1. Swarming is biologically normal, not a fault

This chapter first places swarming in biological context: as the natural reproduction of the honey bee colony by division, not as a direct indicator of poor genetics or a deficient queen.

Swarming is the natural reproduction of a honey bee colony by division. When the prime swarm departs, the mated old queen leaves the nest with a large proportion of the adult workers, while the combs, brood, most of the food stores and part of the worker population remain in the mother colony. The mother colony must then restore continuity around a young queen and may, under some conditions, still issue one or more cast swarms.

This first division is strongly asymmetrical. In the colonies quantified by Rangel and Seeley (2012), about 75% of the adult workers left with the old queen on average, although there was variation among colonies. The swarm therefore takes with it a large share of the mobile workforce while leaving behind almost the entire material infrastructure. The mother colony, by contrast, retains the nest, the food stores and the brood from which new workers will gradually emerge.

The size of the swarm fraction is therefore biologically important. In the study by Rangel and Seeley, a larger swarm fraction was associated with better growth and higher survival of the new colony founded by the old queen. A later model also produced a theoretical optimum close to the observed fraction (Rangel et al., 2013). These findings establish a principle, not a fixed rule that 75% of the bees always leave.

From a biological perspective, swarming is therefore neither “misbehaviour” nor proof of a deficient queen. It is a central reproductive event in the colony life cycle. From a beekeeping perspective, the same behaviour may be undesirable because workforce is lost, potential honey yield is reduced, additional labour is required, and new risks arise for both the swarm and the mother colony.

This distinction is essential for everything that follows: an observed swarm is not, by itself, a genetic diagnosis. It shows that a colony progressed as far as division, but it does not by itself reveal why that state was reached or whether another colony under the same conditions would have responded in the same way.

2. Genetics plays a role — but not as a simple explanation

This chapter explains what heritability means for swarming behaviour — and why it cannot be used as a genetic test for an individual colony.

The fact that swarming is a natural behaviour does not mean that all colonies have the same propensity to swarm. A genetic component has been demonstrated, and the trait can respond to selection (Andonov et al., 2019; Hoppe et al., 2020; Kovačić et al., 2020).

Heritability estimates nevertheless show how strongly this genetic component depends on the population and the statistical model. Andonov et al. (2019), working with Italian honey bees, obtained a heritability of 0.34 for swarming behaviour in the model used. A value of 0.34 was also reported in an Iranian population selected over several generations (Tahmasbi et al., 2015). In the Austrian breeding population analysed by Brascamp et al. (2016), however, the heritability of the combined selection criterion for swarming behaviour was much lower, around 0.08. The correction published in 2018 further showed that the way queen and worker genetic effects were handled statistically substantially affected interpretation.

These values must not be read as though “34% of a swarming event is genetic”. Heritability describes, within a given population and under the conditions studied, what proportion of the observed variation is associated with genetic variation according to the model used. It does not allow a causal decomposition of a single swarming event.

This caution is especially important because swarming behaviour is a trait of the entire colony. Queen, workers, brood demography, adult population, spatial organisation, resources and environment interact. Selection may therefore shift the propensity to swarm or particular response thresholds without it being clear which biological mechanism is altered.

Possible mechanisms include the response to high worker density, the transmission of queen and brood signals, worker sensitivity to those signals, brood demography, or the threshold at which queen rearing is initiated and then maintained. None of these mechanisms has yet been identified as the single genetic target of selection against swarming.

For practice, the implication is clear: a colony that swarms once is not automatically genetically highly swarm-prone, and a colony that does not swarm in one year is not automatically genetically swarm-reluctant. Repeated observations — ideally across several daughters, years or environments — are much more informative than a single event.

3. Location, forage and colony management strongly influence the observed behaviour

This chapter shows why swarming events must always be assessed together with location, forage, weather, demography, genuinely usable space and the timing of interventions.

The pan-European experiment by Uzunov et al. (2014) illustrates the interaction between genetic origin and environment particularly well. A total of 621 colonies representing 16 genotypes were observed at different European sites. Swarming behaviour varied significantly with both genotype and experimental location, but differences among locations were greater than those among genotypes for the traits studied.

This does not mean that the environment “cancels out” genetics. It means that the observed phenotype results from their interaction. A queen from a line selected for low swarming propensity can still swarm under some conditions, while a more swarm-prone line may fail to express the behaviour during a particular season.

