Bee Buzz Box March 2021 The Honey Bee Queen Replacement Enigma — Part I The Colony in Transition

Alan Wade, John Robinson, Jenny Robinson and Peter Robinson

For nearly fifty million years honey bees have successfully replaced their queens. They are well practised and have finessed queen installation as an art form.

Frontispiece: Direct colony requeening in conventional dequeen and requeen procedure

During brief periods of queen changeover, a honey bee colony may support as many as three laying queens and a large number of virgin queens rather than the just the normal single queen. For the beekeeper, however, the presence of these extra queens are likely to frustrate any requeening effort, that is until all rival and potential queens have been removed.

Another requeening problem arises when there is uncertainty as to whether a colony is actually queenright. The absence of brood in a colony is often a poor indicator of queen loss, so misreading its actual queen status may also confound successful requeening.

Uncertainties attendant to colonies undergoing natural queen replacement often perplex all but the most experienced beekeeper. A prime example is the widely held belief that removing queen cells will prevent a colony from swarming. However removing queen cells will not change the colony’s intent to swarm. The colony will promptly raise a new batch of swarm queen cells as indeed they will if you accidentally remove supersedure queen cells. Further, if one swarm queen cell is missed, the colony will swarm anyway.

Finding colonies with sealed or recently emerged queen cells signals a possible need for hive manipulation. For example a colony preparing to swarm can be split – to replicate swarming – or be left alone if there are only a handful of queen cells – to allow the colony to replace its ailing queen. The safest practice under these circumstance is to delay requeening and to wait until the colonies settle to their solitary queen condition.

These observations signal the need to better understand how bees periodically choose to replace their queens.

The normal functioning honey bee colony

A woodland bee colony (Figure 1) has a peak summer population of around 20 000 worker bees and, for much of the year, supports a population of around 200 drones. Except for a few drifting bees, all bees in residence are the progeny of the solitary colony queen.

Members of this honey bee community are routinely replaced and the queen is no exception. All bees age, can be weakened by disease or may die accidentally. Planned queen succession, like other hive member replacements, is essential for colony survival.

Figure 1 Nest entrance to honey bees in a two-to-three hundred year old Red Box (Eucalyptus polyanthemous) – Block 9 Section 253 Kambah. This tree has harboured one and sometimes two honey bee nests intermitently over the past 40 years.

Claims that a beekeeper should never requeen his or her hive and that the same queen has been there for many years are easily refuted. Bees routinely orchestrate queen departure and then install a daughter queen and do this for a variety of reasons. Replacement is normally achieved by swarming – where the old queen departs and a new queen takes up residence – or supersedure – where the queen is more simply replaced.

For wild colonies annual queen replacement is the norm. Sometimes bees will replace their queen more frequently. A colony will change out its queen whenever it performs unsatisfactorily, say due to age or miss-mating or where queen egg-laying capacity drops off precipitously. If, on the other hand, a colony grows quickly it may swarm, or swarm repeatedly, a large proportion of its productive bees departing and taking with them colony stores. The hive will then need a replacement queen, one supplied by a ripe swarm cell, a luxury not extended to the swarm if its queen fails.

In extreme cases, colonies may ‘swarm themselves to death’, that is nearly all bees abscond with swarms. In the more normal circumstance, the bee population will be rebuilt and stores made up to enable the parent colony to survive through winter and into the next season. Swarms – when they survive – make up for colony losses. So the number of wild colonies in an undisturbed landscape remains fairly constant, constrained only by the number of nesting sites and the availability of forage. Seeleyi reports that of forest honey bee colonies in New York State:

…most (97%) survive summers, but only 23% of founder colonies and 84% of established colonies survive winters. Established colonies have a mean lifespan of 5–6 years and most (87%) have a queen turnover, probably by swarming, each summer.

Seeley also noted that the colony density in this local study area had returned to pre-Varroa levels signalling the resilience of bees faced with new diseases and periodic drought. So while a colony may appear to be potentially immortal, in practice any unmanaged colony will succumb within half a decade.

