Latest episode of the podcast is up! This week, I go through chapter 6 of Honeybee Democracy by Thomas Seeley. This chapter focuses on the mechanisms used to reach a unanimous decision on a new nest site. We will learn about how a scout bee's waggle dance enthusiasm correlates to nest site quality; whether scouts assess quality through comparison or instinct; and how recruited scouts choose to follow a dance or not. Most interestingly, we learn why a scout bee stops dancing. So listen over on Podbean, or wherever you get your podcasts!
Homestead Updates
I wanted to give a big shout out of love and support to everyone in Texas! Y’all have been through the ringer with your winter storms. I know some areas are still without power and/or water. It’s easy for people in the Northern states to poke fun but it’s no joke to go through something like this. I was in Georgia during a big ice storm and it’s shocking how quickly things can become desperate when you live somewhere that just isn’t set up to handle extreme cold and snow. Back then, the county I lived in had just 3 snow plows because it made no sense to have more! What a nightmare. Well, we got through it and I know a lot of Texans are already coming out the other side. I’m wishing you all the best. I’m sure beekeepers were scared for their colonies, and I hope the losses haven’t been too bad. This Northern lass is sending you love!
If you’d like to help, there are a number of food banks providing food throughout Texas. You can also make a donation through the American Red Cross, who are setting up shelters for those in need. I also found an Austin based pet charity that is helping people connect with volunteers and supplies for their pets during this difficult time.
Agnes the poorly red hen has gone back outside! She finished her meds and seemed overall better but just wasn’t eating much. As a result, her poop was very green due to bile, which is a sign that a chicken just isn’t consuming enough. My husband suggested I try feeding her dubia roaches, which she loved, and after a few days of getting 2 feedings of those a day, her poop finally returned to normal. Then, on the 24th, we had a mild day (50F, real feel 42F), so I seized the opportunity to put her outside. I wanted an easy introduction so decided to throw a ton of scratch around before I let her out. This worked to distract the flock but Agnes ruined things by immediately marching up to a hen and starting a fight. Two other girls joined in, which got Pepperjack the rooster involved, and then he decided that he had to fight my leg! He wasn’t backing off so I ended up scooping him up and parading him around to humiliate him in front of the hens. After that, things settled down and they all went into the coop together by sunset. As of the next morning, they are a fully integrated flock again, and Agnes is back to her position as one of Pepperjack’s favourites. Since then, she's been foraging like usual and seems to be on the upswing. Let's hope it continues!
Pepperjack is definitely getting ready for Spring and is spicier than usual. I’ve noticed that his favourite hen, Cheddar, has thinning feathers on her back from being mated, and might need a hen saddle soon. He’s also giving me more attitude and attacks me out of the blue. Thank goodness he has no spurs!! But the kicks he gives me are pretty tough. On the extremely rare occasions that children visit my little farm, I will have to put him away because he cannot be trusted. I confess to being ridiculous and have some hurt feelings about it all because I love him ridiculously. I want us to be friends, dammit! But if not, I’ll just keep showing him who is in charge.
We had some consistent above-freezing days, followed by rain, so our snow melted and the garden flooded (as it always does). Of course, this is the time my rain boots finally sprung a leak after 3 very diligent years of service. Godspeed, adorable chicken themed sloggers! I will honour your memory by buying an identical pair (this time in red!).
Some buds are poking up through the earth. Could Spring be on the way? Stay tuned!
Hive Updates
Lots of girls came out to do their cleansing flights on our mild days! A fair amount of bee poop to be found nearby and on the side of the hive. Not an abnormal amount, though, which is always good. This is the time of year when I’m keeping an eye out for signs of Nosema, a fungus that causes diarrhea and so it often first seen as excessive amounts of bee poop on the front of the hive.
I’m rarely on Facebook these days, only really popping in occasionally to check on my whippet rescue group and some local beekeeping ones. Well, I saw on a beekeeping group, someone asking about doing an alcohol wash for mites on a dead hive to assess infestation level, and I am kicking myself for not thinking of this!! If I’d seen the post earlier, I could have mite tested the Saskatraz colony and the 2 nucs I lost. Hey ho!
