Rinse and repeat

Midwinter mite treatment is no substitute for a properly applied late summer treatment that protects your all important winter bees. However, you also need to control mites in the winter or there is a good chance their numbers will reach damaging levels the following season 1.

Mid September

Late summer treatment and no winter treatment – mite levels in red.

OA (oxalic acid-containing) treatments are the ones to use in midwinter (e.g. Api-Bioxal). These can be trickled in syrup onto each seam of bees or they can be vaporised (sublimated), effectively coating everything in the hive with a very fine dusting of crystals.

Trickling damages open brood whereas sublimation is exceedingly well-tolerated by the colony.

If you are certain the colony is broodless then trickling is faster 2 and – because you don’t need power or any more PPE 3 than a pair of gloves – much easier.

If the ambient temperature is consistently below ~6°C and I know the colony is broodless I usually trickle. If the temperature is higher and/or I’m uncertain about whether there is brood present I usually vaporise.

I watch the weather and treat after the first prolonged cold spell of the winter.

Experience over the last few years suggests this is when colonies are most likely to be broodless.

Most likely is not the same as certain 🙁

Count the corpses

After treating I closely monitor the mite drop over several days. I use white Correx Varroa trays that slide underneath the open mesh of my kewl floors.

Easy counting ...

Easy counting …

I don’t count the mites every day, but I do try and count the day after treatment and 2-4 days later. I record the mite drop per hive and, over time, look for two things:

  1. The cumulative mite drop. This indicates the original infestation level of the hive. Usually it’s in the range 10-75 mites (total) for my colonies in midwinter, but – as you’ll see – it can be much higher.
  2. The speed with which the daily mite drop falls to a low single-digit average. OA treatment is very effective at killing phoretic mites. If there’s a continuing high level of mite drop it suggests that more are getting exposed over time.

In my experience, vaporised OA often results in a greater mite drop 24-48 hours post-treatment rather than in the first 24 hours 4. After that I expect (hope) the daily mite drop tails off very quickly.

Vaporised OA remains effective in the hive for several days. Randy Oliver reports studies by Radetzki who claims it remains effective for up to three weeks. I think this is an overestimate but I’m sure it continues working well for four to five days.

OA, whether vaporised or trickled, on broodless colonies is 90-95% effective i.e. if there were 100 mites in the colony you should expect as few as 5 remain after treatment.

Four to five days after the initial treatment I eyeball the numbers across all the hives in an apiary and look at the profile of the mite drop.

Mite drop profiles

I couldn’t think of a better term for this. Essentially, it’s the shape of a graph of mites dropped per day after treatment.

I don’t usually draw the graph – I have a life – but I do look carefully at the numbers.

Here are a couple of sketched graphs showing what I mean. Days are on the horizontal (X) axis, dead mites per day are on the vertical (Y) axis. Treatment applied on day 0. No count (yet) on day 6.

Mite drop profile – this is what you want

In the graph above there are high(er) levels of dropped mites on the first day or two after treatment, but levels thereafter drop to a basal level of perhaps 1-4 mites per day.

Each time I count the mites I clean the Varroa tray (the rinse in the title of the post).

Assuming the day 5 mite drop is very low, the profile above is what I’m looking for. It shows that treatment has worked and no repeat is necessary.

The profile below is much less promising 5.

Mite drop profile – this suggests additional treatment is needed

In this graph (above) the mite drop remains high every day after treatment. Sometimes they even increase over time.

If you assume treatment is equally effective – say 90%+ – on the five days after treatment 6 this must mean that there are mites being killed on days 4 and 5 that were not exposed to treatment on the earlier days.

How can this be?

The most likely explanation is that the colony had some sealed brood that has emerged in the days following treatment, exposing previously ‘hidden’ mites to the miticide.

It’s good that they’ve perished, but are there more hiding? How do you tell?

Enough of my hand drawn idealised graphs with no real numbers … what about some actual data?

Real world data

The graph below shows data for seven colonies in a single apiary. All were treated with Apivar in late summer. All were treated with a vaporised oxalic acid-containing treatment on the 28th of November. 

