An aggregation of the Asian giant rock bee (Apis dorsata) photographed at the IISc campus, Bengaluru, by Thresiamma Varghese. Notice that the bees have abandoned one colony and left behind the naked wax comb.
- Honey bees have been held up for millennia as the prime example of a well-functioning, conflict-free society, a paragon of societal virtues well-worth emulating by humans.
- Karl von Frisch famously said, “The life of the bees is a magic well. The more one draws from it, the more richly it flows.” More recently, we have also drawn more surprises from this.
- Some of the most interesting new knowledge that we have gained by putting honey bees under the scanner of logic, theory and experimentation paints a not-so-rosy picture of honey bee society.
- We have uncovered many examples of strife and dysfunction in this “divine being” and “a prime favorite of the gods”.
This is the first part of a two-part essay. The second part will be published in the next edition of ‘More Fun Than Fun‘.
“Its sustained flight, its powerful sting, its intimacy with flowers and avoidance of all unwholesome things, the attachment of the workers to the queen—regarded throughout antiquity as a king—its singular swarming habits and its astonishing industry in collecting and storing honey and skill in making wax, two unique substances of great value to man, but of mysterious origin, made it a divine being, a prime favorite of the gods, that had somehow survived the golden age or had voluntarily escaped from the garden of Eden with poor fallen man for the purpose of sweetening his bitter lot.”
These immortal words of William Morton Wheeler (1865-1937) have not and cannot be bettered. If Edward Osborne Wilson (1929-2021), who is justly credited with writing like a poet, could not resist the temptation to quote these words, how can a mere mortal like me?
Honey bees can do all that Wheeler gives them credit for and more because of their remarkable anatomy, physiology, behaviour and social organisation, understanding the evolution of which is one of the brightest chapters in modern biology. Honey bees live in populous, rather crowded colonies with 10 to 50 thousand or more individuals. These individuals are of three kinds, a single large fertile female bee (the queen), a few hundred or a thousand male bees (the drones) and tens of thousands of small, nearly sterile female bees (the workers), all jostling on top of each other.
Their nest is a palace of wax; the workers drink honey and secrete wax molecules out of their bodies, which they manipulate with their mandibles, legs, and antennae to fashion two exquisite arrays of hexagonal cells placed back-to-back. The cells are used to rear the brood – queen, drone, and worker brood, all requiring cells of different sizes – and to store honey and pollen. The manner in which the worker bees combine instinct and innovation to achieve the perfect hexagonal shapes and deviate as little as necessary from that perfection to accommodate irregularities in the substrate and to join fragments of comb containing cells of different sizes is one of the most intriguing aspects of honey bee behaviour.
Drone bees do no work on the nest, staying for varying periods of time until they find a mate. Worker bees feed their brothers as long as the food is plentiful, but drag them out of the nest when it is not. Drones visit mysterious (to scientists) drone congregation areas and mate with virgin queens in flight, dying in the process as they leave behind a portion of their genitalia in the queen’s vagina to serve as a mating plug, the function of which is also somewhat mysterious as the queens will mate repeatedly with as many as 20 or more males in succession.
Worker bees perform essentially all the tasks needed to run a colony in an impressively orderly manner according to an age-dependent plan. Young bees start as cleaners, transition into builders as their wax glands mature, then into nurses as their mandibular glands mature and then into guard bees, positioning themselves at the entrance to sniff out intruders of the honey bee or other variety. Finally, they venture out of the nest to gather nectar and pollen and return to the nest to perform a unique dance to recruit naïve bees to the food source, giving them information about what they have found and how much of it and even how to get there. The German zoologist Karl von Frisch won the Nobel Prize in 1973 for deciphering the honey bee dance language.
Nectar-laden foragers will have to queue up at the nest entrance to be unloaded by specialised unloaders. If they have to wait long, they will not curse but will understand that the colony no longer needs what they have brought and try something else, water perhaps. It is this kind of decentralised, self-organised collective behaviour that constitutes the wisdom of the hive and is at the heart of the success of insect societies, including those of ants, termites and wasps.
The bees’ age-dependent change in behaviour (technically called age polyethism) is sufficiently flexible to deal with unexpected demographic changes in the colony. If we experimentally remove old bees, ‘precocious foragers’ will begin to forage at an age inappropriate for a normal colony. If we remove young bees, larvae will be cared for by ‘over-aged nurses’ as foragers revert to nursing duties, regenerating their atrophied mandibular glands as necessary.