Sadly, I lost my OH queen hive, and it was a real surprise. I thought this was my strongest colony! The cluster stayed low in the box and was big and noisy! But, alas, they are no more. I did a necropsy, including a mite test, and I’m confident it was a combination of mites and isolation starvation. The mite test was 7:300, which is high for winter (that is a level I would definitely treat, and would expect to see in the Fall when the levels tend to rise). As for isolation starvation, it was too cold for the outer bees of the cluster to go to the food, which was just 2 frames away. I’m absolutely gutted by this loss but there’s nothing to do other than learn from it and move forward.
I decided to order a nucleus colony from a local keeper (and one of my previous teachers). She is actually who I originally got my OH queen from, and I ordered a nuc with an OH queen as opposed to Georgian. Excited to see how she does!
This loss means I am down to just the 2 colonies: Macha, my Southern US queen, who I got in one of my first nucleus colonies, meaning she’ll be 3 this year! And my Sask-daughter colony. Right now, the latter has a larger cluster and seems more active. I’m worried Macha and her ladies are weak. Just hoping both colonies can hold on until the Spring.
Placed a big supply order for the year, including some supplemental feeds, and mite treatments (namely oxalic acid and formic pro). This hobby is not cheap!!
*
Chapter 6: Building a Consensus
We deprecate division in our Meetings and divisive unanimity.
It is the unity of common fellowship, we believe, that we
Shall most surely learn of the will of God.
-Society of Friends, Book of Disciples, 1934
Looking back at what we have discussed so far, what normally arises from a swarm’s decision making process? A dissent free decision, and that is genuinely amazing if you think about the sheer mechanics of such a feat.
This chapter examines exactly how the bees reach a unanimous decision
Honeybees practice a kind of democracy called ‘unitary democracy’, which involves individuals with shared interests and preferences, as opposed to us humans who practice ‘adversary democracy’, where groups of individuals with conflicting interests try and reach an agreement
We learned in a previous chapter about Martin Lindauer’s Balcony Swarm, which failed to reach a consensus and attempted to fly off when two sites were still being advertised. As a result, the swarm was divided, faced confusion in the air, and ultimately lost their queen, causing the entire enterprise to completely fail
We can see, then, how important it is that a swarm’s scouts are in agreement before flying to their new home
When considering exactly how this consensus is reached, Seeley asks two key questions: what causes the support of scout bees for the best site to grow throughout the debate? And, what causes the support of scout bees for the poorer sites to fade during the debate?
Lively vs Lacklustre Dances
Let’s consider what we know about swarms so far. We know that a swarm consists of approximately ten thousand bees, and that a few hundred of these bees will operate as nest scouts
We can also assume that there are dozens of potential nest sites that are worthy of consideration (waggle dance), and that each site is originally discovered by just one scout bee
This means that, of the hundreds of scout bees, just a few dozen are the discoverers of new sites, while the rest are recruited for support of a location
Of those first recruited, these bees will examine the site advertised and then dance for it upon returning to the swarm should she find it satisfactory
Thus, a swarm’s democratic choice of a new home is an election process; there are multiple candidates (nest sites), competing advertisements (waggle dances), individuals supporting a particular candidate (recruits), and individuals who are still neutral (scouts not yet recruited to a site)
Scout bees can gain support for their site via their dances, and scouts already supporting a site can over time become apathetic and return to the neutral pool
It’s clear that to gain the most supporters, a scout must strongly advertise her site, which begs the question: if she shows the greatest amount of enthusiasm through her dance, does this relate to how good her site is as a new home? If all scouts abide by the premise of strongest dance equaling the best site, does this mean that only the very highest quality of site will be ultimately chosen?
In the previous chapter, we saw that the strength and enthusiasm of a scout’s dance does correlate with nest site quality.
The first evidence of this came in 1953, thanks to Martin Lindauer’s observations of an artificial swarm. He had set up two empty wooden hives 75 meters/250 feet from his artificial swarm, the scouts of which discovered both boxes during the first day of placement.
