Lasting learning?

In one of the recent blogposts I wrote about non-resistant bee colonies against varoa mite, learning from resistant worker bees how to deal with the mites and being resistant. It seems now that these colonies that has learned resistance have to have worker bees from genetically resistant colonies drifting into them continuously to stay resistant.

This resistance went on for at least a couple of years. I referred to experiences of Hans-Otto Johnsen and Terje Reinertsen ( Norway and Magnus Kranshammar ( Sweden.

A.) In the Norwegian cases colonies with queens from non-resistant stock originally were kept in apiaries together with resistant bee colonies. Bees are kept on small cells.

B.) In Magnus’ case it was somewhat similar as the colony of the original queen, which was the result of genetically selected resistance, and splits from it with queens of originally non-resistant stock, were kept in the same apiary. The total number of colonies was not big here. These bees were kept on large cells.

A.) A couple of years ago Hans-Otto started a variety of the first tests he made mentioned in the blogpost above. He introduced virgin queens of non-resistant stock into splits from his Elgon-stock. In this test he didn’t place these new colonies in apiaries together with his resistant stock. Instead he placed them in an apiary of their own in the forest, a lot more than 2 miles (3 km) from other bees. For two years these colonies functioned without problems concerning Varroa.

This autumn they all showed crippled winged bees. He made a test with the Bee Shaker ( About 35 mites from 300 bees. 11-12% Varroa level. Probably there was resistance in the colonies at least the first year as long as the old worker bees were alive.

In an apiary all colonies will share more or less worker bees with each other. The whole apiary works more or less together. This depends of course also on how much worker bees from different colonies drift. In an apiary with resistant colonies, bees from resistant colonies will help keeping non-resistant colonies resistant. (And if the non-resistant colonies are many compared to for example a single resistant, this may create problems for the resistant one.)

B.) Two splits from a colony with a queen of non-resistant stock, which anyway were resistant, were moved to an apiary of their own in a forest about 2 miles (3 km) from other bees.

The mother colony of the splits was kept in the same apiary as the colony with a resistant queen (which was still alive this year). It was this resistant queen’s colony that gave bees to this mother colony of the described splits. So you could expect some drifting had occurred with resistant worker bees now living in the colony with the non-resistant queen. Thus some good resistant worker bees were living in the splits as well to begin with, until they died their natural death.

Drones from the non-resistant queen of the mother colony of the splits were of course present in these. But of course there were also drones in some colonies 2 miles (3 km) away. Those colonies were of Elgon heritage, but not of the very best variety of resistant stock. The splits got virgin queens from an Elgon colony of good resistant quality. The Varroa level was initially very low in the splits, about 0.3% (without treatment this year or the previous year).

The splits developed very well after the queens were mated and laying. In the beginning of September the Varroa level was 2% and 3.3% respectively. It had increased more than expected, but was not alarmingly high. The cell size in those splits was initially 5.4 mm, that is large cell size. (Resistant thus in spite of the large cells.) The story of those bees are told here: But the virgins came from established small cell size colonies since many years. The plastic combs used for increase were small cell. The 3.3% part had relatively somewhat more left of large cell combs.

131mk1 One of the splits in case B. The original large cell combs are of an old Swedish (as well as an American) size, 12”x12” (30 x 30 cm). With the help of adapters these combs are placed in boxes which hold Jumbo sized frames (448 x 286 mm) plus a deeper bottom board. The combs used for increase are small cell plastic Langstroth from Mann Lake (

C.) In a project that has been going for a couple of years now, one apiary was placed a lot more than 2 miles (3 km) from other bees. The queens were of non-Elgon, non-resistant, large cell bees. They were introduced into Elgon colonies on small cells in 2013. In autumn 2014 they were moved to their test apiary and treated with a small amount of Thymol to ensure an in the test initially small and quite even population of Varroa mites. They had during 2014 been standing in apiaries with Elgon colonies. Thus probably containing some drifted Elgon worker bees. The test apiary was poor from a nectar point of view. In 2015 the Varroa level in autumn was low, around 1%. A split was made during 2015 and queens shifted to daughters made in the split in two of the units. The virgins were mated in this apiary. One of the originally queens, all had been sisters initially, were still left in its colony. The new queens were of course very inbred as the available drones were very closely related to the virgins.

In the middle of the summer 2016 the Varroa level was 3-4%. A couple of weeks later all three colonies showed crippled winged bees. They were treated with two pads with 5gr Thymol each (once) after which they recovered.


The experiences given here makes it probable that there are more than one component necessary for a queen of a non-resistant stock to have a resistant colony. First the colony with a non-resistant queen must become resistant. It can become resistant quickly in two ways. (One may ask whether it is desirable to deliberately do this.)

1.) Worker bees from a resistant colony can be united with a non-resistant colony through shifting the places of a resistant and a non-resistant colony. Most of the other colones in the apiary are resistant. It is uncertain how many of the colonies that should be genetically resistant.

