Honey Bee Genetics and Varroa-Resistant Breeding

There are approximately 2.6 million managed honey bee colonies in the United States in 2026. The queen producers who supply those colonies annually produce roughly 1 million daughter queens - enough to requeen about one-third of the nation's hives. In 2004, western US queen breeders used 218 queen mothers to produce 430,000 of those daughters. Each breeder queen produced an average of 1,972 daughters. The genetic diversity of American commercial beekeeping flows through a bottleneck narrower than most people realize, and the consequences of that narrowness show up in every annual colony loss survey.

Only nine of the more than 24 named Old World subspecies of Apis mellifera were ever introduced to the Americas. Initial importations consisted of a few tens to hundreds of queens per subspecies. Limited importations over the last 90 years have further narrowed the base. Colonies with low genetic diversity - specifically, an effective paternity frequency of 7.0 or fewer - are 2.86 times more likely to die than more genetically diverse colonies. The correlation between genetic narrowness and colony failure isn't subtle. It's nearly threefold.

Against this backdrop, a handful of breeding programs have spent decades trying to build bees that can survive what's killing them. The approaches differ. The ambitions are the same. The biology makes all of it extraordinarily difficult.

The Races

Italian bees - Apis mellifera ligustica - arrived in the United States in 1859, imported to New York City and Philadelphia from Germany and Switzerland. They replaced the German dark bee (A. m. mellifera), which had been in America for over 200 years but had developed a reputation for being, by all accounts, foul-tempered and disease-prone. The Italians were bright yellow, gentle, productive, and easy to manage. They became the foundation of American commercial beekeeping and remain it.

Carniolan bees - A. m. carnica - came from Slovenia, southern Austria, and the Balkans. First described scientifically by Pollmann in 1880. Between 1857 and the end of World War I, Slovene merchants exported at least 170,000 swarms - some estimates say 500,000 - making it one of the most active livestock export markets in European history. Carniolans are dusky brown-grey, exceptionally gentle, and build up colony populations in spring at rates that Italian bees can't match. They also swarm at rates Italian bees can't match, which is less desirable. The second most popular race in the US.

Caucasian bees - A. m. caucasica - originated in the high valleys of the Central Caucasus, with Georgia as the central homeland. Their tongues average 7.1 millimeters - the longest of any known honey bee variety - which gives them access to nectar sources that shorter-tongued races can't reach. By 1943, there were 43 Mountain Grey Caucasian breeding apiaries in the US, accounting for approximately 25% of all queens produced. Then they fell out of favor. The propolis production was excessive - they glued everything to everything - and managing hives that were essentially bonded shut with tree resin wore out beekeepers faster than any temperament issue could.

These three races - Italian, Carniolan, Caucasian - plus Africanized hybrids that arrived unbidden, constitute the working genetic palette of American beekeeping. Every breeding program operates within this palette or deliberately reaches beyond it.

The Monk Who Couldn't Stop Traveling

Karl Kehrle was born in Germany and sent at age eleven to Buckfast Abbey in Devon, England. He took the name Brother Adam, joined the Benedictine order, and began keeping bees. He would continue for the next seventy years.

The catalyzing event was the Isle of Wight disease - later identified as tracheal mite infestation - that devastated British bee populations in the early 1900s. Brother Adam noticed something that would shape the rest of his life: crosses between Italian and dark bees survived when pure dark bees died. Hybrid vigor. The mixed-race colonies had something the purebreds didn't.

In 1919, he raised a queen later designated B-1 that embodied, as he described it, "all the desirable qualities of the Ligurian and former Native in an ideal combination." B-1 became the ancestral foundation of all Buckfast bees. In 1925, he established an isolated mating station on Dartmoor - because without controlling which drones a queen mates with, breeding gains dissolve into the background population within a single generation.

Then the traveling started. From 1950 onward, Brother Adam began collecting bees from across Europe, the Near East, and North Africa. Over 37 years of expeditions, he traveled more than 100,000 miles and collected more than 1,500 queens. Algeria, Morocco, Tunisia, Tripoli, Egypt. Israel, Jordan, Syria, Lebanon, Cyprus. Greece, Turkey, Yugoslavia, northern Italy, Spain, Portugal. He incorporated A. m. cypria from Cyprus, A. m. anatoliaca from Anatolia, A. m. sahariensis from the Sahara, and A. m. monticola from East African mountains. In 1987, at the age of 89, he went to Africa searching for scutellata and monticola bees.

The Buckfast bee that emerged from seven decades of this work is not a race. It's a managed hybrid - a deliberate assemblage of genetics from at least six subspecies, selected for disease resistance, productivity, gentleness, and the ability to tolerate English weather, which is its own form of hardiness test. Brother Adam received the Order of the British Empire, the German Bundesverdienstkreuz, and honorary doctorates from Uppsala University and Exeter University. He died in 1996 at 98. The breeding program outlived every other project he could have pursued in those seventy years.