The initial dances for these boxes were not very enthusiastic but slowly more bees came to examine the hives, until Lindauer had a total of 30 dancers for both hives
On the second day of observation, Lindauer spotted an exceptionally enthusiastic dancer on the swarm and realised she was advertising for a location other than the wooden hives he had provided. He was able to ascertain the nest site she danced for, which was an underground cavity beneath a tree stump in a corner of a woodlot. The site was protected from wind due to thick shrubbery, had a small entrance of 3cm/1.2 inches in diameter, and the cavity was approximately 30 liters/27 quarts in volume. The cavity was also dry inside despite recent heavy rain. Simply put, it was the perfect bee home!
Usually, as part of his study, Lindauer killed bees advertising for locations away from his offered nest sites but he chose not to do so in this case. Within just an hour, this dancer’s enthusiasm led to more scouts assessing the potential site and returning to advertise for it.
Lindauer felt that this single bee taught him a great deal. Namely, that a bee has an innate concept of nest site quality that is referenced when investigating a potential site, and that a scout’s dance provides information not just on location but also on quality
He summarized thusly, “The most lively dances indicate a nest-site of the first quality; second-rate homes are announced by lacklustre dances.” Pg.122
Representing Site Quality in Dance Strength
Seeley notes that he procrastinated on truly studying how scout bees transmit site quality via their dances because he had long believed Lindauer’s conclusion was correct, and further that work Seeley had done himself continued to support the correlation between strength of the dance and quality of the advertised site
In the previous chapter, we saw how the scout bees of Appledore Island danced more strongly for the better homesite, and also covered a little of the work Seeley had done observing the strength and duration of a bee’s dance when advertising a bountiful forage site
We learned how the richer the nectar source, the stronger would be the dance
Specifically, a bee will adjust the number of dance circuits in relation to nectar-source richness by adjusting the rate of dance circuit production (R, in dance circuits per second), and the duration (D, in seconds)
The total number of dance circuits produced (C, in dance circuits) is the product of rate and duration so C = R x D
Rich nectar sources elicit a higher R and longer D, just as Lindauer noted the “faint-hearted dances” for inferior nest sites, while superior sites “solicited with a lively and long-lasting dance”. Pg.123
By 2007, Seeley knew he needed to get solid, quantitative data about how nest scouts transmit nest site quality via their dances
Seeley would return to Appledore Island for this study with two collaborators: Marielle Newsome, an undergraduate at Cornell with beekeeping experience, who was about to start her graduate studies in insect behaviour; and Kirk Visscher, a behavioural biologist from University of California Riverside, and longtime collaborator of Seeley’s since their days as students at Harvard
The plan: place an artificial swarm at the center of Appledore Island with two nest boxes set 250 meters/820 ft away from it at 40 meters/130ft apart from one another so that they could be discovered almost simultaneously
One of the nest boxes was high quality at 40 liters in volume, while the other was medium quality at 15 liters volume
The first 5-7 scout bees at each box would be logged and then recorded back at the swarm to capture their dances, allowing them to assess the strength of it
Video analysis would show the precise time each scout danced as well as how many dance circuits she produced
Initially, Seeley was concerned that they would need to laboriously label every bee of the swarm as there was no way to predict which of the thousands of bees would first appear at the boxes. Thankfully, however, they were saved from this by an ingenious idea of Kirk Visscher’s; he had discovered a way to apply individually identifying paint marks to a scout be without causing her distress (and thus effecting her behaviour)
Using a small insect nest, they could gently capture the scout as she exited the nest box she had been investigating, carefully mark her with paint, and then release her in the exact spot where she had been captured. This seemed to be the key to keeping the bee calm, and she would fly off as if nothing had occurred!
Seeley and his collaborators performed 7 trials of their experiment over most of the month of July, and succeeded in capturing how 41 and 37 scouts advertised for the 40 liter and 15 liter nest, respectively
Scout bees reported on their sites for just a few hours, and an individual bee's report was often spread over multiple trips back to the swarm
Looking at one scout bee’s behaviour (named Red for her paint colour), Seeley reports that she discovered the 40 liter box at 9.33am, spent 10 minutes inspecting it inside and out, and then flew to the swarm cluster to announce her discovery. Her first waggle dance lasted for 6 minutes and 162 dance circuits!