2.) Splitting a resistant colony and giving the split(s) virgin or laying queens of non-resistant stock.

To maintain this resistance achieved, still with a queen of non-resistant stock in the colony, when the first resistant worker bees that came along with the splits are worn out and dead, new resistant worker bees have to come into the colony in some way. (Resistant worker bees are bees in colonies that show resistant behavior to such an extent that their colony can rid themselves of mites through different kind of hygienic traits.) It seems the number of resistant worker bees needed to come from other colonyies “per time unit”, are not very many. By just keeping the colony in an apiary with resistant colonies this seems to be achieved, probably through drifting of worker bees between the colonies in the apiary.

In all cases above, A, B and C there were no new resistant worker bees drifting into the colonies. In case B though virgin queens from a resistant queen were introduced, but no new bees from the original resistant stock were re-introduced. Instead new workers with probably better genetic set-up for resistance were born. But there were now probably no, or very few, resistant worker bees left that could teach newborn bees. At least this is a conclusion that is close at hand, to explain the increase of the Varroa level. It will be interesting to see what will happen further on in these two colonies. Hopefully the virgins were mated to other drones than those that came along in the splits from the genetically non-resistant queen.

No shortcut

It is thus no shortcut to get resistant bees, to split genetically resistant colonies and introduce whatever kind of queens into those splits (of non-resistant heritage)! What these experiences tell us is how important it is to increase your number of colonies and replacing the dead outs by splitting the best resistant colonies. The very best might many times be to let those splits raise their own queens. And in addition to this breed queens from good resistant colonies of good heritage in being resistant.

7 thoughts on “Lasting learning?

  1. Nice facts and thoughts putting more focus on genetics and breeding (or at least selection), as essentiell tools to gain and mantain resistant stock.
    The 3 cases you choose as evidence for your conclusions, all demonstrate an increase of mite number plus DWV bees on non resistant queens/stock, after they have been replaced from their initial yards, where they are located together with resistant stock.
    To make your conclusion even more convincing I would mention the controls, that means original resistant stock, that was not shifted and faces more or less the same weather conditions. I assume that these colonies do not increase mite number, but you haven’t mentioned. BTW what is the mite development of a resistant colony during the year. Is the level is always constant, as well in autumn?
    You know that your conclusions weaken the theory of epigenetic resistance, which implys that resistance behaviour could be transferred from older field bees to young bees.
    Could you comment on this please? I mean one could argue that for instance the older bees teach the younger bees only in case enough mites are present. If the old bees die, before hygienic behaviour is needed, they couldn’t teach. Perhaps, this would be in line with your findings and the epigenetic theory. But thats only a thought.
    I’m very curious about the discussion of this topic.

    • We could call the controls those colonies,as you suggest, that are managed as usual at their original sites.
      A.) Varroa level on all colonies with H-O Johnsen (which haven’t been treated at all for 15 years) measured this late summer to be 1-3 mites on 300 bees, more than a couple of hundred colonies checked. Those colonies with non-resistant queens that were staying with his resistant bees functioned for two years as his resistant, after which they were shifted. No DWV during two years. When they were placed alone the varroa level had risen to 11-12% at the end of the second year and a lot of DWV-bees.
      B.) The mother colony of the splits, still at its original place show low varroa level. Daughters of the original resistant queen is present in colonies in the same apiary. But now when the original resistant queen is gone it may change next year, who knows.
      C.) Well, here we have kind of control colonies at another site in forest area, not very productive, but somewhat better. And the colonies there are not inbred as these with raised varroa level. The stock of bees in “control” colonies are Elgon. But the apiary is very isolated as well. In spring the Varroa level was 1-3 mites in 400 bees in the colonies. In late summer 0-6 mites in 400 bees. The colony with 6 mites, considered the “worst” one was moved to another location and due to curiosity was treated. 10 mites was collected from the treatment. Maybe the bees started grooming after the season and I choose the wrong colony as worst, who knows. We have to remember also that this alcohol wash method isn’t 100% accurate, but give a good hint of the truth.
      The mite development of a resistant colony during the year can surely be very different due to the presence of non-resistant colonies more or less close, and the quality/composition of the resistance. I know of only one paper on this, the doctorate thesis of Remy Vandame made in Mexico some years ago now measureing the poplulation of mites in Africanized/AHB (considered resistant) and European bees/EHB. In short the mites had about 800 mites as the lowest and more than 3000 at the most in AHB. In EHB the mite population didn’t go down as much as in AHB and as a result over time continued to grow to a fatal height. To be noted here is that these colonies were placed in the same apiary.
      Yes, my conclusions weaken the epigenetic possibilities, but not to the extent that they are of no or very little importance or impact. Still epigenetics is very important, yes without it resistance wouldn’t bee possible actually. remember that resistans was achieved in about 5 years in both South America and South Africa. Without epigenetics this is impossible to explain. It’s way to short time for “normal” genetic selection to explain it.
      And still the only explanation for non-resistant colonies to become resistant in an “instant” is teaching from resistant worker bees and learning by non-resistant workers.
      If older bees teach younger only if mites are present I have no idea, but it seems plausible as if there are no mites there are no means how to teach mite elimination by grooming for example.