Earlier field observations already pointed in the same direction. Simpson (1957) reported that, in English honey-producing apiaries, approximately 10 to 40% of colonies could swarm in an average year in the absence of specific swarm-control measures, with substantial variation among years and locations. These historical figures should not be transferred quantitatively to present-day Swiss apiaries, but they illustrate how strongly swarming incidence can depend on context.

Available space also requires precise interpretation. The classic experiments show that a shortage of cells for egg laying and a high density of adult bees are not biologically equivalent. Artificial restriction of laying area alone is insufficient to reproduce the state leading to swarming (Simpson & Riedel, 1963; Simpson & Greenwood, 1975).

Simpson and Moxley (1971), however, observed that small colonies allowed to continue growing in small hives developed occupied queen cells and swarmed much more frequently once they had effectively exceeded the capacity of their housing. The conclusion is therefore not that “lack of space triggers swarming”, but that the way a growing colony occupies and outgrows its available space forms part of the demographic and spatial state that can favour the transition towards reproduction.

In practice, brood-nest congestion should be understood as a composite phenotype. It may combine a large adult population, extensive capped brood close to emergence, relatively little young brood, fewer immediately available cells, nectar or pollen stored within the brood nest, and changes in spatial organisation. Brood-nest congestion is therefore an important warning sign, but not a diagnostic test with known sensitivity or specificity.

Nectar flow and liquid feeding should likewise be understood as resource flows that partly compete for the same cells as brood. Fed syrup can be stored, moved and redistributed, and the use of comb space for brood and food stores remains dynamic (Free & Spencer-Booth, 1961; Eyer et al., 2016; Colin et al., 2018).

The complete chain “heavy spring feeding → storage within the brood nest → stable swarm preparation → prime swarm” has not, however, been demonstrated. In an already very strong colony, unnecessary heavy feeding could plausibly intensify an existing congestion pattern, but this plausibility should not be converted into a general rule that spring feeding causes swarming.

Queen age is another influencing factor. Hauser and Lensky (1994) observed more swarm cells in colonies headed by older queens, even though the experimental colonies were regularly expanded. The finding shows that queen age can modify the propensity to swarm even under active space management. It does not prove that this effect is independent of every other dimension of colony state.

This supports a practical heuristic, not a causal test: when many colonies in the same apiary enter swarm preparation at the same time, shared factors such as nectar flow, weather, seasonal development, available space and management should be examined first. When a single colony or several related daughters repeatedly swarm early under comparable conditions, a genetic component becomes more plausible.

4. A threshold model as a conceptual framework — not as a demonstrated mechanism

This chapter uses a threshold model as a conceptual framework: genetics alters the propensity to respond, while several dimensions of the environment and colony state influence whether swarm preparation becomes stabilised.

It is useful to represent swarming propensity as a response threshold, provided that this image is not confused with a demonstrated biological mechanism. A colony does not necessarily possess a single measurable “swarming threshold” that is crossed when one variable reaches a particular critical value.

The model instead helps explain why two colonies under similar conditions can respond differently. Genetic origin may shift particular response probabilities, while brood demography, bee density, resources, the queen and spatial organisation change simultaneously. As several dimensions converge, maintaining queen rearing and eventually proceeding to colony division may become more likely.

Fefferman and Starks (2006) proposed a theoretical model incorporating, among other variables, colony size, congestion, worker age structure and the queen's maximum egg-laying capacity. The model is useful for exploring interactions among several factors. It is, however, a model rather than an experimental demonstration of a causal chain leading to natural swarm departure.

The genetic evidence requires the same caution. Heritability estimates show that swarming propensity can respond to selection, while genotype-environment experiments demonstrate that its expression depends strongly on context (Andonov et al., 2019; Brascamp et al., 2016, 2018; Tahmasbi et al., 2015; Uzunov et al., 2014).

The most robust formulation is therefore: genetics alters probabilities or particular response thresholds, while the environment, colony state and management alter the conditions under which that response is expressed. How these dimensions are fully integrated by the colony remains unresolved.

This framework also avoids premature attribution of blame. A swarm is neither automatically proof of poor genetics nor automatically proof of poor colony management. It shows that a particular combination of conditions and colony properties ultimately progressed as far as division. Determining which were most important requires comparison rather than retrospective attribution based on a single event.

5. Requeening: genetic improvement or age effect?

This chapter separates two questions that are often conflated: requeening can reduce swarming propensity without proving that the previous swarm was caused by poor genetics.