She was sighted again at the nest box at 10am and spent another 10 minutes there before returning to the swarm, although this time she did not dance. Instead, she spent 6 minutes crawling across the surface of the cluster.
Even though Red was noted to visit the box 2 more times (at 10.16am and 10.26am), she only ever danced after the very first visit. After 10.30am, she stopped visiting the site altogether.
So despite her initial, enthusiastic dancing, Red appeared to have lost interest in visiting and dancing for her site
Based on the record of 10 other scout bees in this study, Red’s behaviour is typical; each bee made an initial inspection lasting 5-35 minutes, returned to the swarm to advertise the site via waggle dance, then she would revisit and return whereupon she might dance again or simple settle in to rest
Going back and forth between the site and the swarm was seen consistently and usually lasted for an hour or so before the scout appeared to lose interest in advertising and even visiting the site
The main finding from this study, though, was the strong difference in dance circuits performed for the high quality (40 liter) site and the medium quality (15 liter) site
Although great variation was seen, the average dance circuit per bee was higher for scouts of the 40 liter box vs those for the 15; 89 vs 29 circuits per bee, respectively
It seemed that the scout could tell the quality of the site upon the first visit; upon returning to the swarm cluster after the initial visit, 76% (31 of 41) of the scouts from the 40 liter box advertised it with dance, whereas only 43% (16 of 37) scouts from the 15 liter did so
Initially, Seeley and his collaborators were surprised by the great variation seen, although it became clear that when counting and averaging the dances, one could see that a high quality site elicits a dance with more circuits. Meaning that, even though site reporting quality in an individual level varied, there was a clear swarm-level reporting of site quality
For instance, a single scout from a better site does not always advertise with more enthusiasm than a single scout from an inferior site but, if you were to look at the dances of six bees from each site, you can see that the number of dance circuits for the 40 liter box is greater than that of the 15 liter box 100% of the time!
Although swarm-level reporting results in selection of the superior site, early in the process of discovery there is a greater chance of error due to individual reporting
For instance, if a scout bee fails to advertise a high quality site, it will be overlooked by the swarm unless another scout happens to discover and report it (for which the probability is low)
A solution would be for each scout bee to consistently report on a site discovery, and it does appear as if the bees do this overall
During the experiment, Seeley and his collaborators found that the 2 scouts that first discover the 2 nest boxes almost always (86% of the time) reported to the swarm
In contrast, scouts who come to the boxes later (recruited bees) were less likely to report (55% of the time)
This discovery led to more questions, such as: what prompts the initial scout to dance so strongly? Is it the inspection of the location for the first time that triggers this response? For now, this is unknown.
Looking at the percentages mentioned previously, we can see that this process of reporting new sites to the swarm is not foolproof, and we saw in the previous chapter’s discussion of the ‘best of 5’ choice test how one swarm chose a mediocre site due to 2 scout bees failing to report on their discovery of the superior site (for unknown reasons)
An additional fascinating discovery from their study was the fact that each of the marked scouts visited just one of the 2 nest boxes despite how closely they were positioned to one another
This seems to support Lindauer’s theory that scout bees have an innate scale of nest site quality; she does not estimate quality via comparison to other sites
Even though these scout bees did not visit both sites, scouts for the 40 liter box always danced more strongly, indicating that they do have an innate knowledge of what makes a good nest site
Similar site assessment is seen in foragers; a new forager with no previous experience outside of the hive spontaneously prefers objects with complex shapes, certain colours, and odours, all of which appear to steer her towards flowers
“I should emphasize that almost certainly a scout bee does not consciously think through her evaluation of a site. Instead, she probably does so unconsciously with her nervous system integrating various sensory inputs relating to cavity size, entrance height, and the like, and generating within her a sense of the site’s overall goodness. It may be that finding a desirable tree cavity feels to a homeless scout bee as inherently pleasurable as feasting on a delicious meal does to a hungry human being.” Pg.131
The Strong Grow Stronger
The scout bees reporting on the best site dance enthusiastically with the greatest number of dance circuits