There are also other ways bees respond to the colony’s needs, overriding the etiquette of age-appropriate behaviour, if necessary. If a dead bee is creating a stench, the workers most sensitive to the smell will assume the role of the undertaker. If the more sensitive variety of bees is missing, the stench has only to become a little more intolerable for some other bee to drop what it is doing and attend to the task of ridding the colony of the offensive smell. Having mated with many males, honey bee queens gather a lifetime supply of sperm, which they store and nourish in a pouch called the spermatheca. They simultaneously use sperm from many males to fertilise their eggs and thus produce several patrilines of daughters, with different sensitivities for different tasks.
When the colony has grown large, workers will build large queen cells and feed the larvae in them with a special ‘royal jelly’ to produce new queens. At about the time of the eclosion of the daughter queen/s, the mother queen will depart with a fraction of the workers, bequeathing the remaining workers, the nest and its food stores to her daughter queen. The mother queen and her entourage of workers leave the nest and settle down on a tree branch or rock in a temporary swarm.
A few scout bees will leave the swarm in search of suitable nesting sites to construct a new home. Scouts will return to the swarm and use their dance language to inform the rest of the bees of their find. When several foragers dance to inform naïve bees about different food sites, different groups of bees can go to each profitable site and collect pollen or nectar, as the case may be. When several scouts dance to advertise different potential nesting sites, the bees have to find a way of agreeing to go to a single site, and preferably the best on offer.
In the words of Thomas D. Seeley, who has made the most extensive studies on house-hunting honey bees, “these house hunters evaluate the potential dwelling places they find; advertise their discoveries to their fellow scouts with lively dances; debate vigorously to choose the best nest site, then rouse the entire swarm to take off; and finally pilot the cloud of airborne bees to its home.” In short, “they will hold a democratic debate to choose their new home.”
It is not surprising then that honey bees have been held up for millennia as the prime example of a well-functioning, conflict-free society, a paragon of societal virtues well-worth emulating by humans – what can be a better example of this attitude than the quote from William Morton Wheeler at the head of this article.
Karl von Frisch famously said, “The life of the bees is a magic well. The more one draws from it, the more richly it flows.” In more recent times, we have also come to realise that the more we draw from it, the more surprises we draw from the well. Some of the most interesting new knowledge that we have gained by putting honey bees under the scanner of logic, theory and experimentation paints a somewhat different and not-so-rosy picture of the honey bee society. We have uncovered many examples of strife and dysfunction in this “divine being” and “a prime favorite of the gods”.
The theory of inclusive fitness
When it to comes to the evolution of sociality, inclusive fitness theory is the mother of all theories. In 1964 W. D. Hamilton formalised the idea that the ‘fitness’ that mattered to natural selection could be got not only by having offspring, as was generally assumed until then, but also by helping the survival and reproduction of genetic relatives. Offspring matter because they carry our genes and if that’s all that matters, then our relatives also matter because they too carry copies of our genes. To deal with situations where organisms combine begetting some offspring and also caring for some relatives, Hamilton proposed the idea of ‘inclusive fitness’, the sum of direct fitness gained through producing offspring and indirect fitness gained through caring for relatives. The bottom line is that the evolutionary value of anyone, be it offspring or relative, is measured by the proportion of genes you share with the recipient of your care.
In the world of humans and most other sexually reproducing organisms, there are no relatives to whom we are more closely related than we are to our offspring. We share one-half of our genes with our offspring and our full siblings (this is usually represented by the number r of value 0.5; i.e. r = 0.5) but even less with all other relatives. So, no relatives can be more valuable than our own offspring. But imagine how our social lives would be altered if we were more closely related to our siblings than to our offspring!
Haplodiploidy
Well, there is no need to imagine, for that is precisely what happens in the insect order Hymenoptera which contains the ant, bee, and wasp societies. In their peculiar mode of genetics, which is called haplodiploidy, females develop in the normal way from fertilised eggs. They, therefore, contain two sets of chromosomes, one maternal and one paternal. On the other hand, males develop parthenogenetically from unfertilised eggs and therefore contain only one set of chromosomes, only the maternal set. Strange as it may seem, in haplodiploid species, males neither have fathers nor sons, although they do have grandfathers and grandsons!
Haplodiploidy also creates peculiar asymmetries in genetic relatedness among relatives. Queens can choose to fertilise their eggs or not. They allow sperm to flow from their spermatheca to the oviduct and fertilise their eggs when they need to produce daughters and prevent sperm from flowing to the oviduct and lay unfertilised eggs when they need to produce sons. It is worth reflecting on the fact that these queens have complete control over the sex of their offspring, something that we are still unable to do.
Because females are diploid, they produce haploid eggs by the normal process of meiosis, where there is a reduction in the number of chromosomes from the diploid number characteristic of the species (2N) to half that number (N). A female’s paternal and maternal chromosomes are randomly distributed between different eggs to reduce the number of chromosomes. Therefore, a female’s eggs are related to each other by one-half (r = 0.5). Males, being already haploid, produce haploid sperm by the process of mitosis without any further reduction of the number of chromosomes. The sperms of a male are all clones of each other (r = 1.0).
