Category Archives: Science

Natural vs. artificial swarms

I’ve now covered four of the most frequently used swarm control strategies. These are:

  • Pagden’s artificial swarm – the horizontal splitting of the colony
  • The vertical split – an equipment-frugal variant of the above involving a vertical separation of the colony
  • The nucleus method – in which the queen is removed with sufficient workers to make up a small (nuc) colony, leaving the original colony to rear another queen
  • The Demaree method – which, at its simplest, relocates the queen from the brood and associated nurse bees, but does not physically split the colony

If conducted correctly all should prevent loss of a swarm. However, the individual methods – even the first three which involve the physical separation of the bees in the hive – are not the same.

In addition, these swarm control methods do not recapitulate the separation of bees that occurs when a hive naturally swarms.

The purpose of this post is to contrast the original and new colony composition of the split-based methods of swarm control (i.e. Pagden and vertical) with natural swarms.

Temporal polyethism

I introduced this term when discussing the honey bee colony as a superorganism. It means that adult worker bees have different roles depending upon their age. For the first two and a bit weeks they have duties inside the hive such as cell cleaning, brood rearing and wax production.

They then transition through a period of being guard bees before becoming foragers, flying from the hive and collecting water, nectar and pollen.

For convenience I’ll refer to these two groups of bees as young, nurse or hive bees and flying bees.

Vertical and horizontal splits

The classic Pagden artificial swarm and the vertical split are fundamentally the same process.

If unsealed queen cells are found during a colony inspection the queen, with a frame of emerging brood, is moved to a new box. This box is placed on the site of the original hive.

The remaining bees and brood are moved, either to one side in the case of the Pagden or on top of the queen-containing box (separated by a split board) in a vertical split.

Split board ...

Split board …

Critically, the new box with the brood and bees is provided with a new hive entrance, located off to one side or on the opposite side of the original hive 1.

Flying home

Over the following day or two the flying bees leave the relocated brood box with the new entrance and return to the queen-containing brood box in the original location.

As a consequence of their excellent homing navigational skill, the hive manipulation results in the separation of the bees into two populations:

  1. The flying bees i.e. those over ~3 weeks of age that had orientated to the original hive location, which are now located with the queen.
  2. The nurse bees i.e. those less than 3 weeks old, which remain in the relocated brood box, together with the brood in all stages (eggs, larvae and pupae).
Artificial swarm separation of the colony

Artificial swarm separation of the colony

How does the artificial swarm compare with the age distribution of bees in a real swarm?

Real swarms

I’ve previously discussed prime swarms and casts. The former contain a mated queen. In contrast, casts are produced from very strong colonies after the prime swarm has left. Casts are headed by a virgin queen. These are sometimes called after swarms and are usually smaller than prime swarms.

What about the workers in the swarm? What might be expected?

Perhaps they’re primarily the older flying bees? After all, these are the bees that have finished their hive duties and are now routinely foraging outside the hive. It’s the natural place for them.

Swarm of bees

Swarm of bees

Alternatively, remember that swarms have no ‘homing’ instinct for a day or two after emerging. They can be readily moved and you can safely ignore the less than three feet or more than three miles rule. Perhaps this means that they’re primarily young bees that have yet to go on their orientation flights?

Real experiments and contradictory results

Enough speculation … how do you determine this experimentally?

There have been numerous studies of the age distribution of bees in natural swarms. However, the data tends to be rather contradictory though the methods used are often broadly similar.

How do you determine the age composition of workers in a swarm?

Essentially you ‘spike’ the colony with a set number of marked bees of a known age over about 8 weeks. This is easy to do, but tedious.

Workers are allowed to emerge in an incubator. On the day of emergence (0 days old) they are marked with a colour that distinguishes them from older or younger bees. Every three days 100 identically marked i.e. same age, bees are added to the study hive(s). Over the period May to July this will accumulate red, then yellow, then blue, then mauve, then cyan, then pink etc. cohorts of workers, each representing a known age class.

It must be a nightmare spotting the queen in these hives 😉

The colony is allowed to swarm, the swarm collected and the number of bees of the different age cohorts in the swarm counted.

I missed a step out there. Have you ever tried counting the bees in a swarm? It’s much easier if they don’t move.

1002, 1003, 1004, 1005, er, where was I? Damn!

1002, 1003, 1004, 1005, er, where was I? Damn!

Perhaps it’s best that I missed that step out 🙁

What you end up with is a count of the total number of bees in the swarm and the numbers of bees of each 3 day cohort over the last several weeks. You can therefore determine the age distribution of the workers in the swarm.

Is it as simple as that?

I’ve actually oversimplified things a bit. There’s a possibility that different age cohorts of bees die within the hive at different rates, perhaps depending upon forage availability or weather or something else.

Think about it. Assume there was a dearth of nectar in late May and the blue and red labelled cohorts added during that period were underfed and died prematurely.

If there were very low numbers of blue and red bees in the swarm you might assume that these ages were ‘left behind’ by the swarm … when actually they weren’t able to swarm at all.

The real question is therefore whether the age distribution of bees in the swarm is similar to that in the parental hive.

OK, OK … is it?


Swarms do contain bees of all ages.

However there are significantly more young bees and many fewer old bees than would be expected from the age distribution of workers in the parental colony.

Age distribution of bees in swarms

Age distribution of bees in swarms

The o and e in the graph above represents the position of the observed and expected median age class for the expected distributions. So, in swarm C the observed median age is ~10 days old, whereas the originating hive median age was ~19 days.

The graph above comes from a 1998 study by David Gilley 2 and supports earlier work 3 by Colin Butler 4 which is often cited as one of the definitive studies on the ages of bees in a swarm.

Additional considerations

Is it surprising that young bees predominate in natural swarms?

Swarms usually emerge from the hive late morning or early afternoon on warm, sunny days. In fact, at exactly the time most older bees aren’t in the hive anyway because they’re out and about foraging.

Remember also that swarming is a precarious activity for the colony. Most swarms do not survive 5. Natural selection will have resulted in swarm populations that maximise their chance of survival.

Once bees start foraging their life expectancy is pretty short. It has been estimated that they experience about 10% mortality per day. If only old bees left in the swarm with the queen the newly established colony would very rapidly dwindle in size, perhaps before significant numbers of new brood emerged (which takes 21 days from the first egg being laid). This would likely limit the chances of survival of the new colony.

What has this got to do with artificial swarms?

As beekeepers (or at least as responsible beekeepers) we spend May and June rushing about like headless chickens trying to control swarming in our bees.

Many of us achieve this using a variety of methods which are generically referred to as artificial swarms. I suspect that many beekeepers think that the artificiality is because of our interventions.

Where have all my young girls gone?

Where have all my young girls gone?

It is … but it’s worth remembering that the artificial swarms we generate are very different in composition to natural swarms. Our artificial swarms predominantly leave the older bees associating with the queen, with the young bees remaining with the brood.

These old bees have to draw new comb and rear the new brood. These are activities they last did weeks ago (a long time in the life of a bee).

Final thoughts

There are artificial swarm control methods that were developed to better replicate the age distribution of bees in a natural swarm. One example of these is use of a Taranov board. I’ll cover this in a future post.

It’s also worth noting that the bees of different ages in a natural swarm have different roles even before they occupy a new location. The older bees form a mantle around the bivouacked swarm that protects it from inclement weather (amongst other things) and the oldest bees are the scouts responsible for finding a new nest site.

Again, both topics for another post … I’ve got bait hives to set out 🙂


Superorganism potential

The term superorganism can be used to refer to a colony of honey bees. The term gained prominence in the mid/late noughties having been reintroduced by the world-renowned myrmecologist 1 E.O. Wilson.

Bees, like ants (myrmex, “ant”, from the Greek μύρμηξ), are social insects in which there are divisions of labour. Different individuals within the colony perform different tasks. Some of these roles are defined by the castes in the colony – queen, worker and drone in a colony of honey bees for example – and some are defined by physiological differences between individual members of the same caste.

The term superorganism describes the entirety of the colony and is defined as a group or association of organisms which behaves in some respect like a single organism.

Essentially, a superorganism has characteristics and behaviours that the individuals within the colony – due to caste or physiological specialisation – do not exhibit.

The superorganism operates as a unified entity, collectively working together to maintain and reproduce the colony.

Division of labour and temporal polyethism

Drones and queens have relatively straightforward roles in the colony. Drones, like teenage boys, lounge around eating and thinking about sex. The queens are egg-laying machines.

An egg laying machine

An egg laying machine

Although there’s undoubtedly work involved in laying your bodyweight in eggs at the height of the season, the real work in the colony is – appropriately – done by the workers.

