Chemical communication and genomics of swarming behavior in honey bees (apis mellifera L.)

Open Access
Richards, Jessica Yung
Graduate Program:
Master of Science
Document Type:
Master Thesis
Date of Defense:
November 05, 2012
Committee Members:
  • Christina M Grozinger, Thesis Advisor
  • James Homer Tumlinson Iii, Thesis Advisor
  • Edwin George Rajotte, Thesis Advisor
  • Anita Collins, Thesis Advisor
  • honey bee
  • Apis mellifera
  • swarming
  • chemical communication
  • genomics
Honey bees (Apis mellifera L.) are an outstanding model system in which to examine the communication systems underlying complex social behavior. Swarming behavior is a fascinating example of this communication on a grand scale: when honey bee colonies swarm, a large group of workers will depart from their hive with the old queen, leaving behind the rest of the workers and a number of developing queens. The chemical communication systems regulating swarming behavior, and the neurophysiological and molecular mechanisms underlying the responses to these chemical signals and/or cues, have not been characterized. In Chapter 2, we used solid-phase microextraction (SPME) and gas chromatography/mass spectroscopy to examine the volatile chemical compounds produced by the queen during different phases of the swarming process (in-hive, pre-bivouac, and liftoff). We demonstrated that queens release a significantly more complex blend at swarm liftoff than during the previous phases of the process, and that volatiles produced by swarming queens during liftoff are significantly more attractive to workers than volatiles produced by non-swarming queens. We identified five compounds from this blend, one of which is (E/Z)-beta-ocimene, a compound found in mated queens. There was considerable variation among the queens. This suggests that blends of chemicals, rather than a specific chemical, are associated with the swarming process. In Chapter 3, we examined the physiological and brain gene expression differences between workers that remain in the hive versus those that depart with the swarm. Microarray and quantitative real-time PCR analysis demonstrated that these two groups of bees have distinct gene expression patterns, including significantly higher levels of the egg-yolk and storage protein vitellogenin, which is a marker of the nursing behavioral-physiological state. We also identified 142 transcripts that were consistently, significantly, differentially expressed in the brains of swarming versus non-swarming workers from two different colonies. This gene list overlapped significantly with a list of differentially expressed genes in nurses versus foragers. Subsequent analysis of directional overlap demonstrated that swarming workers had similar gene expression patterns as younger, nurse bees. However, the reproductive potential of swarming workers (as measured by the number of ovarioles) did not differ significantly from that of non-swarming workers. This suggests that individual fitness benefits do not factor into the likelihood to join the swarm. Our results demonstrate that queens could play a role in triggering the initial swarming event through the release of novel pheromones, and swarming and non-swarming workers represent distinct physiological and behavioral states that likely are differentially responsive to these pheromones.