      • Thanks a lot for your fast reply. Indeed the control measures you have given are convincing, supporting your conclusions.

        I’m not sure whether I was correct in naming teaching an epigenetic trait, as by definition epigenetic means, a trigger from outside that affects the gene expression e.g. by activation. Hence, the gene in question is only activated, if this trigger is there. For instance, if small cell bees develope in higher density, a higher brood temperature is claimed. This higher brood temp. might act as gene activator for some genes. Or if hydrophobic pesticides are inside the wax, they kill symbiotic fungi/bacteria which in turn (their chemicals) could act as activators or deactivators of potential genes.
        Teaching of adult bees to young bees should therefore not belong to epigenetic resistance. The same is true for physical measures as well which belong to management system e.g. cell size which affect choice of mites (e.g. 4,9mm has roughly 50% of mite preference compared to 5,4mm according to the paper you have mentioned some time ago).
        However, when discussing the consequences of your conclusions, which again I find very important, that means, one has to put special attention on evaluation of their own colonies. Which means mating, which should be planned in locations or at times where unwanted drones are rare, to increase the enrichment of resistant genes. And selection, which should focus on requeening of non resistant queens and breeding from resistant ones.
        The options to evaluate resistant traits are discussed several times before. I guess this is easy for constitutive expressed resistance like VSH. However, it’s more complicated for epigentic resistant factors or crisis induced resistance. The latter one is integral part of the strategy the resistantbees community is following and means, bees might be forced by a strong mite trigger, to induce or learn the hygienic behaviour or to put that higher on the priority action list compared to before the crisis. I’ m not sure whether this belongs to epigenetic resistance, because their are some genes expressed after the crisis or not. Or whether this belongs to behavioural adaptance according to stresses only.
        When evaluating theri stock, one has to be sure to not outselect a potential resistant stock to early and not to late of course as well in order to do not waste time.

        • Yes, teaching and learning might not or might have epigenetic implications. I would say we don’t know. But I agree it doesn’t seem to be typical for epigenetical effects.
          If higher broodnest temperature is an effect of small cells I don’t know. Do you have a reference? Would be interesting.
          To evaluate totally the own colony does always involve at least some degree of uncertainty. But even if I think we never will be able to totally will get the 100% correct picture of the traits of colony, we will have enough of a correct picture to enable us to make a selection. That’s one reason I think the most important is to shift queens in the least good 1/3 of the colonies (or 1/2 as Brother Adam said), and breed from some of the others. Best from many of them, but for me I have to weed out so many when I select for breeders that not many are left.
          Crisis and stresses I think are important to trigger epigenetic changes.
          I don’t think it’s important to avoid to 100% “mismatings”. They may well contribute, at least to keep away inbreeding for the colony at hand to function well, as the sistergroup created may well conttribute with some impportant traits even if they are not miteresistant. When daughters are bred from a “mismated” queen which may be choosen as a breeder, because it’s good (partly due to the mismating), daughters will be born with different fathers. The queens then will mate in their turn with many resistant drones and a few non-resistant. The totality of the sistergroups may well still create a colony that works very well, in resistance and otherwise. If not it will be weeded out finally, finally, by nature or by the beekeeper.
          So again – shift queens in the worst ones, Make slits from the best ones and breed from the best ones.

          • Regarding the temperature reference, no I don’t have one. I was teached that a management system, which focusses on small cells involve a suitable comb distance which is smaller then the standard one of 35mm, instead 33 or even 32mm. This plus the higher densisty of brood/comb will incease the brood temperature. This has two favourable results, first mites don’t like too high temp. and second bees that develope at higher temp. will live longer. The latter is experimentally proved by Prof Tautz and cited in some of his books and probably in papers of him. There he states that an increase of 0,6°C during the brood, e.g. from 35 to 35,6°C will result in a 10 days longer survival time of worker bees, which is quite significant. This in turn is in line with the finding of a lot of small cell bee keepers that their bees on small cells live longer and the number of bees/stock is higher. Hence, this is indirect evidence. But maybe there exist some paper about that. Maybe Lusby has published it somehwere.

          • There are beekeepers from before the return of the small cells who used narrower distance between combs and reported quicker spring build-up. the argument then was that it was easier for the bees to hold the correct higher temperature for the brood. But that SC and narrower distance should create higher still temperature I’m not aware of, more than speculations, that well could be correct of course. I tend to think that 35 or 35.2°C or something else is dependent more on genetic, or maybe epigenetic, causes.

  2. In principle this could be measured very easily, if one has a thermo sensor with a data logger. But maybe there is some literature already about that. But you are right as long as there is no clear reference or experimental prove, one can challenge that although I found the argument plausible…

Comments are closed.