Consensus or Quorum? by Tom Seeley

Introduction

We are nearly at the end of July, and here in the northern hemisphere the beekeeping season is winding down. In keeping with an early flowering season, the ling heather is blooming on our moors at least three weeks earlier than usual. Swarm season is almost over and any swarms collected now will need careful attention to enable them to store enough honey for winter.

Thinking about swarms and decisions about when they leave the bivouac site for the new home, I’ve always found this process puzzling. I have never really understood what constitutes a quorum and why till now, and this week Professor Tom Seeley explains it all very clearly in his fascinating article. Thank you once again, Tom, for sharing this information with us in a guest blog.

Finding a Swarm

When one finds a swarm of bees that is hanging in a cluster, it is good to know whether these bees are, or are not, about to fly off to their new homesite.  If they are not, then you have time to shake them into a hive.  But if they are about to depart, then you might as well bid them farewell. 

How can you tell if a swarm is about to take flight?  It is easy.  Just put an ear close to the swarm, say 15 cm from it, and listen.  If you hear shrill sounds that last about one second each, coming from within the swarm cluster, then this tells you that the swarm’s bees are preparing to launch into flight to move to their new home.  These high-pitched sounds are called piping signals and they are produced by the nest-site scouts as they walk among their swarm mates.  A nest-site scout bee produces a piping signal by grabbing and pressing her thorax atop another bee, pulling her wings tightly together, and activating her flight muscles.  This sends a 1-second-long vibration into the bee that the scout bee has grabbed.  This piping signal tells the grabbed bee that the time has come for her to warm up her flight muscles, in preparation for the swarm launching into flight.

Fig. 1.  A nest-site scout producing the run-pause-press piping signal, while running over a wooden surface

One of the mysteries that I investigated in the early 2000s was this:  How do the nest-site scouts in a swarm know when to start producing their piping signals?  It was tempting to think that the nest-site scouts do so when they sense a consensus in the advertising of potential homesites, such as what is shown on 22 July, in Fig. 2.  But I had my doubts about this.

Fig. 2.   Pattern of dances produced by the nest-site scouts in a swarm that I watched on 20-22 July 1997.  The scouts in this swarm found 11 potential home sites (A-K), and  there was a prolonged “competition” between the scouts from sites B and G, but ultimately they reached a consensus that site G was their best option.

What made me wonder about how the nest-site scouts know when to start producing piping signals was something that a famous German scientist, Martin Lindauer, reported in the 1950s.  He had patiently observed and recorded the dances produced by nest-site scouts on  17 swarms.  And, as is shown in Fig. 3, on 22 June 1952 he observed a swarm in which the nest-site scouts never reached a consensus.  Its scout bees danced excitedly about two sites, one in the NW and the other in the SW.  Even so, this swarm took off!

Fig. 3.  The announcements made by the dancing nest-site scouts in the “Balcony swarm” that Martin Lindauer observed on 22 June 1952.  These bees never reached a consensus. When the swarm took off, its split itself, with bees flying to the NW and SW.

It turns out that what informs a nest-site scout that it is time to produce piping signals is NOT that she has sensed that a dance consensus has been reached at the swarm.  Instead, what tells her to start making piping signals is something that she has sensed at the prospective homesite:  the presence of a sufficient number (i.e., a quorum) of fellow scout bees at this site.  

This became clear when a fellow investigator of honey bee behavior, Professor Kirk Visscher of the University of California—Riverside, and I tested the hypothesis that nest-site scouts use what is called “quorum sensing” to know when they should excite their swarm mates to take off and fly to their new homesite.  In other words, to know when to activate their swarm mates to fly to their future homesite, the nest-site scouts rely on sensing that a sufficient number (a quorum) of scouts is visiting, and so knows the location of, their new residence.  Kirk and I tested a key prediction based on this hypothesis:  if we slow the formation of a quorum of nest-site scouts at a swarm’s chosen site, while leaving all the rest of a swarm’s decision-making process unaltered, then this should delay when the nest-site scouts start to produce piping signals. 

We conducted this test on a treeless island off the coast of Maine:  Appledore Island.  We went  to this island because here we could delay the formation of a quorum of scout bees at a swarm’s chosen nest site, while leaving all the rest of the bees’ decision-making process unchanged.  We delayed quorum formation by setting out five perfectly good nest boxes close together at one site on the island, as is shown in Fig. 4.  This caused the scout bees that visited this site to be dispersed among  the multiple nest boxes, and this delayed the formation of  a quorum of scout bees at any of the nest boxes.  We also performed  a control trial with each swarm.  In the control trial, we set out just one nest box; it sat in a second site on the island.  Thus, we watched  each swarm go through its home-site selection process twice.

   Fig. 4.  The cluster of five nest boxes (each one is sheltered within an orange lean-to) along a path on the eastern side of Appledore Island.  Every 15 minutes, a count was made of the scout bees visible outside each nest box.

We tested four swarms, one at a time, and with each swarm we observed that the number of scout bees built up rapidly at the single nest box in the one-nest-box trial, whereas their numbers built up quite slowly at each nest box in the five-nest-box trials.  This was because in the latter trials, the scout bees distributed themselves among the five nest boxes.  We also observed that the scout bees in each swarm reached a decision much more quickly in their swarm’s one-nest-box trial (average 196 minutes) than in its five-nest-box trial (average 442 minutes).  Likewise, the scout bees began producing piping signals sooner in the one-nest-box trials vs. the five-nest-box trials.   These results provide strong support for the quorum-sensing hypothesis.  It remains a mystery, however, how exactly the scout bees sense a quorum of their fellow scouts at a site.  They may do so visually (as we did), by seeing a crowd of fellow scout bees at a site.  Or perhaps they do so by touch, i.e. by noticing that they often bump into other scout bees when they go inside a potential nest site to inspect its interior.

Why do nest-site scouts rely on quorum sensing rather than consensus sensing to know when they have reached a decision?  I think that one reason they do so is because a swarm needs to have a sizable group of bees that have visited the new homesite and so are able to guide the non-scouts in a swarm to this site.  Encountering numerous fellow scouts at the new homesite is a reliable indication that this need has been met.  Quorum sensing may also help ensure that a swarm makes a good decision. The requirement of a quorum of at least 20–30 scout bees present simultaneously at a future homesite must help to ensure that a swarm chooses wisely.  The presence of this many bees at a site means that numerous scouts have examined this site and judged it desirable.  

But why is 20–30 bees the critical number of bees rather than, say, 10–15 bees or 40–60 bees?  I suspect that the quorum size of 20–30 scout bees is a part of the honey bee’s house-hunting process that has been tuned by natural selection to give a swarm an optimal balance between speed (favored by a small quorum) and accuracy (favored by a large quorum) in choosing its homesite.  A swarm must choose its living quarters quickly, lest it be caught outdoors in a storm.  A swarm must also choose its dwelling place wisely, so that over the years it will enjoy good protection from harsh weather and fierce foes.