Tag Archives: BEEHAVE

Kick ’em when they’re down

Out, damn'd mite ...

Out, damn’d mite …

Why bother treating colonies in midwinter to reduce Varroa infestation? After all, you probably treated them with Apiguard or Apivar (or possibly even Apistan) in late summer or early autumn.

Is there any need to treat again in midwinter?

Yes. To cut a long story short, there are basically two reasons why a midwinter mite treatment almost always makes sense:

  1. Mites will be present. In addition, they’ll be present at a level higher than the minimum level achievable, particularly if you last treated your colonies in late summer, rather than early autumn.
  2. The majority of mites will be phoretic, rather than hiding away in sealed brood. They’re therefore easy to target.

I’ll deal with these in reverse order …

Know your enemy

DWV symptoms

DWV symptoms

The ectoparasite Varroa feeds on honey bee pupae and, while doing so, transmits viruses (in particular DWV) that can completely mess up the development of the adult bee. Varroa cannot replicate anywhere other than on developing pupae. It’s replication cycle, and the resulting mite levels in the colony, are therefore tightly linked to the numbers and availability of hosts … honey bee pupae.

If developing brood is available the mite can replicate. Under these conditions, newly emerged adult, mated, female Varroa spend a few days as phoretic mites, riding around the colony on young bees. They then select a cell with a late-stage larvae in, enter the cell and wait until pupation occurs. If developing worker brood is available each infested cell produces 1 – 2 new mites (drone cells produce 3+) and mite numbers increase very rapidly in the colony.

In contrast, if there’s no developing brood available, the mites have to hang around waiting for brood to become available. They do this as phoretic mites and can remain like this for weeks or months if necessary.

Therefore, when brood is in abundance and the queen in laying freely mites can replicate to very high levels. In contrast, when brood is limiting and the queen has reduced her egg laying to a   v  e  r  y     s  l  o  w     r  a  t  e     the mite cannot replicate and must be predominantly phoretic.

When does this happen?

Lay Lady Lay … or don’t

Ambient temperature, day length and the availability of nectar and pollen likely influence whether the queen lays eggs. When it’s cold, dark and there’s little or no pollen or nectar coming into the hive the queen slows down, or even stops, laying eggs.

About 8 days after she stops laying there will be no more unsealed brood in the colony. About 13 days after that all the sealed brood will have emerged (along with any Varroa). Therefore, after an extended cold period in midwinter, the colony will have the lowest level of sealed brood … and the highest proportion of the mite population will be phoretic.

Under normal (midsummer) circumstances about 10% of the mite population is phoretic. It’s probably unnecessary to state that, if there’s no brood available, 100% of the mites must be phoretic.

All licensed miticides work extremely well against phoretic mites.

Caveats, guesstimates, global warming and the Gulf Stream

Global warming

Global warming …

Whatever the cause, the globe is warming (irrespective of what Donald Trump tweets). Long, hard winters are getting less common (or perhaps even rarer, as they were never particularly common in the UK). In Central, Southern or Eastern Britain it’s possible that the colony will have some brood present all year. In parts of the West, warmed by the Gulf Stream, I’d be surprised if a colony was ever broodless. Only in the North is it likely that there will be a brood break in midwinter.

Most of the paragraph above is semi-informed guesswork. I don’t think anyone has systematically analysed colonies in the winter for the presence of sealed brood. Sure, many (including me) have opened colonies for a quick peek. Others will have peered intently at the Varroa board to search for shredded wax cappings that indicate emerging brood. The presence of brood will vary according to environmental conditions and the genetics of the bees, so it’s not possible to be dogmatic about these things.

However, it’s safe to say that in midwinter, sealed brood – within which the mites can escape decimation by miticides – is at a minimal level.

Reducing mite levels and minimal mite levels

Within reason, the earlier you apply late summer miticides, the better you protect the all-important overwintering bees from the ravages of viruses, particularly deformed wing virus. This is explained in excruciating detail in a previous post, so I won’t repeat the text here.

However, I will re-present the graph that illustrates the modelled (using BEEHAVE) mite levels.

