The Mite a Certificate Can’t Stop

A parasite that could break commercial beekeeping is moving toward North America. Canada’s main line of defence is a certificate that cannot do the job.

Canada’s canola, blueberries, tree fruit and seed crops all depend on honey bee pollination. That work rests on a healthy beekeeping industry, and that industry is protected from a mite called Tropilaelaps by one thing: an exporting country tests its bees, does not find the mite, and certifies them clean. The Canadian Food Inspection Agency treats that certificate as proof the mite is absent. It is not. It means only that the mite was not found.

Tropilaelaps is a mite of honey bee brood, native to Asia. It is worse than Varroa, the parasite beekeepers already fight every year. It breeds faster, does more damage, and our bees have no natural defence against it. It is not in North America yet. But it is moving west: established in Russia since 2021, and confirmed in Georgia and Uzbekistan in 2024. Britain and Canada have already suspended bee imports from Ukraine because of it.

The weakness is the certificate itself. A country can honestly say it has not detected the mite while the mite is there. Three things make that gap real.

First, the tests do not work. The standard checks wash or shake adult bees, which works well for Varroa. But Tropilaelaps lives in the brood, not on the adults. A colony can be five percent infested and still pass. An honest, well-run inspection can find nothing and certify an infested apiary.

Second, the mite survives the trip. It lives longer on dead bees than live ones, and feeds on them. Every package of bees arrives with dead bees in the bottom of the box. In 2025 a broodless swarm was found on a ship bound for New Jersey still carrying 28 Tropilaelaps after weeks at sea. Transit is not a barrier.

Third, the history runs one way. Every major honey bee pest of the past forty years was found only after it had already settled in. Australia ran the best bee surveillance in the world and still caught Varroa too late to remove it. New Zealand did the same, and put the cost in the hundreds of millions of dollars. Once these mites establish, no country with both wild and managed bees has ever cleared them out.

There is a precedent for what does work. In 1922 the United States closed its border to live bee imports to keep another mite out. Sixty-two years passed before that mite turned up, and when it did it had walked overland from Mexico, not come through the certified trade. Closing the one route you can control buys time.

For Canada the cost of doing this is small. Beekeepers replace most of their winter losses themselves, by splitting their own hives. Bees brought in from outside North America are a small part of the supply. They are also the part that carries the risk.

So the ask is simple. Treat a certificate of “not detected” as what it is: a record that nobody found the mite, not proof it is not there. Reopen the import risk assessment. And pause live bee imports from outside North America until surveillance can actually find this mite before it arrives.

If Tropilaelaps establishes here, the review afterward will ask what we should have done. We already know the answer. The time to act is while the mite is still on the other side of the ocean.

For our detailed report click here

---

The full position paper

Absence of evidence is not evidence of absence. The case for precautionary restriction of live honey bee imports against Tropilaelaps mercedesae. Canadian Beekeepers Federation, 13 July 2026.

A note on what is known, and how fast it is changing. Every factual claim in this paper is accurate to the best of the available evidence as of the date above. That qualification is not a formality. The science of Tropilaelaps is moving quickly, and much of what was confidently believed only a few years ago has since proven false: that broodless consignments pose no transmission risk, that the mite survives only a few days off brood, that adult-bee wash tests detect it as they detect Varroa. Each of these has been overturned by work published in the last two years, and in every case the correction has run the same way. The threat is larger, the mite hardier, and detection weaker, than previously assumed.

There is no basis for assuming this trend has finished. The precautionary case set out here rests precisely on that uncertainty: decisions must be made now, on the evidence now in hand, and cannot wait for a settled science that may arrive only after the mite does.

1. Summary

Canada’s protection against Tropilaelaps mercedesae rests on a single mechanism: an exporting country tests its bees, does not find the mite, and certifies accordingly. The Canadian Food Inspection Agency has confirmed this, stating on 27 November 2025 that there is “no scientific evidence to suggest Canada should ban imports of honey bees from outside North America,” and that exporting countries “must certify they’re free of the mite.”

The unfounded assumption inherent in that phrase is the subject of this paper. What an exporting country can honestly certify is that Tropilaelaps has not been detected. CFIA treats that as a certificate that the mite is absent. These are not the same.