Worker bees exhibit temporal polyethism i.e. they display different patterns of behaviour depending upon their age. They have a maturational schedule in which they sequentially undertake age-correlated roles in the colony:

  • Young bees work in the hive in a series of roles starting with cell cleaning (days 1-2), nursing developing larvae (nurse bees; days 3-11) and wax production (days 12-17).
  • After two to three weeks the workers undergo significant physiological changes (weight loss, changes in immune function, reduced stress resistance) which prepare them for a productive life outside the hive. During this period the bees transition through a period when they act as guard bees.
  • Older bees (the ‘flying’ bees) perform a range of foraging activities including water carrying, pollen collection and nectar gathering.

And then they die in the field 🙁

Behavioural plasticity

This behavioural maturation is controlled by a so-called negative feedback loop between vitellogenin (Vg 2) and juvenile hormone (JH).

Nurse bees have high Vg levels which are reduced at the transition to foraging. Conversely JH levels increase with the onset of foraging (I know this sounds counterintuitive). These changes are responsible for a range of physiological changes in the worker bee.

Behavioural maturation in worker bees

Behavioural maturation in worker bees

But it’s not as simple as that. High Vg levels can block JH synthesis, so delaying maturation and foraging. Similarly, JH may reciprocally inhibit Vg synthesis and induce early foraging.

Clearly that last couple of sentences indicates that worker maturation is not an invariant process. It doesn’t always occur after 2-3 weeks.

In fact, the maturation or ageing process in honey bees is a very interesting phenomenon.

Ageing exhibits seasonal variability and remarkable plasticity.

Nurse bees can survive for at least 130 days and overwintering bees may survive up to 280 days. Clearly ageing in bees is a remarkably variable process. Overwintering bees ‘mature’ into either nurse bees or foragers. Presumably this has evolved as an effective mechanism of allowing spring colony build up (by having sufficient bees for the different roles) once environmental conditions improve.

In addition, there is another striking feature of the maturation process of honey bees.

Under certain social environmental conditions maturation is reversible.

This reversible maturation can be demonstrated by removing the nurse bees from the hive. Under these conditions some of the younger foragers revert, both behaviourally and physiologically, to nursing tasks. JH levels drop and Vg levels increase.

Old foragers are unable to undergo this rejuvenation.

Reversible maturation in worker bees

Reversible maturation in worker bees

Which finally and in a round the houses way gets me to the subject I meant to cover in the first place this week …

Brood and the superorganism

The honey bee colony superorganism not only contains a queen, workers and drones. It also contains brood. In the following text I’ll use the term brood as a collective noun meaning all the eggs, unsealed larvae and sealed pupae in the colony (unless otherwise specified).

Is the brood a component of the superorganism?

It certainly is.

Laying workers ...

Laying workers …

Remember previous discussion of laying workers. These are workers that lay unfertilised eggs which develop into drones. Egg laying by workers is suppressed by pheromones produced from unsealed brood 3. Therefore brood does influence the behaviour of the colony 4.

If the complete colony – brood, workers, drones and a queen – is a superorganism, which components of the colony, individually or together, have the potential to form the superorganism?

And why should this matter?

Swarming and the superorganism

During swarming, either naturally during colony reproduction, or during manipulation by the beekeeper, the ‘superorganism’ is broken up.

During natural swarming the (old) mated queen leaves the colony with 60-75% of the workers to establish a new colony. By the time the swarm leaves, the original colony – which has all the eggs, larvae and brood (obviously) – is usually already well on the way to rearing a new queen. The (new) virgin queen emerges, gets mated, and the colony has successfully reproduced.

Many of the colony manipulation methods that are used to prevent the loss of natural swarms exploit the potential of the components in the colony to form a complete new colony.

Most ‘artificial swarms’ work by breaking the colony – the superorganism – into two parts:

  1. The queen and the ‘flying’ bees. Even young bees can fly, so the term ‘flying’ bees refers to the older bees from the colony that have matured sufficiently to leave the hive.
  2. The nurse bees and all the brood.
Swarms, splits and superorganisms

Swarms, splits and superorganisms

These two parts both have the potential to create a new colony.

The queen and the flying bees that form the swarm (or the queenright part of an artificial swarm) occupy a new site (or hive 5), draw comb in which the queen lays, the larvae are fed 6, pupate and emerge. At the same time, foragers collect the necessary nectar and pollen to maintain the new colony.

The swarmed colony (i.e. the queenless part of an artificial swarm) contains ample stores and the nurse bees. What they don’t have is a queen. But they do have eggs and young larvae. The nurse bees select and feed one or more of these young larvae with copious amounts of Royal Jelly. A few days later a virgin queen emerges, matures, mates and returns to the colony to start laying eggs.

Sealed queen cell ...

Sealed queen cell …

Therefore both natural and artificial swarms exploit the potential in both parts of the original colony to eventually reproduce the colony.

No potential

Not all components of the colony have the potential to give rise to a new colony or superorganism. A solitary queen doesn’t even have the ability to feed herself properly, let alone double up for egg laying and nursing larvae duties.

This comes as a surprise to some people. If you frequent any of the online discussion forums you’ll sometime see questions posted like this:

What sort of hive do I need to buy to put a queen bee in to make honey?

Followed by some polite, or not so polite, responses saying that there’s a little bit more to beekeeping than that 7.

The ‘flying’ bees alone, in the absence of a queen, also have no potential. They can lay eggs (as laying workers, see above), but since the eggs are unfertilised the colony will be doomed. It’s not unusual for a queen from an artificial swarm (or from a cast) to fail to return from a mating flight, so condemning the workers in the hive to oblivion.

Swarms and behavioural plasticity

The classic artificial swarm involves moving the nurse bees and the brood to a new site, leaving the queen and the flying bees in the original location.

You do this so that the flying bees that have orientated to the position of the original hive – whether out in the field actively foraging or in the moved hive – eventually return and so become separated from the nurse bees and the brood.

In doing this you remove the urge to swarm and you weaken the queenless hive.

The majority of those flying bees are foragers.

And this is where behavioural plasticity is essential. remember that the artificial swarm predominantly contains foragers, not the nurse bees needed to feed developing larvae.

Some of these foragers undergo rejuvenation to produce wax or to become nurse bees. These build new comb and, in a few days, feed larvae that have hatched from the eggs laid by the queen.

This behavioural plasticity contributes to the potential of the artificial swarm to produce a new colony or superorganism.

A small swarm ...

A small swarm …

Do the same processes happen in natural swarms?

That requires a discussion of the worker composition of swarms which is not straightforward and will have to wait for another day 😉


Hanging around

I’ve recently discussed the misnamed ‘phoretic’ phase in the life cycle of Varroa destructor. Here we’ll briefly explore some features of this important, but non-reproductive, phase. It’s important because, as I’ll show, it influences subsequent mite reproduction.

I won’t rehash the life cycle of Varroa in detail as I’ve covered it previouslyVarroa is an ectoparasite of honey bees. It reproduces in capped cells, feeding on the developing pupa. A mated female mite enters the cell a few hours before capping and she and her incestuously mated daughters are released when the bee emerges.

The longer pupal development takes, the more progeny mites are produced, so Varroa has evolved to preferentially infest drone brood.

A smorgasbord of viruses

As an ectoparasite of honey bees, Varroa is responsible for the transmission of a smorgasbord of pathogenic viruses to the developing pupa. Subsequent virus replication, particularly by the aptly-named deformed wing virus, can result in developmental deformities.

Worker bee with DWV symptoms

Worker bee with DWV symptoms

Emerging workers with deformities are rapidly ejected from the hive. Other infested workers, with high viral levels, have reduced longevity. This is probably what accounts for the majority of overwintering colony losses. It may also explain so-called ‘isolation starvation‘.

The ‘phoretic’ phase

Mites outside capped cells are termed ‘phoretic’ mites. Recent studies have indicated that these mites are feeding on the workers to which they are attached. The same studies have shaken the long-held assumption that Varroa feeds on haemolymph 1 by rather neatly demonstrating that it is the fat body tissue of the bee that is the plat du jour.

The duration of the ‘phoretic’ phase is dependent upon the state of the colony. Since it is defined as the phase in which mites are not associated with developing pupae, in a broodless colony all the mites are ‘phoretic’. Under these circumstances mites remain ‘phoretic’ until either brood is produced, or they fall (or are groomed) off and drop through the open mesh floor.

The duration of the ‘phoretic’ phase

In colonies with ample brood the ‘phoretic’ phase is, on average, 6 days in length. The range often quoted is 4 – 11 days. The absolute figure must depend upon a number of factors. These include the chance of a mite encountering a late-stage larva. This is presumably influenced by the amount of suitable-aged brood in the colony and – because there is a division of labour in the hive – the type of bee upon which the mite is riding around the colony on.

