How Did Bees Go from Solitary to Social Living?

Solitary Nesting, Primitive Sociality & Eusociality in Bees

The first bees were solitary and many species continue living these lifestyles, so how did honey bees grow so large? The development into group living, task allocations, reproductive priorities and even a simple “mummy said we should share” was built through trial and errors. How did these Hymenopterans go from a solitary lifestyle like that of the blue banded bee, Amegilla cingulata to group nesting (“gregariousness”) to primitive sociality (bumble bees) and into full sociality (honey bees, ants, termites, wasps)?

This article examines the lifestyle of purely solitary bees and those solitary bees who nest in groups, whether as individuals in one location, or by sharing nest entrances. Most wild bees fall into this category. It then examines how environmental conditions cause bees to switch group and solitary living through Halictus rubicundus, the orange-legged furrow bee. Subsociality is explored through the bumble bee lifestyle and the reproductive competition queens have with her daughters. It concludes with Apis mellifera, the model bee species for sociality.

An Overview on the Evolution of Sociality

Social living is rare, and yet there are cases where it has independently evolved (or regressed) in a species. Ants, wasps, termites and bees reached eusocial living through different means, as each group’s ancestor had the right preconditions (and in the right order), mixed in with the right luck- it’s no good to be a pre-social species and then exit by way of the dinosaurs.

With these precursors in mind, the first bees to enter “subsociality” would have shared traits like mothers guarding nests, and those same mothers being present at the same time as her offspring. Some of the offspring would stay in the nest (whether to clean it, defend it, reuse it, or all of the above). These bees were operating independently, and we see evidence of similar cooperation in various solitary bee species (e.g. Ceratina, Mikat et al., 2016).

The next steps towards sociality would have been progressive (rather than “mass”) provisioning of nests. Our ancestral bee returned to her nest often to provide food, increasing the parent-offspring contact time and turning broodrearing into a continuous activity. Why does this matter? Consider the pheromonal changes as an example: The brood can begin producing scents that have a positive or negative effect on the bee, and over evolutionary time these changes become refined into signals. If the brood releases chemical Y it is hungry and the adult is required to forage, if it’s chemical Z then it is cold and heating is required, and if it’s chemical X, it’s a powerpuff girl.

Gregarious Nesting is a Step Above Pure Solitude

Gregariously-nesting bees are females who nest in close physical proximity to one another (be they the same or different species) without cooperating in broodrearing or labour. Each female builds her own nest and supplies her own provision; there is no “direct” advantage from being near their neighbours’ offspring. Gregariousness is most common in ground-nesting genera (Andrena, Nomia, Halictus, Lasioglossum) and some cavity nesters (Osmia, Megachile).

The untrained individual may mistake bees nesting in close proximity for sociality, and upon a closer inspection they’ll realise that gregarious nesting lacks eusocial criteria like cooperative brood care, reproductive division of labour and overlapping generations sharing the nest. (There are other factors quoted in the literature throughout the years, for example Crespi, 1994 stated that sociality required “food-shelter coincidence, strong selection for defence, and ability to defend nests”). Gregariousness still matters for social evolution, as the aggregation of bees can:

• Improve mate finding (boys will have better chances around more girls).
• Reduce predation risk, owing to the dilution effect – if you’re the fifth bee in line, a predator may be sated/give up after the fourth bee, sparing your brood.
  • The flipside could be that more predators/parasites are attracted to your nesting site.
• Signal the suitability/unsuitability of an area for nesting. The more bees in a single spot, the more likely it is a good one – social proof.
  • Caveat: More bees in a spot, the more competition for resources – it’s a tradeoff.
• Increase opportunities for kin interactions, as daughters and siblings nest near their mother (“natal”) sites.
  • Relatedness among neighbouring bees is higher than random and raises the likelihood that kin-selected behaviours emerge, even if they are not (yet) expressed in that population.

