I’ve selectively bred bees that were gentle, then immune to AFB, and then resistant to the deadly tracheal mite. So when I saw my first varroa mite on a sticky board in 1993, I thought “this is gonna be easy.” Boy was I wrong!

COLONY COLLAPSE REDUX?

We’ve all heard about the high rate of colony losses this past winter. This was hardly unexpected [[1]]. As stated by Dr. Zachary Lamas [[2]]:

The universal prevalence of [the amitraz] resistance genotype in all of these [mite] samples is strongly suggestive that amitraz may have been ineffective at controlling these Varroa populations. The removal of amitraz as a viable tool to manage Varroa will destabilize beekeeping operations until a replacement miticide or management strategy arrives.

Due to amitraz’s loss of efficacy, researchers and startups are feverishly trying to come up with new “treatments” for varroa control. But the tragic irony of using any miticide to control varroa is that doing so increases the selective pressure for mites that are resistant to that chemical. Beekeepers have exhibited their success at breeding mites resistant to a “silver bullet” miticide time and again.

Keep in mind that any “treatments” are nothing but stopgap measures (Figure 1).

Fig. 1 There was nothing necessarily wrong with jumping to the top of the IPM pyramid to control varroa when it first invaded, but we needed to keep in mind that miticides were only short-term stopgap measures to hold the fort until we bred bees that could handle the mites themselves.

A SIDE NOTE ON RECENT COLONY COLLAPSES

It wasn’t just varroa or amitraz resistance. Beginning last September, our own operation got hit by a mysterious dwindling — clearly not due to varroa/DWV (we were monitoring by mite washes, and there were no sick brood, deformed wings, and seldom any dead bees on the ground), we’ve never used amitraz, and had no pesticide exposure. But it was like déjà vu — the field signs were very similar to the “CCD” that we observed when Nosema ceranae invaded during the 2000s — the slow dwindling of the adult workers, leading to a shrinking cluster over healthy brood (the well-described field sign of collapse due to Nosema apis [[3]]). So I pulled out the ‘scope, and although there was scattered nosema present, it didn’t differentiate between the healthy and dwindling colonies.

Despite our normal feeding of pollen sub and syrup, we helplessly watched what we expected to be strong colonies for almond pollination dwindle to nothing.

The good news: Although the losses devastated us, there were colonies in each yard that continued to thrive (apparently resistant to the mysterious pathogen), and we are now rebuilding our operation with our fingers crossed.

WASTED YEARS

We’ve had varroa in the U.S. for nearly forty years! When we put our minds to it, it took only eight years for us to put a man on the moon. Had we made the effort, we might have done the same with varroa. But the ready availability of inexpensive synthetic miticides allowed us to keep treating the symptoms, rather than addressing the cause.

Let me be clear, the ARS labs, notably the efforts of John Harbo, Jeff Harris, Tom Rinderer, and Bob Danka (as well as a number of other beekeepers (such as Danny Weaver and Kirk Webster), were doing the work of breeding mite-resistant stocks, but the beekeeping industry itself is to blame for not demanding (and rewarding) our queen producers to adopt and improve them.

Practical application: Our queen producers respond to market demand, and can easily sell every queen that they produce. Until beekeepers demand mite-resistant stock, there is little incentive for queen producers to go to the effort of offering them.

NOVEL HOST-PARASITE RELATIONSHIPS

Before humans were involved, some mites evolved parasitic lifestyles on various honey bee species in Asia — in stable host-parasite relationships. When humans later introduced these mites as novel parasites to the European honey bee, our beloved bees were caught unprepared to deal with them, and varroa and Tropilaelaps run rampant (plus increase virus transmission).

Even after four decades, in North America the vast majority of managed colonies are still stocked with honey bee bloodlines that exhibit scant resistance to varroa infestation (although our feral populations appear to be evolving resistance).

We already know that varroa’s original host, the Eastern honey bee (Apis cerana), is able to keep varroa (and Tropilaelaps) in check as “minor parasites” (present, but not causing serious harm). That means that it’s possible for honey bees to deal with mites on their own — it’s time for us to get serious about handing varroa management over to our bees, rather than trying to come up with flashy new miticides.

And we’ve now got Tropilaelaps knocking on our door. All it’s gonna take is a single swarm on a cargo ship (or some bees imported into Canada or Mexico) to introduce it into North America. Since Apis cerana is highly resistant to both varroa and Tropilaelaps, there’s a good chance that if we breed bees resistant to varroa, that we may be better prepared for the invasion of the devastating Tropilaelaps mite [[4]].