Two full sisters sharing the same father and mother are related to each other by three-quarters (r = (0.5+1.0)/2 = 0.75). Thus, females are more closely related to their full sisters than they are to their daughters. It follows that it is more evolutionarily profitable for a honey bee worker to care for her full sisters than to produce her own daughters. This may well be the reason why natural selection has tolerated that honey bee workers lose their ability to mate and produce daughters; they spend their entire lives caring for the queen’s offspring instead.
But honeybee workers are more closely related to their own sons (r = 0.5) than they would be to the queen’s sons (their brothers; r = 0.25). This is because they share half their maternal genes with their brothers and none of their paternal genes. This may well be the reason why they have not lost their ovaries and can continue to produce sons.
Queen-worker conflict over male production
Haplodiploidy creates so many quirks that even Hamilton did not see all of them. The young Christopher K. Starr, then a graduate student at the University of Georgia, realised that while honey bee workers should prefer their own sons (r = 0.5) over their brothers (r = 0.25), they should also prefer their nephews (r = 0.75/2 = 0.375) over their brothers if the queen is singly mated and the workers are full-sisters of each other. This is not good for the queen. But the queen, as we have seen, mates with several males, and the workers are often half-sisters of each other (r = 0.25). Now the nephews (r = 0.25/2 = 0.125) should be the least preferred option.
Contributing a single-author chapter to the multi-author volume on Sperm Competition and the Evolution of Animal Mating Systems (a rare honour for graduate students to this day), Starr wrote presciently that “workers are related to the three possible classes of male offspring [ignoring their own sons] in the following way: full sisters sons > brothers > half-sisters’ sons. They are, on average, less related to nephews than brothers whenever [queens mate with more than two males] and should prefer that the queen lay all the male eggs. Workers would therefore be expected to interfere with each other’s reproduction.”
The Polish honey bee biologist Michal Woyciechowski in collaboration with his mentor Adam Lomnicki built a mathematical model which confirmed that “if workers are able to perceive the multiple mating of the queen, they can apply a conditional strategy: ‘take care of nephews if the queen mates with one male, take care of brothers if it mates with more than two males'”.
Adam Lomnicki (1935-2021) was the most distinguished ecologist and evolutionary biologist in Poland in his time. His numerous researches into mathematical population ecology were crucial in resolving the controversy between individual and group selection and fostering the modern understanding of the natural regulation of animal populations, as set forth in his monograph ‘Population ecology of Individuals‘ (1988).
More importantly, Lomnicki was a great mentor to generations of ecologists and other scientists in Poland and was an effective bridge between Poland and the West. Lomnicki frequently held numerous meetings, schools and workshops with some foreign invitees, to which students from all over Poland travelled to Warsaw in large numbers. I had the privilege of being invited to one of Lomnicki’s durbars in 2001 and found him to be a brilliant mind with abundant charm and a gracious host.
Worker policing
Francis L. W. Ratnieks, then at Cornell University, took this idea a step further and showed that “Population-genetics simulations of the fate of a “police allele,” which confers any marginal increase in policing behaviour to the workers carrying it, indicate that such an allele will invade and spread to fixation provided that queens mate two or more times”. As he noted, “Reproductive harmony may, therefore, and counterintuitively, result from lowered relatedness among workers”. Labelling any behaviour that the worker bees might use to interfere with each other’s reproduction as “Worker Policing”, Ratnieks made this concept instantly famous as a conflict resolution mechanism.
Ratnieks and Kirk Visscher performed two clever experiments to prove the idea of worker policing. In one experiment, they offered honey bee workers a choice of eggs laid by the queen and the workers of their own colony and found that only 2% of worker-laid eggs but 61% of queen-laid eggs remained after 24 h. Clearly, the worker laid eggs were being destroyed preferentially by the workers in the experiment.
In the second experiment, they offered workers a choice of queen laid eggs and worker laid eggs, but both from a different colony. Here too, they found that less than 1% of worker-laid eggs and 59% of queen-laid eggs survived. The second experiment showed that workers were able to discriminate queen-laid eggs from worker-laid eggs even when both the queen and the workers were from another colony. Obviously, the workers were not destroying eggs based on directly measuring their genetic relatedness to them. We now know that queens mark their eggs with a pheromone which allows the workers to tell them apart.
