Skip to main content Cornell University NBB
more options


thumbnail imagethumbnail imagethumbnail image

Juliana Rangel-Posada

Advisor: Tom Seeley

Co Advisor: Kern Reeve

Start Date: Fall 2004

The mass departure of a honey bee swarm from its nest:  The signals and signalers that induce a synchronized exodus, and the effect of partitioning ratios and genetic relatedness during fissioning


During the reproductive season, large honey bee colonies synchronize the explosive departure of most workers (and the mother queen) as a swarm, in a process taking less than 15 min. As a Ph. D. candidate, I have explored the mechanisms and functional organization of swarming in honey bees as outlined below:

(1) The signals initiating the mass exodus of a honey bee swarm from its nest (Rangel & Seeley 2008).  We examined the signals triggering a swarm’s explosive exodus from the parental nest and documented the concurrent changes in bee density and mobility. We video-recorded three swarming colonies exiting observation hives and analyzed how the bees prepared for and then performed their departures. Over the 60 min before swarm exodus, piping signals gradually increased, and ultimately peaked at the start of the swarm departure.  Also, during swarm exodus, bee density (number of bees in 100 sq cm) dropped markedly, while average bee velocity (mm/s) and the production of buzz-run signals spiked dramatically.  Neither waggle runs nor shaking signals increased before or during swarm exodus.
(2) Effect of genetic relatedness on the distribution of genetic subfamilies in swarms and parental colonies (Rangel, Mattila & Seeley 2009). We explored whether during swarming, the subfamily compositions of the bees in a swarm and the bees staying in the colony differ from a random distribution.  We set up 4 observation colonies headed by super-sister queens that were artificially-inseminated by 10 unrelated drones of known genetic composition. The colonies swarmed naturally, and we randomly collected 125 bees from both the settled swarms and the parental colony. We are currently analyzing DNA microsatellites in each group (using 5 highly variable loci) to determine patriline membership of all the workers and to estimate relative proportions of patrilines in both the swarm and the parental colony. 
(3) Identity of the signalers that trigger the exodus of a honey bee swarm from its nest (Rangel, Griffin & Seeley, in prep.).  In the summer of 2008 we asked (a) what is the identity of the bees that initiate the swarm exodus? (b) does the search for nest sites start before the swarm leaves? (c) and if so, do nest-site scouts recruit to those sites prior to departure? We set up 3 observation colonies on Appledore Island, Maine, and placed a nest box about 200 m from the hives. We labeled all bees visiting the nest box, and monitored the behavior of bees at the nest box and in the hive with video and audio recorders. We found that (1) nest-site scouts were the first producers of piping and buzz-run signals inside the nest in preparation for swarming; (2) nest-site scouts begun searching for sites up to 3 days prior to departure; and (3) nest-site scouts recruited 81-485 bees to the nest box, mostly in the 2 hours prior to the exodus.
(4) Effect of initial swarm size on growth, development, and survival of newly-established honey bee colonies (Rangel & Seeley,in prep.) We established 12 field colonies founded by 3 small (5,000 bees), 3 medium (10,000 bees), and 3 large (15,000 bees) swarms,in mid-May 2007 in Ithaca, NY. Since only 1 in 5 swarms survives its first year, we predicted that larger colonies would grow faster than the rest, and thus would have a higher probability of surviving the winter. By late-September 2007,biweekly measurements showed that large colonies had a larger population, built more comb, stored more food, and produced more brood, than medium or small colonies. Of the 12 original colonies, only 3 large, and 1 medium colonies survived by 1 May 2008.  Our results showed that a colony needs to send out as many bees as it can with the swarm to grow large enough to increase its chances of survival to the next year. 


(5) Effect of a colony’s splitting ratio during swarming on the survival probabilities of both the swarm and the parental colonies (Rangel & Seeley, in progress). We manipulated the splitting ratio of 15 average-sized colonies (≈ 12,500 bees) as they swarmed in May 2008 to determine any differences in growth and overwintering survival if split into a swarm and a parental colony in 3 types of ratios:(a) 30% : 70%; (b) 60% : 40%; and (c) 90% : 10% of the workers going with the swarm, or staying in the parental colony, respectively. We took various biweekly measurements of growth for the swarm colonies, and are currently monitoring overwintering survival of all colonies through 1 May 2009.  By October 2008, large swarm colonies had built more comb, stored more food, and produced more brood, than medium or small colonies.  By October 2008, 2 small swarm colonies died.  The survival probabilities of parental and swarm colonies will help us understand why typical swarms contain ≈ 80% of a colony’s total population. 



Please contact me if you would like a copy of any manuscript.



Rangel, J. & Seeley, T. D. (2008)The signals initiating the mass exodus of a honey bee swarm from its nest. Animal Behavior. doi: 10.1016/j.anbehav.2008.09.004.

Rangel, J., Mattila, H. R. & Seeley, T. D. (2009) No intracolonial nepotism during colony fissioning in honey bees.  Proc. R. Soc. B doi:10.1098/rspb.2009.1072.