In my book Why Evolution is True, I begin the chapter on natural selection (Ch. 5) by describing how native honeybees in Asia defend themselves against the attacks of a predatory Asian giant hornet (Vespa mandarina), a species of huge and fearsome hornets that has now invaded North America. (You’ve surely heard of them as “killer wasps” or “murder wasps”: they are so big, vicious, and venomous that their stings kill several dozen people a year.)
These hornets can completely destroy a honeybee hive, as the hornets are impervious to stings and have powerful jaws that can decapitate bees at a rate of more than one a second. After raiding a nest and killing all the residents in an hour or two, the wasps slurp up the bees’ honey and then carry off the bees’ brood to feed their own larvae, who are ravenous for meat.
Wasps find the honeybee nests via scouts, who, upon encountering a nest, mark it and the surrounding vegetation with a pheromone, which somehow attracts other wasps to the nests, often leading to a big raid. (Wasps may also follow the scouts to the nest.)
But natives bees have evolved a counter-adaptation: “balling”. Sometimes, when a scout wasp lands on a nest of native bees, the residents lure the scout inside, where a swarm of bees awaits. They pounce on the scout, surrounding him with a big ball of bees, and proceed to vibrate their abdomens. This raises the temperature inside the ball to 47°C (117° F), a temperature that cooks the wasp to death within 20 minutes but doesn’t hurt the bees. Here’s a short video of a bee ball killing a hornet, which you can glimpse in the center of the vibrating mass:
It’s a clever and evolved strategy for disposing of a scout before it has a chance to alert other wasps. The wasp’s fearsome behaviors, and the “arms race” that has led to “balling” by the native bees, is the example I use to introduce natural selection.
As you might predict, introduced bees, like the European honeybee (Apis mellifera), haven’t been in Asia long enough to evolve the balling strategy, and thus are much more likely to have their nests destroyed by wasps. Natural selection takes time.
But some native honeybees have evolved (or perhaps learned) another defense against wasps, and this is described in a new paper in PLOS ONE. It in fact involves the first known use of tools in honeybees. Click on screenshot below to read it (pdf is here, reference at bottom):
The title pretty much tells the tale. Here’s the victim honeybee, Apis cerana, a native species (click all photos to enlarge them):
And the main predator studied, Vespa soror, another large fearsome wasp that raids and destroys the bees’ nests. These are big ‘uns, with lengths ranging from 26-46 mm (1-1.8 inches)
The study was done in Vietnam using commercially kept bees with wooden nest boxes. The researchers observed that many of the nest boxes were speckled around the entrances with tiny raised spots of animal dung, as in picture A below. They observed bees collecting dung (B; note that the B is experimentally marked), and immediately conveying it to the nest entrances, where they worked it into tiny mounds around the entrance.
Photo C shows a worker bee (these are of course all females) with a bit of dung in its mouth. They don’t eat it, but take it to the nest entrances. D shows another marked bee, with the orange spot, plastering the dung by the entrance.
Attacks by wasps and the results are shown in photos E (an attack by six V. soror workers), and F, which shows how the wasps actually enlarge the entrance by chewing on the wood, presumably to allow easier access to other wasps. Those wasp jaws are strong!
The researches did a number of manipulative and observational experiments to answer some questions about this behavior. I’ll summarize them briefly:
a.) The fecal spotting is widespread. Over the ten-day observation period, the number of hives accumulating spots went from about 10% to between 40% and 90%. This is, perhaps, because the researchers were working during the wet season, when wasp attacks are most frequent.
b.) Attacks by the wasp prompt more fecal spotting of the hives. The spotting occurs either during an attack of wasps who don’t destroy the hive, after an attack when experimenters prevent further attacks, and for at least three days after an attack. The continuation of spot-making after an attack is used as one clue about how the spotting works (see below).
c.) Attacks by a less dangerous wasp species prompt less fecal spotting. Wasps in the species V. velutina don’t attack nests en masse like V. soror, but simply pick up single bees in flight. Sure enough, even when encounters between the bee and these two species were equally frequent, there were significantly fewer fecal pellets deposited by the bees after V. velutina than after V. soror encounters. This of course makes sense if the fecal pellets are put in place to reduce the deadliness of wasp attacks.