The point to note is that there is scope for queen-worker conflict, but that conflict is managed by worker policing. And it’s not as if workers keep on laying eggs which are continuously eaten by other workers—that would be very wasteful and not qualify as efficient conflict management. As a response to the threat of policing, workers seem to show self-restraint and lay very few eggs in the first place. In another study, Kirk Visscher showed that worker-produced drones accounted only for 0.12% of all males produced in normal colonies. Will the workers lay many more eggs if we somehow eliminate the policing workers? That is hard to do, but it is possible to remove the motivation for policing and thus eliminate policing itself. If the queen dies, there is no incentive to police, as there will be no more brothers to care for, and many more workers will indeed begin to lay eggs.
Thus, in a review of the literature in 2006, Francis Ratnieks, Kevin Foster and Tom Wenseleers conclude that “Reproductive conflicts are widespread, sometimes having dramatic effects on the colony”. However, three key factors (kinship, coercion, and constraint) typically combine to limit the effects of reproductive conflict and often lead to complete resolution.”
In animal societies, the potential for conflict is inevitable because the genetic interests of the interacting individuals are never entirely aligned unless they are clones of each other, like the somatic cells in our body. The consequences of letting conflict simmer and flare up can indeed be serious. Consider the case at hand—queen-worker conflict over the production of males. If workers focus their time and energy on laying their own male-destined eggs because they prefer sons over brothers, colony efficiency will surely suffer.
An interesting twist is that loss of colony efficiency itself can be a selective force favouring workers’ self-restraint. Indeed, it has been recognised from the beginning that the colony efficiency argument can favour worker policing independent of the relatedness argument. In other words, workers might be selected to police each other because they are less related to their sisters’ haploid eggs than to the haploid eggs of the queen (relatedness argument) or because in the absence of such policing, the colony becomes inefficient, leading to loss of fitness to everyone (efficiency argument).
We are not yet in a position to be entirely clear, which is the better argument for worker policing in different species and in different situations. We sometimes resort to the efficiency argument when the relatedness argument fails. This means, of course, that there is much more to learn about how honey bees and other insect societies achieve a balance between cooperation and conflict, which is good news for young researchers.
Questioning the fundamentals
Madeleine Beekman and Benjamin Oldroyd, both now at the University of Sydney in Australia, are fine examples of young people who have taken up the challenge of significantly expanding our knowledge of conflict and its resolution in honey bees. They have succeeded because they have not been afraid of the possibility of overturning long-held beliefs that are comfortably consistent with existing theory – they have not been afraid of rocking the boat.
On one occasion, they wondered, rather philosophically whether, by uncovering the potential for queen-worker conflict and then discovering worker policing as a mechanism to subvert that conflict, we have “come full circle to once again regard insect societies as being comprised of obedient workers, even if only by force?”
Along with graduate student Michael Holmes and two other collaborators, they asked if the long-held wisdom that the workers produce only 0.12% of the drones in honey bee colonies was, in fact, true? Their strategy to verify this was as simple as it can ever get—look harder. They looked at more colonies, more workers, looked at different periods of the honey bees’ reproductive cycle and use molecular markers to determine who the mothers of the drones were, more precisely than ever before.
Lo and behold, they were rewarded with the most unexpected results. They found that worker contribution to male production was 4.2%, a value 45 times higher than Kirk Visscher found in 1989. More interestingly, they found workers were smart enough to step up their cheating when the benefits were highest, i.e. during the reproductive (swarming) period when the drones have the highest opportunities for mating, during which they accounted for 6.2% of the males produced. Dissecting large numbers of workers, they found workers were 100 times more likely to have developed ovaries than previously reported.
Perhaps their most interesting finding was that a small number of patrilines of workers excelled at cheating. It is easy to see how the cheating gene transmits from father to daughter to grandson even though males have neither fathers nor sons. Thus, some honey bee workers specialise in getting direct fitness and not relying only on indirect fitness via caring for the queen’s offspring. That honey bee workers can do so even though they have lost their genitalia and cannot mate to produce daughters is testimony to the power of natural selection.
In retrospect, it is easy to see that natural selection would be powerless to eliminate cheaters until they overwhelm the system and lead to its collapse. So, we should always expect a small proportion of cheaters in any well-organised system – when the cost of eliminating cheaters is more than the cost of tolerating them. That cheating is not a random phenomenon but is a speciality of certain patrilines suggests a genetic basis for cheating and an arms race between policing and cheating.
What I have described here is only the tip of the iceberg: we have since uncovered even more surprising and powerful examples of cheating in honey bees in recent decades. So, I will continue this discussion in the next instalment of More Fun Than Fun.
Raghavendra Gadagkar is a Department of Science and Technology (DST) Year of Science Chair Professor at the Centre for Ecological Sciences at the Indian Institute of Science, Bengaluru.