Haplodiploidy & Cavity Nesting: From Subsociality to Eusociality in A. mellifera

Beyond progressively provisioning nests, another factor in the transition to A. mellifera’s sociality is in the choice of nesting site. A. mellifera nests in cavities with tree hollows being the preferred option. These nests were more defensible and reusable than mudbanks or bare soil, so staying at the nesting site for successive generations “made sense” to our prehistoric bee.

Another (albeit not-as-important) shift was in haplodiploidy (males arising from haploid [unfertilized] eggs and females from diploid [fertilized] eggs), which would have occurred towards the later stages of the eusocial transition. Haplodiploidy is relevant to worker relatedness and worker policing. Relatedness because of the sexual asymmetry: sisters are more closely related to each other than their offspring, so the kin selection theory (Hamilton, 1964) favours the workers prioritising their queen’s eggs over that of a sister worker. Worker policing is when the adult bees remove worker-laid eggs to preserve the queen’s role as the sole egg layer (math below).

By now there is a divide between daughters who “stay” and those who depart in search of new nests. Those that stay have the advantage of safety in numbers, along with the benefits of group defence and food gathering, at the expense of their reproductive fitness. Those that dispersed had the risks associated with establishing a new nest (predation, food, resources, bad luck etc…) but could rear their own offspring. A nest’s survival increased with more defenders, and its foraging capabilities improved with increases to the number of gatherers¹; helping became more profitable than leaving.

The transition from “helpers” into “workers” involved shifts in developmental pathways associated with nutrition, hormones & gene expression. The primitive helpers could still reproduce albeit at a reduced rate (as some social bees still do today, e.g. bumble bees), however there was competition amongst them to attain the “dominant female” status. Some sort of task-based allocation would have had to occur prior to the differentiation of a non-reproductive worker caste. If it happened the other way around and the caste system came before the task system, then the “sterile” worker would have floundered around and performed any task it deemed fit, rather than the one that is in the best interest of her proto-hive.

This is a subheading for explaining the math of haplodiploidy & Apis mellifera worker relatedness. I originally had it as a footnote and later realised it took up an entire page. Enjoy

Math for worker policing was extracted from the eponymous article

As queens mate with multiple drones, each worker bee will have a different father and the same mother queen; they are considered full (or in social bees, “super”) sisters. Each worker receives half of her genes from her mother and half from her father. The queen passes on a random half of her genes to each egg she lays. Since the drones are haploid (providing only the set of genes they received from their mother queen), the sperm from that drone is genetically identical across all its sperm cells. Therefore, for two sisters:

• From the mother (queen), there is a 50% chance of sharing the same allele (“gene variant”) for any gene because they receive a randomly selected half of her genes.
• From the father (drone), they always receive the same alleles because all his sperm are genetically identical.
• This adds up to an average of 0.5×0.5=0.25 (25% from the mother’s side) + 0.5 (50% from the father’s side since the sperm are identical) = 75% relatedness.

Half-sisters share the same mother but have different fathers (since the queen mates with multiple drones). They will still share about 50% of their genes through their mother, as mentioned above. However, because they have different fathers, it is only the mother’s genes that are shared, and the father’s genes are excluded from the calculation. Thus, the relatedness among half-sisters is 0.5×0.5=0.25; 25% from the mother’s side only.

What if that half-sister was to lay eggs that hatch into a drone? Since the egg is unfertilized, there is no male genetic material and it will emerge as 50% of the worker bee, so the relatedness between a worker and her half-sister’s son is 0.25 * 0.5 = 12.5%. There you have it, workers police the eggs of one another since those drones are less genetically related than those laid by their mother queen.

Nest Guarding Determines Offspring Survival: A Sociality Precursor

The standard solitary bee approach is to provision the nest then depart. This poses a risk that the offspring will be parasitised by wasps and/or cuckoo bees or preyed upon by other wasps and ants. Mikat and colleagues found that nest guarding by the mother bee matters most in the survival of the offspring, and propose it as a driver of eusociality (Mikat et al., 2016). In studying bees from the genus Ceratina, the authors found that mothers who guarded their nest had a higher survival than those who did not.