d.) Extracts of the wasps’ pheromone glands prompt the bees to make more fecal spots. The researchers pinned pieces of filter paper to the nest saturated with either an ether control (it would evaporate, though) or an extract of the “van der Vecht glands” from the wasps, which are thought to be where the wasps’ “scouting” pheromones are made. (They should have used an organic non-wasp chemical that wouldn’t evaporate.) This chart shows the significant increase of spots prompted by using the gland extract as opposed to the ether control (caption from paper); the experiment lasted only six hours.
This suggests that the wasps’ marking of the nests, as opposed to wasp presence alone, is a main force leading to fecal spotting. You don’t need wasps to create the bees’ behavior, just pheromones.
e. Fecal spotting keeps the wasp V. soror away from nest entrances. The researchers divided nests into three categories with respect to the level of fecal spotting: light, moderate, and heavy, shown respectively by the three colors of bars in the figure below. When they measured the behaviors of the hornets in several ways (duration of visits, —which include visits in which wasps don’t land on the nest box), duration of time wasps were landed on the hive, duration of time at the entrance, and duration of chewing at the entrance, every single index of aggressive wasp behavior decreased with increased spotting. The asterisks below show significant difference between groups, almost certainly due mostly to the light spotted nests versus the group (moderately spotted + heavily spotted). This shows that the fecal pellets do something to reduce wasp predation.
There are, as the authors note, lots of questions that remain. I’ll deal with only three here:
1.) How, exactly, does the dung reduce wasp visits? There could be two ways. First, the dung could act as a deterrent to the wasps, perhaps via its odor. The researchers could, I suppose, test this directly in the laboratory, but that wasn’t done. The other way is that the dung could mask the “scouting pheromone” deposited by wasps or somehow overpower its odor, reducing wasp attraction. This could in principle be tested by looking at the frequency of wasp visitation on nests where no pheromone had been deposited. But that would be hard to do, as it requires constant monitoring of nests.
The authors favor the idea that the dung itself is a repellent, as the bees continued to add spots around the hive entrance for several days after an attack. (The presumption is that the pheromone odor would have disappeared by then.) But the dung could of course serve both functions.
2.) How did this behavior evolve? I’m assuming here that this is an evolved rather than a learned behavior, though part of it may be learned. There have been reports of this and other wasp species smearing plant juices around the entrances of hives, perhaps using the plants as either a wasp repellent or a pheromone masker. As the authors write:
Fecal spotting is behaviorally analogous to observations of “plant smearing” by A. cerana japonica in Japan, which occurs in response to attack by V. mandarinia . In this recently described behavior, workers carry gnawed plant material in their mandibles and then smear their juices around nest entrances, leaving dark stains. Although we did not study how filth foraging is organized, we observed several workers performing “emergency” dances outside of hive entrances, a behavior that recruits nestmates to smear plant material in Japan. It is fascinating that A. cerana has been observed foraging for plant material in the northern part of its range and for filth (feces) in the southern part of its range to defend nests against attack by different, but equally deadly, mass-attacking Vespa predators.
It’s possible that the ancestral behavior here is smearing plant juice around nest entrances, especially if you think that animal dung was not as prevalent in the time before bees were kept in boxes by humans or didn’t live around domestic animals. From there it would be a simple step to evolve collecting dung instead of plants, which is easier since you don’t need to gnaw the dung. It’s not clear, of course, whether either behavior is genetic or learned, but I suspect both have a least a partial genetic component. That itself could be tested in principle by seeing if the behaviors appear in bees that are raised “naively”, without the chance to learn from other bees. I don’t know if that is possible.