The colony in transition

Regular queen changeover is important. Primarily it ensures colony vigour but it also serves to maintain gene flow and ultimately selection to better adapt the wider bee population to local conditions. So equipped the survival of whole bee populations – not the individual hive – is more or less assured. However a colony preparing to replace its queen faces the hazard of unsuccessful daughter queen establishment. Queen failure will be its death knell.

Despite remarkably reliable installation of a new colony queen, the actual workings of queen transition can be difficult to read. Our experience in operating two-queen colonies, where there may be as many as ninety five thousand bees, has made ascertaining queen condition rather more challenging. For the most part however success in managing colonies preparing to swarm is perhaps the single most challenging aspect of keeping bees.

Brood balance rules

There are several features of overall hive condition that signal a colony is out of kilter, an abnormal ratio of different stages of brood (fairly obvious when inspecting brood frames), by hive restlessness (signalled by hive roar if the colony is queenless), by the presence of active queen cells, or by the presence of too many bees to tend a large amount of brood and too many bees, as well as too little space to cure nectar and store honey. Two rules emerge:

Rule 1 – Brood balance ratio rule A good queen, well provisioned and with ample nurse bees to raise brood, will lay to full capacity and at a steady rate. Here the amount of brood at different stages of development will be reflected in the time it takes for a worker bee to develop: eggs 3 days: larva 6 days: pupa 12 days, a simple 1:2:4 ratioii. You may never stop to count brood cells, but a quick scan of brood nest combs showing only eggs and young larvae or only nearly all brood sealed will tell you that something is awry.

Rule 2 – Colony bee number to brood balance rule A healthy expanding colony having an adequate numbers of bees to provision larvae and to maintain the brood nest will be in balance. It will have a large retinue of house and field bees working actively together to steadily build and fill honey storage combs. When, however, there are an overly large number of bees emerging and a surfeit of nurse and house bees, and the colony becomes overcrowded, the queen may lose control over her progeny. Surplus bees, those that would result in broodnest overheating or that are no longer employed in processing nectar and storing honey will beard outside the hive entrance or hang off combs as replete storage bees. Keeping a close tab on fast expanding broodnests in spring will help you judge whether the amount of brood relative to the number of bees is getting out of balance.

The Scottish master beekeeper Bernhardt Möbusiii read these pointers as either an early sign of swarm preparation, or – out of swarming season – a condition where the colony no longer had room to store incoming nectar. His keen observation of brood-bee imbalance being the major factor – apart from queen age – in initiating swarming deserves wider recognition.

Queen bee signatures A forensic examination of queen cells left behind as a result of queen succession shows that the signatures are different for each method of queen replacement. Whenever a colony replaces its queen, it is simply a matter of finding cells from which queens started, and examining them closely.

Emergency queens originate from modified worker cells so are started deep in the worker cell matrix and are located close to where the queen last laid, that is almost anywhere where there were young brood. There may be few to many such cells.

Supersedure queens are always started on the surface of the comb and are usually, but not always, located within the main brood nest. There are rarely more than four or five such cells.

Swarm cells are similarly formed on the comb surface or are built free-form on comb margins. Characteristically they are many in number, more than a hundred in colonies of the African Honey Bee (Apis mellifera scutellata), and are located mainly at comb margins, but especially along the bottom bar. Their presence is not difficult to distinguish from that of the supersedure condition.

Common expressions of colonies transitioning to a new queen are those of interruption to normal brood raising, or the presence of active queen cell building, exemplified by:

(a) The apparent queenless condition A colony that has recently swarmed or that has accidentally lost its queen may be free of eggs and young larvae, indeed even be entirely broodless. As a virgin queen (or a queen out on mating flight) is easily missed it is never certain that the colony is, in fact, queenless.