The certification mechanisms used now are not up to the task of stopping Tropilaelaps, for three independent reasons.

1. The tests do not work. The adult-bee screening methods on which apiary inspection regimes universally rely, alcohol wash and powdered sugar shake, have low sensitivity for Tropilaelaps. A colony can carry roughly a 5% brood infestation and still return a “not detected” result.
2. The mite survives the journey, on the dead bees in the box. Gill et al. (2024) found Tropilaelaps survives up to eight days on dead adult bees, roughly three times longer than on live ones, feeding on the exudate of decomposition. Every commercial package arrives with a layer of dead bees on its floor. In 2025 a living, broodless Apis dorsata swarm intercepted on a ship bound for New Jersey, after an estimated 16 to 99 days at sea with no brood, was still carrying 28 Tropilaelaps.
3. Non-detection has repeatedly turned out to mean “present but unseen.” Every major honey bee pest incursion of the past forty years was detected only after the pest had established. There is no reason to believe Italy, Australia or New Zealand know more about Tropilaelaps today than Australia knew about Varroa in May 2022.

Against this, there is one strong precedent for what prohibition achieves. The United States Honeybee Act of 1922 closed the legal live-bee import channel. Sixty-two years passed before the tracheal mite the Act was written to exclude was found in an American hive. When it was, it had come overland from Mexico, not through the certified trade channel.

The 1922 Act was passed on the basis of science that later proved wrong. It still bought North American beekeeping sixty-two years. CFIA’s insistence on conclusive scientific evidence before acting inverts the very logic that has protected this industry for a century.

2. What is at stake

Tropilaelaps mercedesae is a parasitic mite of honey bee brood, native to Asia, where it has adapted from giant honey bees to the western honey bee, Apis mellifera. It is a WOAH-listed notifiable pest. It is not present in North America.

2.1 It is worse than Varroa

• Reproduction. A shorter life cycle, no dependence on a phoretic adult phase, and a larger share of mites reproducing each cycle. In one Thai study of co-infested colonies, worker-brood infestation ran at 19.9% for Tropilaelaps against 0.7% for Varroa in the same hives (Khongphinitbunjong et al. 2016).
• Feeding damage. It feeds on both pre- and post-capped brood, opening multiple wound sites per larva where Varroa opens one or two. Surviving adults emerge undersized and deformed.
• Virus load. A competent vector of deformed wing virus (Forsgren et al. 2009).
• Host defence. Apis mellifera lacks the grooming and biting defences Apis dorsata deploys against this mite.

2.2 Colony outcomes in the field

The UK’s Animal and Plant Health Agency rates the economic impact of establishment as “large,” with high confidence. For a Canadian industry underpinning canola, blueberry, tree fruit and forage seed pollination, the equivalent assessment is not seriously in doubt.

2.3 It is moving toward us

• Established in Russia (Krasnodar and Rostov) since at least summer 2021; also detected in Tyumen, western Siberia.
• Confirmed in Georgia (2024) and Uzbekistan (2024). Suspected in Tajikistan, Iran and Azerbaijan.
• Great Britain suspended honey bee and apiculture imports from Ukraine on 7 October 2025. Canada banned Ukrainian imports in 2024.

3. The first failure: the tests do not detect this mite

Import certification is only as good as the diagnostic method behind it. For Tropilaelaps, that method is inadequate, and this is now quantified. A 2026 study in Scientific Reports compared the sensitivity of standard monitoring methods across 26 infested Apis mellifera colonies.

“Unlike for V. destructor, adult bee-based methods (alcohol wash and powdered sugar shake) had low sensitivity to detect T. mercedesae… further research is needed to develop sensitive detection methods that can be employed on a broad scale.” — Scientific Reports, 2026

Tropilaelaps lives and reproduces in the brood; it does not spend a settled phoretic phase on adult bees the way Varroa does. An inspection regime built around shaking adult bees is looking in the wrong place. An exporting country can run a diligent, honest, well-resourced inspection programme using internationally accepted methods, find nothing, and certify an infested apiary.

This is the crux. CFIA is not relying on a bad-faith certifier. It is relying on a good-faith certifier using a test that the peer-reviewed literature says does not work for this pest. The certificate can be entirely truthful and entirely worthless at the same time, because what it truthfully records is a failure to detect, not an absence.