3.) Is this really tool-using by bees? The authors define “tool use” according to four criteria of Benjamin Beck as published in a 1980 book: an animal uses tools if it uses an environmental object (a piece of dung); if it alters the object in a way to make it more efficient to use (the bees mold the feces and apply them to nest entrances); if the user holds and manipulates the tool before or during use (collection and carrying of feces to nest and deposition as spots); and if the user “effectively orients the tool” (bees place the spots almost exclusively around the nest entrance, which is where predatory hornets enter). Since the bees do all of these things, the authors conclude that this is a case of tool use—the first known tool use in honeybees (a lot of other insect species use tools).
I’m not so concerned about whether this behavior falls into the category of tool use, whose criteria vary from investigator to investigator, as I am to understand how the “tools” work and how the spotting behavior evolved.
___________________
Mattila HR, G. W. Otis, L. T. P. Nguyen, H. D. Pham, O. M. Knight et al. 2020. Honey bees (Apis cerana) use animal feces as a tool to defend colonies against group attack by giant hornets (Vespa soror). PLOS ONE 15(12): e0242668. https://doi.org/10.1371/journal.pone.0242668
I’m curious about how this complex behavior evolved too. How did the smearing of plant juice get started. It seems you’d need quite a bit of it to deter predators. The behavior resembles the way mud wasps collect mud and smear it to build a nest. Perhaps the bees originally used smearing to enhance their hive structure.
Is it a particular plant-containing dung, or will any old dung do? And, how long does the effect last? Why not as a matter of course place dung there as a constant safety feature? Why wait to do it? Are tge bee stings ineffective against hornets?
Claiming that a bee “effectively orients the tool” by putting a dung pellet in the right place seems dubious. My sense of effectively orienting a tool means grabbing it by the handle in order to bring the business end to bear on a task. This doesn’t apply to dung pellets, whose wasp-repellent properties are spatially symmetric; there’s no sense in which their orientation matters. So the authors have construed “orientation” to mean “location” in order to shoehorn this behavior into the category of tool use.
I had thought that an easy way to prevent the hornets from attacking European (commercial) bee hives would be to make the entrance of the hive too small for the hornets to get in. It was interesting to see that if a beekeeper was going to do that, that he would have to use a metal barrier rather than a wood one.
That certainly rings a bell – though possibly from apiarist’s efforts to exclude other predators.
Hives have other requirements. In the buzz of other bells in my head from bee-keeping lore is that a number of species (or other regional groups) position workers at the entrances to air-condition the colony. By inference, predator-exclusion is not the only effective control on colony entrance size.
I’m wondering how that would play in wild colonies, which build their own nests from scratch. Or from a wall and a roof, then build the other walls to desired sizes.
Interesting! Of course the counter to this, for the wasp part, is selection to be attracted to dung.
That could lead to a shitload of wasted time, though.
But as the post says, the use as a tool is not the important thing here as compared to how the tool works, how it evolved.
I include here my comment that I enjoy your science post so you will not think they are not appreciated or read! Love this.
I would like to second Catherine’s comment. I always read and love the science posts, even though I seldom have any relevant comment to make. I learn a lot. Please don’t stop!
There’s a pre-print that I’ve heard is in review at Science that I think would make a good blog post, it’s called “Predicting future from past: The genomic basis of recurrent and rapid stickleback evolution”.
Sub
Twenty seconds to “cook” the hornet to death? Why wouldn’t hornet immediately start fighting his way out?
The balls are apparently hundreds of bees so they should mass a lot and likely mechanically keep the hornet from using its wings as well.
The kill rate of a hornet is estimated to be one bee every 14 s (if it needs to hunt through the hive).
[ https://www.nytimes.com/2020/05/03/us/murder-hornets-asian-giant-hornet-bees.html ]
Somewhere I read that it was one bee kill every four seconds, not 14, that would be a big difference. Thank you!
In the middle of a melee … could you find your way “out”. Assuming that there was an “out” (if the ball extends to the daylit entrance slot, then your eyes would give a choice of two, incorrect, directions to travel in.
Also, would the hornet recognise overheating as an existential threat? I know humans who have failed to appreciate the danger of overheating fast enough to retain consciousness, and they have (allegedly) better powers of forethought than the average insect.
