A balanced and detailed look at the science surrounding the issue of non-treatment of honey bees, including a comparison between the two main approaches: selective breeding and using survivor stock.
Hello and welcome to the Walrus and the Honey Bee, a podcast based on Steve Donohoe's blog of the same name. Steve is a UK based small scale bee farmer who produces honey for sale and has a passion for raising great queens. He is assisted by his son Alex, the Mole.
We've been having some lovely sunny spring weather which has been great for the bees. Steve has already started his first cell builder, and will be grafting next week. This, of course, is guaranteed to make it rain for a month so that queen mating is impossible. But you never know, we might get lucky. You have to be in it to win it.
This episode is an article by Kirsty Stainton called Varroa: Why Treat? Kirsty is the author of the book titled Varroa Management, published by Northern bee Books. It's a great guide to all of the available treatments for varroa mites in the UK. She is a scientist with a PhD in Insect Molecular Genetics from Oxford University. She is also the technical manager for the National Bee Unit here in the UK - they are the folks responsible for honey bee health across the country.
Okay, so here goes. Buckle up and listen to this.
The arrival of Varroa destructor to the UK in 1992 changed the practice of beekeeping forever. In order to prevent colonies from succumbing to the combined effects of the parasite and the virus it vectors called deformed wing virus, beekeepers and bee farmers must continually monitor and treat, or otherwise manage colonies to reduce mite populations. This adds extra expense and time to the beekeeping process, and many are concerned about the impacts of chemical treatments on the colonies and the environment. Concerns about chemical residues and potential off-target effects on colony health have led some to abandon chemical treatments altogether.
When varroa was introduced into the UK, the first treatments used were based on synthetic pesticide chemistries. The mites quickly become resistant to synthetic pesticides (especially tau-fluvalinate and flumethrin) and some synthetic pesticides accumulated in beeswax with unknown repercussions for the colony.
Although we now have varroa treatments based on non-synthetic chemistries such as thymol, oxalic and formic acid, these are not always differentiated from the synthetic pesticides and so suffer from the same bad reputation. This is despite the fact that, to date, there is no evidence of varroa becoming resistant to their active ingredients. The organic acids, formic acid and oxalic acid, occur naturally in the hive at low concentrations owing to their natural presence in some plants. The excess dissipates quickly after treatment.
In addition to the view that it is better for bees and the environment if chemical treatments are avoided, is the idea that our overreliance on these products will prevent us from developing mite resistant honey bee strains that do not need chemical intervention. This can be done in one of two ways: selective breeding for resistance traits or breeding from honey bees that have survived without mite treatments.
How will bees ever become resistant to mites if we continue to use chemical treatments?
If we continue to apply chemical treatments to colonies, the bees have little chance of becoming resistant to the mites. This is because we remove the selective pressure for them to do so by dealing with the problem ourselves. Without treatment, some fraction of honey bee colonies may survive. Assuming the surviving population is viable, resistance can develop if the survivors possess beneficial alleles that contribute to the fitness benefit of surviving the parasite and associated viruses; they are then passed on from parent to offspring. The word ‘allele’ is used to refer to different versions of a gene.
However, we cannot assume after withdrawing varroa treatments that the population will be viable or that a stable host-parasite relationship will develop. Extinction of a population can occur, especially where non-native organisms are concerned. Data has repeatedly shown high overwintering colony losses in European honey bees when treatments and management techniques against the mite are withdrawn.
There have been many attempts at creating resistant honey bee strains that are manageable and productive and do not require chemical treatments. An excellent review published in 2020 evaluates 153 peer-reviewed studies of mite resistant honey bee strains developed through selective breeding or survival of non-treated bees between 1980 and 2020. In the study, they describe nine strains developed using selective breeding and eleven strains that were derived from colonies surviving non-treatment. What were the factors involved in these successes?
Natural resistance
The honey bee species Apis cerana (the Asian honey bee and original host to varroa) has natural resistance to varroa; they have co-evolved for hundreds of thousands, maybe millions, of years. The reason for their resistance is the presence of naturally existing traits such as a shorter post-capping period, entombment of infested drone brood, increased hygienic and grooming behaviour and reduced mite fertility in worker brood.