Time of treatment and mite numbers

Time of treatment and mite numbers

The gold arrow (days 240-300 i.e. September and October) indicates when the winter bees are being reared. These are the bees that need to be protected from mites (and their viruses). Mite numbers (starting with just 20 in the hive on day zero) are indicated by the solid coloured lines. The blue, black, red, cyan and green lines indicate modelled mite numbers when the colony is treated with a miticide (95% effective) in mid-July, August, September, October or November respectively.

The earlier you treat, the lower the mite levels are when the winter bees are being reared. Study the blue and black lines.

This is a good thing.

In contrast, by treating very late (the cyan and green lines) the highest mite numbers of the season occur at the same time as the winter bees are being reared. A bad thing.

But … look also at mite numbers after treatment

Look carefully at the mite numbers predicted to remain at the end of the year. Early treatment leaves higher mite levels at the start of the following year.

This is simply because mites escaping the treatment at the end of summer have had an opportunity to reproduce during the late autumn.

This is why the additional midwinter treatment is beneficial … it kills residual mites and gives the colony the best start to the new calendar year§.

Kick ’em when they’re down

Early treatment protects winter bees but risks exposing bees the following season to unnecessarily high mite numbers. However, in midwinter, these residual mites are much more likely to be phoretic due to a lack of brood in the colony. As I stated earlier, phoretic mites are relatively easy to target with miticides.

So, give the mites a hammering in late summer with an appropriate and effective miticide and then give those that remain another dose of the medicine in midwinter.

But not another dose of the same medicine

Since the majority of mites in a colony with little or no brood will be phoretic, you can easily reduce their numbers using a single treatment containing oxalic acid. This can be administered by sublimation (vaporisation) or by trickling (dribbling).

There’s no need to use any treatment that needs to applied for a month. Indeed, many (Apiguard etc.) are not recommended for use in winter because they work far less well on a largely inactive colony.

Trickle 2 - £1

Trickle 2 – £1

I’ve discussed sublimation previously. However, since this requires relatively expensive (£30 – £300) specialised delivery and personal protection equipment it may be inappropriate for the two hive owner. In contrast, trickling requires almost no expensive or special equipment and – reassuringly – has been successfully practised by UK beekeepers for many years. I did it for years before I bought my Sublimox vaporiser.

Therefore, in two further articles this autumn (well before you’ll need to treat your own colonies) I’ll describe the preparation and storage of oxalic acid solutions and its use.

Be prepared

If you want to be prepared you’ll need to beg, borrow or steal the following – sufficient oxalic acid (or Api-Bioxal), a Trickle 2 bottle sold by Thorne’s, a cheap vacuum flask (Tesco £2.50), granulated sugar and a pair of thin disposable gloves.

Do this soon. Don’t leave it until midwinter. You need to be ready to treat as soon as there’s a protracted cold spell (when brood will be at a minimum). Over the last few years my records show that this has been anywhere between the third week in November and the third week in January.

More soon …

† Only MAQS is effective against mites sealed in cells. This is why most miticides are used for extended periods in the late summer or early autumn … the miticide must be present as Varroa emerge from sealed cells.

‡ I’ll repeat the caveat that this is an in silico simulation of what happens in a beehive. Undoubtedly it’s not perfect, but it serves to illustrate the point well. It’s freely available, runs on PC and Mac computers, and is reasonably well-documented. In the simulations shown here the virtual colony was ‘primed’ with 20 mites at the beginning of the year. BEEHAVE was run using all the default settings – climate, forage etc. – with the additional application of a miticide (95% effective) in the middle of the months indicated. Full details of the modelling have already been posted.

§ The National Bee Unit recommend Varroa levels are maintained below 1000 throughout the season. Without treatment, 20 mites at the start of the season can easily replicate to ~750 in the autumn. If you start the season with 200 mites then levels are predicted to reach ~5000 in the following summer. The colony will almost certainly die that season or the next. There’s a more detailed account of the consequence of winter brood rearing and the level of mite infestation written by Eric McArthur and reproduced on the Montgomeryshire BKA website that’s worth reading.