Actually, considering one particular part-cooked caver … I’m inclined to credit the hornets in his comparison. Nice guy; mad as a hatter; thick as a brick.
If being “cooked to death” feels similar to a blood choke in jujitsu (as opposed to an air choke, immediate danger) that would make sense. Thank you!
She can’t get out–there are too many bees on her.
I’d be too busy decapitating bees to notice the temperature!
Had never seen an actual video before; it looked like less bees to make a bee ball than I imagined, I didn’t think the wasp would be visible at all. Thank you!
I hoped to see an article here when I read about the putative tool use – interesting analysis!
A context to the bee trait is this discovery from 2016 in bumble bees [ https://www.sciencemag.org/news/2016/10/hints-tool-use-culture-seen-bumble-bees ]:
That’s really cool!
Fascinating! As for learned vs evolved, understanding that there is similar behaviour (plant smearing), do other groups in other locations use dung? And can one mix in other bees from other groups, or would they get rejected? I am assuming that waiting to see if the behaviour evolves in introduced species would just take too long, but if it is learned, would it pass to newcomers who, if then isolated, would pass it to their offspring/new group?
The Guardian had a more “layperson” take on the PLOS One paper: https://www.theguardian.com/environment/2020/dec/09/honey-bees-use-animal-poo-to-repel-giant-hornet-attacks
Poo! That’s what the NHS say for the over 60s bowel cancer screening – poo! Why not say turd?
Ok, which animals supply the dung? That might make a difference depending on their diet?
If we experiment with spraying hives with faeces can we deter hornets from attacks? Presumably the honey scent is disguised? Would that harm ability of honey bees to find their way back to the hive?
Lots of interesting questions arise!
The balling behavior already exists in domestic honeybees. When a honeybee queen begins to run out of the sperm stored from her youthful mating, she has trouble laying fertilized worker eggs, producing mostly unfertilized drone eggs. At this time the workers will begin rearing some of the worker eggs to be queens and will eliminate the old queen by balling. It doesn’t seem too much of a stretch to apply that same behavior to a hornet invader.
Oh that makes perfect sense!
Yesterday, I had read WEIT early in the day and was kept busy so I didn’t read this until today. I was aware of the balling vs Asian Giant Hornets already. It was very interesting to see dung being used for an alternative approach.
I must admit I’m slightly disappointed after reading the title (well only slightly, the article is fascinating enough). I thought we would have seen bees slinging shit onto the hornets (bomber beetle-like) 🙂 , but my imagination ran away with me.
It would be interesting to find out whether it is the shit itself or the shit masking the marking pheromones that gives some protection.
OT, but I read “it is the shit itself or the shit masking the marking pheromones” as marketing promotions.
As in wasps saying “Stay away from that hive, I’ve heard they’re slinging some heavy shit!”
Me too, read it the same way. I had saved this article to get back to and it took a while more than I thought. It was well worth the read.
Just read this after Jerry pointed out not many people reading it.
Very interesting, though “tool use” does seem a bit of a stretch. Its a bit as though the bees have developed a moat to go around their castle walls, now they just need to invent crossbows and boiling oil as well.
I know we shouldn’t moralise species like this, but wasps are truly vile creatures and anything that protects other species against them is worth a cheer.
I love the phrase “Although we did not study how filth foraging is organized”!
I wonder if the animal dung contains some of the same plant extracts as what they are smearing on the hive entrances. The pandas rolling in animal dung are apparently doing so to coat themselves in plant extracts that make them less sensitive to the cold, so maybe similarly useful extracts are also passed through to dung?
Smearing hives with animal dung by honey bees as defence mechanism against hornet seems less convincing as evolutionary adaptation as the artificial wooden movable frame hives were started to be used only less than 150 years ago, and the study has been conducted on the artificial modern huves rather than on feral colonies. Further, in India, upto 50-60 years ago, almost all rural households and presently in remote rural houses, people used to/have been using cattle dung for plastering mud floors and walls of their houses it being providing temperature regulation / insulation, moisture proofing, insect repellant and antimicrobial.