Only a few strains of Apis mellifera (European honey bee) possess enough natural resistance traits to allow them to resist infestation without high colony losses. Among them are Apis mellifera scutellata and Apis mellifera capensis (both honey bee subspecies found in southern Africa), Apis mellifera intermissa (which is native to north Africa) and Africanised honey bees that are derived from scutellata.
Caution should be used when drawing parallels with these examples and resistance to varroa in UK honey bees (a mongrel mix of Apis mellifera subspecies), as those strains naturally possess more of the traits observed in Apis cerana that protect against mite infestation, such as a shorter worker post-capping period, increased levels of grooming and hygienic behaviours. The longer post-capping period in workers found in British bees increases varroa reproductive success. They also have lower levels of grooming and hygienic behaviour, although at least these traits can be enriched through selective breeding.
Surviving non treatment
Outside of these naturally resistant examples, there are a number of well-known strains that have arisen from susceptible Apis mellifera populations that managed to survive varroa without intervention. Some of these have been discovered ‘in the wild’, such as one from feral colonies in France and another from feral colonies in the Arnot forest in the USA, while others involve breeding from surviving colonies in experimental set-ups.
One major problem with the ‘breeding from what you’ve found’ technique is that you can end up with small, unproductive colonies. For example, the feral Arnot bees were found to have smaller nests and higher levels of swarming than managed colonies. This seems to be an obvious outcome, as swarming, in addition to being reproductive, can be seen as an avoidance behaviour. It results in a swarm with a reduced mite burden when it founds its new colony that has left behind a colony that benefits from a brood break disrupting mite reproduction. Conversely, the population from France derived from surviving feral bees is found to actively suppress mite reproduction using hygienic behaviours. Mite infestation rates are three times lower than in untreated controls, however the honey yields were over 40% less than that of treated control colonies.
Deformed wing virus
We cannot talk about varroa resistance without talking about deformed wing virus. Varroa parasitisation weakens bees and causes various sub-lethal effects, but it’s the heavy virus infection that kills them and eventually the colony. Deformed wing virus plays a role in some of the populations surviving without treatment. One example from Fernando de Noronha in Brazil shows that honey bee colonies there can survive extremely high mite loads of 18% to 20%, even up to 42%, of worker brood infestation — but this is due to an absence of any virulent strains of deformed wing virus circulating in the population. Without the virulent, dominant strains of the virus commonly found in Europe, honey bees seem to be able to survive with high mite levels, although there are only a few examples of this scenario.
We have virulent strains of deformed wing virus circulating in the UK, but another example of honey bees developing resistance due to avirulent deforemed wing virus strains arose in Swindon, of all places. Avirulent is a cool word for a pathogen which does not cause disease, or causes very mild symptoms. By breeding from colonies of natural mite survivors in the apiary, Ron Hoskins initiated a closed breeding programme from colonies surviving varroa and created an isolated population that was not treated and purportedly had hygienic behaviours that conferred resistance. The properties of this population were tested and it was found that the predominant virus in circulation was a unique, avirulent B variety of deformed wing virus. Mite infestation numbers were not reported and sadly, no further data are available since this 2016 study.
It is hard to reproduce this result as it is not known how this unique situation came about or how we might manipulate virus strain types to our benefit. This mechanism does not give any insight into a reproducible method to help develop resistant bees.
Tolerance versus resistance
One further problem with an unselective method is that you can end up with tolerant bees rather than resistant bees. The difference is that resistance is the ability for the bees to suppress mite reproduction, while tolerance is the ability for the bees to endure high levels of mites in the colony. This distinction is important because honey bees that can survive high levels of mites can still have high levels of deformed wing virus. A recent study suggests that this could have serious ramifications for virus spill-over into wild bee species that share an environment with honey bees.
Furthermore, resistance has a fitness cost on the parasite while tolerance does not. Tolerance can result in parasites becoming more virulent, resulting in higher parasite burden and host death, particularly among neighbouring susceptible colonies. Selecting for resistant bees is detrimental to the parasite but selecting for tolerant bees risks increasing the parasite’s virulence and/or prevalence.
