Antibiotics, air pollution, and winter losses in Canadian honey bees

Canadian beekeepers cut their antibiotic use in half after 2018. Over the same years, more colonies died over winter. A 2025 study in Nature Sustainability looked at why. It found the story is bigger than antibiotics. The strongest predictor of winter losses was nitrogen dioxide, a pollutant from diesel exhaust.

The lead author is Dr. Brendan Daisley of the University of Guelph. We work with Dr. Daisley on other research, so we read this one closely. This page is our plain summary of it. Wild Antho was not part of this study.

What the researchers asked

In 2018, Canada changed the rules on livestock antibiotics. Beekeepers could no longer buy them over the counter. They now need a veterinary prescription. The goal was to slow antibiotic resistance, which is a threat to human health worldwide.

The researchers wanted to know two things. Did beekeepers use fewer antibiotics after the change? And what happened to the bees?

Where the data came from

The team used the national survey run each year by the Canadian Association of Professional Apiculturists. They looked at 2015 to 2023.

Between 443 and 626 beekeepers answered each year. Together they managed about half to two thirds of all the honey bee colonies in Canada.

The team then added weather records and air quality records from across the country. In all, the analysis drew on 119,244 data points from 234 locations.

What they found

Antibiotic use fell by about half. Before the rule change, about 47% of beekeepers reported using oxytetracycline. After it, about 24% did. Oxytetracycline is the main antibiotic used against foulbrood, a bacterial disease of bee brood.

Fumagillin use fell too. It dropped from about 48% of beekeepers to about 19%. Fumagillin treats nosema, a gut disease. It was not restricted in the same way. Its only Canadian supplier closed in 2018, and the product was hard to get for years.

Mite treatments did not change. About 69% of beekeepers treated for varroa before 2018, and about 74% after. The difference was not significant.

Winter losses rose. Across the nine years, Canada lost an average of 27.6% of its colonies each winter, and the trend went up. Prince Edward Island, Alberta, and Ontario had the highest losses. Nova Scotia had the lowest, at 15.9%.

Less antibiotic use tracked with more winter loss. The link held even after the team accounted for weather and air pollution.

The surprise: diesel exhaust

The team built a model to see which factors best predicted winter losses. The best model explained about 72% of the variation.

Four factors did most of the work:

1. Nitrogen dioxide in the air
2. Winter temperature
3. Antibiotic use
4. Snowfall

Nitrogen dioxide was the strongest single predictor. More of it went with more dead colonies. Carbon monoxide showed the same pattern.

Why would exhaust hurt bees? Bees find flowers by smell. Nitrogen dioxide reacts with flower scents and breaks them down. The paper cites earlier research showing this can cut foraging efficiency by 83% to 90%.

Timing makes it worse. Nitrogen dioxide builds through the summer and peaks in late summer and fall. That is when a colony raises less brood and packs away food for winter. A colony that cannot find flowers in August goes into winter light.

Colder winters and heavier snowfall also went with higher losses.

What beekeepers themselves blamed

The survey also asks beekeepers what killed their colonies. Their top answers, in order:

1. Poor queens
2. Weather
3. Varroa mites
4. Starvation
5. Weak colonies in fall

Poor queens ranked first, well ahead of everything else. That is the part of this study we can do something about. It is why we test our queens every year and breed from colonies that have already come through winters here.

What the study does not prove

The authors are careful about this, and so are we.

It shows a link, not a cause. Fewer antibiotics and more winter losses happened together. That does not prove one caused the other.

The survey has gaps. It does not say whether a beekeeper used antibiotics to prevent disease or to treat it. Queen quality, mite levels, and starvation were not measured in a way the model could use.

Antibiotics have costs of their own. They kill helpful gut bacteria along with harmful ones. Dr. Daisley’s earlier work shows this can weaken a bee’s immune system. The authors suggest that years of routine use may have left colonies dependent on the drugs. Taking them away suddenly, with nothing to replace them, may be what hurt.

What the authors recommend

They call for a balanced approach: careful antibiotic use, plus better tools, so that beekeepers are not left choosing between resistance and dead colonies. They name three:

1. Probiotics, using helpful bacteria that crowd out disease
2. Breeding programs for disease resistant bees that can tolerate varroa
3. Immune priming, such as queens exposed to inactivated foulbrood bacteria, though the authors say it is not yet clear how long that protection lasts

They also say air pollution deserves urgent attention as a cause of colony loss.

What we take from it

No single change will fix winter losses. Queens, mites, weather, forage, and air all pull on the same colony.

Our own view is that local queens help. A queen raised and overwintered here has already had to conserve resources through a winter. This study did not test that, so take it as our opinion.

The second recommendation is the work we do. Our hives were part of Dr. Daisley’s Canadian Bee Gut Project at the University of Guelph, and we are still working with Guelph on queen cell research. You can read how we breed varroa resistant bees, or see our queen bees.

Source

Daisley, B., Macpherson, C. V., Brettingham, D. J. L., Moore, A. E., Thompson, G. J., & Allen-Vercoe, E. (2025). Impacts of antibiotic use, air pollution and climate on managed honeybees in Canada. Nature Sustainability, 8, 1087-1099. https://doi.org/10.1038/s41893-025-01603-y