Practical application: The only long-term solution to varroa and Tropilaelaps is to selectively breed bees that have the capacity keep these parasites in check without our help!

BREEDING MAY BE SIMPLE, OR MORE COMPLICATED

When the tracheal mite invaded California, it killed 70% of our colonies. In a few years Steve Taber had selectively bred bees that were resistant to the mite, and since no one was breeding from dead colonies, soon the only bees left were those that were naturally resistant to tracheal mite.

Practical point: The “breeding population” of bees that existed when tracheal mite invaded already contained genetic alleles for traits that conferred resistance. Similar to how a single allele can confer amitraz resistance to a varroa mite, it appeared that it didn’t take much genetic change for bees to gain resistance to the tracheal mite [[5]]. Unfortunately, it appears that it takes far more than one allelic variation to confer resistance to varroa.

SELECTING FOR TOLERANCE OR RESISTANCE?

I often hear of people suggesting that we breed for “mite tolerant” bees, but is that what we really want? (Figure 2).

Fig. 2 The above bee is mite tolerant (since it is still alive). But varroa and its associated viruses extract metabolic, physiological, and immunity costs from the bees’ finite resources. I suggest selecting for mite resistance rather than tolerance.

Mite tolerance is mainly about virus resistance [[6]]. Back in the day, honey bees were virus resistant — Dr. Leslie Bailey found that it was difficult to infect a bee with viruses by mouth. And when varroa first infested our colonies, our bees were amazingly mite tolerant, since you could allow the mite infestation rate to climb to sky-high levels, and then save the colony in fall with a single treatment with Apistan. But then the viruses evolved, using varroa as their vector.

We still see some colonies that can tolerate high mite levels, with healthy brood. But although they don’t die, their bees are not going to be as healthy as bees that haven’t had their fat bodies sucked out by a mite.

Practical application: When I first started seriously breeding for mite resistance, my sons asked me whether we should include the queens of mite-tolerant colonies (lots of mites, but no signs of virus) as breeders. But since those colonies did not control their mite population, that would have maintained non-resistant genetics in our breeding population, so I decided “no.” On the other hand, mite-resistant colonies keep mites to a nearly undetectable level, so don’t have any more virus issues than our bees did back before varroa invaded (no one even talked back then about any viruses other than sacbrood).

This brings up the obvious question:

HOW DOES VARROA’S NATIVE HOST DEAL WITH THE MITE?

I’ve written about this previously [[7]], and we are not completely clear on the traits they use, [[8]] but in short, Apis cerana wrote the rules for its host-parasite relationship with the mite, focusing upon limiting the mite’s opportunities for reproduction. As best I can tell, here is the contract:

Don’t waste your time trying to reproduce in our worker brood—our larvae will sacrifice themselves if you do. (So far, we haven’t seen this to any extent in Apis mellifera.)
We’ll allow you to reproduce solely in our drone brood, under the condition that we’ll monitor for any signs of stress to that pupa (through a sniffing hole in the capping). Harm that pupa, and we will entomb you in a waxen grave. (Apis mellifera has yet to evolve a sniffing hole, and generally doesn’t uncap infested drone brood, but does remove sick worker brood.)
And we’ll further restrict your reproduction by rearing drones only from time to time. (Apis mellifera unfortunately rears drones so long as there is a pollen flow going on.)
At all times we will make your life miserable by aggressive grooming and biting. (Despite this, some mites survive for a long time in Apis cerana colonies.)

Apis cerana likely also uses additional traits to keep varroa in check as a minor parasite, and some tools that our bees don’t have in their arsenal. But there‘s more than one way to skin a cat, and Apis mellifera has come up with other tricks — notably forms of disruption of mite reproduction (uncapping), “suppression of mite reproduction” (SMR)[[9]], or hygienic removal of infested pupae (VSH), apparently triggered by “sacrifice me” signals from the pupae.

SOCIAL APOPTOSIS

In societal species (notably eusocial insects with a soldier caste), individuals can sacrifice themselves for the benefit of the colony as a whole (as a honey bee does when it stings in defense of the hive) [[10]]. This is analogous to the altruistic death of the cells that line your gut in response to getting infected by a pathogen (they sacrifice themselves for the good of your body before the pathogen can reproduce). In Apis cerana, the worker larvae appear to do this in response to a protein in varroa saliva [[11]]; and drone pupae by emitting a scent signal.

