Category Archives: Woodpeckers

Winter covers and colony survival

Synopsis : A recent study shows increased overwinter colony survival of ‘covered’ hives wrapped in Correx and with insulation under the roof. What provides the most benefit, and are the results as clear cut as they seem?

Introduction

A recent talk by Andrew Abrahams to the Scottish Native Honey Bee Society coincided with me catching up my ¹ backlog of scientific papers on honey bees. I’d been reading a paper on the benefits of wrapping hives in the winter and Andrew commented that he did exactly that to fend off the worst of the wet weather. Andrew lives on the island of Colonsay about 75 km south of me and we both ‘benefit’ from the damp Atlantic climate.

The paper extolled the virtues of ‘covered’ hives and the data the researchers present looks, at first glance, compelling.

For example, <5% of covered hives perished overwinter in contrast to >27% of the uncovered control hives.

Wow!

Why doesn’t everyone wrap their hives?

However, a closer look at the paper raises a number of questions about what is actually benefitting (or killing) the colonies.

Nevertheless, the results are interesting. I think the paper poses rather more questions than it answers, but I do think the results show the benefits of hive insulation and these are worth discussing.

Bees don’t hibernate

Hibernation is a physiological state in which the metabolic processes of the body are significantly reduced. The animal becomes torpid, exhibiting a reduced heart rate, low body temperature and reduced breathing. Food reserves e.g. stored fat, are conserved and the animal waits out the winter until environmental conditions improve.

However, bees don’t hibernate.

Winter cluster 3/1/21 3°C (insulation block removed from the crownboard)

If you lift the lift the roof from a hive on a cold midwinter day you’ll find the bees clustered tightly together. But, look closely and you’ll see that the bees are moving. Remove the crownboard and some bees will probably fly.

The cluster conserves warmth and there is a temperature gradient from the outside – termed the mantle – to the middle (the core).

If chilled below ~5.5°C a bee becomes semi-comatose ² and unable to warm herself up again. The mantle temperature of the cluster never drops below ~8°C, but the core is maintained at 18-20°C when broodless or ~35°C if they are rearing brood. I’ve discussed the winter cluster in lots more detail a couple of years ago.

The metabolic activity of the clustered winter bees is ‘powered’ by their consumption of the stores they laid down in the autumn. It seems logical to assume that it will take more energy (i.e. stores) to maintain a particular cluster temperature if the ambient temperature is lower.

Therefore, logic would also suggest that the greater the insulation properties of the hive – for a particular difference in ambient to cluster temperature – the less stores would be consumed.

Since winter starvation is bad for bees (!) it makes sense to be thinking about this now, before the temperatures plummet in the winter.

Cedar and poly hives

I’m not aware of many comparative studies of the insulation properties of hives made from the two most frequently used materials – wood and polystyrene. However, Alburaki and Corona (2021) have investigated this and shown a small (but statistically significant) difference in the inner temperature of poly Langstroth hives when compared to wooden ones.

Poly hives were ~0.5°C warmer and, perhaps more importantly, exhibited much less variation in temperature over a 24 hour period.

Temperature and humidity in poly and wood hives

In addition to the slight temperature difference, the humidity within the wooden hives was significantly higher than that of poly.

The hives used in this study were occupied by bees and the temperature and humidity were recorded from sensors placed in a modified frame in the ‘centre of the brood box’. The external ambient temperature averaged 0°C, but fluctuated over a wide range (-10°C to 20°C) during the four month study ³.

Temperature anomalies

Whilst I’m not surprised that the poly hives were marginally warmer, I was surprised how low the internal hive temperatures were. The authors don’t comment on whether the ‘central’ frame was covered with bees, or whether the bees were rearing brood.

The longitudinal temperature traces (not reproduced here – check the paper) don’t help much either as they drop in mid-February when I would expect brood rearing to be really gearing up … Illogical, Captain.

The authors avoid any discussion on why the average internal temperature was at least 5-8°C cooler than the expected temperature of the core of a clustered broodless colony, and ~25°C cooler than a clustered colony that was rearing brood.

My guess is that the frame with the sensors was outside the cluster. For example, perhaps it was in the lower brood box ⁴ with the bees clustered in the upper box?

We’ll never know, but let’s just accept that poly hives – big surprise 😉 – are better insulated. Therefore the bees should need to use less stores to maintain a particular internal temperature.

And, although Alburaki and Corona (2021) didn’t measure this, it did form part of a recent study by Ashley St. Clair and colleagues from the University of Illinois (St. Clair et al., 2022).

Hive covers reduce food consumption and colony mortality

This section heading repeats the two key points in the title of this second paper.

I’ll first outline what was done and describe these headline claims in more detail. After that I’ll discuss the experiments in a bit more detail and some caveats I have of the methodology and the claims.

I’ll also make clear what the authors mean by a ‘hive cover’.

The study was conducted in central Illinois and involved 43 hives in 8 apiaries. Hives were randomly assigned to ‘covered’ or ‘uncovered’ i.e. control – groups (both were present in every apiary) and the study lasted from mid-November to the end of the following March.

Ambient (blue), covered (black) and control (dashed) hive temperatures

There were no significant differences in internal hive temperature between the two groups and – notably – the temperatures were much higher (15°-34°C) than those recorded by Alburaki and Corona (2021).

All colonies, whether covered or uncovered, got lighter through the winter, but the uncovered colonies lost significantly more weight once brood rearing started February. The authors supplemented all colonies with sugar cakes in February and the control colonies used ~15% more of these additional stores before the study concluded.

I don’t think any of these results are particularly surprising – colonies with additional insulation get lighter more slowly and need less supplemental feeding.

The surprising result was colony survival.

Less than 5% (1/22) of the covered hives perished during the winter but over 27% (6/21) of the control hives didn’t make it through to the following spring.