A colony in transition takes quite some time, sometimes almost a month, to establish a new queen. We’ve often found a new laying queen during followup inspections a few weeks after finding a colony broodless. Had we assumed the colonies were queenless and gone ahead introduced queens anyway, those caged queen would have all been lost.

There is a simple remedy to the presumptive queenless condition. It involves going to a neighbouring healthy hive and removing a frame containing eggs and young larvae. Place this frame in the centre of the brood box, mark the top bar well so it can be easily recognised, and come back a day or two later to see whether any emergency queen cells have been started (Figure 2). Note that large diameter queen cells have encroached on neighbouring worker cells and that larvae in the queen cells are noticeably larger than those of neighbouring worker larvae. If the colony is indeed queenless, it will have been conditioned by the addition of young brood and will more readily accept a caged queen. Alternatively, if a new queen is already present and searching had failed to find her, no harm will have been done.

We never introduce a new caged queen (or a select queen cell) to a colony unless we are certain that it is queenless. One of the great advantages of making up a nucleus or baby colony a day before introducing a new queen is that you can be assured that the nucleus is queenless as new queen cells will have been started.

The most obvious case of bees being queenless is that of a colony that has long lost its queen. Very often the colony will be weak and there will be many small drones present. Such colonies have laying workers, so called gynecoid queens, and may have small patches of drone brood laid in worker cells.

It is a surprisingly easy condition to recognise when you first open hives in early spring. These so called hopelessly queenless colonies are difficult to requeen and the bees are best shaken off frames well away from the hive. The gear is removed or united with a normal functioning hive after a mandatory disease hive check.

Figure 2 Emergency queen cell development following failed queen introduction at Jerrabomberra Wetlands Apiary 21 October 2016

There are other situations where a queen may be present but no brood is present. Most notable of these are conditions where nothing attractive to bees is flowering when the queen will stop laying. Most of us will have observed this low-brood condition in late autumn when there is little or no forage and there is little value in the colony raising new forager bees. This said, it is still important to have a good laying queen to supply the colony with young long-lived (diutinous) bees needed to maintain the colony over winter. Small patches of brood will appear periodically even in the depths of winterivv despite the heavy drain on colony stores to raise them. We have found that taking the time to work out why a colony is broodless is more a matter of common sense than adopting the expedient practice of dropping in a new queen that is unlikely to survive.

The situation is more complicated when, during the actual period of natural queen changeover, there are a number of incipient queens (queen cells and virgin queens, so-called gynes) present. Requeening a colony while the colony is actively replacing its queen will likely result in loss of the queen you are trying to introduce. Let’s look at these quirky conditions as they may also confound your well mentioned effort to requeen a more normal failing or cantankerous single-queen, queenright hive.

(b) The swarming colony condition

That swarming preparation has begun is obvious as there will be many queen cells present. Most often such cells are located on brood comb margins that a quick check of bottom bars of brood boxes will quickly reveal (Figure 3).

Once swarming preparations are well underway, the colony queen will have stopped laying and there will be no eggs and little or no unsealed larvae. As well, foraging and comb building will also have ceased and, if you are particularly observant, you may find a slimmed down queen ready to fly and bees gorging themselves on honey and pollen making ready to swarm. Further salient features of swarming are that the colony queen never departs until sealed queen cells are present and that the swarm that departs without a queen (or that accidentally loses its queen) will almost invariably return to the hive.

Figure 3 Queen cells on broodnest bottom bar at Jerrabomberra Wetlands Apiary, 12 November 2016

A colony may temporarily abandon swarming and tear down queen cells when weather conditions suddenly deteriorate, though most often it will resume its effort to swarm once conditions improve. Counterintuitively, and under the same poor weather conditions, overcrowding resulting from confinement, can be the trigger to commence swarming preparations.