For the purpose of this post we’ll consider bees of three ages – newly emerged, nurse bees and foragers. Newly emerged bees (days 1-2 post emergence) clean cells and nurse bees (days 3-11) feed developing larvae. Older bees are involved in wax production (days 12-17) and foraging (>18 days until death).

Logic would dictate that mites would ‘choose’ 2 to associate with bees that bring them into contact with developing larvae of the right age to infest.

Do they?

Hanging around with nurses

Xie et al., (2016)3 assembled artificial colonies containing equal numbers of new bees, nurse bees and foragers, all suitably marked so their age was known. These colonies were provided with a queen, open brood and stores. At the start of the experiment the colonies had 1500 bees and low Varroa levels.

The scientists then introduced 200 ‘phoretic’ mites from another colony 4 and left the colonies for 48 hours. They then age-sorted the bees and harvested all the ‘phoretic’ mites by washing them off in alcohol.

'Phoretic' mites prefer nurse bees

‘Phoretic’ mites prefer nurse bees

On average, ~16% of the nurse bees had ‘phoretic’ mites attached. In contrast, only ~10% of the foragers and ~5% of the new bees had mites. These studies involved seven individual experiments, in two countries and two separate years, using a different source colony for the mites. The statistics are significant.

So mites prefer nurse bees.

Is this ‘simply’ 5 because the nurse bees are more likely to bring the mite into close proximity with a suitably-aged larvae?

Or does associating with, and presumably feeding on, nurse bees have other benefits for the mite?

Mite fecundity and fitness

Fecundity is the reproductive productiveness 6 of an organism.

We’ve obliquely tackled this subject recently. Using an in vitro artificial ‘feed packet’ system, Ramsey and colleagues demonstrated that mites fed on the fat body of bees laid more eggs i.e. they had higher fecundity.

Xie et al., also tested fecundity of ‘phoretic’ mites from newly emerged bees, nurse bees and foragers. They did this by manually harvesting mites after a 3 day ‘phoretic’ phase, adding them to a pre-pupa and then counting the number of progeny female mites 9 days later.

Mite fecundity and fitness

Mite fecundity and fitness

‘Phoretic’ mites from nurse bees exhibited higher fecundity (more female offspring), higher fitness (more mature female offspring) and lower infertility (female mites that did not generate offspring).

So evolution has elegantly resulted in ‘phoretic’ mites associating with the right type of bee to bring them close to developing larvae (upon which they reproduce) and made them better able to reproduce once they get there.

Why do Varroa mites prefer nurse bees?

This is the title of the Xie et al., paper.

Xie et al., sort of answer the question they posed in the title. What they don’t do is explain why ‘phoretic’ mites on nurse bees are more fecund. However, the recent Ramsey paper suggests that this may be because nurse bees have a larger fat body and higher levels of vitellogenin.

If they’re better fed perhaps they produce more viable offspring? 7.

Another known unknown (semiochemicals)

How do mites detect the differences between new, nurse and older bees?

Perhaps they ‘smell’ different?

Mites preferentially infest drone brood because it produces a range of methyl and ethyl esters of straight-chain fatty acids, in particular methyl palmitate.

Similarly, the preference for nurse bees might be explained by their production of another semiochemicals 8. If we could identify this semiochemical it might be possible to create a ‘sponge’ soaked in it that attracted all the mites in the colony.

A bit simplistic, but you get the idea.

In reality, it’s likely that nurse bees are identified by the relative strengths of a range of semiochemicals produced by bees of different ages.

In reality it’s also likely that dropping a ‘sponge’ soaked in eau de nurse bee into the colony could unbalance all sorts of other events in the hive … 🙁

I told you it wasn’t simple.

Rattus norvegicus

Rattus norvegicus

Hanging Around is the fifth track on Rattus norvegicus, the 1977 debut album of one of the finest rock/punk bands of all time, The Stranglers. The album also includes the incomparable Peaches and (Get a) Grip (On yourself), both of which were released as singles.

No more heroes from the same year, but a different album, is also a classic by the same band.

You had to be there … I was 😉



Pedantically not phoresy

The life cycle of the ectoparasitic mite Varroa destructor essentially consists of two stages. The first is within the capped cell, where reproduction takes place. The second occurs outside the capped cell when the recently-mated female progeny mites matures while riding around the colony attached to a nurse bee.

Almost without exception this second stage is termed the phoretic phase.

It isn’t.


Phoretic is an adjective of the word phoresy. Phoresy is derived from the French phorésie which, in turn, has its etymological origins in the Ancient Greek word φορησις.

And φορησις means being carried.

Which partly explains why the correct definition of the word phoresy is:

An association between two organisms in which one is carried on the body of the other, without being a parasite [OED]

Phoresy has been in use for about a century, with the word phoretic first being recorded in the Annals of the Entomological Society of America (25:79) in 1932:

It is possible, as suggested by Banks (1915), that such young mites are phoretic, being carried about from place to place on the host’s surfaces.

And, no, they weren’t discussing Varroa.

“Without being a parasite”

These are the critical words in the dictionary definition of phoresy which makes the use of the word phoretic incorrect when referring to mites on nurse bees.

Because mites on nurse bees are feeding – or at least a significant proportion 1 of them are.

They are therefore being parasitic and so shouldn’t be described as phoretic.

Om, nom, nom 2

Last week I discussed the recent Samual Ramsey paper presenting studies supporting the feasting of Varroa on the fat body of bees.

In the study they harvested bees from a heavily mite-infested hive and recorded the location on the bee to which the mite was attached.

The majority were attached to the left underside of the abdomen. More specifically, the mite was wedged underneath the third abdominal tergite 3.

What were they doing there? Hiding?

Yes … but let’s have a closer look.

Ramsey and colleagues removed some of the mites and used a scanning electron microscope to examine the attachment point on the bee. Underneath the tergite there is a soft membrane. The imprint of the body of the mite was clearly visible on the membrane.

Varroa feeding location on adult bee

Scanning EM of Varroa feeding location on adult bee

The footpads of the mite were left attached to the membrane (left image, white arrows), straddling an obvious wound where the mouthparts had pierced the membrane (black arrow). Between them, the inverted W shape is presumably the imprint of the lower carapace of the mite.

The close-up image on the right even shows grooves at the wound site consistent with the mouthparts of the mite.

These mites were feeding.

Extraoral digestion

Varroa belongs to the order (a level of classification) Mesostigmata. Most mesostigmatids feed using a process termed extraoral digestion.

Extraoral digestion has also been termed ‘solid-to-liquid’ feeding. It involves the injection of potent hydrolytic enzymes which digest solid tissue, converting it to a semi-solid that can be easily ingested. It can reduce the time needed to feed and it increases the nutrient density of the consumed food.

If Varroa fed on haemolymph it wouldn’t need to use extraoral digestion. Instead it would need all sorts of adaptations to a high volume, low nutrient diet. Varroa doesn’t have these. It has a simple tube-like gut parts of which lack enzymatic activity … implying that digestion occurs elsewhere.

A picture is worth a thousand words

Do the images of feeding mites support the use of extraoral digestion?

EM cross-section of Varroa feeding

EM cross-section of Varroa feeding

The image above 4 shows the cross-section of a Varroa (V), wedged under the tergite (Te), feeding through a hole (arrow in the enlargement on the right) in the membrane (M). The fat body (FB) is immediately underneath the membrane. The scale bar is incorrectly labelled 5.

A close-up of the wound site shows further evidence for extraoral digestion.

Feeding wound at higher magnification

Feeding wound at higher magnification

Beneath the wound site (C, arrow) are remnants of fat body cells (white arrow) and bacteria (black arrow; of two types, shown in D). A closer look still at the remnants of the fat body (E and F) shows cell nuclear debris (blue arrows) and lipid droplets (red arrows).

These images are entirely consistent with extraoral digestion of fat body tissue by feeding Varroa. The presence of bacteria near the wound suggests that bacterial infection may result from Varroa feeding, possibly further contributing to disease in bees.

So, pedantically it’s not phoresy

So-called phoretic mites, unless they’re on the thorax or head of the bee, are not really phoretic. They are being carried about, but they are also likely feeding. By definition that excludes them from being phoretic.

Instead they are ectoparasites of adult bees.

What are the chances that beekeepers will stop using the term phoretic?

Slim to none I’d predict 6.

And, of course, it doesn’t really matter what the correct term for them is.

What’s more important is that beekeepers remember that it’s at this stage that mites are susceptible to all miticides.

The June gap

But it’s also worth thinking about the potential impact of brood breaks.

During brood breaks all the mites in the colony must be ‘phoretic’.

Generally, the majority of the mites in a hive are in capped cells. Depending upon the stage of the season, the egg-laying rate of the queen and other factors, up to 90% of the mites are associated with developing pupae.