Unless mite populations are monitored while selecting for ‘resistant’ honey bees, you would not know if you’ve got resistance or tolerance.
Mating isolation, inbreeding and geographic translocation
Small survivor populations can suffer deleterious effects due to inbreeding or they may only develop partial resistance due to the lack of sufficient genetic diversity to contribute resistant alleles. Strains of survivors have occurred in areas with some reproductive isolation. The famous ‘varroa resistant’ Gotland bees were created in an experiment from a group of 150 colonies (a mix of Buckfast, ligustica & carnica) placed on the isolated Swedish island of Gotland. They were whittled down to less than ten colonies after ten years without treatment. In 2014, it was found that colonies had mite loads in excess of 30 mites per 100 bees (suggesting that these bees are tolerant rather than resistant, although later papers describe that bees could inhibit mite reproduction). Yet the colonies survived the winter; although they survived as small colonies, not ideal for productive beekeeping.
Provided there is sufficient genetic diversity in a population, it can retain resistance alleles and its isolation also prevents continual mite reinfestation. When a population lacks sufficient genetic diversity, for example, if it’s very small to start with, there may be insufficient genetic resources to create a fully resistant population and it may suffer inbreeding effects. This was found to be the case in the Gotland bees. It was discovered in 2015 that the Gotland population was suffering from an extreme loss of genetic diversity and it was reported that the population needed varroacide treatment after two decades to ensure its survival. In cases such as these, it may be possible with careful selection to breed in more diversity while retaining resistance but this would not be a trivial undertaking.
Although there are a number of documented resistant strains, they are rarely available for purchase. Part of the reason for this is that during investigations of resistant strains, some strains that were resistant at one location were no longer resistant when moved to a new location. This is because the genome is not a static system; genes act in response to the environment and gene expression can change under different environmental conditions.
Local adaptation can have a surprising impact on honey bee health. One study comparing five honey bee strains at twenty locations across Europe found that non local strains performed less well than local strains when translocated to other countries. Despite their resistant properties, Gotland bees performed no better than local bees when translocated to a new location and the same was true of a resistant strain from Avignon when it was moved to another region.
Colony survival of high varroa loads can be due to (or partly due to) favourable environmental factors, rather than bees being resistant, and would therefore experience high losses upon moving. These data seem to support a system of developing resistant strains using local genetics, rather than attempting to import them.
Selective breeding and genetic markers
The high colony losses that are incurred when treatment is withdrawn is one reason that selective breeding may be a more preferable avenue for deriving resistant strains than selecting from survivors. Another is that selective breeding gives a chance for breeders to retain desirable characteristics, while selecting from ‘survivors’ often seems to result in more undesirable traits. In theory, we can select for one or more of the traits that benefit naturally resistant strains of honey bee. It would increase varroa resistance in susceptible populations to create strains with increased resistance to the mite. This has been done many times, especially selecting for increased grooming behaviour and hygienic behaviour; behaviours that might reduce mite fertility. In this way we are selecting for a trait with a known mechanism to inhibit mite reproduction and hopefully avoiding ending up with a strain that is merely tolerant to high levels of parasitism.
What would be even better would be if we can identify the behaviours at a genetic level using genetic markers to guide the selection process. The Mondet review assessed all of the genetic studies on varroa resistance and reported 159 genes had been identified by that point that are involved in honey bee resistance to mites. However, there was little overlap in the genes involved, between different resistant strains, suggesting there is no set of universal resistance genes. Unfortunately, this means that it may not be possible to create a universal set of genetic markers. Although we can quantify and measure the differences in these 159 genes, we do not fully understand how they work or interact with one another, or how they contribute to the resistance genotype.
Unfortunately, the process of selective breeding may also be hindered if working with populations that have been extensively selectively bred for large productive colonies, calmness and reduced swarming. Selective breeding for a given behaviour affects hundreds of genes rather than a single gene. Over many generations of beekeeper selection, this reduces the number of alleles in the population for the genes involved in those behaviours. This is harmful as a diverse array of alleles, which will include some number of rare alleles, is an important resource for developing full resistance, or for adapting when another pathogen or pest comes along. So a depleted number of alleles reduces the genetic resources and makes the population less robust. One solution is to introduce new genetic stock, for example by outcrossing with different strains, or by selectively breeding from a range of different strains to ensure a diverse population. All of these issues may be why only nine strains of selectively bred resistant honey bees have been reported in peer-reviewed literature over a period of 40 years.
