the hive became the positive patterns in the afternoon. Honeybees could learn the tasks very well. The results show that trained bees could reverse their pattern preference following midday breaks, as well as overnight breaks, at the feeder and at the hive. The trained bees also had significantly different pattern preferences at the feeder and at the hive entrance within each testing period. The terms MF1 and MH1 (or MF2 and MH2) denote the early morning (or late morning) tests at the feeder and the hive, and AF1 and AH1 (or AF2 and AH2) denote the early afternoon (or late afternoon) tests at the feeder and the hive, respectively. See text for more details.

• Movie 1 -

  Movie 1. Indoor experiments carried out in an All Weather Bee Flight Facility (AWBFF) at the Australian National University’s Research School of Biological Sciences.

• Movie 2 -

  Movie 2. Experiments were carried out during July and August 2005 at the Bee Station of the Wuerzburg University. Bees returning to the hive had to choose between two patterns to access the hive, as only the entrance behind the positive pattern was open during training but both entrances were open during the test periods.

             

  Methods

  Training and learning tests for Series 3 experiments

  In the learning tests, blue/white gratings were used at the feeder, and black/white gratings at the hive. All of the tests (MF1, MH1, MF2, MH2, AF1, AH1, AF2 and AH2) were carried out individually. Fig.^.S4 shows that the preference for the positive pattern in all tests was significantly greater than the random choice level of 0.50 (P<0.001; Fig.^.S4,a,h). The average choice frequency at the feeder in favor of the blue horizontal gratings in the early morning test (MF1, Fig.^.S4,a) was 0.61±0.04, N=4; however, the average choice frequency in the late afternoon test (AF2; Fig.^.S4,b) was 0.22±0.02, N=17. The preference for the blue horizontal gratings was therefore significantly reversed between the MF1 and AF2 tests (P<0.001). At the hive, the average choice frequency in favor of the vertical gratings in the early morning test (MH1, Fig.^.S4,c) was 0.34±0.05, N=16; however, the average choice frequency in the late afternoon test (AH2, Fig.^.S4,d) was 0.83±0.04, N=14. The preference for the vertical gratings between the MH1 and AH2 test was thus also significantly reversed (P<0.001). The trained bees clearly reversed their pattern preference between late afternoon and early morning. Similar conclusions can be drawn from data from the feeder when we compare the performance between the late morning (0.68±0.02, N=17; MF2, Fig.^.S4,e) and the early afternoon tests (0.359±0.02, N=15; AF1, Fig.^.S4,f); as well as at the hive, when we compare the performance between the late morning (0.21±0.03, N=17; MH2, Fig.^.S4,g) and the early afternoon tests (0.71±0.03, N=16; AH1, Fig.^.S4,h).