Mite drop profiles – real world data

I counted the mite drops on the 29th (T+1), the 2nd (T+4) and 3rd (T+5). The figures for 30th to the 2nd were averaged, which is why the bars are all the same height.

  • Colonies 3 and 6 had very low mite levels. Though not the lowest in the apiary 🙂
  • Colonies 2 and 7 had pretty good mite drop profiles, with low single-digit numbers on day T+5. None of these four colonies (2, 3, 6, 7) need treating again.
  • Colonies 1 and 5 have high mite levels 7 and – despite the pretty good levels on T+5 in colony 1 – were both re-treated.
  • Colony 4 was also treated again as the profile was flat and I suspected they had low levels of mites but were rearing brood..

And repeat

Note: The instructions for Api-Bioxal specifically state that the maximal dose of 2.3g/hive should be made in a single administrations with only one treatment per yearPrior to the VMD licensing and approval of Api-Bioxal there was effectively tacit approval for beekeepers to use unadulterated oxalic acid by trickling or vaporisation, without any particular limitations on frequency of usage.

It’s worth stressing that you should not repeat oxalic acid trickling 8.

Here is some real data for repeat treatments of another colony in the same apiary.

Repeat treatment for brood-rearing colony

The average mite drop per day over the first 5 days was ~60. This justified an additional treatment. Over the next 6 days 9 the average drop was ~20. I considered a third application was needed after which the mite drop per day was in the low single digits.

And again

Repeated treatment is needed if there is sealed brood in the colony.

The likelihood is that two additional treatments will be required.

Why two?

Here’s a reminder of the development cycle of the Varroa mite in developing worker or drone brood.

Repeated oxalic acid vaporisation treatment regime.

Worker brood occupies capped cells for 12 days (days 10 – 21 of development, shown above). Vaporised oxalic acid-containing treatments show a drop in efficacy after 4-5 days 10.

Therefore, to cover a complete cycle of capped brood, you need 3 x 5 day treatments to be sure no mites emerge without them being greeted with a lethal dose of something really, really unpleasant 😉

There should be no drone brood in your winter hives 11 but, if there was, 3 x 5 day treatments should just be enough to cover the complete cycle of capped drone brood as well. However, a fourth treatment might be needed.

Note (again): The instructions for Api-Bioxal specifically state that the maximal dose of 2.3g/hive should be made in a single administrations with only one treatment per year

Not all hives are equal

There are 15 hives in the apiary containing the bee shed. Colony 1 had just about the highest mite levels. However, as shown in one of the graphs above, adjacent colonies can have markedly different mite levels.

There is no clear correlation between mite drop after treatment and colony size. Colony 1 is a double brood monster, but the others in the bee shed are all single brood 10 and 11 frame Nationals 12.

Some colonies need repeated treatment, others did not.

To maximise efficient treatment and minimise unnecessary miticide usage it is necessary to monitor all the colonies.

It’s also worth noting that monitoring only a single hive in an apiary may be misleading; compare colonies 1 and 6 above in the graph of real data from the bee shed.

This monitoring takes just a few minutes. I usually do it after work. In the bee shed this is easy as I now have LED lighting and it’s nice and dry.

Easy conditions to count mites

In my out apiaries I have to do it by headtorch … under an umbrella if it’s raining 🙁

Checking mite drop by torchlight

That’s the last job of the winter completed … time now to review the season just gone and plan for next year.


Colophon

Rinse and repeat

Rinse and repeat is a truncation of instructions often found on the side of shampoo bottles – Lather, rinse and repeat. Other than potentially resulting in an endless loop of hair washing, it also means that a process is (or needs to be) repeated.

In The Plagiarist by Benjamin Cheever, a marketing executive becomes an industry legend by adding one word – REPEAT – to shampoo bottles. He doubles sales overnight.

For Varroa treatment the instructions should be amended to Repeat if necessary … and note again the instructions on Api-Bioxal which, at the time of writing, is the only oxalic-acid containing VMD approved miticide that can be administered by vaporisation.