4. The mite survives the journey, including in a box of dead bees

4.1 The package bee graveyard

Gill and colleagues (Chiang Mai, 2024) ran survival assays using situations the mite would actually encounter in a shipment. The results invert the conventional wisdom.

The mite survives roughly four times longer on dead bees than on live ones. Gill’s team observed the mites spending most of their time on the corpses and apparently feeding on the exudate produced by decomposition. Dead bees are not a hostile environment for this mite. They are a food source.

Every package carries a layer of dead bees on the floor of the box: ordinary transit mortality, present in every consignment, expected and unremarkable. Gill’s work establishes that this layer is a survival substrate capable of carrying Tropilaelaps through the entire shipping period.

“This finding suggests that T. mercedesae could be transported in scenarios where live A. mellifera brood and adults are not present, such as in used beekeeping equipment containing decaying brood and bees, or in queen shipments and packages where bees have died during the caging / packaging and transportation process.” — Gill et al., 2024

Gill’s paper notes that it “has been assumed that broodless colonies and colonies that have died at sea do not pose a transmission risk for Tropilaelaps spp. (EFSA, 2013).” That assumption is what allows a broodless consignment to be treated as low-risk. It is now known to be wrong.

4.2 A broodless swarm carried the mite for weeks

In July 2025, a living bivouac of several hundred adult Apis dorsata dorsata was found aboard a cargo vessel inbound to Port Elizabeth, New Jersey. There was no comb and no brood. The crew contained the swarm and froze it aboard ship at −6 °C; on later inspection, 28 Tropilaelaps mercedesae were recovered and confirmed by morphology and DNA barcoding.

CO1 phylogenetics placed the swarm in a southern India lineage. On the establishment scenario the genetics support, the swarm had been aboard, alive and broodless, for between 29 and 99 days. Even the most conservative alternative gives 16 to 18 days. The relevant number is therefore not “a few days” but weeks.

4.3 The package is the pathway, and there is no way to check it

• It arrives with its own dead-bee substrate. That layer will keep Tropilaelaps alive for the length of the journey, longer, in fact, than the live bees will.
• The survival window comfortably exceeds the shipping period. A package moves from an offshore apiary to a Canadian one in days.
• There is no practical way to certify a given package free of the mite. A live package cannot be alcohol-washed or brood-uncapped without destroying the product being sold, and those are precisely the tests that fail on Tropilaelaps in any case.

This is the mechanism by which live mites could enter the country. Not a stowaway swarm on a ship, but an approved, sealed, unremarkable package of live bees, carrying live Tropilaelaps on the dead ones at the bottom.

4.4 Why molecular testing does not provide a reliable answer

PCR detects DNA, not live mites. A clean PCR result cannot certify a package free of the mite: a cryptic, low-level, brood-dwelling parasite is easily missed by a sub-sample, so a negative describes what was tested, not the box. And live bees cannot be held indefinitely awaiting a laboratory result.

In the other direction, Tropilaelaps DNA persists in wax and comb long after the mite that shed it is dead, generating positives that overstate the live-mite risk. Molecular testing detects genetic material, not live infestation.

5. The second failure: the surveillance record

If a certificate of non-detection were a reliable proxy for absence, the historical record would show pests being intercepted at borders, or at least detected before establishment. It shows the opposite.

5.1 What the pattern shows

1. Detection lags introduction, and by how much is never precisely known. Up to five years in New Zealand; three years in Russia; roughly a decade for Nosema ceranae. Every one of these figures is an inference drawn after the fact. Nobody has measured the lag directly, because measuring it would have required detecting the pest, which is the thing that failed.
2. Non-detection is recorded; absence is inferred. A country reports “no detections.” An export health certificate treats that as “no mites.”
3. Even a detection does not ensure eradication. There is no example anywhere of a successful eradication of an established Varroa or Tropilaelaps population in a country with contiguous managed and feral bee populations.

Canada has no sentinel apiary network for Tropilaelaps at all. Australia had the best in the world, detected Varroa only after establishment, and still lost. Already struggling with Varroa, it is likely the combination of both mites will push hive size below populations strong enough to survive the winter in most regions of the country.