The “protection” arose in response to the presence of a species’ natural enemies as a way to prolong parental care. If a bee lived in a perfect world with no downsides, she would provision her nest, seal it and then move onto a new nesting site to further spread her genes without change. If however, the bee becomes aware that her nest is a target for predators and parasites, and that it is more effective to stay protecting one nest rather than starting multiple, then she would begin shifting strategies.

Not all guarding strategies are created equal: in this study, C. cucurbitina bees were more active in protecting the nest, crawling through the perimeter, whereas C. chalybea either guarded the cell at its outermost chamber, or plugged it and departed like other solitary species (Mikat et al., 2016). As you may have figured out, the C. cucurbitina strategy is most effective for offspring survival.

The Solitary Nesting Process

The solitary bee is a strong, independent woman who need no man. She starts by selecting a nesting site (ground/trees/hollows/cavities), constructing brood cells, provisioning them with pollen-nectar loaves (not quite “bee bread”), or liquid-ish pollen mass (or oils, depending on the species), sealing the nest and off she goes elsewhere. There is no brood care after oviposition & survival of her offspring depends on nest architecture, provisions, timing and luck, assuming no predator or parasite gets to it first. Now the details:

Nest Construction & Excavation Precision for Maximal Offspring Survival

Females assess their nesting substrate for drainage (avoids flooding), thermal buffering for temperature control, structural stability and parasite access post-sealing. Get one of those wrong and the offspring will not survive, and worse, the mother would not have known of their demise, as she is not around them. Bees from solitary families mainly nest in the ground (e.g. the mining bees or sweat bees), leafcutter bees prefer pre-existing cavities like hollow stems and wood-beetle borings, and a variety of genera will be attracted to constructed cavities in the form of bee hotels.

Once the nest is selected, excavations and/or renovations begin; the soil is loosened and cavities are cleared of debris. This is energetically costly and an examination of the female’s nest reflects her health: deeper nests with more cells denote a stronger foundress. Once cleared, the cells are lined to create a sealed microenvironment that’s waterproof, protects the eggs from microbes, regulates oxygen transfer, provides mechanical stability to the cavity and delays (but not prevents) parasite/predator penetration. Nesting materials include:

• Soil/mud (many Andrenidae, Halictidae).
• Leaves (cut or macerated; Megachilidae).
• Plant resins (resin bees).
• Glandular secretions (cellophane-like polyester linings in Colletes species).

Solitary Bees Provide Larval Food En masse

Now that the nest is built, it is stocked with food (“mass provisioning”) in the form of nectar and pollen (some bees like Macropis fulvipes collect floral oils – see Dotterl & Schaffler, 2007) Again, what you find in each cell reflects the mother’s health and foraging success (which also ties into the season). One added factor is that female larvae get larger provisions, and this is a common theme in bees (and other insects) as they are the ones who do most of the reproductive work.

Eggs are laid atop the provision mass or in suspension right above it. As in other bees, fertilised eggs develop into females and unfertilized eggs male (yes yes Cape Honey bees can produce females from unfertilized eggs – that’s beside the point). Once the larva hatches it does not need to venture far to reach the food it needs to complete the larval phase. Male eggs are located closer to the nest entrance and with less provisions – this is an adaptive sex allocation strategy since the loss of the males (due to weather, predators, parasites and/or bad luck) is not a big deal to the solitary species – males are there for reproduction and nothing more.

Nest Plugging & Departure of the Adult Bee

The final step in the solitary nesting process is the sealing and/or compartmentalisation of the cells – the same material used in nest construction is used to create partitioning walls between neighbouring cells, and for later sealing the nest entrance. Nest sealing is done for pathogen control, hindering parasite/predator movements, buffering the elements and localising failure points if one of the brood dies due to the aforementioned conditions. Sealing is final, and the female does not reopen cell. The larvae hatches, eats and undergoes pupation until it’s time to break the seal and emerge as an adult to being the cycle anew.

