BRENNEN

Industries / Food, farming and forestry / Agriculture and farming / Beekeeping

Bee health and pests

Finding and treating the diseases, parasites and pests that kill honeybee colonies, and counting the colonies lost.

7

Milestones achieved, from 1906 to 2025.

2

To watch for: announced, not yet confirmed outside.

3

Open deliverables. Nobody has shipped them.

The road

AchievedWatch forOpen

Achieved

  1. 1906

    Two foulbroods told apart

    "Foul brood", a name the naturalist Schirach gave in 1769 to rotting bee larvae, was being used for two different diseases. In a 1906 USDA bulletin, the bacteriologist Gershom Franklin White named them European foulbrood, after a bacillus Cheyne had found in England in 1885, and American foulbrood, after a new germ White had found constantly present in diseased larvae from New York State since 1903 and named Bacillus larvae (now Paenibacillus larvae). Beekeepers and inspectors still use the two names for two separate diseases.

    archive.org

    Credit

    Gershom Franklin White, USDA (Bureau of Entomology Technical Series No. 14, with a preface by E. F. Phillips)

  2. 1987

    Varroa reaches the United States

    Varroa is a mite that feeds on bees and their brood. It lived on the eastern honeybee, Apis cerana, and switched to the western honeybee that beekeepers keep in most of the world. In 2000 D. L. Anderson and J. W. Trueman showed that the mite harming western bees was a separate species, which they named Varroa destructor, and that only 2 of 18 mite lineages found on Asian bees had made the switch. In the United States it was first found on 25 September 1987 in Florida, on colonies owned by a Wisconsin beekeeper, and within two years it was in 19 states, by the American Bee Journal's account; it is now on every continent except Antarctica, by Washington State University's.

    LessonA mite carried in hives that travel for work can cross a large country within two years.

    americanbeejournal.comebi.ac.uk

    Credit

    discovery by US apiary inspectors; species named by Anderson and Trueman (Experimental and Applied Acarology, 2000)

  3. 2004

    Asian hornet lands in France

    The yellow-legged hornet, Vespa velutina, catches honeybees as they hover in front of the hive entrance. Its first European colony was recorded in 2004 near Agen in southwest France; by 2013 nests had been destroyed in Spain and Italy and hornets seen in Portugal and Belgium, according to a review by Karine Monceau and colleagues (Journal of Pest Science, 2014). A genetic study by M. Arca and colleagues (Biological Invasions, 2015) traced the likely source to the area between China's Zhejiang and Jiangsu provinces and concluded that very few, possibly only one, queen mated with several males started the invasion.

    LessonOne stowaway queen was enough to start an invasion that spread across several countries in under ten years.

    sante-agroecologie-vignoble.bordeaux-aquitaine.hub.inrae.frebi.ac.uk

    Credit

    Karine Monceau, Olivier Bonnard and Denis Thiéry (review); M. Arca, J. F. Silvain and colleagues (genetics)

  4. 2007

    Colony collapse disorder named

    In the winter of 2006/2007, US beekeepers lost colonies on a large scale. The adult worker bees disappeared quickly and no cause was known, so the syndrome was named colony collapse disorder (CCD). A team led by D. vanEngelsdorp sampled colonies in 13 apiaries of 11 beekeepers in Florida and California in January and February 2007 and measured 61 variables, including pathogens and pesticide residues. No single measure emerged as the likely cause; bees in collapsed colonies carried more pathogens, and more kinds at once, than bees in healthy ones.

    LessonLarge losses can come from several stresses acting together, so a search for a single culprit can stall.

    pmc.ncbi.nlm.nih.gov

    Credit

    vanEngelsdorp, Evans, Pettis and colleagues (PLoS ONE, 2009)

  5. 2023

    First licensed bee vaccine

    American foulbrood has no cure. Clinical cases must be reported in the United States and Canada, infected bees, hives and equipment are burned, and antibiotics only hold it back; strains of the germ resistant to oxytetracycline, the long-standard antibiotic, were verified in the United States in 2000 (Miyagi and colleagues). In January 2023 the USDA gave Dalan Animal Health a conditional license, for two years at first, for a vaccine made of killed P. larvae cells. It is mixed into the queen's feed; worker bees pass it to the queen in royal jelly, and her larvae hatch with some immunity. The Microbiologist magazine called it the world's first vaccine to protect honeybees.

    the-microbiologist.comebi.ac.uk

    Credit

    Dalan Animal Health (chief executive Annette Kleiser); made by Diamond Animal Health

  6. 2023

    Australia stops chasing varroa

    Australia was the last major beekeeping country free of varroa until June 2022, when the mite was found in surveillance hives in Newcastle, New South Wales. A 15-month eradication response failed. On 19 September 2023 the National Management Group, the national body that decides on pest emergencies, agreed that eradication was "no longer technically feasible", citing non-compliance by some beekeepers and a spike in detections over a wider area. A national program to manage the mite and slow its spread followed and ended on 6 February 2026.