After a swarming event, replacing the queen is frequently recommended. This can be a sensible measure, but two statements must be distinguished:

• A young queen can reduce the probability of renewed swarm preparation.
• The previous queen was genetically too prone to swarming.

The first statement has experimental support. The second requires additional observations.

Simpson (1957, 1960) already observed a relationship between queen age and the tendency to build queen cells or to swarm. Hauser and Lensky (1994) later found a pronounced age effect in a subtropical experiment: colonies headed by older queens built about 2.5 times as many swarm cells in one experimental year and about 3.9 times as many in the other as colonies headed by young queens, even though the colonies were regularly expanded.

These ratios should not become a universal quantitative rule. They come from a particular experimental design and climate. They nevertheless show clearly that a young queen can shift swarming probability without requiring the assumption of genetic improvement.

Forster (1969) likewise observed a marked reduction in swarm preparation in colonies headed by young queens. Planned queen replacement is therefore among the few preventive measures with direct experimental support. It reduces a probability; it does not make swarming biologically impossible.

The precise mechanism behind the age effect is less certain. Hauser and Lensky discussed, among other possibilities, age-related changes in signals produced by the queen. This hypothesis is plausible, but it should no longer be reduced to the simple model “old queen = less pheromone = swarming”. Current knowledge indicates that reproductive information from the queen is distributed through contact networks, complemented by signals associated with eggs and very young larvae, and that its effects also depend on the state of the receiving workers.

Requeening may therefore act through several possible pathways: age, reproductive physiology, laying quality, social signalling and — if the new queen comes from another line — genetics. The success of requeening alone does not reveal which of these pathways was decisive.

There is also a practical complication: requeening in a colony already preparing to swarm is rarely the only change made. Queen cells may be removed at the same time, brood or bees may be taken away, space may be altered, or egg laying may be temporarily interrupted. If the colony remains calm afterwards, the independent effect of the new queen usually cannot be isolated.

Replacing an older or poorly performing queen, or a queen from a line in which several daughters show a pronounced tendency to swarm, may therefore be technically justified. Automatically replacing a young queen after a single swarming event, by contrast, assumes a genetic cause without demonstrating it.

6. When is requeening technically justified?

This chapter translates the scientific evidence into a practical diagnosis: when is requeening justified, and when does it risk becoming only a reflex?

Requeening is an important tool in colony management. It allows an old, poorly performing or undesirably disposed queen to be replaced. The problem begins when every swarming event is automatically treated as proof of a genetically unsuitable queen.

Before replacing a queen, four questions are particularly useful:

1. Does the behaviour repeat? Does the colony repeatedly enter swarm preparation early under management comparable with colonies that remain stable?
2. Do related colonies show the same behaviour? Do several daughters, sisters or colonies of the same origin show similar swarming propensity under different or comparable conditions?
3. Are other breeding traits also problematic? Excessive defensiveness, restlessness, poor performance, weak hygienic behaviour or other traits should be considered within multi-trait selection.
4. Is the queen simply old or declining? Her age, laying activity and the general brood pattern should be assessed separately from the genetic question.

There is no validated score in which a particular number of positive answers automatically leads to requeening. These questions are intended to structure diagnosis.

Repeated early swarming behaviour in several related colonies under good management strengthens the hypothesis of a genetic component. By contrast, a single event during exceptionally rapid colony growth, a strong nectar flow or marked brood-nest congestion provides much less information about the genetic value of the queen.

Queen age must likewise remain a separate explanation. An older queen may justify replacement because of her age and its effect on swarming propensity without implying that her lineage is genetically unsuitable.

The central rule is therefore: requeening is a targeted measure when there is a diagnosis; without diagnosis, it risks masking the underlying question.

7. What breeding can achieve — and what it cannot

This chapter shows why low swarming propensity is a legitimate breeding goal without providing complete control over the behaviour.

Low swarming propensity is a legitimate breeding goal. It makes colony management easier, reduces labour and swarm losses, and may improve the predictability of production. Evidence from several populations and breeding programmes supports the conclusion that the trait has a genetic component and can respond to selection (Andonov et al., 2019; Hoppe et al., 2020; Kovačić et al., 2020).

This possibility must not be confused with complete control. Heritability estimates vary among populations, scoring systems and statistical models. Genotype-environment experiments also show that location can strongly influence the behaviour that is expressed (Uzunov et al., 2014).

For breeding, repeated observations are therefore far more informative than a single result. The value of a queen or line should ideally be assessed from several descendants and in several environments, or at least over several seasons where possible.