100 Queens From Russia

In the far eastern Primorsky region of Russia, European honey bees had been living alongside Varroa destructor for over a century. The mite originated in Asia, parasitizing the Asian honey bee Apis cerana. When A. mellifera was introduced to the region, varroa jumped hosts - and the European bees that survived were the ones that had developed resistance through natural selection. Not through breeding programs. Not through human intervention. Through decades of dying until the survivors had something the others didn't.

Thomas E. Rinderer, Research Leader of the USDA Honey Bee Breeding, Genetics and Physiology Research Unit in Baton Rouge, Louisiana - a position he held for nearly 40 years before retiring in January 2016 - recognized the potential. In autumn 1994, his team explored whether varroa resistance existed in Primorsky bee populations. By June 1995, they had established a test apiary in the region: 50 colonies, mite levels equalized, no treatments. Monthly data collected through September 1996.

The results led to the importation of 100 queen honey bees to the United States in 1997.

The field trial data told a clear story. In 1999 trials, Primorsky colonies averaged about half the mites found in domestic control colonies. In a 2000 Louisiana trial, Russian colonies showed a 2.5-fold mite population increase over 91 days - compared to a 17.3-fold increase predicted for domestic colonies under the same conditions. In Iowa and Mississippi, 150 Russian colonies had zero detectable varroa three months after being deliberately inoculated with approximately 100 mites each. Forty-eight had very few. The mites were there. They just weren't reproducing.

The mechanism was measurable: Russian colonies averaged 0.85 viable female mite offspring per foundress mite, versus 1.23 for Italian colonies. A difference of 0.38 offspring per reproduction cycle. It doesn't sound like much until you multiply it across thousands of reproductive cycles across thousands of cells across months of brood rearing. The exponential math of parasite reproduction means small differences in reproductive success compound into enormous differences in mite population growth.

Honey production wasn't sacrificed. Mississippi 2000 overall average: 125 pounds per colony. Selected Russian breeders averaged 185 pounds, ranging from 149 to 238 pounds.

The Russian Honeybee Breeders Association incorporated in 2007, using a barrier island in Louisiana as an isolated mating site. Fifteen members, each assigned two genetic lines within one of three blocks. Members must operate a minimum of 200 colonies. DNA testing requirement: bees must score above 0.50 on the Russian ancestry metric. Most score 0.75 or higher.

The Bees That Throw Out Their Dead

In 1995, Dr. John Harbo and Dr. Roger Hoopingarner at Michigan State University discovered a behavioral trait that would become one of the most important tools in varroa-resistant breeding: Varroa Sensitive Hygiene.

VSH bees detect and remove pupae infested by varroa mites. Not dead pupae - living pupae that happen to have mites reproducing on them. The behavior targets worker brood aged four to six days post-capping, triggered by mite oviposition or something associated with it. The bees uncap the cell, pull out the pupa, and carry it out of the hive. The mites, now homeless, lose their reproductive opportunity.

The numbers from Harbo and Dr. Jeffrey Harris at the USDA lab in Baton Rouge: VSH bees removed 91% of pupae containing reproductive mites and 58% of pupae with infertile mites. They could tell the difference. Reproductive mites - the ones actually multiplying - were removed at nearly twice the rate of non-reproductive ones. The bees were not randomly cleaning house. They were targeting the threat.

The genetic architecture is favorable: a colony needs only 50% of the VSH alleles to express high mite resistance. Queens carrying VSH genetics, mated to random drones in open mating, still produce colonies that manage varroa without chemical treatment. The trait doesn't require purity to function. It works at half strength. This is unusual and important - most desirable traits in honey bees wash out rapidly in open mating. VSH persists.

The Ankle-Biters

In 1997, Dr. Greg Hunt at Purdue University began a breeding program based on a different anti-varroa behavior: grooming. Specifically, the tendency of certain bees to use their mandibles to bite, chew, and damage varroa mites - amputating legs, puncturing body walls, generally making life on a bee's body an uncomfortable and occasionally fatal experience for the parasite.

The program ran for almost a quarter century of selective breeding. The bees became known as "ankle-biters" for obvious reasons. A gene called neurexin emerged as a top candidate indicator of grooming propensity - the same gene linked to grooming behavior in mice and social disorders in humans, which is the kind of cross-species parallel that makes geneticists simultaneously excited and cautious.

The performance data: 70% winter survival for mite-biter colonies (21 out of 30) versus 33.3% for control packaged bees (6 out of 18). When distributed to commercial beekeepers for evaluation, twice as many mite-biter colonies survived winter compared to unselected stock. That survival advantage translated to nearly four times greater honey production - because colonies that survive winter are colonies that produce honey the following season. Varroa population growth rates in mite-biter colonies ran at one-third of unselected Italian bee colonies.

Three different approaches to the same problem - Russian resistance through reproductive suppression, VSH through brood removal, Purdue through physical combat - all producing measurable results. All operating against the same constraint that limits every honey bee breeding program ever attempted.

