High Levels of Miticides and Agrochemicals in North American ApiariesImplications for Honey Bee Health

If you eat almonds, apples, and blueberries, you’re in quiet partnership with bees. For the past two decades, that partnership has been in trouble. Colonies have been dying over winter, and the strange, unsettling pattern known as Colony Collapse Disorder pushed a hard question to the center: what are bees actually exposed to out there? Not just one spray on one field, but the chemical weather they live in day after day. Mullin and colleagues decided to measure that weather. Not with a single apiary or a single crop, but with a sweep across North America. During the two thousand seven and two thousand eight growing seasons, they drew eight hundred eighty-seven samples from migratory and commercial beekeepers in twenty-three states and a Canadian province. They sampled the hive’s living archive: brood nest wax, foundation wax, stored pollen known as beebread, trapped pollen coming in on foragers, and the bees themselves. Then, they ran those samples through a modified QuEChERS extraction—think of it as a catch-a-lot solvent shake—followed by liquid chromatography tandem mass spectrometry and gas chromatography mass spectrometry. On average, each run screened about two hundred chemicals, including roughly one hundred seventy-one pesticides and toxic metabolites. It was the most expansive residue survey of its kind at the time, and it wasn’t cheap; the analytical effort alone cost around one hundred seventy-five thousand dollars. Here’s the headline. They detected one hundred twenty-one different pesticides and metabolites across those eight hundred eighty-seven samples, and the hive matrices—wax, pollen, bees—were loaded with mixtures. Across the combined bee, pollen, and wax datasets, two or more compounds showed up in ninety-two percent of samples. The average sample carried about six point five detections. Only a handful of samples were clean—one wax, three pollen, and a dozen bee samples had no detectable pesticides. When you stack all the detections, you end up with several thousand residues—an exposure landscape that isn’t one spike, but a dense field of overlapping signals. About half of all samples contained at least one systemic pesticide, the kind that moves inside plant tissues and can show up in nectar and pollen. Not every matrix tells the same story. Wax, the comb that frames the brood nest and stores honey and pollen, reads like a long-term diary. It held some of the highest concentrations and, crucially, persisted. Almost all comb and foundation wax samples—ninety-eight percent—carried the miticides fluvalinate and coumaphos, with maxima of two hundred four parts per million for fluvalinate and ninety-four parts per million for coumaphos. Even the foundation wax sold to beekeepers, before it ever housed brood, was uniformly contaminated in the batches tested: up to ten point one parts per million fluvalinate and fourteen point three parts per million coumaphos, with means of roughly two point six and three point three parts per million. That’s a lot of background before any new exposure even arrives. Pollen tells a different story. It’s the protein pipeline feeding larvae and nurse bees, and it was chemically diverse and dominated by fungicides. Chlorothalonil stood out as a marker compound. In pollen, it reached up to ninety-nine parts per million. When the team plotted pollen fungicides against pollen chlorothalonil, the relationship was almost one-to-one—chlorothalonil practically defined the fungicide load. Systemic insecticides were also present in pollen, including neonicotinoids: twenty-three detections of thiacloprid, fourteen of imidacloprid, eleven of acetamiprid, and a single thiamethoxam hit. Neonicotinoids, however, were notably absent in the bees themselves in this dataset—no residues detected in bee tissues—even as they appeared in pollen and wax. Bees, finally, looked like a recent-contact snapshot. Their residues were typically lower than in pollen or wax, shaped by metabolism and by what transfers into and off the bee’s cuticle. But when bees ran into something acutely toxic, it was stark. In two mortality events, dead-bee samples contained permethrin at nineteen point six parts per million, compared with a twenty-four to seventy-two-hour adult L D fifty of about one point one parts per million for a bee, and fipronil at three point one parts per million against an L D fifty of zero point zero five parts per million. Those are smoking guns—levels twenty-fold and more above the lethal dose for a single bee. Most of the time, though, residues in bees fell below acutely lethal benchmarks, setting the stage for sublethal and mixture effects rather than outright poisoning. Let’s pull on that wax thread for a moment, because it matters for how exposure accumulates. Fluvalinate and coumaphos are lipophilic; they like fat. Beeswax is basically a lipid matrix, so those compounds sink in and stick around across seasons. Mullin’s team quantified the connection. When they compared fluvalinate in wax to fluvalinate in bees from the same colonies, wax levels rose with bee levels along a line with a slope of about eight point five and a baseline around five thousand nine hundred parts per billion. The fit wasn’t perfect—an R squared of zero point five two—but it was solid and statistically significant. Coumaphos showed a similar wax-to-bee pattern, again significant, with a steeper slope. Translate that: colonies living on contaminated comb tend to carry those miticides in their bees. Pollen didn’t predict bee residues for these miticides nearly as well. In other words, for fluvalinate and coumaphos, wax is the main reservoir, not the incoming food. There’s a broader point hiding in the math. When the researchers summed up miticides in wax and compared them to all pesticides in wax, the relationship was almost a straight line with a slope of one. That means, in comb, the miticides