Demuth provides an exceptionally lucid account of the swarming processvi:

About a week after the issuing of the prime swarm the first of the young queens in the parent colony emerges from her cell. Instead of destroying the other young queens and establishing this first emerged young queen as the new mother of the colony, the bees usually swarm again about eight days after the prime swarm has issued, this after-swarm being accompanied by one or more young [sic virgin] queens. Other afterswarms, each one smaller than the preceding, may issue with an interval of one or two days between until the colony is so reduced in numbers that further swarming is given up and all but one of the remaining young queens are killed. About ten days after emergence the surviving young queen usually begins to lay, and normal brood rearing is again established in the parent colony after an interval of at least sixteen days during which no eggs have been laid. Each after-swarm establishes itself in a new abode, begins building its combs, and the young queen begins to lay about ten days after emergence. If sufficient food is available such colonies may build up to normal strength for winter. This is the natural method of reproduction of colonies in the honeybee.

So far we have recorded the condition of colonies making preparations to swarm. Notable features of colony that has already swarmed are that a laying queen may not appear for several weeks, that the whole hive will be sparsely populated and that worker bee recruitment will be delayed and limited to that of any remaining emerging brood. In the worst case scenario the colony may be so depleted it may never recover or the remaining virgin queen may fail to mate.

When swarming is over, or after you have taken effective swarm control measures, you can go back to your requeening plan, this time more assured of a successful effort to requeen. The problems of requeening during the normal swarming season – late August through to the end of October – prompts another question: ‘What is the best time to requeen?’

Most beekeepers prefer to conduct requeening by mid spring before colonies become overly large and difficult to handle. There is, nevertheless a very strong case for routine autumn requeening when colonies very rarely swarm. Instead of allowing really strong colonies to down-regulate their numbers at the end of the season, they can be split and both parts requeened, or if of average strength, just requeened. This will provide a much needed supply of colonies headed by young queens long before queens can be raised or are commercially available in spring.

Another advantage of autumn requeening is that such queens make an earlier start to laying than do older queens. Well provisioned, colonies come away quickly as early as late winter and will rarely swarm during the spring and early summer buildup. This commends late season, say March, requeening though you will need to place a queen order well in advance to be assured that queens can be supplied.

A further fallback position is to consider summer requeening. While large numbers of bees makes finding and removing queens more difficult than in autumn, or indeed in a normally functioning hive in spring, queens tend to be most readily available for purchase in summer.

(c) The queen replacement condition

The supersedure condition is indicated by the presence of a handful of queen cells located in the brood nest. Note eggs in adjacent worker cells (Figure 4) absent in colonies preparing to swarm. It is important to distinguish this condition from that of swarming, where large numbers of queen cells are present, and emergency queen replacement, where a few to many queen cells may be present.

Figure 4 Second batch of supersedure cells in the presence of an old and previously particularly productive Carniolan queen. Jerrabomberra Wetlands Apiary 31 October 2016

Colin Butler clearly differentiated the process of supersedure from that of swarmingvii:

Queen supersedure, the process by which a colony of honeybees replaces its queen without swarming, is of frequent occurrence… the process of queen supersedure differs radically from that of swarming by which colony reproduction is achieved, and in which one or more new queens are also reared and the old queen leaves the parent nest with a swarm if she is still alive.

Simple queen replacement can occur at any time of the year but is most prevalent when colonies are under worker breeding stress, that is whenever the queen cannot lay fast enough for number of nurse bees available to raise brood. This is well illustrated by a lesser know facet of supersedure identified by John Hoggviii:

Queen supersedure is never the cause of swarming. But supersedure may occur concurrently with the swarm whenever the queen’s failure was caused by being overtaxed while generating the bees for that swarm. Significantly, it is the virgin queen that is then selected to accompany the swarm. The failed queen is retained in the parent hive to be superseded in turn…

Apparently, just as the survival instinct of the bees in control won’t allow a swarm to leave a hive without a replacement queen in the parent [hive], they won’t allow a failing queen to issue with that swarm.