Hypothetical example: Imagine a strain of honey bees in which any worker that detected a mite on its body immediately flew away not to return [[12]]: varroa would be unable to exist in that colony.

The second-best choice would be to limit the mite’s ability to reproduce (varroa’s Achilles’ heel). The worker larvae of mite-resistant strains of Apis mellifera may not directly commit altruistic suicide, but appear to instead emit a “sacrifice me” pheromone that triggers uncapping and perhaps hygienic removal [[13]].

RE-EVALUATING SELECTIVE BREEDING STRATEGIES

Over the years, beekeepers and researchers have tried a variety of different approaches to selectively breed for mite-resistant, mite-tolerant, or “survivor” bloodlines [[14]^{, [15], [16], [17]. [18]}]. The simplest is to copy the ruthless hand of Mother Nature by letting the bees and mites duke it out and then breeding from the queens of any colonies that survive [[19]].

There are a few problems with this method:

If you don’t start with a breeding population that already carries some alleles for resistance, your entire population may die (as happened on Santa Cruz Island [[20]]).
Or you may wind up with bees that maintain small colonies and swarm a lot, as happened in the Gottland program [[21]].
Or the resulting stock may be more “spicy” than desired (notably with Africanized stock).
But the main problem is that so few colonies may survive the initial vetting process, that you don’t have enough colonies left to continue the program (or stay in business).

This method has worked informally in South Africa [[22]], and Cuba [[23]], and more formally in John Kefuss’ operation in France (Kefuss coined the name “The Bond Method” after the James Bond movie “Live and Let Die”). It’s also having results in many feral populations, and with some of the “treatment free” beekeepers who catch feral swarms.

Practical application: The Bond Method is very simple, and can work, but results in the unnecessary death of many colonies. There is no need to punish colonies with an ugly death (and to lose most of your stock each year) when all you need to do is to demote and replace the queens of non-resistant colonies.

PROGRAMS SELECTING FOR SPECIFIC TRAITS

Rather than the live-and-let-die approach, others have selected for specific traits observed to be associated mite resistance:

Varroa-sensitive hygiene (VSH) — bees that remove mite-infested pupae
Mite non-reproduction (MNR or SMR) — proportion of mites that fail to reproduce (Harbo assay, Figure 3)
Hygienic behavior — freeze-kill or unhealthy brood odor (UBO) assays for pupal removal
Grooming behavior — mite-biting assays, typically on sticky boards
Uncapping-recapping behavior — using Veet waxing strips
Post-capping stage duration — effective for Apis capensis
Marker-assisted selection — screening for genetic markers associated with resistance [[24]]

Fig. 3 Some proxies for mite resistance are time-consuming and tedious to perform, and select only for a specific trait, not for a colony’s overall success at resisting mite buildup.

There’s a problem with selecting for only a single, specific trait — no one trait alone is likely to dependably confer resistance. Resistant colonies hit the mite from several directions, and the more different blows, weapons, and techniques in their genome, the better!

REVIEWS OF THE SUCCESS OF SELECTION METHODS

Over the past few years, a number of research groups have come to the conclusion that taking the shortcut of using the assessment of only a single trait as a proxy for varroa resistance has had poor success.

As stated by Eynard [[25]]:

A recent extensive review of varroa resistance traits has revealed that there is no clear evidence for significant correlations between the standard traits measured as proxies for varroa resistance, such as VSH, grooming, MNR, hygienic behaviour or recapping behaviour.

Practical consideration: By using a specific trait as your selection criterion, you are limiting the bees as to the method to use to fight the mite! Although a right punch may be effective as a first blow, it takes a combination of techniques to win a long-term fight. Single-trait selection also means that you could easily overlook other promising traits that random mutation may come up with, such as the molecular warfare exhibited in some European mite-resistant bloodlines [[26]].

And as eloquently concluded by Sprau [[27]]:

Our results of a 4-year selection program clearly demonstrate that the currently suggested evaluation methods for important traits of Varroa resistance are highly variable, extremely time-consuming, and thus hardly feasible under practical conditions of animal husbandry. Successfully selected traits for MNR and VSH are not stable in subsequent generations as recombination between polygenic traits, mutations, mating behavior, and environmental influences introduce variability. Therefore, evaluation of every generation is critical to identify colonies with the desired traits. Considering these limitations, we fundamentally and critically question the current rationale and methodology of selecting and breeding for Varroa-resistant colonies.

COULD THERE BE A BETTER WAY TO RUN A SELECTIVE BREEDING PROGRAM?