In spite of all these constraints, it has been possible to create a resistant honey bee strain with desirable characteristics. The context for why and how it works is important. In Cuba, a 110,000 square kilometer island with a 60 year ban on honey bee imports, there are 200,000 managed honey bee colonies. All colonies are managed by 1,900 government registered beekeepers and are selected for productivity, hygienic behaviour and calmness under Cuban Center for Apicultural Research ‘Bees Selection Program’. They have been managed without varroacide treatment for 20 years. The bees are a mongrel mix of Apis mellifera mellifera, ligustica, caucasica and carnica and deformed wing virus is present. As a result of the selection process, colonies were found to have developed increased hygienic behaviour and increased removal of cells experimentally infested with mites. Varroa reproductive success reduced to 0.87 in worker cells (compared to approx 1.5 in UK bees). They are also reported to be calm and productive with an average of 45 to 70 kilograms of honey per colony.
The key reasons why this worked are:
1. mating isolation due to being on an island with a ban of imports, which allowed resistance alleles to form and spread without being diluted out
2. a large, highly-diverse starting population of multiple different strains to prevent inbreeding effects and loss of diversity
3. high compliance and co-operation from beekeepers and
4. active selection for hygienic behaviours that are shown to reduce mite fertility.
Unfortunately, as the evidence seems to suggest that we may not be able to simply import these resistant populations, we would have to create them ourselves to ensure the resistance traits are compatible in the genetic context of UK mongrel bees. The evidence suggests that a honey bee population with robust genetic mixture is a good starting point and the UK population is just such a population, but without some selective breeding, we are more likely to end up with undesirable bees.
There are various examples of untreated bees that have become varroa resistant without selective breeding but each case appears to be a biological fluke and no data exist on how to reproduce them — short of abandoning treatments and hoping for the best — and data are often incomplete or lacking on temperament and productivity. A leap of faith is not an option if your livelihood depends on your bees. Selective breeding seems to be a more viable option but must encompass a large population of honey bees and requires a high level of compliance among beekeepers within that area. A lot of resources are required for such an endeavour.
From my perspective, these are problems to be solved at the level of research programmes, organised breeding co-operatives and government policies. My options as an individual are more limited, I can:
A. continue to treat for varroa in a considered and co-ordinated way to maximise mite reduction and minimise damage to the bees or
B. stop treating and select from surviving colonies with low mite counts and other desirable traits.
If I take the second option, it may eventually give rise to a varroa resistant strain of honey bees, it will not necessarily have all the desirable traits that I am accustomed to, it may not remain resistant (as there are over 110 registered apiaries within 10 km of my apiaries) and I am almost guaranteed to cause the death of a large proportion of my colonies.
Colony loss is not a minor issue to me. In one paper I reviewed while researching for this article, I was struck by a comment about causing the death of a honey bee colony. It said “It should also be noted that the deliberate induction of a honey bee colony death is considered unethical and against the standards of animal welfare (World Organisation for Animal Health, 2018), with some countries considering it illegal (for example Germany & Switzerland)”. My motivation for treating my bees is simply so that they do not die.
Okay, that's the end of the article by Kirsty, and I'm very grateful to her for writing it. If you visit my blog post you will find a huge number of references relevant to this subject. It's very helpful to have such a comprehensive list of scientific papers on this topic.
Steve is not a big fan of the current trend across many of the beekeeping associations in the UK in the direction of going treatment free, especially for beginner beekeepers who live in areas where many other beekeepers also operate. He rotates treatments between amitraz, formic acid, thymol, and oxalic acid, to suppress mite levels in his colonies. He has great bees - very healthy and productive, gentle, and prolific. At least for now - who knows what the future holds?
Thanks everybody for visiting. Please share this with other beekeepers - spread the word so that the knowledge can help others. Have a great week, and bye for now.