Swarming propensity should also not be selected independently of other traits. Andonov et al. (2019) estimated genetic relationships between swarming behaviour, honey yield and defensive behaviour. Such relationships are population- and model-dependent and should not be generalised without caution. They nevertheless illustrate a fundamental breeding principle: low swarming propensity is only one trait among several.

A colony that swarms little but is unhealthy, excessively defensive or poorly productive does not automatically become good breeding stock. Conversely, selection for low swarming propensity does not mean “correcting” a biological defect of the honey bee. Swarming remains the colony's natural form of reproduction; selection here serves beekeeping objectives such as manageability and production.

Breeding can shift probabilities. It does not eliminate the biological programme. Even a line strongly selected for low swarming propensity may enter swarm preparation when several conditions favourable to reproduction converge.

8. Practical conclusion: from reflex to diagnosis

This chapter summarises the consequences for the apiary, training and advisory work: after a swarm, the colony's state should be reconstructed rather than immediately assigning a single cause.

A swarm should not automatically be interpreted as a genetic defect of the queen. The observed behaviour arises on several levels: genotype influences swarming propensity, environment and resources alter the conditions under which that propensity is expressed, demography and brood-nest congestion describe the state of the colony, and queen age is an additional influencing factor.

After a swarm, comparative questions are therefore crucial: How strong was this colony compared with the others? How much capped brood was close to emergence? Was there marked brood-nest congestion? How was the nectar flow developing? Had liquid feed been given and where was it stored? How old was the queen? Were several colonies in the apiary preparing to swarm at the same time? Do related colonies repeatedly show the same behaviour?

Weather must also be interpreted cautiously. An analysis of 1,335 swarming events in Germany showed that departures were rare on rainy days and fell sharply during cold periods (Henneken et al., 2012). Colonies already ready to swarm may therefore wait for a suitable flight window.

Another mechanism is plausible: in poor weather, more foragers remain inside the hive and the spatial organisation of the colony changes rapidly (Riessberger & Crailsheim, 1997). This could temporarily intensify congestion in an already very strong colony. The causal chain poor weather → increased bee density → triggering of swarming has not, however, been demonstrated.

Finally, the stage of intervention must also be distinguished. Providing genuinely usable space early enough to a growing colony belongs to prevention. Altering population, brood, space or egg laying once queen rearing is already under way means intervening in an existing process. Removing queen cells does not demonstrate that the reproductive state has disappeared. Physically preventing the loss of a swarm is no longer biological swarm prevention.

The central question is therefore not only “Which method should I use?”, but: What state is the colony actually in, what result is being sought, and how well supported is the proposed intervention by evidence?

---

Learn more:

• The mechanisms of natural swarming
• Does selection in beekeeping allow for heritability?
• Queen rearing and honey bee genetics
• Swarm prevention
• Why does a bee colony replace its queen?