¶ The cumulative (year upon year) effect of late summer treatment with no midwinter treatment has been discussed previously. I’ll simply re-post the relevant figure here – 5 years of bee (in blue, left axis) and mite (in red, right axis) numbers with only one treatment per season applied in late September. Within two years the higher mite numbers that are present at the start of the year reproduce to dangerously high levels.

Mid September

Mid September

Get dribbling

There has been a prolonged spell of cold weather in Eastern Scotland. Temperatures have rarely risen above 5°C, with hard frosts overnight. However, a warm front moved in on Tuesday night and the last few days have been significantly warmer. The lack of activity at the hive entrances and a quick peek under the insulation through the perspex crownboards (where fitted) indicated the bees were all tightly clustered during the cold spell. Furthermore, the absence of debris on the removable Varroa monitoring trays fitted to many of the open mesh floors, suggested that little or no brood was being reared.

Ridiculous to the sublime

Ridiculous to the sublime

Varroa counts

Varroa trays ...

Varroa trays …

There was another clue that the colonies are likely broodless. I had been recording the natural Varroa drop of a few colonies over the last month. I did this by simply counting Varroa at each visit, calculated on a mites/day basis. Although generally low (and very low in a few colonies), it had been steadily increasing. This is a good indication there were more phoretic mites in the colony … again, presumably due to the absence of suitable brood for them to parasitise.

It’s worth noting that the natural mite drop is a notoriously unreliable method of accurately determining mite levels in a colony. For example, it’s dependent upon the amount of sealed brood in the colony. With no sealed brood all mites must be phoretic. In contrast, with limitless sealed brood 80-90% of the mites are within cells. However, although estimates from mite drop are not hugely accurate, they are a lot better than doing nothing. The National Bee Unit has published a Varroa calculator. This allows you to use a combination of the mite drop per day, the time of year, length of season and level of drone brood to predict the total numbers of mites in the colony. For some inexplicable reason this asks for the level of drone brood in December … with 0% not being an available option  🙁

Time to treat

With little or no brood in the colonies, now is a perfect time to treat with an oxalic acid-containing preparation to hammer down the remaining mite population. I’ve previously discussed the importance of this midwinter treatment (see Two treatments … a double whammy). In many ways it’s preparation for the season ahead, rather than for the protection of the bees already present in the colony. The lower the mite levels are at the beginning of the season, the longer it will take for the mite population to reach dangerously high levels.



You can model these events using BEEHAVE. This is an interesting in silico model of a beehive. With mite numbers of ~10 at the beginning of the year, maximum levels reached are low to mid-hundreds by late summer, reducing to a couple of hundred the following winter. This assumes no intervening treatment and runs the model using all the default settings. In contrast, using the same parameters but starting the year with ~100 mites, levels peak at between 3000 and 4000 mites, returning to about 1800 in December.

Remember that the National Bee Unit recommends mite levels should not exceed 1000 or there is a risk of “significant adverse effects on the colony”. Therefore, the midwinter treatment is an important preparation for the year ahead, delaying the point at which these dangerously high mite levels are achieved.

Have your hives got less than 100 mites in them now?

Remember also that, with no sealed brood, midwinter is also the ideal time to expose as many mites as possible to the treatment. With the exception of prolonged treatment with hard chemicals like Apistan or Apivar, it’s probably the only time you’ll achieve greater than 95% reduction in mite numbers. With little or no brood present there’s nowhere for the mites to hide.

Dribbling or vaporisation?

An oxalic acid-containing treatment is recommended in midwinter. This can be delivered by dribbling or sublimation (vaporisation). Under optimal conditions, efficacy of the two methods is broadly similar (90%+) though there is some evidence that dribbled oxalic acid is slightly detrimental to colonies (when compared with sublimation, but not when compared to doing nothing).