But as the laying rate dwindles more and more mites are released from cells and become ‘phoretic’, unable to find a suitable late-stage larva to infest.

And which bees do the mites associate with?

Nurse bees primarily, for reasons I’ll discuss in the future. But – spoiler alert – one of the reasons is likely to be that they have a larger fat body.

So, a mid-season brood break (e.g. the ‘June gap’) is likely to result in lots more nurse bees becoming both the carriers and the dinner of the mite population.

Some or many of the nurse bee cohort may perish, perhaps from damage to the fat body or from the viruses acquired from the mite. However, bees exhibit phenotypic plasticity, meaning that older bees can revert to being nurse bees when the queen starts laying again.

Late season brood breaks

In late summer mite levels are usually at their highest in the hive. A brood break occurring now will release a very large number of mites to parasitise the adult bee population.

Presumably these mites select the bees best able to support them 7.

And which bees are these? The nurse bees of course. But it’s also worth remembering that there are key physiological similarities between nurse bees and winter bees. Both have low levels of juvenile hormone and high levels of vitellogenin (stored in the fat body).

So I’d bet that the ‘phoretic’ mites during a late season brood break would also preferentially associate with any early-produced winter bees.

Furthermore, once the queen starts laying again – perhaps in early/mid-autumn – the winter bees being produced would be subjected to the double-whammy of high levels of mite infestation and potential damage from ‘phoretic’ mites.

Practical considerations

More work is required to model or actually measure the impact of late season brood breaks, high levels of ‘phoretic’ mites, nurse bee numbers and winter bee development.

Compare two colonies of a similar size with a similar mite load, treated at the same time in early autumn with an appropriate miticide. If one of them experienced a late summer brood break (pre-treatment) and consequent high levels of ‘phoretic’ mites, does this reduce the chances of the colony surviving overwinter?

Who knows? Lots and lots of variables …

Fundamentally, it remains important to treat colonies early enough to protect the winter bee population. You’ve heard this from me before and you’ll hear it again.

However, it’s something to think about and I can see ways in which it might influence the strategy and timing of mite control used. I’ll return to this sometime in the future.


Chewin’ the fat

A little over a year ago reports started to circulate of a study showing that Varroa feed on the fat body of bees rather than on haemolymph.

Having worked in Glasgow through the early noughties the title of this post was a no-brainer and an outline draft was written in December 2017. However, the peer-reviewed paper wasn’t published until last month, so it’s only now we’ve got the chance to judge the study and consider its implications.

Varroa feed on hameolymph, right?

Historically this was the accepted dogma. However, the experimental data supporting this conclusion – based upon labelling bees with radioactive isotopes and seeing what the mites acquired after feeding – was really not definitive. The experiments had been done in the 1970’s and the specificity of the labelling was a bit dubious. In addition, during the intervening period scientists had determined that, unlike vertebrate blood which is rich in cells and nutrients 1, haemolymph has little of either and is actually a pretty lousy food source.

In addition, and somewhat more circumstantially, Varroa control using chemotherapeutics fed to bees (and subsequently taken up by the mite during feeding) had been relatively disappointing.

Perhaps these chemicals weren’t getting to the right tissues of the bee?

Perhaps Varroa don’t feed on haemolymph after all?

The Ramsey study

This new study reports three independent experiments that, together, indicate that Varroa actually feed on the fat body of bees, rather than on haemolymph. The paper is so-called ‘open access’, so anyone can access it and therefore I’ll just provide a synopsis of the important bits.

The questions Samual Ramsey and colleagues attempted to answer were:

  1. Where on the bee do mites feed? Is it primarily or exclusively near the fat body?
  2. When Varroa feeds, what host tissues are ingested?
  3. What sort of diet is required to maintain Varroa and allow their reproduction in vitro2.

Location, location, location

The authors counted phoretic mites on 104 bees. Over 95% of them were located on the underside of the body, predominantly on the left side of the bee, under the tergite or sternite3 on the third metasomal segment (i.e. the second visible segment of the abdomen).

Mite location on nurse bees

Mite location on nurse bees

This position is consistent with feeding on the fat body tissues which are most abundant under the inner ventral surface of the metasoma.

Seeing red

Bees were fed with Nile red, a lipophilic fluorescent stain that preferentially accumulates in the fat body. They co-fed bees with uranine, a differently coloured fluorophore that accumulates in the haemolymph. They then allowed mites to feed on the fluorescently labelled bees and subsequently photographed the mites under fluorescent light.

The rationale here was straightforward. If the mites fed on the fat body they would stain red due to taking up the Nile red stain.

Mites visualised after feeding on fluorescently labelled bees

Mites visualised after feeding on fluorescently labelled bees

Which they did.

It was notable that the red stain predominantly accumulated in the rectum and gut of the mite (image O above). The authors conducted all sorts of controls to confirm that the stains actually stained what they were supposed to – you can view these in the paper.


In the final part of the study the authors maintained mites in vitro (in an incubator), feeding them on a diet containing increasing amounts of fat body or haemolymph. These are tricky experiments and in some way the least satisfactory part of the study.

Two results suggest that fat body was beneficial or essential to the mites. Firstly, only mites that had 50% or more fat body in the diet survived for 7 days. Secondly, there was a dose response to the amount of fat body in the diet and fecundity. Mites on a 100% fat body diet exhibited 40% fecundity, the highest level observed in the study.

What can we conclude from the Ramsey study

Of the three experiments presented, the Nile red fat body stain uptake by mites is reasonably compelling.

The feeding position study is essentially correlative, but there could be other interpretations of the data. For example, that location on the bee might be the least accessible to a ‘grooming’ bee. Perhaps it’s a survival mechanism?

Survival and fecundity in in vitro studies wasn’t great. However, in defence of the authors, fecundity of mites under natural conditions can be as low as 40% and is not higher than 80%. Not all mites have baby mites. Thankfully.

Only 20% of the mites survived one week under in vitro conditions, even on a 100% fat body diet. In contrast, mites fed haemolymph alone died within 48 hours. This poor level of survival was surprising and suggests other essential components of the diet were probably missing.

Other published studies have shown reasonable survival of Varroa for at least 3 days, with at least one report of mites surviving on flowers for up to 7 days. I’m also aware that other laboratories can maintain mites in vitro for longer than 7 days without using any honey bee-derived components in the diet.

Hang on … what is the fat body anyway?

The fat body is multi-functional. It has been compared to the vertebrate liver and adipose tissue. It acts as a major organ for nutrient storage, energy metabolism and detoxification of things like pesticides.

Vitellogenin made by and stored in the fat body reduces oxidative stress and is associated with extending the longevity of overwintering bees. The fat body also has critical roles in metamorphosis.

So, not only multi-functional, but also very important.

Significance of the results … is this a game changer?

This paper has been discussed online as a ‘game changer’. That’s probably a bit strong. Whilst the fluorescent stain uptake study is reasonably convincing it must be remembered that it was conducted on adult bees.

Do mites on pupae also feast on the fat body?

This will have to be determined in the future. It’s a more difficult experiment of course.

The other two studies, and a number of additional small observations I’ve not discussed here, are certainly supportive, but not alone hugely convincing. The in vitro study in particular will be interesting to compare with (currently unpublished) studies from other laboratories that do not use honey bee fat bodies in their mite feeding and maintenance diet.

Practical matters

Does it matter what part of the bee the mite feeds on?

Clearly it does for the mite, but what about the beekeeper?

I think this study is significant for the beekeeper for two reasons – the first will only be relevant if and when lipophilic miticides are developed, the second matters right now.

  1. Strategies are being developed to add highly specific miticides to the diet of bees which are then delivered to Varroa when the mite feeds. To date, these have been rather underwhelming in their performance. If Ramsey is right, modification of these miticides to make them lipophilic (like the Nile red fluorphore) will concentrate them in precisely the right place to ensure the mites get a lethal dose.
  2. A key product of the fat body is vitellogenin. The long-lived overwintering bees have high levels of vitellogenin. Mites feeding on, and depleting, the fat body would be expected to result in reduced vitellogenin levels in the bee 4. This would explain why high Varroa levels are associated with reduced longevity of winter bees and consequently increased overwintering colony losses.

The most important take home message

To prevent mites that feed on fat bodies from damaging vitellogenin production miticides have to be used early enough to protect the winter bees.

In the paper Ramsey makes the statement:

Simple reduction of mite loads late in the season to decrease the overwinter parasite load may not be enough, as it still allows for the mites to damage tissue critical to the process of overwintering …

Instead …

A treatment schedule that includes treatment in late summer or early fall before mites can significantly damage fat body in developing winter bees would likely be more effective.

Which is precisely the point I’ve made previously about treating early enough to protect winter bees.

What the Ramsey paper adds is the piece of the jigsaw possibly explaining why late summer treatment is so important.