References

• Andonov, S., Costa, C., Uzunov, A., Bergomi, P., Lourenco, D., & Misztal, I. (2019). Modeling honey yield, defensive and swarming behaviors of Italian honey bees (Apis mellifera ligustica) using linear-threshold approaches. BMC Genetics, 20, 78. https://doi.org/10.1186/s12863-019-0776-2
• Brascamp, E. W., Willam, A., Boigenzahn, C., Bijma, P., & Veerkamp, R. F. (2016). Heritabilities and genetic correlations for honey yield, gentleness, calmness and swarming behaviour in Austrian honey bees. Apidologie, 47, 739–748. https://doi.org/10.1007/s13592-016-0427-9
• Brascamp, E. W., Willam, A., Boigenzahn, C., Bijma, P., & Veerkamp, R. F. (2018). Correction to: Heritabilities and genetic correlations for honey yield, gentleness, calmness and swarming behaviour in Austrian honey bees. Apidologie, 49, 462–463. https://doi.org/10.1007/s13592-018-0573-3
• Colin, T., Bruce, J., Meikle, W. G., & Barron, A. B. (2018). The development of honey bee colonies assessed using a new semi-automated brood counting method: CombCount. PLOS ONE, 13, e0205816. https://doi.org/10.1371/journal.pone.0205816
• Eyer, M., Neumann, P., & Dietemann, V. (2016). A look into the cell: Honey storage in honey bees, Apis mellifera. PLOS ONE, 11, e0161059. https://doi.org/10.1371/journal.pone.0161059
• Fefferman, N. H., & Starks, P. T. (2006). A modeling approach to swarming in honey bees (Apis mellifera). Insectes Sociaux, 53, 37–45.
• Forster, I. W. (1969). Swarm control in honey bee colonies. New Zealand Journal of Agricultural Research, 12, 605–610. https://doi.org/10.1080/00288233.1969.10421245
• Free, J. B., & Spencer-Booth, Y. (1961). The effect of feeding sugar syrup to honey-bee colonies. The Journal of Agricultural Science, 57, 147–151. https://doi.org/10.1017/S0021859600047614
• Hauser, H., & Lensky, Y. (1994). The effect of the age of the honey bee (Apis mellifera L.) queen on worker population, swarming and honey yields in a subtropical climate. Apidologie, 25, 566–578. https://doi.org/10.1051/apido:19940607
• Henneken, R., Helm, S., & Menzel, A. (2012). Meteorological influences on swarm emergence in honey bees as detected by crowdsourcing. Environmental Entomology, 41(6), 1462–1465. https://doi.org/10.1603/EN12139
• Hoppe, A., Du, M., Bernstein, R., Tiesler, F., Kärcher, M., & Bienefeld, K. (2020). Substantial genetic progress in the international Apis mellifera carnica population since the implementation of genetic evaluation. Insects, 11(11), 768. https://doi.org/10.3390/insects11110768
• Kovačić, M., Puškadija, Z., Dražić, M., Uzunov, A., Meixner, M. D., & Büchler, R. (2020). Effects of selection and local adaptation on resilience and economic suitability in Apis mellifera carnica. Apidologie, 51, 1062–1073. https://doi.org/10.1007/s13592-020-00783-0
• Rangel, J., & Seeley, T. D. (2012). Colony fissioning in honey bees: Size and significance of the swarm fraction. Insectes Sociaux, 59(4), 453–462. https://doi.org/10.1007/s00040-012-0239-5
• Rangel, J., Reeve, H. K., & Seeley, T. D. (2013). Optimal colony fissioning in social insects: Testing an inclusive fitness model with honey bees. Insectes Sociaux, 60(4), 445–452. https://doi.org/10.1007/s00040-013-0309-3
• Riessberger, U., & Crailsheim, K. (1997). Short-term effect of different weather conditions upon the behaviour of forager and nurse honey bees (Apis mellifera carnica Pollmann). Apidologie, 28(6), 411–426. https://doi.org/10.1051/apido:19970608
• Simpson, J. (1957). The incidence of swarming among colonies of honey-bees in England. The Journal of Agricultural Science, 49, 387–393.
• Simpson, J. (1960). The age of queen honeybees and the tendency of their colonies to swarm. The Journal of Agricultural Science, 54, 195.
• Simpson, J., & Greenwood, S. P. (1975). Results of restricting the brood space of honeybee colonies. Journal of Apicultural Research, 14(1), 51–55. https://doi.org/10.1080/00218839.1975.11099801
• Simpson, J., & Moxley, E. (1971). The swarming behaviour of honeybee colonies kept in small hives and allowed to outgrow them. Journal of Apicultural Research, 10(3), 109–113. https://doi.org/10.1080/00218839.1971.11099681
• Simpson, J., & Riedel, I. B. M. (1963). The factor that causes swarming by honeybee colonies in small hives. Journal of Apicultural Research, 2(1), 50–54. https://doi.org/10.1080/00218839.1963.11100056
• Tahmasbi, G., Kamali, M. A., Ebadi, R., Nejati Javaremi, A., Babaei, M., Gharadaghi, A. A., & Bahraini, R. (2015). Genetic trends and parameters of honey production, swarming and defense behavior in Iranian honeybee (Apis mellifera meda) colonies. Journal of Agricultural Science and Technology, 17, 1735–1742.
• Uzunov, A., Costa, C., Panasiuk, B., Meixner, M., Kryger, P., Hatjina, F., Bouga, M., Andonov, S., Bieńkowska, M., Le Conte, Y., Wilde, J., Gerula, D., Kiprijanovska, H., Filipi, J., Petrov, P., Ruottinen, L., Pechhacker, H., Berg, S., Dyrba, W., Ivanova, E., & Büchler, R. (2014). Swarming, defensive and hygienic behaviour in honey bee colonies of different genetic origin in a pan-European experiment. Journal of Apicultural Research, 53, 248–260. https://doi.org/10.3896/IBRA.1.53.2.06