Based upon the results of our own breeding program, I believe there is (Figure 3).

Fig. 4 Lots of teams have tried shortcuts to select for mite resistance. But there may not be any. Better to simply measure whether they get the job done!

Ask yourself two simple questions (to be answered further down):

Which selection criterion is most informative?
Which assessment method takes the least amount of time and effort?

SELECT FOR THE OBJECTIVE, RATHER THAN THE MEANS TO THE END

In a previous life, I built houses for many years. Our clients didn’t care what methods, tools, or tricks we used to do the job, but rather that we got the job done! Similar to building a house, there are a lot of different things involved in mite resistance (Eynard [[28]] found some 60 different genetic “associations” that may be involved in resistant bloodlines).

As stated by Guichard [[29]], regarding the improvement of selection strategies:

Given that the traits described as conferring resistance to honey bee colonies may exhibit regionally variable efficiency in terms of improving survival or at least in limiting colony, an a priori choice of resistance traits to be selected in any given susceptible population is a hazardous strategy to obtain surviving colonies … A further challenge to choosing relevant traits to be selected is that in all naturally-resistant populations investigated to date, various combinations of resistance mechanisms are thought to contribute to colony survival … Thus, a limited response to selection is expected when only one trait is selected, which seems to be a common procedure in current selection programs.

Not only are some of the proxies (such as brood inspection for the Harbo Assay, microscopic inspection of mauled mites on sticky boards, testing for hygienic behavior, counted recapped cells, or genetic analysis) time consuming and sometimes costly, using them doesn’t necessarily lead to producing a resistant stock that is workable and productive.

Practical application: So rather than selecting for an associated proxy (such as VSH or mite biting), just select for bees that “get the job done.” In a selective breeding program to solve a problem (colony death due to the varroa-virus complex), the question to ask ourselves is: Do we want to use indirect selection (selecting for a particular criterion that we think may be associated with solving the problem), or instead use direct selection of colonies that keep their mites under control by whatever means?

ARS researchers Harbo and Harris laid this out clear back in 1999 [[30]]

We define mite resistance as the ability of a colony of honey bees to impede the growth of a population of varroa.
With this definition, a highly resistant colony of bees would cause a mite population to decline and then to either disappear or be maintained at a very low level. This is the breeding objective.
By comparing the growth of mite populations in each colony, one can determine which bees are more resistant to mites.
There is no need for colonies to die in a breeding program that selects bees for resistance to mites.

Twenty-five years later, Eynard’s group reports [[31]]:

An important experimental aspect that limits varroa resistance studies, including this one, is that most of the currently applied measures of varroa resistance are difficult to scale to a large number of samples. For example, they can involve tedious and potentially subjective scoring, induced or artificial varroa infestation, multiple measures in time, estimation of ratios and applying heuristic thresholds for minimum detection. Nevertheless, direct estimates of varroa infestation remain the simplest traits to score for quantifying varroa resistance. Indeed, a low varroa infestation can be explained by multiple phenomena either linked to the environment, to beekeeping practices, to varroa biology or to a property of the honeybee colony itself. Boldface mine].

Back to the two questions that I began this section with:

Which selection criterion is most informative?
Which assessment method takes the least amount of time and effort?

The answer to each is to perform mite wash counts, and breed from the mothers of those colonies that control varroa on their own.

NEXT MONTH

I’ll dive into what I’ve learned about selective breeding of honey bees, and my suggestions for how to run a program.

CITATIONS AND NOTES

[1] A New Era in Mite Management. ABJ June 2025

[2] Lamas, Z, et al. (2025) Viruses and vectors tied to honey bee colony losses. https://doi.org/10.1101/2025.05.28.656706

[3] White, GF (1919) Nosema-Disease. USDA Bulletin No. 780.

[4] Shrestha, M, et al. (2020) Individual-level comparisons of honey bee (Hymenoptera: Apoidea) hygienic behavior towards brood infested with Varroa destructor (Parasitiformes: Varroidae) or Tropilaelaps mercedesae (Mesostigmata: Laelapidae). Insects 11(8): 510.

[5] We still don’t know exactly what mechanisms bees use for resistance to the tracheal mite.

[6] Thaduri, S, et al. (2018) Temporal changes in the viromes of Swedish Varroa-resistant and Varroa-susceptible honeybee populations. PLoS One 13(12): e0206938.