 

Footnotes

  1. Because, if you treat early enough to protect the winter bees, mite levels increase again before year-end as they reproduce in late season brood.
  2. No … don’t bother trying to convince me trickling is slower. It is not. I’ve done both, timed, using a Sublimox vaporiser which is one of the fastest on the market. For one person working alone trickling is faster.
  3. Personal Protection Equipment.
  4. Though, as with all things beekeeping, there’s lots of variation. Some hive drop more in the first 24 hours, others start slow and then quickly build up before – just as quickly – falling again.
  5. Except there are lots of dead mites involved … so there is some good news.
  6. I’m not sure this is a valid thing to do but will return to discuss it in a future post.
  7. I stop counting at ~100 mites unless I’m averaging over several days. The colony 1 count on day T+1 could have been significantly higher.
  8. Surprisingly, Api-Bioxal allows two treatments per year by trickling, but only one by vaporisation. Where’s the logic in that?
  9. I usually treat at 5 day intervals but the weather on the 8th was awful and I didn’t want to use my 240V vaporiser outside in heavy rain! I’m happy to go the extra mile for my bees, but not to risk electrocution.
  10. Not completely, but enough to warrant a repeated treatment. I have some data on this and will cover it in a future post. It’s worth mentioning here that Pete Little, a bee farmer in Somerset, determined that a 5 day interval was optimal by empirical observation. This was first ‘published’ to my knowledge on the Beekeeping Forum.
  11. Or if there is your queen has failed and the colony is doomed.
  12. What makes the mite counts so different? I suspect it’s because colonies rear autumn brood at different rates with higher counts from colonies that rear more brood.

16 thoughts on “Rinse and repeat

  1. Stan Thornton

    I have six hives which are all top bar hives. I have been using OA on blue paper towels for several seasons, with good results. According to Randy Oliver, I don’t believe this has been approved by the USDA , yet. Living in northern California along the coast , mites have been quite a concern. Some of the comments I’ve been reading in the Bee Culture journal articles have been about …having a high humidity in the brood area to reduce Sucessful gestation by varroa mites. This new information contradicts the concept of good ventilation up high in the hive.

    Your articles are great! Thanks from coastal Northern California.

    Reply
    1. David Post author

      Hello Stan

      I’ll admit to being pretty ignorant on the effect of high humidity and mites. I know there’s an upper limit above which mite reproduction is limited. There’s a recent thread on Beesource discussing a new paper by Derek Mitchell which touches on this subject. Other than the open mesh floor my colonies have no additional ventilation. In a recent post I discussed the oft-quoted recommendation to prop the crownboard up with matchsticks to increase ventilation. This dates back to pre-history (almost!) and I think there’s good evidence that the bees do not appreciate ventilation above the cluster. Exhibit 1:
      Ventilation - no thanks
      Pleased you enjoy the posts.
      Cheers
      David

      Reply
  2. Steve Riley

    Hi David ,
    A few thoughts:-
    1) Repeated Oxalic treatments are illegal in the UK
    2) Winter treatment is prophylactic
    3) This stops any natural selection of drones mating from successful colonies in the spring. This results in the propagation of genes from bees that cannot cope with varroa (Rinse & repeat)
    Regards
    Steve

    Reply
    1. David Post author

      Hello Steve

      ‘Prophylactic’ as in intended ‘to prevent disease’. Absolutely! High levels of winter mites will reduce spring build up of colonies and will mean they are much more likely to struggle with high mite levels mid-season.

      I’m not sure why you think repeated midwinter treatments are illegal. It doesn’t say that on the instructions on the last 350g of Api-Bioxal I purchased. It recommends single treatment, but that is very different from it being a crime (which appears to be what you are saying) to repeat a treatment if it is needed.