Solitary-to-Social Transition Species via Geography: Halictus rubicundus

Evidence for a genetic control to sociality is found in the orange-legged furrow bee, Halictus rubicundus. The sweat bee family Halictidae has multiple origins of sociality, and these conclusions were drawn based on climate: solitary species live in cold climates where there is not enough time to produce multiple broods required for sociality, unlike those who live in warmer regions in lower latitudes (multiple sources – see for example Danforth 2002).

For H. rubicundus, social populations are found in areas with longer growing seasons (e.g. New York or southern Ontario) and solitary ones in colder climates like Scotland, Alaska or the Italian mountains. If you’re thinking “hang on, these are all cold places!” you are correct, this bee is a holarctic species whose natural range spans the colder regions.

The 2002 study by Soucy & Danforth showed a genetic element to the expression of social and solitary behaviour in Northern American H. rubicundus populations (Soucy & Danforth, 2002). This expression was distinct from the European populations, owing to geographic barriers across their native range (Michels et al., 2024). The species is flexible, as when solitary populations in cool climates are transported to warmer climates, they begin expressing more social behaviour and producing multiple generations of brood (e.g. Field et al., 2010).

Modelling work found similar results, with longer season lengths (and subsequent faster brood development owing to higher temperatures), favour the emergence and maintenance of a social strategy, whereas shorter seasons and cooler climates favour solitary lifestyles (Ruttenberg et al., 2024). The authors’ model suggests that a “sweet spot” for season length is between 135-155 days, and this supports both solitary and social species coexisting. They tied this to a real-world example of Lasioglossum baleicum females in Japan, who produce both eusocial and solitary nests within the Goldilocks climate of Hokkaido.

Primitive Sociality in the Bumble Bees: A Nesting History

I’m told that calling bumble bees “primitively social” is a misnomer because “in most species, castes are clearly recognizable by size and thus are morphologically distinct, and females also store food in cells different from the brood cells, both characteristics of highly eusocial bees” (Amsalem et al., 2015). This classification could have stuck because bumble bees have both a solitary and social phase, with B. terrestris as an example:

• The Solitary phase begins when the founder queen emerges from the insect-equivalent of hibernation (diapause) and searches for a nesting site. The queen builds several wax cells, provisions them with pollen and lays the first batch of eggs. She does not leave the nest, and continues broodrearing activities (foraging, cleaning, guarding, nursing etc..) while the larvae develop. When the larvae commence pupation, she builds more wax cells to prepare for the next generation of brood.
• The social phase begins when the first generation of daughter “worker” bees emerge. The queen stops foraging and spends the rest of her life laying eggs, whilst the workers take over all nest duties. This phase continues until the end of the foraging season, when the colony begins producing the reproductive generation (“gynes”).

Another reason for the “primitively social” moniker is anarchy within the latter stages of the social phase. When the queen starts laying male reproductives, the workers that possess activated ovaries also begin laying eggs (which also hatch into males as they are unmated)². These rogue layers become aggressive towards the queen and other reproductive workers, and is termed the “competition phase” of the colony cycle (see Hammond & Keller, 2004 for more on male parentage in social insects).

Is B. terrestris Primitively Social Because Her Colonies Do Not Overwinter?

From the math above, the lack of polyandry in B. terrestris makes it favourable for workers to raise their own male offspring (“sons”), and while worker policing is present, it is not always sufficient. Towards the end of the foraging season, the bumble bee workers scramble over one another to ensure the next generation is mated and dispersed at the cost of nest harmony, since colonies do not survive winter. Compare that to (relatively) more orderly reproductive process of swarming in honey bees and you can appreciate how their sociality may be viewed as “primitive”.