(Un)acceptable losses

To put these last figures into context the authors quote a BeeI Informed Partnership survey where respondents gave a figure of 23.3% as being ’acceptable’ for winter colony losses.

That seems a depressingly high figure to me.

However, look – and weep – at the percentage losses across the USA in the ’20/’21 winter from that same survey ⁵.

Bee Informed Partnership 2021 winter colony losses (preliminary data)

This was a sizeable survey involving over 3,300 beekeepers managing 192,000 colonies (~7% of the total hives in the USA).

If hive covers reduce losses to just 5% why does Illinois report winter losses of 47%? ⁶

Are the losses in this manuscript suspiciously low?

Or, does nobody use hive covers?

I don’t know the answers to these questions, but I also wasn’t sure when I started reading the paper what the authors meant by a hive ‘cover’ … which is what I’ll discuss next.

Hive covers

The hives used in this study were wooden Langstroths and the hive covers were 4 mm black corrugated polypropylene sleeves.

This is what I call Correx … one of my favourite materials for beekeeping DIY.

These hive covers are available commercially in the USA (and may be here, I’ve not looked). At $33 each (Yikes) they’re not cheap, but how much is a colony worth?

Significantly more than $33.

I’ve not bothered to make the conversion of Langstroth Deep dimensions (always quoted in inches 🙁 ) to metric and then compared the area of Correx to the current sheet price of ~£13 … but I suspect there are savings to be made by the interested DIYer ⁷.

However, knowing (and loving) Correx, what strikes me is that it seems unlikely to provide much insulation. At only 4 mm thick and enclosing an even thinner air gap, it’s not the first thing I’d think of to reduce heat loss ⁸.

4 mm Correx sheet

Thermal resistance is the (or a) measure of the insulating properties of materials. It’s measured in the instantly forgettable units of square metre kelvin per watt m².K/W.

I couldn’t find a figure for 4 mm Correx, but I did manage to find some numbers for air.

A 5 mm air gap – greater than separates the inner and outer walls of a 4 mm Correx hive cover – has a thermal resistance of 0.11 m².K/W.

Kingspan

It’s not possible to directly compare this with anything meaningful, but there is data available for larger ‘thicknesses’ of air, and other forms of insulation.

An air gap of 100 mm has a thermal resistance of about 0.17 m².K/W. For comparison, the same thickness of Kingspan (blown phenolic foam wall insulation, available from almost any building site skip) has a thermal resistance of 5, almost 30 times greater.

And, it turns out, St. Clair and colleagues also added a foam insulation board on top of the hive crownboard (or ‘inner cover’ as they call it in the USA). This board was 3.8 cm thick and has somewhat lower thermal resistance than the Kingspan I discussed above.

It might provide less insulation than Kingspan, but it’s a whole lot better than Correx.

This additional insulation is only briefly mentioned in the Materials and Methods and barely gets another mention in the paper.

A pity, as I suspect it’s very important.

Perspex crownboard with integrated 50 mm Kingspan insulation

I’m very familiar with Kingspan insulation for hives. All my colonies have a 5 cm thick block present all year – either placed over the crownboard, built into the crownboard or integrated into the hive roof.

Two variables … and woodpeckers

Unfortunately, St. Clair and colleagues didn’t compare the weight loss and survival of hives ‘covered’ by either wrapping them in Correx or having an insulated roof.

It’s therefore not possible to determine which of these two forms of protection is most beneficial for the hive.

For reasons described above I think the Correx sleeve is unlikely to provide much direct thermal insulation.

However, that doesn’t mean it’s not beneficial.

At the start of this post I explained that Andrew Abrahams wraps his hives for the winter. He appears to use something like black DPM (damp proof membrane).

Hive wrapped in black DPM (to prevent woodpecker damage)

Andrew uses it to keep the rain off the hives … I’ve used exactly the same stuff to prevent woodpecker damage to hives during the winter.

It’s only green woodpeckers (Picus viridis) that damage hives. It’s a learned activity; not all green woodpeckers appear to know that beehives are full of protein-rich goodies in the depths of winter. If they can’t grip on the side of the hive they can’t chisel their way in.

When I lived in the Midlands the hives always needed winter woodpecker protection, but the Fife Yaffles ⁹ don’t appear to attack hives.

Here on the west coast, and on Colonsay, there are no green woodpeckers … and I know nothing about the hive-eating woodpeckers of Illinois.

So, let’s forget the woodpeckers and return to other benefits that might arise from wrapping the hive in some form of black sheeting during the winter.

Solar gain and tar paper

Solar gain is the increase in thermal energy (or temperature as people other than physicists with freakishly large foreheads call it) of something – like a bee hive – as it absorbs solar radiation.

On sunny days a black DPM-wrapped hive (or one sleeved in a $33 Correx/Coroplast hive ‘cover’) will benefit from solar gain. The black surface will warm up and some of that heat should transfer to the hive.

And – in the USA at least – there’s a long history of wrapping hives for the winter. If you do an internet search for ‘winterizing hives’ or something similar ¹⁰ you’ll find loads of descriptions (and videos) on what this involves.

Rather than use DPM, many of these descriptions use ‘tar paper’ … which, here in the UK, we’d call roofing felt ¹¹.

Roofing felt – at least the stuff I have left over from re-roofing sheds – is pretty beastly stuff to work with. However, perhaps importantly, it has a rough matt finish, so is likely to provide significantly more solar gain than a covering of shiny black DPM.

I haven’t wrapped hives in winter since I moved back to Scotland in 2015. However, the comments by Andrew – who shares the similarly warm and damp Atlantic coastal environment – this recent paper and some reading on solar gain are making me wonder whether I should.