      With open mated queens and the quality of my beekeeping I do not think Varroa-resistance is a reasonable or attainable goal. I’m also far from convinced that the resistant bees some discuss are actually resistant at all. The evidence needed is to move them away from the region they were ‘selected’ and expose them to a novel Varroa and virus population. Those I know that have done this are just as susceptible as unselected stocks. Selection selects both for bees and the mite/virus population …

      Cheers
      David

      Reply
  3. Vince poulin

    A very informative post. Most useful is seeing the variability in mite drops over time and amoung hives. Especially valuable are comments on what look to be controlled values – “sub digit” numbers. This season none of my mite drops have exceeded more than 30. Post treatment counts using OA vapour follow the same pattern seen here. Interesting that I haven’t seen big differences in numbers between size of colonies (3 NUCs and 1 two brood box hive). My target is 0 – 3 mites in 4-5 days post treatment followed by a long stretch of 0 – 1 or 2 mites before thinking numbers are on the rise. I don’t hesitate to do multiple treatments if numbers do not drop. I treat every new split to get the brood break advantage. OA is an especially effective treatment when properly timed.

    Reply
    1. David Post author

      Hi Vince

      30 or less is pretty good. The bees will appreciate that. It’s worth noting that if colonies are treated late in the season – say late October, rather than early September – the mite numbers never build up again before midwinter. It all depends on how long mites have to reproduce in autumn brood following the late summer treatment. I ensure I treat early enough to protect the winter bees and therefore know I’m going to need to treat again midwinter.

      Cheers
      David

      Reply
      1. Vince poulin

        David I was out of the country when I last posted. While away on Dec 4 family did an OA treatment (vapour). Unfortunately no daily counts after day 1 which were low. Two weeks later I returned to: 132, 135, and 66 total count on boards. I decided to do an OA2 treatment. My post treatment numbers over 4-days are 169, 71 and 80, for an average drop of 42, 18, and 20. Post treatment numbers vary little per day. Look like your hives 1 and 5. These are not down to single digits. I’m two OA treatments into winter bees. Clustering likely has played a role. My thoughts are to repeat the treatment. It’s been relatively warm. 4-8 C most days. Really cold weather has not hit yet. Bees flying most days, just a few. Dead bees on landing boards but lots taking winter candy from cakes. Would you treat again?

        Reply
        1. David Post author

          Hi Vince

          Clustering may play a part as you say. However, if OA remains active for only a week or so (which I suspect, but don’t have full data for) it’s likely the additional drop on OA2 was because more mites had emerged. In the meantime, mites not killed by OA1 will now have reinfested brood if there’s any available (phoretic period is ~6-11 days through choice – which they might not have in midwinter of course). I would be tempted to treat again … with the caveat that, as I’ve now clarified, the instructions for Api-Bioxal clearly (and illogically!) state that only one treatment by vaporisation should be used per year.

          I know that’s not a concern in British Columbia (where you also don’t have to wrestle with cleaning the caramelised gunk out of your vaporiser) but I thought it best to avoid ambiguity and add it for UK readers.

          Happy Christmas
          David

          Reply
  4. Archie McLellan

    Hello David

    Thanks for this. Am I right in thinking that it is a foregone conclusion (for you) that all your colonies will need a winter treatment? Your monitoring charts above are for checking effectiveness *after* treating.

    I treated my two colonies with Apivar in mid August. Though I had horrendously high counts this time last year (even though I had treated in August), the DMDs this winter are 0.7 and 2. If I were to treat, I would be aiming to get down to low single figures, so I guess I shouldn’t treat these colonies for the present.

    The general advice is not to treat unless required. However, the instructions in ALL the approved medicines include the direction to treat all colonies in an apiary at the same time. Is this because of the spread of mites to nearby colonies, or a marketing/sales ploy – because, as you noted above, next door colonies can have very different mite counts.

    Archie

    Reply
    1. David Post author

      Hi Archie

      In reverse order … coordinated treatment (as recommended on approved medicines) is probably there as a ‘catch all’. In the late summer it is essential to be maximally effective as that is when robbing is occurring. Why have one nice clean hive next to something that’s still possibly riddled? In mid-winter I see no reason why it’s needed – there’s almost no brood being reared so no drifting. The subtlety is probably lost in the hope the labelling will be understandable … though many still don’t read it!