Back in Chapter 4, we saw that site G emerged as the winner of one swarm’s chosen home
The bees dancing for this site produced the greater number of dance circuits consistently. Even when fierce debate raged between sites A, B, D, and G, the average number of circuits per bee showed that site G received the greater number (59, 29, 42, and 74 respectively)
The following morning, when the sites reported were narrowed to just B and G, the average number of dance circuits remained high for G at 42, compared to just 16 for B
“Because the best site stimulates its supporters to dance the most strongly, its supporters have the highest per capita success in converting neutral scouts into additional supporters.” Pg.133
New supporters in turn dance strongly, which recruits yet more dancers until eventually one group has overwhelmed all others with the number of dancers/support
Let’s look at an example of how this process works for a situation of 2 sites that differ in quality: the high quality site stimulates supporters to advertise with an average of 90 dance circuits, compared to an average of 30 dance circuits from supporters of the inferior site
During the first 3 hours after discovery, with one scout producing 90 dance circuits and the other 30, we can see that the total amount of advertising for the 2 sites is 3:1 (favouring the superior site)
Assuming that 8 neutral scouts are recruited to the 2 sites, there will be 6 new scouts supporting the high quality site, and just 2 supporting the lesser quality site
Over the next 3 hours, the high quality site will lead to a total of 540 dance circuits (6 bees x 90 dance circuits per bee), while the lower quality site will produce 60 dance circuits (2 bees x 30 circuits per bee)
Now the level of advertising, or force of persuasion, is 9:1
Looking at this increase from just 1 bee discovering each site, we can see how the higher quality site grows its support exponentially over time
Seeley points out that the neutral bees who are recruited do not monitor the multiple advertisements and rule out the weaker ones; instead, they appear to randomly pick up a dance after encountering it on the swarm cluster
This was witnessed during an experiment conducted by Kirk Visscher and Scott Camazue, a physician, photographer, and fellow bee fanatic
In December 1995, in the desert east of Indio, California, where natural nest sites for bees are scarce, the 2 men set up artificial swarms and 2 nest boxes
They labeled each scout bee that danced for a site with individual ID markers, and recorded all of their dancing throughout the full decision making process of the swarm
They wanted to learn more about the process of those bees that followed (were recruited) a dance. Did she visit both sites to compare? She did not. Instead, they saw that the dancers who became dance followers did so purely in proportion to the amount of dancing for the 2 boxes
Simply put, they were choosing at random. So how does this lead to the stronger site winning if a bee could randomly select for any advertised site?
We know that a good site will reach a point with more dancers and that increases the chance of other bees choosing that dance
Recall that we have the discovery scouts who find the site and then recruits others; these additional recruits inspect the site for themselves and then return to dance for it if they find it high quality. If it is a high quality site, more bees will dance (and for longer) and so more of the neutral scouts will ‘chance’ upon these dances, join in, and thus increase the overall number and thus chance of site selection
But complete agreement involves not just one site dominating but also for any competing sites to eventually fade and then cease altogether. We will learn about how support for these sites ends in the next section, aptly entitled. . . .
The Expiration of Dissent
In order to come to complete agreement, the debate must end, and this means that supporters of inferior sites must either switch their support to the most popular site, or simply stop supporting any site at all
We’ve seen so far that scouts supporting a losing site will eventually stop doing so but we have yet to learn about the mechanism of this key aspect of the process
In the 1950s, Martin Lindauer recognized the importance of this dance cessation but struggled to quantify it. He favoured the idea that a scout ceases her support for one site when she learns about a superior one; this is the ‘compare-and-convert’ hypothesis, which posits that a scout compares her first danced for site with a new site, and then converts her allegiance should she find it superior
Despite his hypothesis, Lindauer noticed behaviour that seemed to go against it. For instance, he observed a scout bee start dancing for a new location without ever leaving the cluster to inspect it; she was simply picking up the dance from other dancers.
Why would she switch from a site she had seen herself to one that she had not inspected?