Fortunately, I never throw anything away, so should still have the DPM 😉

Winter losses

Illinois has a temperate climate and the ambient temperature during the study was at or below 0°C for about 11 weeks. However, these sorts of temperatures are readily tolerated by overwintering colonies. It seems unlikely that colonies that perished were killed by the cold.

So what did kill them?

Unfortunately there’s no information on this in the paper by St. Clair and colleagues.

Perhaps the authors are saving this for later … ’slicing and dicing’ the results into MPU’s (minimal publishable units) to eke out the maximum number of papers from their funding ¹², but I doubt it.

I suspect they either didn’t check, checked but couldn’t determine the cause, or – most likely – determined the cause(s) but that there was no consistent pattern so making it an inconclusive story.

But … it was probably Varroa and mite-transmitted Deformed wing virus (DWV).

It usually is.

Varroa

There were some oddities in their preparation of the colonies and late-season Varroa treatment.

Prior to ‘winterizing’ colonies they treated them with Apivar (early August) and then equalised the strength of the colonies. This involves shuffling brood frames to ensure all the colonies in the study were of broadly the same strength (remember, strong colonies overwinter better).

A follow-up Varroa check in mid-October showed that mite levels were still at 3.5% (i.e. 10.5 phoretic mites/300 bees) and so all colonies were treated with vaporised oxalic acid (OA).

Sublimox vaporiser … phoretic mites don’t stand a chance

In early November, mite levels were down to a more acceptable 0.7%. Colonies received a second OA treatment in early January.

For whatever reason, the Apivar treatment appears to have been ineffective.

When colonies are treated for 6-10 weeks with Apivar (e.g. early August to mid-October) mite levels should be reduced by >90%.

Mite infestation levels of 3.5% suggest to me that the Apivar treatment did not work very well. That being the case, the winter bees being reared through August, September and early October would have been exposed to high mite levels, and so acquired high levels of DWV.

OA treatment in mid-October would kill these remaining mites … but the damage had already been done to the ’diutinus’ winter bees.

That’s my guess anyway.

An informed guess, but a guess nevertheless, based upon the data in the paper and my understanding of winter bee production, DWV and rational Varroa management.

In support of this conclusion it’s notable that colonies died from about week 8, suggesting they were running out of winter bees due to their reduced longevity.

If I’m right …

It raises the interesting question of why the losses were predominantly (6 vs 1) of the control colonies?

Unfortunately the authors only provide average mite numbers per apiary, and each apiary contained a mix of covered and control hives. However, based upon the error bars on the graph (Supporting Information Fig S1 [PDF] if you’re following this) I’m assuming there wasn’t a marked difference between covered and control hives.

I’ve run out of informed guesses … I don’t know the answer to the question. There’s insufficient data in the paper.

Let’s briefly revisit hive temperatures

Unusually, I’m going to present the same hive temperature graph shown above to save you scrolling back up the page ¹³.

Ambient (blue), covered (black) and control (dashed) hive temperatures

There was no overall significant difference in hive temperature between the control and covered colonies. However, after the coldest weeks of the winter (7 and 8 i.e. the end of February), hive temperatures started to rise and the covered colonies were consistently marginally warmer. By this time in the season the colonies should be rearing increasing amounts of brood.

I’ve not presented the hive weight changes. These diverged most significantly from week 8. The control colonies used more stores to maintain a similar (actually – as stated above – marginally lower) temperature. As the authors state:

… covered colonies appeared to be able to maintain normal thermoregulatory temperatures, while consuming significantly less stored food, suggesting that hive covers may reduce the energetic cost of nest thermoregulation.

I should add that there was no difference in colony strength (of those that survived) between covered and control colonies; it’s not as though those marginally warmer temperatures from week 9 resulted in greater brood rearing.

Are lower hive temperatures ever beneficial in winter?

Yes.

Varroa management is much easier if colonies experience a broodless period in the winter.

A single oxalic acid treatment during this broodless period should kill 95% of mites – as all are phoretic – leaving the colony in a very good state for the coming season.

If you treat your colonies early enough to protect the winter bees there will inevitably be some residual mite replication in the late season brood, thereby necessitating the midwinter treatment as well.

I’m therefore a big fan of cold winters. The colony is more likely to be broodless at some point.

I was therefore reassured by the similarity in the temperatures of covered and control colonies from weeks 48 until the cold snap at the end of February. Covered hives should still experience a broodless period.

I’m off for a rummage in the back of the shed to find some rolls of DPM for the winter.

I don’t expect it will increase my winter survival rates (which are pretty good) and I’m not going to conduct a controlled experiment to see if it does.

If I can find the DPM I’ll wrap a few hives to protect them from the winter weather. With luck I should be able to rescue an additional frame or two of unused stores in the spring (I often can anyway). I stack this away safely and then use it when I’m making up nucs for queen mating.

I suspect that the insulation over the crownboard provides more benefit than the hive ‘wrap’. As stated before, all my colonies are insulated like this year round as I’m convinced it benefits the colony, reducing condensation over the cluster and keeping valuable warmth from escaping. However, wrapping the hive for solar gain and/or weather protection is also worth considering.

References

Alburaki, M. and Corona, M. (2022) ‘Polyurethane honey bee hives provide better winter insulation than wooden hives’, Journal of Apicultural Research, 61(2), pp. 190–196. Available at: https://doi.org/10.1080/00218839.2021.1999578.

St. Clair, A.L., Beach, N.J. and Dolezal, A.G. (2022) ‘Honey bee hive covers reduce food consumption and colony mortality during overwintering’, PLOS ONE, 17(4), p. e0266219. Available at: https://doi.org/10.1371/journal.pone.0266219.

Feral facts and fallacies

20 Replies

Synopsis : Are feral colonies recently lost swarms or a self-sustaining ‘wild’ honey bee population? The latter must reproduce faster than they perish. Measuring rates of colony loss and nest occupancy provides a good indicator of the likely origin and independence of feral populations.