      All my colonies need treatment in the winter because I treat early enough in the autumn to allow mite reproduction to occur during late season brood rearing. I don’t do a mite count before treatment. It’s a bit hit and miss if you don’t know whether there is brood being reared. If 2/day and active brood rearing then there’s a chance the mites in capped cells may be significant.Not hundreds, but more than you’d want. I’m reasonably confident of the brood status … but far from certain. I think it’s arguable that the colonies #3 and #6 didn’t need treating at all, but I didn’t know that in advance of treatment as phoretic mite drop is so variable and low. If I waited until it increased to a measurable level I know that the mite levels in the hive would be much higher, so treatment would be both required and potentially less effective … 95% of 50 mites vs. 95% of 250 mites, for example.

      I also suspect that mite drop in broodless colonies varies according to the activity of the cluster. I don’t have any good evidence for this but am pretty certain that I see a lower drop during prolonged cold period, presumably because the cluster is packed together tightly and dislodged mites can reattach more easily. Just a guess.

      I’d prefer to be starting the season with very low mite numbers. Since I know they’re present anyway I treat … but only enough to get them down to a very low level.

      Cheers
      David

      Reply
  5. Jane

    Like Archie, I was interested to see that there are no ‘before’ counts. I think that for new beekeepers, that could be very useful. This was my first year, and having monitored in July and found low counts, I wasn’t too worried when I put the Apiguard on in late August. However, when I did the counts during and post-treatment, I had very high levels (hundreds). A combination of that treatment being in place for many weeks, a lack of information about for how long the treatment would cause mortality after removal, and not knowing what the counts would have been before treatment, meant I was slow to realise that the treatment had not been effective.
    Hopefully I’ve now got the mites under control with multiple Api-Bioxal vapourisations – thanks to your reply to my comment on your previous blog post – but I have missed that key ‘winter bee production’ period that you emphasise. In the future, I will monitor before treatment so I know if high counts are treatment mortality or natural mortality with a failed treatment.
    Whether or not multiple Api-Bioxal treatments are a crime, as a new beekeeper I definitely felt I shouldn’t be doing that due to the packet instructions, even though all the internet-based advice talked about multiple treatments with generic oxalic acid when colonies aren’t broodless. It seems there is quite a mis-match between what lots of respectable beekeepers recommend and what the Inspectors can advise. As far as I can tell, the only approved action, with a high chance of being effective, that I could have taken when Apiguard failed was to vapourise or trickle oxalic acid having made the colony broodless by destroying the brood (which is still present now according to inspection board), as I’ve read is done in some research on treatments. I didn’t like that idea.
    The information you provide about the duration of Api-bioxal efficacy is really important and, in my opinion, should be on the packet. Do you know what that would be for thymol treatments such as Apiguard? If it has worked, would you expect counts to fall to acceptable levels while the treatment is still in place for the recommended time? Could they continue to fall after that or does the thymol dissipate quickly?
    As I said, I’ve done multiple vapoursations now, with a peak of 5 on the last one (300 on the first!) and only one mite counted today. Thanks again for you advice and your blog.
    Jane

    Reply
    1. David Post author

      Hi Jane

      ‘Before’ counts are only useful if they’re meaningful and I suspect they aren’t in midwinter (which is what this post is about). At this time of the year many colonies will have no brood, some will have some and a few might have more (but probably not lots). It’s all dependent upon the weather, the bees and the strength of the colony … and probably other things. In the absence of any brood the natural mite drop is influenced by how active the cluster is – on a warm winter day you often see more mites than on a really cold day when they’re clustered tightly.

      Working through all these variables is nearly impossible. It ends up being a best guess. The only way to get a meaningful phoretic mite count is by doing an alcohol or sugar roll. But how does that relate to the total mite load? Again, it depends upon the amount of brood, the rate at which the queen is laying etc. At steady state i.e. X frames of brood with the proportion emerging being replaced by new eggs, you can have make a reasonable guesstimate. I think that’s what the National Bee Unit Varroa calculator does.