Chewin’ the Fat was a four-series Scottish comedy sketch show. It was broadcast from 1999 to 2002, with further Hogmanay specials until 2005. The show had a recurring cast of characters and sketches including The Big ManThe Banter BoysThe Lighthouse KeepersBallistic Bob and Taysiders in Space.

Gonna no' dae that

Gonna no’ dae that – The Lighthouse Keepers

Chewin’ the Fat was filmed in and around Glasgow (where I worked at the time) and the characters parodied a range of local ‘types’ … pretentious Kelvinsiders, Glaswegian gangsters, narcissistic golfers, The man from Kilmacolm, and shellsuit-wearing, chain-smoking, hard-drinking Glaswegian neds.

It was a bit rude and definitely an acquired taste. Without subtitles, some of the scenes would probably have been unintelligible south of the border.

Mites equal viruses

Healthy bees are happy bees 🙂

Sounds good doesn’t it?

Actually, there’s no evidence that bees display or perceive most of the emotions often attributed to them 1.

Happy? Who knows? But certainly not healthy ...

Happy? Who knows? But certainly not healthy …

A more accurate statement might be “Healthy bees are more productive, they are less likely to die overwinter, less likely to be robbed out by wasps or neighbouring strong colonies and their parasites and pathogens cannot threaten the health of other honey bee colonies or, through so-called-pathogen overspill, the health of other pollinators.”

More accurate?

Yes … but it doesn’t exactly trip off the tongue 😉

Whether it makes the bees happy or not, beekeepers have a responsibility to look after the health of their livestock. This includes controlling Varroa numbers to reduce the levels of pathogenic viruses in the hive.

How well are virus levels controlled if mite levels are reduced?

I’ll get to that in due course …

Midwinter mite massacre

The 2018 autumn was relatively mild through until mid/late November. In the absence of very early frosts colonies continued rearing brood.

We opened colonies in mid-November (for work) and found sealed brood, though it was clear that the laying rate of the queen was much-reduced.

These are ideal conditions for residual mite replication. Any mites that escaped the late summer/early autumn treatment (the ideal time to treat to protect the overwintering bees) continue to replicate, resulting in the colony starting the following season with a disappointingly high level of mites.

I’ve noted before that midwinter mite levels are paradoxically higher if you treat early enough in the autumn to protect the all-important winter bees.

Consequently, to start the year with minimal mite levels, I treat in midwinter with a trickled or vaporised oxalic acid-containing (OA) treatment.

A combination of colder weather (hard frosts in late November) and brood temperature measurements 2 indicated mid-December was a good time to treat.

Midwinter mite massacre

Midwinter mite massacre

18th December

In one of my apiaries ten colonies were treated. Some were definitely broodless (based upon Arnia hive monitoring). Others may have had brood, but colonies were not routinely checked.

Over the four day period after vaporising these ten colonies dropped a total of 92 mites. More than 50% of these were from just one double-brooded colony. Overwintering nucs 3 dropped no mites at all in the 12 days following treatment.

This was very encouraging. These are lower midwinter mite levels than I’ve seen since returning to Scotland in 2015.

The one colony with ‘high’ mite levels received two further treatments (on the 22nd and 27th) in an attempt to minimise the mite levels for the start of the season. Going by the strength of the colony and the debris on the Varroa tray it was presumed that this colony was still rearing brood.

Mite drop following the third treatment was negligible 4.

Why are mite levels so low?

I think it’s a combination of:

  • Luck
  • Use of natural, organic, bee-centric and biodynamic beekeeping methods
  • Varroa-resistant bees
  • Very tight control of mite numbers in the 2017/18 season, primarily by correctly timing the winter and the late-season autumn treatments. This is simply good colony management. Anyone can achieve this.
  • A brood break midseason and/or a broodless period when splitting colonies (both give opportunities for more phoretic mites to be lost through grooming). Undoubtedly beneficial but season-dependent. I’ll be discussing ways to exploit these events in posts next year.
  • A low density of beekeepers in Fife, so relatively little drifting or robbing of poorly managed colonies from neighbouring apiaries. Geography-dependent. Much easier in Fife than Warwickshire … and easier still in Lochaber.

And what do less mites mean?

Varroa is a threat to bee health because it transmits pathogenic viruses when feeding on developing pupae.

The most important of these viruses is deformed wing virus (DWV).

Generally, the higher the level of infestation with mites, the higher the viral load 5. This has been repeatedly demonstrated by studies from researchers working in the UK, Europe and the USA.

It is well-established that colonies with high viral loads have an increased chance of dying overwinter, due to the decreased longevity of bees infected with high levels of virus.

DWV symptoms

DWV symptoms

In our work apiaries we regularly measure DWV levels. For routine screening our limit of detection is around 1,000 viruses per bee.

We don’t actually count the viruses. They’re too small to see without an electron microscope 6.

Instead, we quantify the amount of the virus genetic material present 7, compare it to a set of standards and express it as ‘genome equivalents (GE)’.

Many of the bees tested this year contained ~103 (i.e. 1000) GE, which is extremely low. Bees from Varroa-free regions (e.g. Colonsay) carry similar levels of DWV.

Most of our colonies were at or close to this level of virus much of the 2018 season. This is 100-1,000 times lower than we often see even in apparently perfectly healthy colonies in other years or other apiaries.

For comparison, using the same assay we usually detect about 1010 (ten billion) DWV GE per bee in symptomatic adult bees from heavily mite-infested colonies.

So, less mites means less viruses which means healthier bees 🙂

And they might even be happier bees 😉

And your point is?

It’s worth remembering that the purpose of treating a colony with miticides is to reduce the transmission of viruses between bees. This transmission results in the amplification of DWV. This is why the timing of treatments is so important.

Yes, it’s always good to slaughter a few (or a few thousand 🙂 ) mites. However, far better massacre them when you need to protect particular populations of bees.

This includes the overwintering bees, raised in September, that get the colony through to the Spring.

Remember also that it ‘takes bees to make bees’ i.e. the rearing of new brood requires bees. Therefore strong colony build-up in Spring requires healthy workers rearing healthy brood.

This is why it’s important to minimise mite levels in midwinter when colonies are broodless.

What do most beekeepers do?

Fifteen months ago I published a post on the preparation of oxalic acid solutions for trickling colonies in midwinter.

Whatever the vapoholics on the online forums claim, trickling remains the easiest, quickest and least expensive way to treat colonies in midwinter 8.

The best time to treat in the winter is when the colony is broodless. Here in Fife, and often elsewhere, I believe that this usually occurs earlier in the winter than many beekeepers treat (if it happens at all … or if they treat at all).

I usually treat between the end of the third week in November and mid-December, at the end of the first extended cold period.

Oxalic acid preparation recipe page views

Oxalic acid preparation recipe page views

Looking at the page views for these oxalic acid recipes it looks as though many beekeepers treat after Christmas 9 … which may be suboptimal if colonies had a broodless period and now started rearing brood again.

Mine have.

This winter has been quite mild (at least at the time of writing) so there may yet be opportunities to treat really effectively during a broodless period.

Or the chance may have gone …


Know your enemy

What less appropriate time is there, as we enter the festive season of goodwill, to provide a brief account of the incestuous and disease-riddled life cycle of the Varroa mite?

Happy Christmas 🙂

Scanning electron micrograph of Varroa destructor

Scanning electron micrograph of Varroa destructor

Varroa is the biggest enemy of bees, beekeepers and beekeeping. During the replication cycle the mite transfers a smorgasbord of viruses to developing pupae. One of these viruses, deformed wing virus (DWV), although well-tolerated in the absence of Varroa 1replicates to devastatingly high levels and is pathogenic when transferred by the mite.

Without colony management methods to control Varroa, mite and virus replication will eventually kill the colony.

I’ve written extensively on ways to control Varroa. Most of these have focused on early autumn and midwinter treatment regimes. However, next season I’m hoping to discuss some alternative strategies and will need to reference aspects of the life cycle of Varroa … hence this post.

What is Varroa?

Varroa destructor is a distant relative of spiders, both being members of the class Arachnida … the joint-legged invertebrates (arthropods). It was originally (and remains) an external parasite (ectoparasite) of Apis cerana (the Eastern honey bee) and – following cross-species transfer a century or so ago – Apis mellifera, ‘our’ Western honey bee.

Apis cerana, having co-evolved with Varroa, has a number of strategies to minimise the detrimental consequences of being parasitised by the mite.

Apis mellifera doesn’t. Simple as that 2.

One hundred years is the blink of an eye in evolutionary terms and, whilst there are bees that have partial solutions – largely behavioural (small colonies and very swarmy) – they’re probably unable to collect meaningful amounts of honey 3.