When you find supersedure queen cells, simply close the hive as the colony needs a new queen. Removing such cells risks either the old queen dying – when the colony will become queenless – or at best delaying the inevitable process of queen replacement. A father of commercial queen raising Gilbert Doolittleix made famous this discovery in 1889:

After I had this experience with the colony that had two Queens in a hive, (which was a surprise to so many fifteen years ago, when it was thought that no colony ever tolerated but one laying queen at a time), I began to watch for a like circumstance to occur, which happened about a year from that time. In the latter case, as soon as I found the cells, they were sealed over, and not knowing just when they would hatch [sic emerge], I at once cut them out and gave them to nuclei. In a few days I looked in the hive again, when I found more cells started, which were again cut off and given to nuclei, just before it was time for them to hatch. In this way I kept the bees from their desired object for some two months, or until I saw that the old Queen was not going to live much longer, when I left one of the cells, which limey had under headway, to mature. By this plan I got about sixty as fine Queens as I ever reared, and laid the foundation for my present plan of securing Queens, Which is about to be given in this book.

In the aftermath of supersedure, a new daughter queen (and occasionally two daughter queens) may be present and actively laying alongside the old queen. We have found two queens – the queen being superseded and her daughter – in a handful of colonies accurately reflecting the observation made by the British Bee Journal authors exactly one hundred years agox:

It is quite possible that there had been two queens in the hive, a condition that has been prevalent this year, or beekeepers are becoming more observant.

In the normal course of events the old queen disappears, either of natural causes or is removed by the daughter queen. For all practical purposes, queen supersedure will not interrupt apiary operations, though replacing a failing queen should always involve a search of all brood combs for the not unlikely presence of an additional queen.

In Part II we will outline the impact of colonies in transition on requeening practice. We will conclude with recommendations to requeen regularly, to adopt autumn requeening and to queen into nucleus colonies – rather than to directly requeen strong hives. These measure will help overcome some of the more regularly encountered difficulties attendant to spring requeening.


iSeeley, T.D. (2017). Apidologie 48:743–754. Life-history traits of wild honey bee colonies living in forests around Ithaca, NY, USA.

iiCollinson, C. (2019). Applying the basics of honey bee biology. UK National Honey Show 14 January 2019.

iiiMöbus, B. (1987). The swarm dance and other swarm phenomena Part I. American Bee Journal .[April 1987].
Möbus, B. (1987). The swarm dance and other swarm phenomena Part II. American Bee Journal [May 1987).
Parts l and 2 of a Two-part series published in the American Beekeeping Journal by Bernard Möbus NDB Scotland can be found at

ivMöbus, B. (1998). Brood rearing in the winter cluster, Part I. American Bee Journal 138(7):511-514.
Möbus, B. (1998). Brood rearing in the winter cluster, Part 11. American Bee Journal 138(8):587-591.
Parts 1 and 11 are on-line at:

vBees cluster when the ambient temperature drops to 14 0C, maintain a broodless cluster at 18-21 0C and maintain a broodnest temperature of 35 0C.

viDemuth, G.S. (1921). Swarm control. Farmers Bulletin 1198: 1-45.

viiButler, C.G. (1957). The process of queen supersedure in colonies of honeybees (Apis mellifera Linn.). Insectes Sociaux 4(3):211-223.
Butler, C.G. (1960). The significance of queen substance in swarming and supersedure in honey bee (Apis mellifera L.) colonies. Proceedings of the Royal Entomological Society of London A35:129-132.

viiiHogg, J.A. (2006). Colony level honey bee production: The anatomy of reproductive swarming. American Bee Journal 146(2):131-135.

ixDoolittle, G.M. (1889). Scientific queen-rearing as practically applied being a method by which the best of queen-bees are reared in perfect accord with nature’s ways, 184pp., Chicago, Ills. Thomas G. Newman & Son, 923 & 925 West Madison Street.

xBrichter, G. (1921). Two queens in a hive. British Bee Journal and Beekeeper’s Adviser 49:384.