Introduction

Most colonies try to swarm every year. Most – not all – but if your colonies are strong and healthy they are likely to swarm. That’s why swarm prevention and subsequent swarm control are such important skills for the tyro beekeeper to master. Without swarm control the majority of the workforce is ‘lost’, the residual colony will be left temporarily queenless and the potential honey crop is probably much reduced.

A small swarm …

It is not difficult to become competent at swarm prevention and control. However, any beekeeper who claims to never lose swarms is probably being ‘economical with the actualité’ as the late Alan Clark once said.

What happens to those ‘lost’ swarms?

Some forward-thinking beekeepers set out bait hives. Any swarms that end up being attracted to these ‘swarm traps’ will eventually find their way back to a managed apiary. Some swarms end up in the church tower where ‘there have always been bees’, according to local parishioners.

Others move into the roof space above the entrance to the local nursery school, causing fascination, irritation and consternation in equal measure. Their fate depends upon whether the head teacher contacts a beekeeper or a pest controller … but their arrival reinforces the importance of swarm control and the use of bait hives.

A bait hive deployed in mid-April in good time for the swarming season ahead

And other swarms disappear over the apiary fence, across the field and into the local woods, eventually establishing a new colony in a suitable hollow tree.

No risk, no reward

Swarming is a risky business. The swarm leaves with the majority of the flying bees and the mated queen. However, it takes more than that to establish a functional colony. They need to draw comb, rear brood and collect sufficient stores to get through the winter.

That’s a tall order and most swarms fail.

Data from Thomas Seeley in The Lives of Bees suggests that only about 23% of swarms survive the winter.

In contrast, the swarmed colony has about an 80% chance of survival. They’ve got drawn comb, stores, eggs and larvae … ‘all’ they need to do is rear a new queen.

And then they’re likely to swarm again the following year ¹. In fact, without swarm control, the average number of swarms produced by a colony is two per year – presumably a prime swarm (headed by the old queen) and a cast (headed by a virgin queen).

So, swarming is risky, but those swarms that succeed in establishing a new colony and overwintering can themselves attempt to reproduce again the following year.

That’s the reward.

Where are all these bees?

Even taking account of the exemplary swarm control by the UK’s ~25,000 beekeepers ² I’m reasonably certain that a lot of swarms are lost every year.

Where do all these bees go?

I’ve been told of lots of churches or schools or trees with resident bees.

A swarm magnet … or just an old church?

However, it’s certainly not every church, or school or hollow tree that’s occupied. Even when there’s a surplus of suitable nest sites, those that are occupied by a colony are the exception, not the rule.

The main reason of course is Varroa.

In the absence of intervention to reduce the mite population, the developing winter bees get parasitised by Varroa, and the resulting high levels of deformed wing virus (DWV) reduces the longevity of these necessarily ³ long-lived bees.

Consequently the winter cluster shrinks in size, from that of a football (early October) to a honeydew melon (late December) to a large orange (early February).

And then it freezes to death during a cold snap 🙁 .

The apiary in winter …

Numerous studies have shown that untreated colonies, in the absence of any natural resistance or tolerance to Varroa or DWV (though the latter is rarely discussed, and even less frequently tested for), almost always perish within a year or two of Varroa infestation.

Look back at the recent post on Biological control with Varroa for a reminder of the devastation wreaked on an island population of honey bees after the introduction of mites.

Wild? They’re livid feral …

Technically, swarms lost by beekeepers (that become established in the environment) are probably best termed feral colonies.

Originally feral meant simply ‘wild or untamed’, but the more common usage these days means ’animals or plants that have lapsed into a wild form from a domesticated condition’.

Bees aren’t domesticated, but I think feral conveniently encompasses their origin.

However, I’m more than happy to accept that a colony, initially feral, that becomes well-established in the church tower and throws off a swarm or two every year, that requeens every two or three seasons, surviving without intervention or management, must be considered ‘wild’ at some point.

It’s not worth discussing when a colony transitions from feral to wild.

It’s semantics, though I think the distinction between ‘recently arrived from a swarmed managed colony’ and ‘self-sustaining’ is an important one.

Notwithstanding the ravages of Varroa, whether feral or wild, there are colonies in the environment – churches, schools, trees – and probably rather more than many beekeepers are aware of.

The missing bees

Periodically there’s a little flurry of interest in the press about ‘long lost’ or ‘missing’ wild bees discovered in the woods.

Late last summer there were articles in all the newspapers about bees found on Blenheim Estate. The Observer reported this discovery with the headline ”No one knew they existed”: wild heirs of lost British honeybee found at Blenheim.

‘Blenheim bees’ article in the Observer, 7-11-21

As an aside – as this isn’t the real topic for discussion today – there are at least three challenging claims made in that headline; how can you be sure that no-one knew they existed? Is the British honey bee (it is not honeybee) actually lost? How do you know that these bees are their heirs?

Pedantic is my middle name.

But the 2500 hectare Blenheim Estate ⁴ isn’t the only location with apparently self-sustaining populations of honey bees. There are trees, churches and (I dare say) even nursery schools up and down the country that appear to have a ‘resident’ colony or two of bees.

Periodically they’re observed swarming. Sometimes things seem a bit quiet in the spring, but perhaps it’s too cold for the bees to be flying strongly anyway.

By May there’s a lot of activity so all must be well.

Right?

Perhaps 😉

Citizen science

These wild/feral colonies are infrequent but widely distributed. They are therefore difficult for one person to regularly observe. As a consequence there are several ‘citizen science’ projects monitoring some of these sites. Magnus Peterson regularly reports in The Scottish Beekeeper on the one he coordinates for the University of Strathclyde.