One more note: bumble bee queens have a distinct “switch” to reproductive offspring production (Amsalem et al., 2015), whereas honey bees do not – the latter colony continues producing worker bees alongside drones and queens, it does not put a stop on worker production as the honey bee colony is expected to persist for year-on-year, with the periodic dispersal of swarms for the purposes of reproduction. So, one reason for egg-laying competition in Bombus species could be because of the sudden switch from workers (from “nest maintenance”) to reproduction (and subsequent “prepare to abandon ship nest”).

Why are Honey Bee Colonies so Big?

Successful eusocial living can be achieved when populations are in the hundreds, so how did A. mellifera colonies grow into the tens of thousands? Several reasons, the first being honey as a food storage. This enabled the species to overwinter in the colony rather than shutting down shop and entering diapause. As long as food is available, the colony remains viable in times where other insects are dormant. What’s more, an overwintered colony is still active in the centre of its cluster, with the “winter bee” caste being a long-lived and generalist denomination.

With food and shelter taken care of (as mentioned above), we have reproduction. The reusable tree cavity that’s filled with honey gives bees optionality in reproducing according to their requirements, not just that of the season. Solitary bees are at the mercy of the climate, and the number of brood periods is dependent on daylight availability (e.g. Lasioglossum malachurum, Richards, 2000). Honey bee swarming is delayed reproduction, and can occur when the colony is overcrowded, not just when the weather is right³.

Division of Labour & No Reproductive Competition Makes for Greater Cooperation

With regards to division of labour, the departing swarm leaves the colony with sufficient resources for the next generation to build upon, preserving the eusociality arrangement and ensuring that the reproductive hierarchy does not breakdown. Yes, in the early days after a swarm departs there are anarchist worker bees who lay eggs, however after a week or so, those anarchists are gone and the colony resumes normal function (Oldroyd et al., 1999). The swarm arrives at its new home and the workers prioritise comb building until the queen can restart egg laying. From there, the workers resume a typical division of labour process, and the new colony is deemed “established”.

Lastly, honey bees have extreme (and flexible) task specialisations. The same workers who do broodrearing can be turned into foragers in times of extreme need (or artificial manipulation, as in experiments; Lourenco et al., 2019). Combining this with division of labour and you have an efficient colony where the individuals work towards the betterment of the whole rather than themselves, giving us the “superorganism”. Sub-social bees like bumble bees have this to a lesser degree as you’ve read above, since the reproductive competition has workers laying eggs in antagonism to their queen.

Read the whole post and found puns, typos mistakes? Voice your displeasure with a strongly-worded message to admin@tgbee.com.au, or by filling this form.

Footnotes

1. I’m hesitant to use “division of labour” as this is a set of “rules” bound by hormones and pheromones – our pre-social-bee-cluster would have been unlikely to possess the traits that mark a true division of labour, so these bees would have been acting as “solitary units in one big nest”. ↩︎
2. The math on B. terrestris’ worker relatedness, according to Amsalem et al., 2015:
   B. terrestris queens are monandrous and do not mate with multiple males like Apis mellifera. As with other hymenopterans, bumble bee males are haploid and females are diploid. The queen is thus equally related to her male and female progeny (r = 0.5) but is less related to worker-born males (r = 0.25). She is therefore selected to monopolise the production of all progenies, males and females, as well as develop mechanisms that suppress worker reproduction. Workers, on the other hand are more related to their sister gynes (r = 0.75) than to their would-be daughter gynes (r = 0.5). It is therefore advantageous for the workers to rear sister gynes (i.e. queen-laid eggs), which aligns with the queen’s interest, resulting in harmony for gyne (and worker) production.
   In the case of male production, however, a worker is more related to her sons (r = 0.5) than to her nephews (another nestmate worker’s son r = 0.375) or brothers (queen-born sons r = 0.25) Therefore, kin selection [theory] predicts that workers are selected to compete with the queen and their sisters over male parentage. ↩︎
3. Swarming is more complex than that and research is ongoing. There are also “migratory swarms” as in the Asian honey bees. However in this case it’s the same queens that are migrating to a new nesting location, rather than splitting the parent nest for reproduction. ↩︎