5.2 What it cost New Zealand

MAF put the cost of Varroa at between NZ$400 million and NZ$900 million over 35 years. An attempted eradication would have cost NZ$55–70 million, and the Auditor-General records that it “would therefore have been worthwhile from an economic perspective if it were technically feasible.” It was not feasible, because it was found too late.

“the available surveillance tests might not be sufficiently sensitive to enable new infestations to be detected and treated before they spread further.” — Varroa Technical Advisory Group, New Zealand, 2000

That was said of Varroa, for which the alcohol wash works well. For Tropilaelaps the standard tests are materially worse.

“Data of Tropilaelaps geographical distribution should be considered with caution and call for increased vigilance in the context of imports of bees from territories considered ‘officially free’ of Tropilaelaps spp. mites but geographically close to infested areas.” — EU Reference Laboratory for Bee Health, March 2025

The EU’s own reference laboratory is telling regulators, in terms, that “officially free” is a status conferred by not having found the mite, and that for this pest it cannot be trusted. That is the exact assurance on which CFIA is relying.

6. The precedent: prohibition works

Between 1905 and 1919, roughly 90% of honey bee colonies on the Isle of Wight were destroyed by what became known as Isle of Wight disease. In 1921 Rennie described the tracheal mite Acarapis woodi from those colonies and concluded it was the cause. The United States responded within a year: the Honeybee Act of 1922 prohibited the importation of live adult honey bees except under USDA authorisation, expressly to keep Acarapis woodi out.

Acarapis woodi was not found in the United States until 3 July 1984. Sixty-two years passed between the statute and the first detection. The mite was found in Mexico in 1980 and in Texas four years later; the route was overland. No certificate failed. No approved exporter was at fault.

Across the entire operative life of the Act, no major exotic bee parasite has ever been documented entering the United States through the permitted, certified import channel. Every breach for which a pathway has been proposed runs through a route the Act does not govern.

Restricting imports removes the one high-volume, high-risk route we can actually control.

• There is no land pathway. Tropilaelaps is in Asia, the Caucasus and Russia. It cannot walk to Canada. Unlike Acarapis woodi in 1984, it has no Mexico.
• The illegal pathway is real but small. It should be addressed by enforcement, not used as a reason to leave the legal channel open.
• The legal, certified import channel is the principal realistic entry route into Canada, and the one route we can close.

Modern scholarship largely holds that Acarapis woodi was not in fact the cause of Isle of Wight disease. The 1922 Act was legislated on a premise that turned out to be wrong, and was vindicated anyway. Congress acted on incomplete and partly mistaken science to close a pathway, and in the six decades that followed no exotic bee parasite is known to have entered through that channel. The precaution held even though the premise was wrong and the threat milder than feared, which is precisely how precaution is supposed to work when the science is unsettled and the downside is irreversible.

Congress acted in 1922 on incomplete and partly mistaken science, and was vindicated. CFIA declines to act in 2026 on clear and converging science, because it wants more of it. That is not caution. It is the inverse of caution.

7. Canada’s approved sources are not a safe harbour

CFIA’s assurance rests on the claim that Tropilaelaps “has not been found in Australia, New Zealand, Italy or Chile.” Note the verb. Not found is precisely what CFIA said, and precisely what the exporting countries can support. It is being used to mean not there.

7.1 CFIA has already accepted this reasoning once

In 2024 CFIA banned honey bee imports from Ukraine. Its stated reason was not that Tropilaelaps had been found there. It was that CFIA “couldn’t be sure that Ukraine was completely mite-free.” CFIA declined to treat non-detection as absence, and acted. Great Britain took the same step on 7 October 2025, on identical logic.

The principle is already established in Canadian practice. We are not asking CFIA to adopt a new standard. We are asking it to apply the standard it has already used, consistently, to every source whose testing cannot detect this mite, which is currently all of them.

8. What restriction would actually cost

The standard objection is that Canadian beekeeping depends on imported bees to replace winter losses. The numbers do not support it.

