Comparison of drone ecology and behavior between Apis mellifera mellifera and the hybrid Buckfast in southwestern Sweden

Abstract: Most research on the biology of the western honey bee (Apis mellifera) focuses on the worker bees, but knowledge about drones is lacking despite their important role in mating with a virgin queen. Available information about their ecology and behavior are mainly based on direct observations in need of intensive experience and knowledge about honey bees. Only two recent studies conducted in France and Argentina have monitored drones continuously, but on the scale of a single observation hive or during a short time period only. Therefore, studies that have continuously monitored drone flight activity during the whole mating season are still lacking. Further, we are not aware of any studies that have compared the ecology and behavior of drones between different subspecies of Apis mellifera.  Besides the aim of analysing the ecology and behavior of drones in spring and summer in Sweden, where the climate poses challenging conditions for honey bees, the objective of this project was to compare the native, but threatened, subspecies Apis mellifera mellifera (Mel) with the hybrid ’Buckfast’ (Apis mellifera x, Buck). In Sweden, the latter is most commonly used in beekeeping today.  Activity data at the entrance of the hives was collected with the use of Radio Frequency Identification (RFID) technology at two different apiaries with 8 colonies each. RFID technology enables the continuous monitoring of bees tagged with a microchip on their thorax. Hereby, only newborn drones with a maximum age of 3 days were tagged. Further, drones’ activity and behavior at the entrance of the hive was studied from video recordings. Weather parameters were measured by weather stations installed within the apiaries, enabling a direct analysis of the activity in relation to weather conditions (temperature, wind speed, rain, humidity).  Drones survived at most 70 days (M = 22.67, SD = 17.9) with a shorter average lifespan in summer (M = 20.8, SD = 15.4) than in spring (M = 25.2, SD = 20.7). Mel drones showed longest survival at both locations and in both seasons. We discovered most first activity of drones five and six days after tagging (max. 11 days), but also registrations directly on the day of tagging were made with indications that these do not all display evictions from the hive. Compared to the spring trial, in the summer trial earlier first registrations were made (p = 0.022), with Mel drones becoming active significantly earlier than Buck drones (p = 0.027). Flight durations of more than 3 minutes, even longer than 10 minutes, were already performed right after tagging. This was especially the case in the summer trial. Further, RFID data and video records revealed that peak activity occurred in the afternoon, but differences between spring and summer were revealed with an earlier peak in spring. Buck’s peak preceded Mel’s peak of activity in both seasons. However, also registrations during the morning and evening occurred such that drones were mainly active during 7-8 hours of the day. Rain and wind speed negatively affected the activity, but increasing activity was present the higher the temperature and light intensity. Hereby, temperature had a significantly higher positive effect on Mel drones. Moreover, our registrations show that 40% of the drones drifted to other hives, whereby hives of the same subspecies were preferred. Mel drones drifted relatively more often than Buck drones. In addition, video records showed that upon departure, drones pay special attention to cleaning their antennae. This was especially observed at Mel hives. Also drifting drones showed this behavior.  To conclude, the present thesis confirms but also contradicts existing information, especially in terms of the age at first activity and lifespan. In addition, unknown aspects about drone ecology and behavior were revealed with differences between both seasons and subspecies. However, the applied method in the form of RFID technology to monitor the bees showed its limitations and needs further optimisation.