The criticism of these types of studies – certainly not Magnus’s specifically – but any study the largely relies upon infrequent observation by volunteers, is that stuff gets missed. A visit doesn’t happen because it’s raining hard. Or it does happen in heavy rain and no activity is observed and the colony is recorded as dead.

Or worse, recorded as alive, but not flying because of the heavy rain.

With more systematic observation, though not necessarily more frequent, you can have increased certainty that the site that was occupied last autumn is still occupied this spring.

The timing of these observations is important. Three per season is probably the minimum, early, mid and late, but they have to be at particular times of the season – see below.

Crowdfunding

So, let’s assume a colony is found in the autumn and the same hollow tree is occupied in late-April the following year … yippee, the colony is still alive.

Feral – or are they now wild? – bees living successfully with Varroa (at least presumably living with Varroa if they’re almost anywhere in mainland UK).

Perhaps they’ve evolved to have some interesting and useful trait(s) that renders the colony resistant to or tolerant of the dreaded parasitic mites?

These are valuable bees.

They are an important genetic resource.

They must be protected at all costs.

Perhaps it’s time to set up a ~~web~~ colony cam to record their activity? That’s going to cost a pretty penny, so some crowdfunding is needed.

A website is created … a dozen mini-nucs are purchased for the ambitiously planned queen rearing programme and – inevitably – there’s a misquoted article or two in the Guardian.

But hold on …

Are they really the same colony in April that were there the previous autumn?

How can you be sure?

How can you be certain that it’s not an unseasonably early swarm that was missed by the – usually eagle-eyed – local beekeepers? ⁵

It’s not unusual to find the odd charged queen cell during the first colony inspection of the season. At least, I’ve sometimes found queen cells during that first inspection. I’m sufficiently experienced to not go rummaging about in the boxes too early in the season, and so I am sometimes surprised at how well developed the colony is when I open the box.

Charged queen cell

But what if it had been raining, so I’d postponed the inspection?

On the next warm spring day – well before I was able to return to the apiary – the colony could swarm.

I’ve regularly seen April swarms in Scotland and there are many reports of even earlier swarms on social media every year.

Perhaps the active ‘overwintered’ colony is nothing of the sort.

Maybe it’s just been occupied by a very early swarm?

To be sure it’s the same colony you need to do some genetic testing. If the colony is the same the genetic testing will show identity. If the testing shows significant variation then it’s a different colony.

And, if you combine some genetic testing of overwintered colonies with three carefully-timed visits – late season, very early season and mid-season – to a large number of wild/feral colonies, or likely sites that they would occupy, you can determine their longevity and whether they are a self-sustaining population.

Bee trees

And I wouldn’t have given that long and rambling introduction if there wasn’t a recent scientific paper where they’ve done exactly that (Kohl et al., 2022). I’ll describe it briefly as it’s a nicely written and compelling story. The paper is open access, so you can read it if you want to check my interpretation of the data.

Importantly, I think it provides a very good guide to both the quality and quantity of data that are needed to be sure a population of bees are truly wild and self-sustaining ⁶ … or just regularly boosted by careless local beekeeping!

Feral colonies are few and far between. It’s hard work walking around the woods looking for hollow trees that may (but probably won’t) contain a colony. You find lots of trees with holes, but they need to lead to a suitably-sized cavity to be of any use to a colony of bees. Binoculars help (the holes are often 15 metres off the ground) … but perhaps there are better ways of doing this?

A bee tree?

Bee-lining – as described by Seeley in Following the Wild Bees – is an effective way of tracking down wild colonies, but needs good weather, good forage and ample time. It works well when locating a few colonies, but probably takes too long if you want 100+ to produce a statistically compelling set of results.

But what if you also wanted to record how many new nest sites are occupied? You would need to know where the empty cavities were before they were occupied. That’s not something you can determine by bee-lining, so you’re back to traipsing around the woods with a pair of binoculars.

Woodpeckers

But in Germany they have some very large woodpeckers.

The black woodpecker (Dryocopus martius) is a crow-sized bird that excavates correspondingly large holes for nest sites in old-growth forests. The average volume of a black woodpecker nest is about 10 litres, smaller than optimal for a swarm, but appreciably larger than most ‘natural’ tree cavities.

Black woodpecker

Conveniently, there are high-resolution maps of (historical) woodpecker nesting trees in old-growth forests in Swabian Alb, Weilheim-Schongau and the counties of Coburg and Lichtenfels. 98% of these woodpecker nest sites are in large beech trees, most are 10-12 metres above ground and with an entrance of ~10cm diameter (again, not optimal, but better than no nest site for a swarm).

Kohl and colleagues surveyed about 460 of these ‘cavity’ trees three times per season; in July (after the main May/June swarming season) to determine peak occupancy rates, in mid/late September to determine late summer survival and in early/mid April to determine winter survival.

‘Occupancy’ was determined by visual inspection and regular forager activity and/or pollen loads (i.e. they ignored scout bees checking empty cavities). In addition, for some colonies, a dozen or so workers were collected for genetic analysis.

With these data, the mathematical calculation of annual survival rates could be determined, as could the prediction of the annual numbers of swarms needed per colony for the population to be self-sustaining ⁷. In addition, it was possible to determine the average lifespan of a colony.

There were a bunch of perfectly reasonable assumptions made, based upon the known biology of honey bees – all are listed in the paper.

Yo-yoing colony numbers

The scientists counted colony numbers, but could also determine colony densities per km². By making observations over a 3-4 year period it was strikingly obvious that the largest number of ‘cavity’ trees were occupied after swarming in summer, but that numbers dropped dramatically overwinter. This ’recurring temporal pattern of population fluctuations’ is very obvious in the major data figure in the paper.

Temporal population fluctuations of feral honey bee colonies in Germany; A) occupancy rates, B) population density

The average maximum occupancy rate and population density was 11% and 0.23 colonies per km². This ‘dropped massively’ over the winter to just 1.4% and 0.02 colonies per km².