Canadian beekeepers replace between 220,000 and 310,000 colonies a year. Offshore package imports amount to roughly 47,000 kg, at 1 to 1.5 kg per package, so we are realistically accounting for 30,000 to 35,000 hives out of the 250,000 to 300,000 that die over winter. These replacement colonies are well known to be highly susceptible to the Canadian disease environment and often require requeening with North American stock. Colony replacement in Canada is overwhelmingly a domestic operation carried out with domestic splits.

This reframes the trade-off. The choice is not between an existential biosecurity risk and the survival of the industry’s replacement model. It is between an existential biosecurity risk and a modest commercial convenience.

The great majority of Canada’s imported queens already come from within North America, where Tropilaelaps is not present. A precautionary restriction therefore falls almost entirely on the offshore trade, and does not depend on any change to the North American arrangements in either direction.

9. Recommendations

Four measures follow from the evidence above. None of them requires waiting for the mite to arrive.

1. Convene a Tropilaelaps preparedness table. A dedicated working group of federal and provincial regulators, provincial apiarists, the Canadian Honey Council and provincial associations, to deliver a national emergency response plan. Canada has no published hazard-specific plan for this pest. Given that neither New Zealand nor Australia could eradicate an established mite, the effective response window is measured in weeks from first incursion. A plan written after detection is a plan written too late.
2. Treat certification-based assurance as not fit for this pest. The diagnostic evidence shows exporting-country certification cannot reliably exclude Tropilaelaps. The import risk assessment should be reopened on that basis.
3. Precautionary suspension of live bee imports from outside North America. Pending a reopened risk assessment, suspend imports of packages and queens from all non-North American areas. It is the position Great Britain already occupies in respect of Ukraine, and the position Canada already occupies in respect of Ukraine.
4. National brood-based surveillance. Sentinel apiary surveillance at ports of entry and in high-risk regions. Australia’s sentinel programme did not prevent Varroa establishing, but it did detect it and it did date it. Canada currently has neither capability.

10. The question this leaves

If Tropilaelaps establishes in Canada, what will the review conclude that we should have done, and why are we not doing it now?

Australia and New Zealand have already run this experiment, with Varroa rather than Tropilaelaps, but the lesson carries directly. Both were mite-free, both trusted detection to protect them, and both found the mite only after it had established and could no longer be eradicated. New Zealand asked the question in July 2000 and put the bill at NZ$400–900 million. Australia asked it in September 2023, after the most intensive surveillance programme in the world caught Varroa too late to matter. Neither got a second chance.

On Tropilaelaps, all three countries are still in the same position: none has it, and each is trying to keep it out. What separates Canada is how it is spending that chance: with no sentinel network, no hazard-specific response plan, a certification regime resting on tests that cannot find this mite, and an open offshore package pathway that carries live bees, and any mite riding the dead ones beneath them, directly into Canadian hives. The time to change the defence is while the mite is still on the other side of the ocean.

Timeline of Tropilaelaps westward spread

Sources

The full source list, with detailed notes on each citation, is in the PDF of this paper. Principal references:

• EU Reference Laboratory for Bee Health (ANSES). Geographical Spread of the Exotic Mite Tropilaelaps spp., March 2025.
• Animal & Plant Health Agency (UK). Rapid Pest Risk Analysis for Tropilaelaps spp., September 2025.
• Sensitivity of current Tropilaelaps mercedesae monitoring methods in Apis mellifera colonies. Scientific Reports, 2026.
• Gill, M. et al. (2024). Survival of Tropilaelaps mercedesae on a range of matrices. bioRxiv 2024.12.10.627759.
• Interception of an Apis dorsata swarm with Tropilaelaps mercedesae on a cargo vessel inbound to the United States. Frontiers in Insect Science, 2026.
• Brandorf, A. et al. (2024). First report of established Tropilaelaps mercedesae populations in Europe. Journal of Apicultural Research.
• Khongphinitbunjong, K. et al. (2016). Apidologie 47: 1–11.
• Office of the Auditor-General (New Zealand). Response to the Incursion of the Varroa Bee Mite (2002).
• Agriculture and Agri-Food Canada. Statistical Overview of the Canadian Honey and Bee Industry, 2024.
• CAPA. Statement on Honey Bee Wintering Losses in Canada, 2025.

Media enquiries: see Media contact.