The Mating Problem

A queen honey bee mates with 10 to 20 drones. The average effective paternity frequency is 13.6, plus or minus 6.76. A queen captures nearly 90% of her breeding population's allele potential by her tenth effective mating. She stores six to eight million sperm in her spermatheca and uses them for the rest of her reproductive life - which, at 2,000 eggs per day, is a feat of biological economy that human reproductive medicine can only observe with envy.

The mating happens in drone congregation areas - airborne gatherings where drones from as many as 200 colonies accumulate, with estimates of up to 25,000 individual drones per congregation area. Queens fly up to 16 kilometers to reach these gatherings, which increases the probability of outbreeding but makes controlled mating effectively impossible under natural conditions.

This is the fundamental challenge. Breed a perfect queen. Give her perfect genetics. Release her into the mating flight, and she'll mate with whatever drones happen to be flying that day, from whatever colonies happen to be within 16 kilometers. Any breeding gains can be diluted to near-zero in a single generation of open mating.

The solutions are limited: isolated mating stations - barrier islands, remote mountain valleys like Brother Adam's Dartmoor - where the only drones available carry the desired genetics. Or instrumental insemination.

Dr. Susan Cobey, former manager of the Harry H. Laidlaw Jr. Honey Bee Research Facility at UC Davis and later at Washington State University, spent more than three decades as the leading authority on instrumental insemination of honey bee queens. The procedure involves collecting semen from mature drones under a microscope, anesthetizing the queen with CO2, opening the abdomen with specialized tools, and depositing 8 to 10 microliters of semen past the valvefold into the oviduct. Over 24 hours, sperm migrates from oviduct to spermatheca.

It's the only reliable method to fully control honey bee mating. Without it, closed breeding populations can't be maintained. Every trait that breeders select for - varroa resistance, gentleness, productivity, disease resistance - is one uncontrolled mating flight away from regression toward the mean.

The Saskatraz Experiment

In 2004, Albert Robertson - a molecular biologist and CEO of Meadow Ridge Enterprises near Saskatoon, Saskatchewan - started a breeding program using the most straightforward method imaginable: put colonies under varroa pressure without treatments. See which ones survive. Breed the survivors. Repeat.

The Saskatraz program selects for honey production, wintering ability, temperament, tracheal mite resistance, varroa tolerance, and brood disease resistance simultaneously. Research collaboration with the University of Saskatchewan identified molecular mechanisms by which varroa suppresses honey bee immune systems - the parasite doesn't just feed on bees; it actively disables their immune response, which is why varroa-weakened colonies are vulnerable to every secondary infection that follows.

Saskatraz hybrids are produced exclusively by Olivarez Honey Bees in Northern California, using breeder queens from Saskatchewan. The geographic separation - breeding stock selected under Canadian prairie conditions, commercial production in California's Central Valley - creates an interesting question about how traits selected in one environment perform in another. The migratory beekeeping industry already moves bees across climate zones constantly. Genetics selected for Saskatchewan winters pollinating California almonds is just another version of the same displacement.

CRISPR and the Future That Isn't Here Yet

Gene editing with CRISPR-Cas9 was first used in honey bees in 2016. Recent methods have achieved 73 to 77% bi-allelic knockout mutant efficiency - a dramatic improvement from the initial 5-12%. The technique involves microinjecting guide RNA and Cas9 protein into eggs within two hours after the queen lays them.

It remains entirely in the research stage. No commercially available CRISPR-modified honey bees exist. No regulatory framework for their release has been established. The technology is a tool looking for a permitted application, and the beekeeping industry's relationship with genetically modified organisms is approximately as warm as its relationship with neonicotinoid manufacturers.

Marker-assisted selection - using genetic markers to predict traits without modifying the genome - is closer to practical application. A 2025 review described it as "poised to usher in a new era of accelerated and precise genetic improvement." A 2024 study combined AI-driven protein structure modeling with marker-assisted selection, focusing on vitellogenin - the same protein that controls lifespan in winter bees and hypopharyngeal gland development in nurse bees - as a selection marker.

The technology moves faster than the breeding. The breeding moves faster than the regulation. And the mites move faster than all of it. The 2024-2025 colony loss survey recorded 55.6% total annual losses - the highest ever documented since tracking began. The 14-year running average is 41.4%. The rate beekeepers consider acceptable is roughly 19%.

Two hundred eighteen queen mothers produced 430,000 daughters. The genetic diversity of the American honey bee population flows through a keyhole. The programs that are trying to widen it - Russian imports, VSH selection, Purdue's ankle-biters, Saskatraz survivors, Cobey's insemination work - are each pushing against a mating biology that evolved to maximize outbreeding and a commercial industry that inadvertently minimized it.

Brother Adam walked 100,000 miles looking for the right genes. The right genes, it turns out, were always there. The challenge was never finding them. It was keeping them.