The majority of nest sites (n = 112) occupied in late summer were unoccupied the following spring, before swarming started. 90% of colonies survived the summer (from July until late September), but only 16% of colonies survived the following winter.

The spring survival rate was calculated as 74% based upon genetic testing of colonies in early spring and mid-summer

Knowing the summer, winter and spring survival rates enables the annual survival rate to be calculated.

This was a sobering 10.6%.

Therefore, to maintain a stable population, each surviving colony would need to produce an average of 8.4 swarms per season.

That’s an unachievable amount of swarming.

The average lifespan of a feral colony in these three German forest regions was just 0.619 years … a little over 32 weeks.

Clearly, these honey bee populations are not self-sustaining.

Are these German forests typical?

There are two other regions where similar quality data exists for wild/feral honey bee populations. These are the Arnot forest in the USA, studied for decades by Thomas Seeley, and Wyperfield National Park in Australia.

There are striking differences between these two regions and the German forests, both in terms of colony lifespan and swarm numbers needed to be self-sustaining.

For the Arnot forest and Wyperfield National Park, lifespan was calculated as 1.34 and 1.53 years respectively (cf. 0.62 years for Germany), with annual survival of ~50% (cf. 11% in Germany). Annual swarm numbers per colony for the population to be self-sustaining was 0.94 and 0.85 for the the Arnot forest and Wyperfield National Park respectively (cf. 8.43 for the German forests).

Other than these obvious differences in the related figures for survival/longevity and ‘swarms needed’ the other significant difference between self-sustaining populations (like the Arnot forest and Wyperfield National Park) is the colony density.

In areas where feral/wild honey bees are self-sustaining the colony density is at least 1 per km². In contrast, in Germany and a large number of other studied feral populations in other parts of Europe (including Ireland, Spain, Serbia, Poland and other regions of Germany), the colony density is usually much lower, at 0.1-0.2 per km².

So, these German forests are seemingly typical of honey bee populations that are not self-sustaining. These are regions where the feral population is boosted annually (and is essentially dependent upon) an influx of swarms that become temporarily established in natural nest sites.

Environmental colony density

Where do all these swarms come from?

The average managed honey bee colony density in the areas of Germany studied is 4 per km², appreciably higher than either the Arnot forest or Wyperfield National Park. Precise figures for these two were not quoted, but in both locations the feral colonies (remember, these were at ~1 per km²) outnumber managed colonies.

It therefore seems very likely that managed colonies from farmland areas surrounding the German forests acts as the source for swarms, and the latter – because of the paucity of suitable nest sites in the arable land (relatively few buildings, few mature trees etc.) – gravitate towards the forests looking for suitable nest sites.

Feral and managed colonies may therefore be spatially separated, though not very widely. In contrast, in urban environments – where nest sites are probably common – it might be expected that feral and managed colonies are intermixed in the environment.

A by-product of the study by Kohl and colleagues is that they could also calculate the difference in the relative attractiveness of woodpecker nests that had previously, or had never, been occupied by bees. When new colonies occupied woodpecker nest sites there was a strong preference of 5 to 15-fold for sites that had previously been occupied by bees.

This, of course, is why it makes sense to include a single old, dark comb in your bait hives.

That seems like a good place to stop …

I think this German study is interesting. It shows the quantity and quality of data needed to make a compelling case that a location has a self-sustaining population of feral/wild honey bees.

Such locations are likely to exhibit colony densities of at least 1 per km² and to be physically separated from higher density managed colonies. This physical separation could be in the form of simple geographic isolation – just a long way from other apiaries – or something more complex like being surrounded by high hills or water etc.

Self-sustaining wild/feral populations are likely to exhibit >50% annual survival rates, to live for an average of ~1.5 years and to produce about 0.8-0.9 swarms per colony per year ⁸.

If survival rates are lower, or the life expectancy of a colony is much less, then the number of swarms needed to maintain the population rapidly becomes so high that they are unattainable.

In which case, large numbers of feral/wild colonies cannot be self-sustaining, but instead must be present because the area acts as a ‘sink’ for lost swarms from nearby managed colonies.

This post is already longer than my self-imposed-but-regularly-exceeded 3000 word limit so I’ll save further discussion of the Blenheim bees and other feral colonies for another post.

However, I hope the study shows that a healthy scepticism is perhaps sensible when considering any claims made about self-sustaining feral colonies.

That church tower in which ‘there have always been bees’ may well have had bees in it every year.

But that’s not the same as having the same bees in it.

In fact, with an ~90% attrition rate of feral colonies annually it’s very unlikely to be the same colony in successive years.

Note

In the final stages of completing this post – very, very late at night – I re-discovered an article (Moro et al., 2018) on citizen science and feral colonies that I’ll return to sometime in the future.

References

Kohl, P.L., Rutschmann, B. and Steffan-Dewenter, I. (2022) ‘Population demography of feral honeybee colonies in central European forests’, Royal Society Open Science, 9(8), p. 220565. Available at: https://doi.org/10.1098/rsos.220565.

Moro, A. et al. (2021) ‘Using Citizen Science to Scout Honey Bee Colonies That Naturally Survive Varroa destructor Infestations’, Insects, 12(6), p. 536. Available at: https://doi.org/10.3390/insects12060536.

Preparing for winter

The beekeeping season is fast receding into the distance as the first frosts of autumn appear and, finally, the wasp numbers start to diminish. By now colonies should be heavy with stores, either collected by the bees or provided by the beekeeper.

Winter is coming … be prepared

There is relatively little actual beekeeping to be done this late in the year.

Colonies do not need to be disturbed unnecessarily. They certainly don’t require the usual weekly inspection … they’re not going to swarm, you’ve already applied your miticide of choice and fed them with fondant or syrup ¹.