Sublimox in use

Sublimox in use …

Api-Bioxal is the VMD-approved oxalic acid-containing treatment. If used for dribbling be aware that the suggested concentration on the side of the packet is higher than conventionally used in the UK. It’s also worth noting that it’s not available pre-mixed so has to be made up from powder. In this regard it’s a less useful product than the pre-mixed oxalic acid solution that Thorne’s (and possibly other suppliers) sold each winter. The one- or two-hive beekeeper needs to weigh out very small amounts accurately, or get together with others to make a large batch. Hardly what I’d call progress. Furthermore, the inclusion of glucose and powdered silica (as an anti-caking agent) in Api-Bioxal means it leaves a caramelised mess if used for vaporisation. Although a scouring pad and elbow grease will get rid of this mess, it’s another example of how the “approved” commercial product is actually less good – and no more effective – than the oxalic acid dihydrate that beekeepers have been using for 20 years or more.

Notwithstanding these negative comments, Api-Bioxal works well and is less expensive (per treatment) than most of the other VMD-approved Varroa treatments.

Don’t delay, get out and get dribbling …

The forecast for the next 7-10 days is for significantly warmer temperatures. This means that the queen – if she was having a break from egg-laying – will start laying again. There will be open brood by this weekend and sealed brood in your colonies by about the 15th of December. Dribbled oxalic acid is detrimental to – and may kill – open brood so if this is your preferred method of treatment then don’t delay. If you sublimate you’ve got a few days leeway, but don’t delay any longer than that.

Here are a couple of old videos showing trickling (dribbling) oxalic acid onto a large and small colony in the middle of winter. The Trickle bottle from Thorne’s makes administering the treatment very quick and easy.

Of course, sublimation using an active vaporiser like a Sublimox is even faster and doesn’t involve opening the colony. Here’s an example showing treatment of a recently hived swarm in midsummer … I could have removed the Sublimox after about 30 seconds.

The Daily Mail may be predicting the coldest winter since the last ice age (so perhaps there will be another broodless period§) but I wouldn’t rely on them to influence something as important as the midwinter treatment for reducing Varroa levels.

Here’s a perfect example of the problems encountered by the ‘topical blogger’. I wanted to write about midwinter Varroa treatment in the middle of winter, at a time when others – particular new beekeepers – should be treating their own colonies. Typically these treatments are made in late December or early January. However, the long-range (10 day) forecast in late November suggested the second week of December might be suitable. Some of this was therefore written in very late November, the Varroa drop comments added once I’d completed counting around the 4th to the 6th, and the post finished off the following day once I’d treated my own colonies.

This assumes that the queen started laying on the 7th, the first full day with elevated temperatures.

§ I didn’t open any colonies to confirm they were broodless. I was happy enough to take the clues from the increased mite drop on the Varroa trays and the absence of debris indicating uncapping of brood cells. However, I was told by friends that other colonies they opened on the 7th were broodless.


The Drifters cont.

The Drifters ...

The Drifters …

Not the legendary American doo-wap/R&B vocal group but instead a quick follow-up to a recent post on drifting in honey bees. I discovered an interesting article in a 2011 issue of American Bee Journal in which Wyatt Mangum (Mangum, W. [2011] Varroa immigration and resistant mites ABJ 151:475) quantified mites introduced with bees from other colonies. The experiment was straightforward and quite clever … a number of colonies were prepared with very low mite numbers, overwintered and then miticides (unspecified, but from the remainder of the article I’m assuming Apistan) were applied continuously for the rest of the season. This would kill all the mites present. With a Varroa tray in place it was therefore possible to count newly introduced mites throughout the season. These must arrive with drifting workers, drones (not sure if drones ‘drift’ as such … perhaps there’s a better term for their itinerant wandering?), bees that have abandoned other colonies or potentially robbers. The newly infesting mites would of course be killed by the miticide after introduction and before reproduction. They could therefore easily be counted on the Varroa tray under the open mesh floor.

The results were striking … in one year between mid-May and early October an average of 1415 and 1001 mites were introduced to each of the seven ‘recipient’ colonies in two separate apiaries. Mite arrivals weren’t evenly spread, but peaked during a late summer dearth of nectar … perhaps, as suggested by the author, as other colonies started to run out of stores. The source colonies were not identified, but were not within the test apiaries. Whatever the cause, this represents a very significant influx of up to 7-10 mites per day. In Mangum’s experiment these mites could not replicate (due to the miticide that was always present). Had they been able to do so the impact on the recipient colony, in terms of numbers of mites transmitting viruses within the hive, would have been much greater.