[7] https://scientificbeekeeping.com/the-varroa-problem-part-8/ (ABJ June 2017)

https://scientificbeekeeping.com/guessing-the-future-of-varroa-part-2/ (ABJ January 2019)

[8] Grindrod, I & S Martin (2023) Varroa resistance in Apis cerana: a review. Apidologie 54(2): 14.

[9] Scaramella, N, et al. (2023) Host brood traits, independent of adult behaviors, reduce Varroa destructor mite reproduction in resistant honeybee populations. International Journal for Parasitology 53 (10): 565-571.

[10] Mullen, E & G Thompson (2015) Understanding honey bee worker self-sacrifice: a conceptual–empirical framework. Advances in insect physiology 48: 325-354.

[11] Zhang, Y & R Han (2019) Insight into the salivary secretome of Varroa destructor and salivary toxicity to Apis cerana. Journal of Economic Entomology 112(2): 505-514. It may be direct self-sacrifice, or the larvae may emit a “remove me” pheromone.

[12] Rueppell, O, et al. (2010) Altruistic self‐removal of health‐compromised honey bee workers from their hive. Journal of evolutionary biology 23(7): 1538-1546.

[13] Zhao, H, et al. (2015) Behavioral responses of Apis mellifera adult workers to odors from healthy brood and diseased brood. Sociobiology 62(4): 564-570.

[14] van Alphen, J & B Fernhout (2020) Natural selection, selective breeding, and the evolution of resistance of honeybees (Apis mellifera) against Varroa. Zoological Letters 6(1): 6.

[15] Guichard, M, et al. (2023) Prospects, challenges and perspectives in harnessing natural selection to solve the ‘varroa problem’ of honey bees. Evolutionary Applications 16(3): 593-608.

[16] Guzman-Novoa E, et al. (2024) Honey bee populations surviving Varroa destructor parasitism in Latin America and their mechanisms of resistance. Front. Ecol. Evol. 12:1434490.

[17] van Alphen, J (2025) The downside of selection: a forgotten cause of honeybee decline. Archives of Microbiology & Immunology 9(1).

[18] Lin, Z, et al. (2025) Honey Bee Breeding and Breed: Advancements, Challenges, and Prospects. Animal Research and One Health.

[19] Locke, B. (2016) Natural Varroa mite-surviving Apis mellifera honeybee populations. Apidologie 47: 467-482.

[20] Wenner, A, et al. (2009) Biological control and eradication of feral honey bee colonies on Santa Cruz Island, California: A summary. In Proceedings of the 7th California Islands symposium (pp. 327-335). Arcata, C. A: Institute for Wildlife Studies.

[21] Locke, B & I Fries (2011) Characteristics of honey bee colonies (Apis mellifera) in Sweden surviving Varroa destructor infestation. Apidologie 42 (4): 533-542.

[22] Allsopp, M. (2006) Analysis of Varroa destructor infestation of southern African honeybee populations. University of Pretoria (South Africa).

[23] Luis, A, et al. (2022) Recapping and mite removal behaviour in Cuba: Home to the world’s largest population of Varroa-resistant European honeybees. Scientific Reports 12(1): 15597.

[24] Kaskinova, M, et al. (2020) Genetic markers for the resistance of honey bee to Varroa destructor. Vavilov Journal of Genetics and Breeding 24(8): 853.

[25] Eynard, S, et al. (2024) Sequence-based genome-wide association studies reveal the polygenic architecture of Varroa destructor resistance in Western honey bees Apis mellifera. bioRxiv 2024-02.

[26] Conlon, B, et al. (2019) A gene for resistance to the Varroa mite (Acari) in honey bee (Apis mellifera) pupae. Molecular Ecology 28(12): 2958-2966.

[27] Sprau, L, et al. (2024) The selection traits of mite non-reproduction (MNR) and Varroa sensitive hygiene (VSH) show high variance in subsequent generations and require intensive time investment to evaluate. Apidologie 55(5): 68.

[28] Eynard, S, et al. (2024) Op. cit.

[29] Guichard, M, et al. (2020) Advances and perspectives in selecting resistance traits against the parasitic mite Varroa destructor in honey bees. Genetics Selection Evolution 52: 1-22.

[30] Harbo, J & J Harris (1999) Selecting honey bees for resistance to Varroa jacobsoni. Apidology 30 (2-3): 183-196.

[31] Eynard, S, et al. (2024) Op. cit.

Category: Topics

ScientificBeekeeping.com

Selective Breeding for Mite Resistance: Part 1

Contents

Selective Breeding for Mite Resistance

Part 1

Randy Oliver