    LessonUniversity of Sydney geneticist Emily Remnant pointed to almond pollination, where hives from many places gather, as "potential super-spreader events" that an eradication team could not contain.

    nre.tas.gov.auscimex.org

    Credit

    Australia's National Management Group; account by Biosecurity Tasmania

  7. 2025

    Mites shrug off amitraz

    US commercial beekeepers lost more than 60% of their colonies between summer 2024 and early 2025, about 1.7 million colonies, by the USDA's count. USDA Agricultural Research Service (ARS) scientists sampled collapsing colonies in January 2025 and found deformed wing virus A and B and acute bee paralysis virus at unusually high levels. Every varroa mite they screened from the collapsed colonies was resistant to amitraz, the miticide most US beekeepers rely on.

    LessonUS varroa had already become resistant to tau-fluvalinate and coumaphos after years of heavy use; amitraz, first allowed in US hives under a 2010 emergency permit in South Dakota, followed in about ten years, by Washington State University Extension's account.

    ars.usda.govpubs.extension.wsu.edu

    Credit

    Z. S. Lamas, Frank Rinkevich, J. D. Evans and colleagues, USDA ARS

Watch for

  1. RNA treatment for mites

    Norroa's active ingredient, vadescana, works by RNA interference: it carries double-stranded RNA that blocks a gene the mite needs, which by the company's account stops the mites from reproducing. GreenLight Biosciences says "rigorous, replicated field trials" across several US regions showed mite control lasting "up to 18 weeks". EPA registered the product on 25 September 2025 and says its effects are highly specific to varroa and do not affect bees.

    Achieved whena university or government trial run apart from GreenLight publishes mite counts and colony survival over a full season, set beside an existing treatment.

    croplife.comamericanbeejournal.com

    Claimed by

    GreenLight Biosciences (chief executive Andrey Zarur), September 2025

  2. Phone photos sort brood disease

    USDA ARS scientists at the Carl Hayden Bee Research Center in Tucson trained AI models on 2,759 photos of sick larvae, each matched to lab tests, to tell European foulbrood from viral diseases that look like it. They report accuracy of about 72% to 88% depending on the model, falling to 28% to 68% for viral cases from farm landscapes missing from the training photos. They plan a free phone app; it has not been released.

    Achieved whenthe app is released and apiary inspectors outside the team publish how often it agrees with lab diagnoses on colonies it has not seen.

    scientificdiscoveries.ars.usda.govfreshfruitportal.com

    Claimed by

    Kirk Anderson, Duan Copeland and colleagues, USDA ARS (Scientific Reports), August 2025

Open

  1. 01

    Cause of death from photos

    A public, labeled photo set of dead colonies, which beekeepers call deadouts, and a phone tool trained on it that names the likely cause of death (varroa and its viruses, starvation, queen failure, robbing or foulbrood) from photos of the frames and the hive floor.

    The brief
    Credit · unclaimed
  2. 02

    Tropilaelaps on sticky boards

    A model that reads smartphone photos of sticky boards, the sheets slid under a hive to catch falling mites, and flags Tropilaelaps mites apart from varroa and lookalikes, as an early-warning screen for countries the mite has not reached.

    The brief
    Credit · unclaimed
  3. 03

    A public amitraz resistance map

    An open form and map where US beekeepers post the results of the standard field test for amitraz resistance in their mites, by place and season, with the raw counts downloadable.

    The brief
    Credit · unclaimed

Open 01 · the brief

Cause of death from photos

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A public, labeled photo set of dead colonies, which beekeepers call deadouts, and a phone tool trained on it that names the likely cause of death (varroa and its viruses, starvation, queen failure, robbing or foulbrood) from photos of the frames and the hive floor.