Seeley calls this the ‘retire-and-rest’ hypothesis; a bee does not compare her old site to a new one, and does not convert to a new site, she merely loses motivation for her site and becomes quiet/still upon the cluster
Comparing these hypotheses, Seeley notes their different predictions about when a scout bee will stop supporting one site to when she follows a dance for another
The compare-and-convert hypothesis predicts that a bee will cease dancing for her site only after following a dance for another site, which she then inspects and returns to dance for if it is found superior
The retire-and-rest hypothesis predicts that a bee will stop dancing for a site even before she has followed a dance for another
To test these different predictions, Seeley set up swarms, one at a time, and labeled with paint the first few dancers that arrived on each swarm. He would then observe them consistently to note when they started dancing, when they stopped, and when (if at all) they followed the dances of others
Seeley had seen previously that the first few dancers on a swarm tend to advertise losing sites so he focused on these as they had the best chance of demonstrating cessation or converting behaviours
He decided to limit himself to labeling 4-8 scout bees for each swarm and so would need to repeat the experiment with multiple swarms in order to collect sufficient data. In total, Seeley watched 37 scout bees in 6 swarms, which totaled 66 hours of observation
As he had predicted, 84% of the scouts initially reported on sites that were eventually rejected, with just 16% first dancing for a site that was ultimately chosen
27 of 31 bees stopped advertising their site before the swarm’s decision making process had concluded, with the remaining 4 scouts appearing close to doing the same based on how weak their dances had become
So, did those 27 scout bees stop dancing before or after following dances for another site?
Of those 27 scout bees, 26 (96%) stopped dancing before following a dance for another site, with just 1 bee (4%) ceasing to dance after following another dance
These results invalidate the ‘compare-and-convert’ hypothesis, and support the ‘retire-and-rest’ hypothesis
But what causes a dancer of a losing site to stop dancing? One possibility is that the bees have an internal, neurophysiological process that causes eventual loss of motivation. This makes sense if we consider that the eventual fading of interest in a specific site fosters consensus building as it prevents stubborn holdouts, which could lead to an impasse of debate
Potential support of this theory was seen by Seeley during the aforementioned experiment; of the 37 scout bees observed, each bee reduced the strength of her dance over time spent going to and from the site to the swarm
Looking at 3 example scouts (labeled Red, Pink, and Orange), we can see that their dance circuits per trip dramatically reduced over time, with Red going from 49 to 0 abruptly, while Pink and Orange had more gradual declines; Pink going from 74 circuits to 31 to 0, and Orange from 87, 60, 56, 10, and then 0.
“On average, there is a remarkably regular decline in the number of dance circuits produced per trip back to the swarm, and the rate of this decline is approximately 15 fewer dance circuits per trip.” Pg.142
Interestingly, this decline in dance strength is seen whether a scout is reporting on a high quality site or a lower quality one
The difference is that a high quality site is reported with a greater number of dance circuits than those for a lower quality site. This higher number of dance circuits is important because it means that, even with a steady and relative decline in dance circuits for a good site and a poor one, the strength of the dance for the better site will last longer and be ‘louder’ than the lower quality one.
Think of it like this: if I was giving away $1 every 15 minutes, I would be doing so for longer if I started with $500 over $100
Since each scout bee eventually falls ‘silent’, new scout bees will advertise. Seeley noted how the active dancers at 10am had retired by 1pm, and those dancers active at 1pm had retired by 4pm
Seeley likens this to how scientists conduct their social decision making on scientific theories; as one generation of scientists retire and eventually pass away, new scientists emerge who are familiar with the previous generation’s work, are convinced by the most compelling claims, and then adopt new theories of their own.
“One difference between aged scientists and aged scouts, though, is that the people tend to drop out of debate reluctantly, sometimes not until death, whereas the bees do so automatically. I cannot help but wonder whether science would progress more rapidly if, in this regard, people behaved a bit more like bees.” Pg.145
*
And that’s it for this chapter! In 2 weeks, I’ll be back to cover Chapter 7 (‘Initiating the Move to New Home’).
Until then, stay warm, stay safe, and keen an eye out for Spring! Hopefully, it's just around the corner.