Varroa-resistant honey bees will probably evolve (as much as anything is predictable in evolution) but not in my time as a beekeeper … or possibly not until Voyager 2 leaves the Oort Cloud 4.

And there’s no guarantee they’ll be any use whatsoever for beekeeping …

The replication cycle of Varroa

Varroa has no free-living stage during the life-cycle. The adult mated female mite exhibits two distinct phases during the life-cycle. It has a phoretic phase on adult bees and a reproductive phase within sealed (‘capped’) worker and drone brood cells. Male mites only ever exist within sealed brood cells.

I’m going to discuss phoretic mites in a separate post. I’ll concentrate here on the replication cycle.

The mated female mite enters a cell 15-50 hours before brood capping. Drone brood is chosen preferentially (at ~10-fold greater rates than worker brood) and entered earlier. Depending upon the time of the season and the levels of mites and brood, up to 70-90% of mites in the colony occupy capped cells.

The first egg is laid ~70 hours after cell capping. This egg is unfertilized and develops into a haploid male mite. Subsequent eggs are fertilised, diploid, and so develop into female mites. These are laid at ~30 hour intervals.

The replication cycle of Varroa

The replication cycle of Varroa

Worker and drone brood take different times to develop. Therefore a typical reproductive cycle involves five eggs being laid in worker brood and six in drone brood. Not all of these eggs mature, their development being curtailed by the bee emerging as an adult.

There are all sorts of developmental stages involved in getting from an egg to a mature unfertilised mite, but these are not important in terms of the overall outcome. Mite-geeks love this sort of detail 5, but we need to cut to the chase …

Keeping it in the family

The foundress ‘mother’ mite and her progeny all share a single feeding hole through the cuticle of the developing pupa.

What a lovely scene of family ‘togetherness’. 

Male and female mites take 6.6 and 5.8 days respectively to develop to sexual maturity. Therefore the male mite reaches sexual maturity before the first of his sisters.

He then lurks around the attractive-sounding “faecal accumulation site” and mates with each of the (sister) females in turn.

What a little charmer 😉

Male mites are short lived and the eclosion of the adult worker or drone curtails further mating activity, releasing the foundress mite and the mated mature daughters 6.

Reproductive rate (mites per cell)

The three day difference in the duration of worker and drone development means that more mites are produced from drone cells than worker cells. Depending on conditions the reproductive rate is 1.3 – 1.45 in worker brood and 2.2 – 2.6 in drone brood.

Remember that the foundress is also released from the cell. She can go on to initiate one or two further reproductive cycles (or up to 7 in vitro). Consequently, the average yield of mature, mated female mites from worker and drone cells is a fraction over 2 and 3 respectively.

Before entering a fresh cell containing a late stage (5th instar) larva the newly-mated mites need to mature. They do this during the phoretic phase which lasts 5-11 days. Therefore the full replication cycle of the mite probably takes a minimum of about 17 days.

Exponential growth

Two to three mites per infested cell doesn’t sound very much. However, under ideal conditions this leads to exponential growth of the mite population in the colony. Assuming 10 reproductive cycles in 6 months, a single mite would generate a population of >1,000 in worker brood and >59,000 in drone brood 7.

Fortunately (for our bees, not for the mites), ideal conditions don’t actually occur in reality.

Lots of things contribute to the reduction in reproductive potential. For example, only 60% of male mites achieve sexual maturity due to developmental mortality, drone brood is only available at certain times in the season, brood breaks interrupt the availability of any suitable brood and grooming helps rid adult bees of phoretic mites.

Out, damn'd mite ...

Out, damn’d mite …

However, these reductions aren’t enough. Without proper management mite levels still reach dangerously high levels, threatening the long-term viability of the colony.

In the next few months I will discuss some additional opportunities for reducing the mite population.

In the meantime, as we reach the winter solstice, colonies in temperate regions may well be broodless and – as emphasised last week – this is an ideal time to apply a midwinter oxalic acid-containing treatment. This will effectively reduce mite levels for the start of the coming season.

Happy Christmas … unless you’re a mite 😉


Today is the winter solstice in the Northern hemisphere. This is actually the precise time when the Earth’s Northern pole has its maximum tilt away from the Sun. However, the term is usually used for the day with the shortest period of daylight and the longest period of night. In Fife, sunrise is at 08.44 and sunset at 15.37, meaning the day length is 6 hours and 53 minutes long.

With increasing day length queens will start laying again … but there’s a long way to go until winter is over.


The eyes have it

We’re entering the not beekeeping end-of-season phase of the beekeeping year. There’s been a marked reduction in visitor numbers to ‘The Apiarist’ over the last few weeks and – with the weather gradually deteriorating1 – the ‘shack nasties‘ are starting to develop. The online forums (fora?) are filled with increasingly bad-tempered arguments discussions and it might be too soon to be thinking about 2019 (it isn’t).

Wasted words

So, rather than write a series of erudite, well-argued, coherent, logical and persuasive posts about evolution of Varroa resistance in Apis mellifera, or rational mid-season mite-management strategies, or the 75:25 rule for queen and stock-improvement, or an exhaustive review of Swienty vs. Abelo poly Nationals …

Some chance !2

… I’m instead going to spend the next few weeks on a variety of odds and ends. Some interesting and amusing science, some ‘teasers’ on grow-your-own-denim-knit-your-own-yoghurt beekeeping, an introduction to why people keep bees and why they shouldn’t and a long and apologetic explanation of where the site disappeared to when I tried to move it to another server.

Actually, of these only the first exists (see below). The last I hope not to use, though I will be switching servers to accommodate changes in security and to speed things up and to introduce site-wide intrusive advertising and a subscription model to fund my seemingly-unstoppable purchasing of essential beekeeping stuff from Brian at Thorne’s of Newburgh.


The eyes have it

The eyesight of bees is remarkable.

Actually, eyesight alone is not enough. It’s the combination of eyesight with the neuronal processing of the received images that’s truly remarkable.

Remember that the brain of a bee is about 1mm3 and contains about one million neurones 3. With this brain the bee is able to undertake a series of complex mental tasks involving learning and memory, image processing and visual generalisation.

Bees soon learn that particular flowers yield lots of pollen or nectar. They can return to them time and again, recognising them at relatively short distances from their appearance. How do they determine that other flowers – of different shapes, sizes or colours – might also have valuable pollen or nectar? What about tree flowers that have a different appearance again?

It turns out that bees are generalists, at least where flowers are concerned. They recognise things that are flower-like. They have evolved to associate reward (pollen, nectar) with things that have the appearance of flowers.

More general generalists?

Do they only have the ability to identify flower-like ‘things’. Are bees generalists when just identifying flowers and flower-like things, albeit of different colours, sizes and shapes? Is there some sort of hardwiring in the brain of the bee that has evolved this exquisite combination of flower-recognising sensitivity and flexibility?

Alternatively, perhaps bees have a more adaptable image processing capability? For example, we know through simple experiments that bees can rapidly learn to associate very unflower-like shapes with a syrup ‘reward’.

You can train bees to repeatedly return to a distinctively coloured/shaped item with a syrup reward. Over short distances you can move the item and the bees return to the new location, with the final approach being guided by vision, pattern recognition and associated image processing.

The item doesn’t need to look much like a flower.

They can also identify ‘diamond-shaped things’ (again, for example, other shapes are available at a bee lab near you) of a different colour, or even no colour, to those they’ve been trained on.

Shape recognition by bees

Shape recognition by bees …

This suggests that – at least for simple ‘not-flower’ shapes like these – a degree of generalisation is still possible.

But bees operate in a very busy and variable environment filled with shapes and colours that form wildly variable complex images. Generalisation might well be a problem with all of this variation and complexity.

Facial recognition

Bees are good at discriminating between images of simple shapes. How good are they are at recognising the sorts of complex shapes and images that are found in the environment?

What about something that our bees see every week?

Something they might associate with disruption and/or reward?

Like your face …  😀

It turns out that bees can distinguish between faces. Very well. When trained to a syrup reward on one face (top left in image below), they can distinguish it from a different face (second row) about 80% of the time (graph).

Facial recognition by bees ...

Facial recognition by bees …

There’s a distinct possibility your bees recognise you. An interesting twist on the comment many non-beekeepers make about whether we can identify ‘our’ bees.

However, bees trained to recognise a face the ‘right way up’ failed to identify the face if it was inverted (column v above).

Don’t do your hive inspections standing on your head.

Complex image generalisation

Bees can certainly discriminate well between complex images like faces. Are they also able to generalise when it comes to complex image analysis?

For example, could you train bees to associate reward with a range of female faces? Then challenge them with discriminating between a pair of new (never seen before) male and female faces? Would they pick the female face significantly more than 50% of the time?

That’s a pretty tough test. Without the labels how well do you cope with this training set?