Why it is still open

Beekeepers judge a deadout by eye, helped by reference galleries such as the photo guide to causes of colony death kept by AGES, Austria's agency for health and food safety. A November 2026 review of 58 machine-learning studies of colony loss found collapse modeled directly in only two and winter deaths in four, and called for open data and multi-apiary datasets with confirmed loss outcomes. Searches found no labeled set of deadout photos with confirmed causes.

ages.atresearchsquare.com

Built with what exists

Phone cameras and a fixed shot list; Label Studio for labeling; Ultralytics YOLO or a small open vision model such as Liquid AI's LFM3, fine-tuned on one graphics card (see the Note "One GPU to fine-tune"); lab results from the samples beekeepers already send to bee disease laboratories.

  1. First step

    Ask one state apiary inspection service or beekeeping association to have members photograph every colony that dies this winter with the same shot list, alongside the samples they send for testing.

  2. Done looks like

    At least 1,000 deadouts from at least 50 beekeepers in at least three countries, each with photos and a lab or inspector diagnosis, published under a CC-BY license; a model that names the confirmed cause for at least 80% of deadouts from beekeepers whose colonies were not used in training, with agreement between two inspectors reported beside it.

Open 02 · the brief

Tropilaelaps on sticky boards

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

A model that reads smartphone photos of sticky boards, the sheets slid under a hive to catch falling mites, and flags Tropilaelaps mites apart from varroa and lookalikes, as an early-warning screen for countries the mite has not reached.

Why it is still open

Tropilaelaps mercedesae, a brood mite about a third the size of varroa, has been documented on western honeybees in Georgia, Russia and Uzbekistan and is not thought to be in North America yet; it is easily confused with pollen mites, bee lice (Braula) and male varroa, and boards are read by eye under a magnifier or microscope, by the Honey Bee Health Coalition's 2024 primer. A 2026 study in Scientific Reports found bottom-board sticky traps the most sensitive way to find it, Apimondia reports. Open software for sticky-board photos, such as the University of La Rioja and University of Zaragoza model (Sensors, 2024), counts varroa only, and searches found no image model for Tropilaelaps.

honeybeehealthcoalition.orgapimondia.orgzaguan.unizar.es

Built with what exists

Smartphones with a clip-on macro lens; the open varroa-counting model from La Rioja and Zaragoza as a starting point; Ultralytics YOLO; Label Studio; microscope confirmation of each board by a bee laboratory.

  1. First step

    Write to a bee research group or national bee laboratory in a country where Tropilaelaps is present and ask for board photos with their microscope tallies, then run the open varroa model on them to see what it misses.

  2. Done looks like

    At least 500 microscope-confirmed board photos from apiaries where Tropilaelaps occurs, published openly; a model that flags at least 95% of boards carrying any Tropilaelaps, with its false-alarm rate on 500 Tropilaelaps-free boards from Europe or North America reported beside it.

Open 03 · the brief

A public amitraz resistance map

Credit · unclaimed

The first to publish is credited on this line, and the deliverable moves from Open to Achieved.

What to build

An open form and map where US beekeepers post the results of the standard field test for amitraz resistance in their mites, by place and season, with the raw counts downloadable.

Why it is still open

The Honey Bee Health Coalition and the USDA ARS lab in Baton Rouge led by Frank Rinkevich have sent test kits to beekeepers since 2019, but results go to participants and conferences, with no public data release. US-wide figures appear only as a chart given to Washington State University Extension by personal communication in March 2026, showing resistant mites rising from under 20% in 2019 to nearly 50% in 2025. A 2025 review of 74 studies found test methods vary widely and called for national monitoring programs and standard protocols. Australia is ahead: the Australian Honey Bee Industry Council reproduced a Queensland Department of Primary Industries map of resistant properties dated 4 June 2026, though it shows only government-lab-confirmed sites and the council expects more.

honeybeehealthcoalition.orgpubs.extension.wsu.eduebi.ac.ukhoneybee.org.au

Built with what exists

The screening instructions in Washington State University Extension's fact sheet FS411E, which use registered amitraz strips as labeled; KoboToolbox or ODK forms on a phone; a Leaflet or QGIS map; a public CSV under a CC-BY license.

  1. First step

    Put the FS411E steps into a phone form and seed the map with the located, dated results from the 74 studies in the 2025 review.

  2. Done looks like

    Results from at least 300 apiaries in at least 10 US states or three countries in one season, each with raw mite counts, published openly, with the share of resistant results per region compared against USDA ARS figures when those are published.