      I don’t want to disturb the colony more than necessary midwinter and certainly wouldn’t do a alcohol/sugar roll to count phoretic mites. However, retrospectively you can count the mite drop and monitor the Varroa board after treatment. Interpreting those results allows you to draw some conclusions about the state of the colony and whether additional treatments are required.

      It’s a few years since I used Apiguard. My recollection was that the mite drop tailed off pretty quickly after treatment (the one month 2 x 2 weeks) was removed. Remember that OA is distributed around the hive, whereas Apiguard remains in the tray and is transported around by bee activity. I cannot remember whether the mite drop always increased when the second tray was inserted – if it did it would suggest that the efficacy of the first tray was already dropping.

      I used to regularly use Apiguard but have abandoned it in favour of OA or Apivar. Apiguard is very temperature-dependent and often puts the queen off laying. The very last thing I want to do in early autumn is apply a partially effective treatment (because it’s too cool) and stop the queen from rearing those winter bees. In Scotland I suspect there’s nowhere you could reliably use Apiguard in August/September. Midsummer perhaps … but some years it’s not even consistently warm enough then. You don’t say where you’re located but I suspect you treated too late in the season, so reducing the efficacy of the Apiguard.

      Pleased you enjoy the writing.
      Happy Christmas
      David

      Reply
  6. Jane

    Hello David
    Thanks for that response. Fair points re counts, and perhaps mine was a rather extreme situation (and not mid-winter, as you say) when pre-treatment counts would have been useful.
    I agree that it was probably was too cool for the Apiguard. I had made a note to put it on earlier next year but now I’ll consider the alternatives you suggest (assuming the bees make it through the winter after this year’s mistake!).
    Happy Christmas to you too.
    Jane

    Reply
    1. David Post author

      Hello again Jane

      If you don’t mind using a ‘hard’ chemical then Apivar (Amitraz) has a lot going for it. Easy to apply, not wax soluble, very effective, well tolerated by the bees. Just remember to recover the strips after 6-10 weeks. If you’d prefer an organic treatment then many use formic acid (MAQS) but take careful note of the instructions as some beekeepers report queen problems when using it.

      Good luck
      David

      Reply
  7. vince poulin

    David – Have since concluded a 12/18/19 OA treatment. I gave it 6-days of continuous monitoring. Results mirror several of your examples. Two hives had non-declining daily mite drops. One hive demonstrated a nice fall in numbers. Same as your hive 7. My two non-declining hives mirror your hives 1 and 5. Not big numbers but no drop to single digits (daily average 24 and 41 over 6-days). There is no evidence of winter brood hatching as suggested by cap debris on drop boards. All mites should have been phoretic at time of treatment. We did do a 12/04/19 OA treatment but no continuous daily monitoring (I was away). Total numbers have been about the same between the two treatments. Needless-to-say your showing us real data has been super helpful. Because of the numbers I have done a rinse and repeat on all 4-hives (12/24/19). Let’s see how it goes. I’m looking for single digit numbers before stopping treatments.

    Reply
    1. David Post author

      Hello Vince

      I think it’s worth emphasising that miticides, used properly, have ~95% efficacy. It might be unrealistic to achieve single figure drops if the mite numbers start very high.

      I’m in the fortunate position of having a relatively isolated apiaries. Once I get on top of the mites, they stay low with intervention at the right time. Some of my colonies only ever drop <50 mites in total. For our research colonies we also measure the viral loads - these are often 100-1000 times lower than randomly sampled workers from ‘normal’ colonies.

      I’ve not measured winter mite drop from very heavily infested colonies (we have done this from summer colonies riddled with mites and virus and had extraordinary success in ‘rescuing’ colonies with catastrophic Varroa loads). When I get the chance I will. I think there’s a lot we (or at least I) don’t understand about the interaction of colony activity (clustered or not), level of brood rearing and miticide activity/efficiency.

      Happy Christmas
      David

      Reply

Leave a Reply

Your email address will not be published. Required fields are marked *