OK, if that’s too hard, how about an analysis of image generalisation based upon the style of the image?

Humans are pretty good at this sort of thing. We can easily discriminate between the Impressionist painters (e.g. Degas, Monet, Manet, Renoir) and those in the early 20th Century Cubist art movement (e.g. Picasso, Metzinger, Braque, Gleizes, Léger).

For example, is the Picasso (below) Cubist or Impressionist?

We can tell … can the bees?

Monet or Picasso?

Wu and colleagues4 recently attempted to answer this question.

They took pairs of paintings matched for luminance, colour and spatial frequency information, one by Monet (Impressionist) and one by Picasso (Cubist). Bees were trained to associate either the Monet or the Picasso with a syrup reward.

When subsequently tested, bees were able to easily identify the painting they had been trained on from one by the other artist. After 30 training blocks bees made the correct choice about 75% of the time … approximately the same accuracy with which they identify faces (above).

This is not fundamentally a different experiment to the face recognition study as both involved the discrimination between just one complex image and another.

Monets or Picassos?

Using five different pairs of luminance, colour and spatial frequency-matched paintings from Monet and Picasso – with five days of training – they demonstrated that bees could simultaneously discriminate between them up to 75% of the time.

The more training the bees received, the better they were at picking the correct painting each time.

Impressionist or Cubist?

Having trained the bees on multiple Picasso or Monet paintings they then challenged them with new (to the bees … you’ll appreciate that these artists no longer produce new work 😉 ) paintings by the same artists.

Could bees that were able to discriminate between The Cliff at Étretat after the Storm (Monet) from Le Rêve (Picasso) and between Water Lilies and the Japanese bridge (Monet) from Girl before a Mirror (Picasso) and three other pairs correctly select a previously unseen Picasso or Monet?

In all honesty, not very well 🙁

The statistics are poor. For one of the novel pairs tested it appeared as though the bees could discriminate as well as they could one of the training pairs. However, the most positive statement that could be made by the authors was “Notably, for both groups the percentage of correct choices for novel pairs was above chance (i.e., above 50 %) in six out of the eight tests, indicating that a weak generalization may have occurred”.

Underwhelming … in this sort of science the stats wins every time, and this type of statement isn’t very compelling.

Finally, the scientists repeated the entire training regime with greyscale versions of the same training pairs of images. With this training, generalization to the novel pairs was quite a bit better with “only marginal or no significant difference between training pairs and most novel pairs”.

Better, but still not really statistically compelling. However, don’t underestimate the complexity of the task. The results showed that insects with a sesame-seed-sized brain could often discriminate between previously unseen Cubist or Impressionist paintings after a few days training on only 5 pairs of paintings of the same style.

That’s remarkable.

Bird brains

Pigeons live in the same visually complex environment as bees. They have to undertake similar visually demanding tasks during foraging. They can discriminate between Monets and Picassos. They can correctly (and statistically convincingly) determine whether a new Monet or Picasso is more Impressionist-like or Cubist-like.

In addition, when challenged with other paintings of similar styles by different artists (e.g. a Degas or a Braque), pigeons can again generalise in their selection of Cubist or Impressionist 5.

However, to achieve this remarkable visual feat, pigeons need to be trained to hundreds of exemplar paintings over many, many weeks.

Could bees do as well if trained for the same period?

We don’t know.

And we’re unlikely to find out as the lifespan of a worker bee is probably too short 🙁


This post was written as the political fallout of the draft Brexit deal was occupying 110% of the news. By the time it appears online it’s not clear the UK will have a Prime Minister or even a functioning Government.

Be that as it may, there will be a Parliamentary vote on it.

Historically, there is a division of the assembly into those that support the motion (the ‘ayes’ i.e. ‘yes’) and those that do not (the ‘noes’). Once the vote is taken – typically by members of parliament traipsing into the appropriate division lobby – the Speaker counts the votes and announces The Ayes have it … assuming the motion was supported.

Considering the timing, a pun on The Ayes have it seemed appropriate.

Resistance is not futile

Apivar ...

Apivar …

Amitraz-containing miticides are sold in the UK as Apivar and Apitraz.

Until recently they were only available with a veterinary prescription. I expect – though I have not yet seen data to support this – that their usage in the UK will increase now they are off-prescription. These miticides are now widely available and so there is greater opportunity to use – and misuse – them.

If you’re using Apivar 1 for the first time this year you will soon have to remove the strips from the hive.

That’s assuming you started treating early enough to protect the all-important winter bees from Varroa and its deadly viral payload.

This post is a reminder to remove the strips at the right time. The alternative – leaving them in place until the first Spring inspections – risks help the development of resistance to amitraz, so further reducing our opportunity to control mites effectively.

Leave and let die

Without careful integrated pest management (IPM) 2 Varroa levels build up in the hive. Varroa transmits viruses – most important of which is deformed wing virus (DWV) – to developing pupae. High levels of DWV either kills the pupa or results in emergence with or without significant developmental defects. Even those bees that are apparently normally developed have a reduced lifespan 3.

Winter bees with a reduced lifespan prevent the colony from surviving through the winter until the queen starts laying again. I’ve also proposed recently that high levels of DWV, and the resulting increased rate of winter bee die-offs, probably accounts for some cases of isolation starvation.

So … intervention is needed to reduce mite levels, protect your bees and save your colonies.

Follow the instructions!

Apivar is one solution to reduce mite levels. It is an easy-to-apply chemical treatment that is very effective in reducing the Varroa load by ~95%. For a National hive it is applied as two polymer strips, each containing 500mg of slow-release Amitraz. Strips are hung between brood frames for 6-10 weeks and – for maximum efficacy – should be scratched with a hive tool and repositioned half way through the treatment period.


Amitraz …

Unlike some other miticides (e.g. Apiguard and MAQS) there are no temperature restrictions for Apivar usage. The colony does not need to be broodless (a requirement for trickled oxalic acid-based treatments) as the treatment period covers multiple brood cycles.

Other than not using it with supers present the only contraindication for Apivar is to not use it if Amitraz-resistant mites are present.

How does resistance develop?

When discussing parasites and pathogens, resistance 4 is a consequence of two things:

  1. A selective pressure that kills the pathogen
  2. A population which exhibits genetic diversity

The selective pressure could be anything … heat for example, antibiotics prescribed by your GP, an antiviral against HIV or – of relevance here – Apivar against Varroa.

Killing – at the population level – is not absolute. Some individuals within the population survive longer than others. They could be exposed to a slightly lower dose, or be located in a protected niche for example. However, treat for long enough and the majority will be killed.

But there’s more …

Pathogen populations are not genetically invariant. Actually, many are quite diverse and have replication cycles that – deliberately 5 – generate diversity.

Therefore some pathogens are genetically slightly less resistant and some are genetically slightly more resistant to a selective pressure. We can ignore the former as they’ll rapidly be killed off … but we must be concerned about the more resistant ones.

Keep taking the pills

All of this is a ‘numbers game’, better represented with graphs and equations. However, the take-home message is simple … to effectively control a pathogen you need to treat for long enough and with a high enough dose to kill the vast majority of the population.

That’s why you’re encouraged to “complete the course” of antibiotics … or to remove the Apivar strips after 10 weeks and not leave them in over the winter.

Because both courses of action result in selection of more resistant pathogens.

If you stop taking antibiotics too soon, you won’t have treated for long enough and with a high enough dose. You end up selecting for the more genetically resistant pathogens.

Similarly, if you leave Apivar strips in overwinter you’ll be “treating” the remaining mites 6 with a lower dose of the miticide, which is an ideal situation to favour the growth of the slightly more genetically resistant mites.

How does Amitraz resistance develop?

Resistance to Amitraz in Varroa is well documented. It’s been described in a number of countries including the USA and Europe, Mexico and Argentina 7. Generally resistance is defined in terms of a reduced level of mite killing, or – in laboratory experiments – an increased dose required to kill a certain proportion of mites.

However, I’m unaware of any studies defining the genetic basis of Amitraz resistance in Varroa.

But Amitraz is a widely-used acaricide 8 and the genetic basis of resistance in cattle ticks is well understood. In these, ticks resistant to Amitraz carry a mutation in the RMβAOR gene 9.

What 10 is the RMβAOR gene?

I’m glad you asked 😉

This gene encodes the β-adrenergic octopamine receptor protein and readers with good memories will recall that this is one of the targets that Amitraz binds to and inactivates 11.

If the protein carries a mutation the Amitraz cannot bind to it and so the mite – or more correctly the tick as it’s yet to be formally demonstrated in mites – is therefore resistant.

(Bad) practical beekeeping

What does all this mean in terms of practical beekeeping?

It means use the correct number of Apivar strips for the colony and leave them in for the right length of time.