Late queen mating

With temperatures during the day in the low to mid-teens (°C) it is still warm enough to open a colony if you need to.

One of the few reasons I’d open a colony in very late September/early October would be to check if a new queen that had emerged at the end of August had successfully mated. If she had, then all is good. She will continue to lay late into the autumn and should produce sufficient winter bees to get the colony through to the following Spring.

When I lived in the Midlands I would regularly get queens successfully mated in early/mid September. It was pretty dependable, and in good years I’d be actively queen rearing through much of August.

Now, back in Scotland, late queen mating is not something I would want to rely on. I’m certain it happens now and again, but only in very exceptional years.

It’s a tough life being a drone in late August … but not for much longer

This year, many of my colonies turfed their drones out a month ago, and queen mating is not going to happen unless there are plenty of drones about.

A quick peek

It takes just minutes to check whether the queen is mated and laying. Although you don’t need to see the queen, it’s worth using just a whiff of smoke so you have the option of searching for her if needed. If you smoke the colony heavily she’ll end up rushing about or buried under a mass of disturbed bees.

Just a whiff …

You will need to remove the feeder (if using syrup) or the queen excluder and fondant block. Place these aside gently and remember that there are likely to be large numbers of bees adhering to the underside, so balance them on the rim of an upturned roof. This is the time you realise the benefit of using framed rigid wire QE when feeding fondant … removing the block on a flexible plastic QE is a right palaver.

The hive should be busy with bees. Gently remove the dummy board and outer frame. This should be full, or in the process of being filled, with stores. There’s no need to shake the bees off. Just stand it aside out of the way.

‘Guesstimate’ the approximate centre of the brood nest, based upon the density of bees in the seams. Gently lever the frames apart a centimetre or so, then release one of the frames adjacent to the gap you’ve created from its neighbours.

Lift the frame and look for sealed brood, open brood and eggs. By knowing the development cycle of workers bees (3E,5L,13P ²) you can determine approximately when the queen started laying ³.

If she started laying …

Snatching victory from the jaws of defeat

… if there are no eggs or larvae by very late September I would assume that the queen had failed to mate.

You need to use your judgement here. If the weather was poor in the first half of September, but excellent since then, it remains a distant possibility that she has only just mated and has yet to start laying.

Look carefully for polished cells where the centre of the broodnest should be.

And cross your fingers.

Polished cells are a sign that the nurse bees are preparing the comb for egg laying. However, in my experience, they do this even if the queen remains unmated, so it is not a reliable sign that all is well.

You therefore need to use your judgement and be realistic.

Miracles do happen, but you can’t depend upon them ⁴.

If the weather has been consistently poor – windy, low temperatures (for queen mating, which really needs ~18-20°C) or wet – then assume the worst and ‘save’ the colony by uniting it with a nearby strong colony.

A colony without a laying queen in late autumn will not survive the winter in any state that will make it a viable colony the following year ⁵.

In Scotland, I routinely unite colonies that do not have a laying queen at the end of August. As described in the last couple of weeks, I do my final colony checks with feeding and miticide treatment.

I know the chances of a queen getting successfully mated after that are effectively zero.

Quick uniting – air freshener

If you need to unite two colonies quickly, without the usual week long wait while they gently mingle after stacking them separated by a sheet of newspaper, you can use a few squirts of household air freshener.

Open the queenright recipient colony, removing the feeder and carefully placing it aside to avoid crushing bees (see above)
Find the unmated/unlaying/uncooperative queen in the broodless box and remove her (permanently I’m afraid)
Spray the top of the recipient colony with a a few squirts of air freshener
Do the same with the underside of the now queenless broodless colony
Stack the latter on top of the recipient colony
Add the feeder back, again giving a squirt or two of air freshener at the interface to stop the bees from fighting

The air freshener masks the distinctive pheromone ‘smell’ of the two colonies, allowing the bees to mingle without fighting.

That’s it.

Job done.

Caveat emptor

Like everything else on this site, I only write here from direct experience. I have successfully united quite a few colonies like this, though nothing like the number I’ve united using newspaper ⁶.

Given time and the choice I’d always use newspaper ⁷.

But this late in the season you might not have time.

A day after uniting with air freshener you can, if needed, revisit the hive and go through the double brood box to reduce it to a single box for the winter.

Does it matter which air freshener you use?

I have no idea.

I use Glade Citrus Sunny Beat as it was the cheapest I could find at the time I needed it ⁸.

Securing the queenright overwintering colony

If you consult the COLOSS records for overwintering colony losses they include a small percentage that are lost to ‘natural disasters’. COLOSS record queen failures and things like that separately, and – in an earlier paper – they define natural disasters as:

… rather loosely defined, as the causes can be very different in participating countries, including fire, storm, flooding, vandalism, bears, martens, woodpeckers, falling trees, suffocation from snow and many more.

The small percentage (0.1 – ~5%) lost to natural disasters vary from country to country, and from year to year.

What is notable about several of these natural disasters is that they should be avoidable.

If your colonies are strong and queenright, and if you’ve fed and treated them to give them the best chance of surviving the winter, it makes sense to do what you can to avoid these natural disasters.

The hive

I use a combination of polystyrene and cedar hives. Sometimes I even combine the two together in a single hive. The majority of my poly hives are from Abelo or Swienty which, for reasons explained elsewhere, are compatible with all the woodenware I own.

The apiary in winter …

I see no difference in the overwintering colony success between poly and cedar hives.

This doesn’t mean there isn’t one.

I’ve only run about 20 colonies for the last decade. That’s ~200 overwintered colonies. If there were wildly different survival rates I would have noticed. Since I haven’t noticed it either means there is no difference or there is a subtle difference but my sample size is too low ⁹.

All my colonies overwinter on open mesh floors, usually with the Varroa tray removed. The hives in the photo above are being monitored for mite drop in early December following oxalic acid treatment.