The impact of drifting and mite reinfestation

The impact of drifting and mite reinfestation

Using BEEHAVE this impact can be modelled. In untreated colonies (solid lines), primed with 20 mites at the beginning of the year (and default conditions as previously described), the average mite level at the year end is ~430 (n=3) having reached a maximum of ~600. Using the same infestation period as reported by Mangum¹, with a mite infestation rate of 7/day (the lowest he observed), the average mite levels at the year end were ~2700 (n=3), with maximum levels reaching ~3800 in late summer (dotted lines). In this simulation the introduced mites can reproduce. Therefore, within just a few months, phoretic mites carried on workers and drones from other colonies, have the potential to raise mite levels in the recipient colony to dangerously high levels – significantly higher than the maximum recommended level of 1000/colony. This is potentially of fundamental importance in strategies to effectively control Varroa.  It should be noted that in a repeat of his study this large scale infestation was not observed. This suggests that this type of infestation – from outwith the apiary – may only be a problem in certain years or under specific conditions. One possibility that comes immediately to mind would be a collapsing feral colony or abandoned (or potentially not abandoned, but just completely ignored and untreated … or ‘abandoned‘ as some might say 😉 ) hive within foraging distance.

Ample opportunity ...

Ample opportunity …

Interestingly, a recent study has looked at the influence of a number of honey bee pathogens on drifting (or inter-colonial transmission as they rather long-windedly call it) behaviour. Of the viruses, Varroa and Nosema tested, only the presence of high mite levels influenced drifting … but not in the direction that might be expected. Distance between colonies in an apiary was the major factor that influenced drifting and ~17% of tested workers had drifted (with a third to half of these being apparently unrelated to other colonies in the test apiary). Surprisingly, colonies with high Varroa levels were more likely to acquire drifting workers, though the mechanism for this was unclear. The increased mixing through drifting would ensure that these colonies would likely end up with a greater diversity of viral and other pathogens though whether these colonies could, later in the season, act as a source rather than a sink for mites was not tested.


Drone …

Finally, returning to the subject of drifting bees and the ABJ … in the February 2016 issue there’s an interview with Tom Seeley (of Honeybee democracy fame … Sharashkin, L [2016], ABJ 156:157) in which he states that, when quantified, 34% of drones in his apiary colonies were from other hives. This article – on Surviving without Treatments: Lessons from Wild Bees – also discusses the importance of colony separation to coping with Varroa. The feral colonies Seeley studies are located at least half a mile apart in woodland. When recovered and relocated together in apiaries (‘beeyards’ as they’re called in the US) they rapidly succumb to mite-transmitted viral diseases, whereas those maintained some distance apart (30+ metres) survive. Seeley makes the point that pathogens evolving in closely-spaced colonies are likely to be more virulent, whereas those that are in distantly spaced colonies should be less virulent (or they’ll kill the host colony before being transmitted). Seeley is referring to the virulence of Varroa but I think his comments apply better to the viral payload carried by the mite. This is a relatively minor distinction but these observations further emphasise that drifting in honey bees is clearly a major factor in mite, and consequently disease, transmission … and therefore needs to be considered in control.

STOP PRESS – A recent Bee-L post highlighted a further study on the influence of re-infestation. Greatti et al., (1992) showed that ~2-14 mites/day/colony were acquired in their test apiary during June-August, and that this number rose to up to 75 mites/day/colony in September and October². This type of re-infestation can occur by drifting as already discussed, or by workers in the sentinel colonies robbing out mite-infested collapsing nearby hives or feral colonies.

¹In the Mangum study the mites did not infest the sentinel colonies at an even rate of 7+/day. Instead there was a marked peak in mid-season. I’ve not attempted to model this. Clearly if mites don’t arrive earlier in the season the overall levels would be lower (as they wouldn’t have the chance to reproduce). However, an influx of mites in mid/late-season might just arrive at the wrong time to damage the all-important winter bees … the topic of a future post.

²Greatti, M., Milani, N. and Nazzi, F., (1992). Reinfestation of an acaricide-treated apiary by Varroa jacobsoni Oud. Exp. Appl. Acarol., 16: 279-286