Do not …

  • Use one strip on a full colony mid-season to ‘knock back the mites a bit’ 
  • Re-use the strips in another colony (yes really!)
  • Use improperly stored strips (or out of date strips) in which the effective Amitraz dose is reduced

I’ve heard examples of these types of misuse this season. All increase the chance of selecting for Amitraz-resitant mites.

And (the real reason for posting this at this time of year) …

  • Do not leave the strips you added in late summer in the colony throughout the winter

Removing the strips takes seconds. Prize off the crownboard, grab the tab projecting above the top bars, gently withdraw the strip and close the hive up again.

Finally, because of the incestuous lifestyle 12 of Varroa the genetic diversity (and therefore potential presence of more resistant mites) in the population is likely to be increased by the high mite levels that prevail late in the season.

All the more reason to use the effective treatments we currently have in a way that helps ensure they remain effective.


Resistance is futile

Resistance is futile

Resistance is futile is the title of a 2018 album by the Welsh rock band the Manic Street Preachers.

More specifically, in the context of this post, it was the phrase routinely used by the Borg – the alien cyborgs sharing a collective mind – in the Star Trek franchise. Borgs rarely speak, but when they do they usually include this phrase. For example “We are the Borg. Lower your shields and surrender your ships. We will add your biological and technological distinctiveness to our own. Your culture will adapt to service us. Resistance is futile.” The warning about resistance being futile was usually accompanied by the threat that the target would be assimilated”.

I’d started writing this post using the title ‘Resistance is futile’ but realised late on that – as far as Varroa are concerned – resistance is anything but futile 13.

Resistance – to miticides – gives Varroa a reason to live. Literally.

Let’s not help them 🙂

Ouch, that hurt

If you keep bees you’ll inevitably get stung.

Not necessarily often and not necessarily badly, but getting stung goes with the territory.

You’ll probably get stung more often if your bees are stroppy, or if you are clumsy. But even if you’re careful and the bees are calm there’s always the chance of being stung.

I moved a very feisty colony late one evening last week 1. The hive was sealed, moved and re-located to an out apiary. Knowing they were, er, rather temperamental I let them settle for 15 minutes, then gently lifted the entrance block.

Out they boiled … as I beat a very hasty retreat 🙁

I thought I’d got away with it, but driving home 20 minutes later I was stung on the ankle by a stowaway in my boot.

Ouch! That hurt.

I’ve only been stung a few times all season. Most didn’t hurt much at the time and were forgotten within minutes. That sting on the ankle hurt like hell and was sore for a further 48 hours.

Why does it hurt when you’re stung? Furthermore, assuming stings are inevitable, which parts of the body hurt more when stung … and so deserve additional protection?

Why do bee stings hurt?

The honey bee sting is a hollow barbed tube used to deliver the venom. About 50% of bee venom by weight is the small protein mellitin.

It’s fair to say that mellitin is small but potent. It’s only 26 amino acids 2 long and forms a tetramer in aqueous solution. The ‘noughts and crosses’ shape it adopts hides the hydrophobic parts of the peptide and therefore allows it to ‘dissolve’ in venom. However, the tetramer dissociates at or near cell membranes into which monomeric mellitin embeds itself.



And this is where the pain and damage start …

Membrane-association causes cell lysis 3. This results in the release of all sorts of cytokines from the cells which signal ‘damage’ to the body, leading to the inflammatory response usually associated with bee stings. That’s the long-term effect of a bee sting. However, simultaneously, mellitin triggers the expression of proteins known as sodium channels in pain receptor cells. These allow large amounts of sodium to flow across the membrane. It is this that is directly responsible for the pain sensation when you are stung.

So, if being stung is almost inevitable and if bees have evolved stings to cause pain (which they have), in which parts of the body is the pain sensation most marked?

Measuring pain

Pain is a subjective response. What’s painful to me might hardly be noticed by someone with a higher pain threshold. Two individuals receiving the same sensory input can experience very different sensory responses 4.

As an aside it’s well documented that there are differences in the pain felt by males and females 5. All the pain reported in this article is from studies – or personal experience – by males.

Therefore, to meaningfully determine how much pain a sting causes, from a particular insect or at a particular location for example, it’s essential that the studies are properly controlled. This includes taking account of variation between individuals and variation within an individual on a day to day basis.

These are not the sorts of studies that attract large numbers of volunteers 😉

Perhaps unsurprisingly, the scientific work in this field is often published in single author papers in which the author alone is the ‘volunteer’.

The Schmidt Sting Pain Index

Before discussing honey bees specifically a brief diversion must be made to introduce the seminal studies by Justin Schmidt.

Schmidt is an entomologist at the Carl Hayden Bee Research Centre in Arizona. He’s interested in haemolysis (the cell lysis caused by mellitin and other constituents of insect venom) and whether the evolution of sociality in hymenopterans (bees, ant and wasps) required the evolution of toxic and painful stings.

Over about twenty years Justin Schmidt published a number of papers on hymenopteran venoms and the pain that they cause. In his early papers he rated stings on a scale of 1 – 4 (actually 0 upwards, but 0 was totally painless to humans).

Only a very few insects scored 4, including bullet ants about which Schmidt comments “Paraponera clavata stings induced immediate, excruciating pain and numbness to pencil-point pressure, as well as trembling in the form of a totally uncontrollable urge to shake the affected part“.

You can’t fault his commitment, but you might question his sanity.

Schmidt published his magnum opus in 1990 in which he ranked stings by 78 hymenopteran species covering 41 genera 6. His descriptions of the pain induced are often entertaining.  The aforementioned bullet ant is “pure, intense, brilliant pain…like walking over flaming charcoal with a three-inch nail embedded in your heel“.

Another sting scoring 4, that of Synoeca septentrionalis (the warrior wasp) is accompanied by the statement “Torture. You are chained in the flow of an active volcano. Why did I start this list?”.

Why indeed?

Standing on the shoulders of giants

Bees and wasps scored 2 on the Schmidt Sting Pain Index. What Schmidt didn’t investigate was the influence of the location of the sting on the pain experienced.

Which brings me to Michael Smith. In 2014 Smith published an entertaining paper entitled Honey bee sting pain index by body location. It’s published in the journal PeerJ and the full paper is available for download.

It’s a well controlled study written in an engaging style that most readers will appreciate.

Building on the landmark studies by Schmidt, Michael Smith rated the pain endured from honey bee stings in 25 different locations.

Some of these locations should really be protected with a bee suit.

Sting locations tested

Sting locations tested

Smith controlled the study by always including an “internal control” i.e. comparing two test locations with three stings to his forearm.

Every time.

All locations were tested in triplicate (randomly). This meant that Smith was stung a minimum of five times for 38 consecutive days. Ouch.

There are some excellent quotes in the paper … “Some locations required the use of a mirror and an erect posture during stinging (e.g., buttocks)“. Scientists involved in studies that require ethical approval will appreciate the comments made in the paper on self-experimentation.

And the results? To quote directly from the paper “The three least painful locations were the skull, middle toe tip, and upper arm (all scoring a 2.3). The three most painful locations were the nostril, upper lip, and penis shaft (9.0, 8.7, and 7.3, respectively)” 7. Interestingly, skin thickness did not correlate with the pain experienced.

My experience of stings is limited. Those I’ve had to the face (including the lower lip) have been relatively painless, but the subsequent inflammatory response has been dramatic. Smith only scored immediate pain … I think a follow-up study on inflammation and its duration is needed.

I’m not going to conduct it.

Points pain means prizes

You can’t fault the dedication shown by Justin Schmidt and Michael Smith 8. That sort of dedication should be recognised with prizes and honours.

And it was.

Schmidt and Smith shared the 2015 Ig Nobel prize for Physiology and Entomology. The Ig Nobels – a parody of the Nobel prizes – recognise unusual or trivial achievements in scientific research.

The list of Ig Nobel prizes awarded is eclectic and highly entertaining … Medicine (2013) “for treating “uncontrollable” nosebleeds, using the method of nasal-packing-with-strips-of-cured-pork“, Economics (2005) for the inventors of “Clocky, an alarm clock that runs away and hides, repeatedly, thus ensuring that people get out of bed, and thus theoretically adding many productive hours to the workday” and Psychology (1995) for “their success in training pigeons to discriminate between the paintings of Picasso and those of Monet9.

Sir Andre Geim received the Physics Ig Nobel in 2000 for levitating a frog by magnetism. Yes, really. Ten years later he was awarded the Nobel prize in Physics for his studies on graphene. He’s the only holder of a Nobel and Ig Nobel.

Marc Abrahams, the founder of the Ig Nobel awards, regularly tours giving talks on Improbable Research and the Ig Nobel prizes.

Go if you get the chance … it’s highly entertaining.