DIY insulation over a perspex crownboard

In addition, all of my hives have a 50 mm thick block of Kingspan under the roof, integrated into the roof, or integrated into the crownboard. In the bee shed my hives have no roof, and are just capped with a block of Kingspan over the crownboard.

Look, no roof … but insulation present all year round

Make sure the stack of boxes in the hive are stable and secure. If the apiary is exposed, strap everything together securely. A colony might survive a week or two of summer showers with no roof, but will surely perish if exposed for any length of time to cold, wet winter weather.

Apiary security

It is unlikely that you will visit the apiary much in the winter. Once a fortnight is more than enough.

It might therefore be worth considering whether it is sufficiently secure from the attention of unwanted human visitors. Unfortunately, incidents of vandalism occur throughout the season, but a hive kicked over in midwinter has less chance of being detected quickly.

Or of surviving.

Although it should probably be included within the ‘Varmints’ section below, large animals – cows, deer, elk, bear, rhino, kangaroo ¹⁰ – might also inadvertently, or deliberately, overturn a hive.

Safe and secure

Fences, either a couple of strands of barbed wire, an electric fence or a full-blown razor-wire topped security barrier, are usually sufficient to keep large two and four-legged visitors at bay.

COLOSS mention both falling trees and flooding as natural disasters.

Winter storms can and do wreak havoc in some years, though I always associate the summer with storm-toppled trees because they’re in full leaf and therefore offer more resistance. It’s certainly worth looking to see if trees adjacent to your apiary might threaten the hives.

Where did Noah keep his bees? In his Ark hive.

Flooding appears to be on the increase. I have experienced minor flooding in one of my apiaries. None of the hives were threatened, but it made access inconvenient for weeks at a time. Again, it’s worth imagining the worst and preparing for it.

Hives often float, but not necessarily the right way up 🙁

Varmints

Having dealt with the threat of large animals ¹¹ it’s also worth considering the damage some small animals can do to hives.

The two main culprits are woodpeckers and mice. Both can be a menace once the frosts set in, but rarely before that.

Woodpeckers, and specifically green woodpeckers (yaffles ¹²), can learn that beehives contain a wonderful bounty of pupae and larvae. It is learned behaviour. Some green woodpeckers never go near hives, others routinely target them.

In Warwickshire, hives needed to be protected from yaffles. Here in Fife the bird is very much less common and I’ve never had any hives targeted.

Wrapped for winter …

Protection is straightforward. If needed, I simply wrap the hives in a single sheet of DPM (damp proof membrane), pinned in place with drawing pins. The bird need to cling onto the vertical side of the hive to easily burrow through to the brood. The DPM stops them doing this. Leaving bits of the roof or sides of the floor exposed is therefore not a problem ¹³.

Pixie or Dixie?

Mice access hives through overly large entrances. I only have problems with the stupidly cavernous maw of my preferred Everynuc. Mice eat pollen and stores, destroy the brood and wee everywhere 🙁 Thoroughly unpleasant.

Open wide …

A standard mouseguard pinned in place throughout the coldest months of the winter prevents them accessing the hive. Alternatively, on a full-sized colony, the kewl-style underfloor entrances are very effective at excluding rodents.

Kewl open mesh floor showing L-shaped entrance slot

Kewl floor entrance …

That’s not the end of winter-related tasks, but it’s just about all you need to do for your colonies before winter proper starts.

There are some midwinter checks that are needed, but we’ll deal with them nearer the time.

Note

We also have pine martens at one of my apiaries. They are reported to vandalise hives and steal honey (and presumably brood) in late winter. Pine martens are incredibly agile and no fence exists that could keep them out. Time will tell whether they are a problem.

In the meantime, here’s one living up to its name, stealing a pine offcut used to slow down the rate at which they empty the squirrel feeder of peanuts 🙂

Belt and braces

Not even the most determined woodpecker was going to get to this colony!

Keep off Woody

I took this photo on the 4th of January and I’m pleased to say that the hive remained untroubled by woodpeckers throughout the winter.

Woodpeckers

Chicken wire cage …

In very cold weather green woodpeckers can cause damage to beehives when they try and get at the bees and brood. Until the temperate is below 0ºC for an extended period this usually isn’t an issue, but in a prolonged cold snap beehives can represent easy pickings for a hungry bird. I believe that this is learned behaviour and appears to vary in different areas. Around here they know that cedar, or even better poly, hives can be ransacked so these need to be protected during the winter. There are two easy ways to achieve this. One approach is to build a cage of chicken wire, supported on canes. This can be lifted away reasonably easily and ensures the hive doesn’t get too damp. However, it’s more difficult to construct and takes up more storage space during the summer. The alternative approach is to wrap the hive body in old plastic fertiliser or compost sacks. Damp proof membrane (DPM) is often available in skips and is even better as it it more robust and doesn’t get shredded in high winds.

Wrapped for winter …

Using drawing pins I attach the DPM around the top of the brood box. Don’t fix it to the crownboard as you’ll probably have to lift this mid-winter to apply oxalic acid. If the plastic is fixed to the brood box it can usually be left in place during oxalic acid treatment. A single layer thick is sufficient … it presumably works by preventing the woodpecker perching on the side bars and breaking through. Whenever I’ve seen woodpecker damage it’s always to the sides of the brood box where the frame lugs are.

Polystyrene hives need additional protection for the roof as well. I’ve even had blue or great tits damage the roof of an overwintered Paynes nuc box. I tend to move these nucs to a sheltered spot in the garden which is rarely visited by woodpeckers, rather than leave them unattended in out apiaries. Finally, if you do use plastic, use sufficient pins to hold it securely in place in bad weather and don’t leave any gaps for Woody to get through …

: Incomplete protection …

: Polyhives and polythene …