Question 1. For the results on the accuracy of recruitment (Fig. 4), it seems to be critical that not a single Acc bee is dancing in the mixed hive to indicate the location of the food source to its conspecifics there. How were Acc bees prevented from dancing during the time of training for the Aml bees and especially during the phase when recruited Acc bees were allowed to return to the hive before their second foraging trip?

Answer: Acc bees were marked at the feeder and allowed to return to the hive only in the early, pre-training phase on the study, simply to ensure that the bees visiting the feeder were indeed from the observation hive. This phase was brief, and the observation hive was monitored throughout this period, to ensure that no Acc bees were dancing (in any case, bees normally require a few trips to a food source before they start to advertise it through a dance). As the Acc bees were only allowed a total of two trips to the feeder (with one intervening trip back to the hive), no dances occurred during this phase of the study. In all subsequent phases of the actual experimental procedure, new Acc recruits were captured at the feeder and killed.

Question 2. From the regression in Fig. 3 we calculated that Aml bees from the mixed colony trained to the feeder in 500m distance (Fig. 4) would dance for just under 1000ms (assuming the regression still holds when data for 500m are included). This would indicate a distance of about 250m for the Acc bees. How many Acc bees were searching for a food source at 250m?

Answer: In our published experiments, we did not monitor the recruits’ search behavior at 250 m. However, this is a very good question for the future investigation. Thanks!

Question 3. If the recruited Acc bees did not find food at 250m, could they have continued in the indicated direction (along a prominent landmark, the Da-Mei-canal) until they found the feeder with food (either alone or by following Aml animals)?

Answer: See response to Question 2.

Question 4. We wondered whether Acc bees recruited by Aml bees required more time to find a food source than if they had been recruited by Acc bees, i.e., was flight time ever recorded?

Answer: No, we did not record flight times. In this regard, we hypothesize that communication would be more efficient between same species than between different species. However, there are some other factors that could affect the flight time of recruits, such as differences in the motivation of the two species. We observed in our experiments that the Acc foragers in the mixed colony with Acc queen are more likely than Aml to follow a dance. This is also a good question, and will be investigated in future experiments.

Question 5. Apis cerana is slightly smaller than Apis mellifera and the allometry of behavior is a complex issue. Could the size difference have had any conceivable effect on the data (e.g., in influencing how optic flow is measured or how many animals can follow a dance or how different large vs. small bees dance)?

Answer: We cannot answer this question at this stage.

Question 6. Because bees do not change their dance in a mixed colony, each following bee would have to be able to tell the species of the dancing bee in order to consistently arrive at the food source. Presumably, the Aml bees in the mixed colony only survive because they are not recognized as a different species from Acc. How can these two insights be reconciled?

Answer: We are currently carrying out experiments to investigate the existence of social learning with respect to the honeybee dance language in species-mixed hive. The experiments that need to be carried out to answer this question are really quite complex (as is the interpretation of the results), and we hope to be able provide you the full picture in a future publication.

Thank you Dr. Brembs! You and your colleagues have raised many interesting questions that have prompted us to explore intriguing new avenues of future research.