essentially were the residue burden. It’s a grim efficiency: the very chemicals used in-hive to control mites now dominate the long-term contamination of the comb the colony raises its young on. Pollen’s story is different because it’s an intake flow. Foragers bring it in from fields, orchards, and roadside weeds, and it carries what’s on the landscape that week. In the dataset, fungicides—especially chlorothalonil—dominated pollen residues. The link was so tight that you could almost predict a sample’s total fungicide load from its chlorothalonil number alone. Pollen rarely carried a single compound. Across the pollen and wax set together, ninety-eight percent of samples had at least two pesticides, ninety-one percent had three or more, and eighty percent had four or more. The most common pairing in that pollen-plus-wax world was fluvalinate with coumaphos, found together in about eighty-three percent of samples. Add chlorothalonil to those two and you get a very frequent trio, showing up in roughly thirty-nine percent of cases. Push it to four, and combinations like fluvalinate, coumaphos, and amitraz degradates with chlorothalonil appeared in about a quarter of samples. That’s not a curiosity—it’s the routine chemical context of beebread, the food baby bees grow on. Now, an important calibration step. If you’re going to say what’s risky, you need to compare residues to what kills a bee. Different labs report adult L D fifties as milligrams per bee over twenty-four to seventy-two hours. Mullin’s group converted those to a common yardstick—nanograms per gram of bee—by assuming an average bee weighs a tenth of a gram and multiplying by ten thousand. That lets you put, say, a pollen permethrin residue next to a body-weight-scaled lethal dose and ask: are we near trouble? In this survey, most pollen residues sat comfortably below a tenth of the L D fifty. The exceptions were those dramatic, acute cases I mentioned. The center of gravity here is sublethal exposure to mixtures, not single-hit lethality. If we zoom back to bees themselves, there are some nuances worth lingering on. Lipophilic breakdown products of amitraz—the miticide often used as strips in hives—accumulated more in bees relative to pollen than coumaphos did, based on bee-to-pollen ratios. Chlorothalonil, by contrast, was about a hundred-fold lower in bees than in pollen or wax, probably because bees metabolize or excrete it quickly. That explains a puzzle: pollen can be drenched in chlorothalonil while adult bees don’t show much of it, yet brood and nurse bees consuming that pollen may still be exposed during sensitive developmental windows. In colonies associated with Colony Collapse Disorder, bees often still carried sublethal fluvalinate and amitraz or chlorothalonil, and chlorothalonil in their tissues was hundreds of times lower than in their beebread—about two hundred twenty-one-fold lower in one comparison—underlining that rapid turnover. All of this complexity—differing matrices, lipophilicity, metabolism—plays out against the larger landscape of co-occurrence. Not only were combinations nearly universal, but certain pairs kept turning up. Across the broader sample set, the binary duo of fluvalinate and coumaphos was a staple, found together in more than three-quarters of cases. Add chlorothalonil or chlorpyrifos and you arrive at the ternary and quaternary combinations that many colonies lived with week in and week out. The takeaway isn’t just “lots of chemicals,” but “lots of the same families together,” which is exactly where interaction risks—like fungicides potentiating insecticides—often sit. There are, of course, caveats. This was an observational snapshot across two seasons, not a randomized experiment. Sampling wasn’t uniform in time or place. Some statistical relationships, especially those trying to tie a specific pollen level to a specific wax level for a given compound, were noisy or not significant. That reflects the messiness of real apiaries: bees forage across crops, residues move between wax, food, and bee tissue, and detection limits vary by matrix. The study was very clear on this point: it doesn’t prove that pesticides caused Colony Collapse Disorder or any particular colony decline. What it does is document, with unusual breadth and specificity, the chemical backdrop against which those declines happened. So, what does it all mean for how we think about risk? First, wax is not just furniture—it’s a reservoir. Contaminated comb can seed bee exposure long after a spray is gone from the field. Second, pollen is the mix zone. It concentrates what’s blooming nearby into the colony’s protein bank, and in this survey, it was often heavy on fungicides and seasoned with systemics, pyrethroids, and organophosphates. Third, bees themselves are the moving meter—good for catching acute hits, less reliable for long-term burdens because their bodies process and shed certain chemicals quickly. The policy and research message from Mullin’s team is straightforward. Single-chemical safety assessments miss the world bees live in. When forty-nine point nine percent of samples contain at least one systemic pesticide and almost all pollen and wax carry multiple residues—often in repeating combinations—risk has to be evaluated at the mixture level. That means more monitoring that spans matrices, not just honey or bee tissue, and toxicity tests that ask what a fungicide-insecticide pair does together at sublethal doses over weeks, not just what each one does alone over forty-eight hours. If there’s a hopeful angle, it’s that the fingerprints point to places we can act. Foundation wax is often contaminated before it sees a colony; cleaning up that input could lower the baseline. Fungicides like chlorothalonil anchor a lot of pollen residue; tweaking application timing during bloom could cut a major source. Because a few miticides dominate comb residues, rotating or reformulating those tools might pay outsized dividends in the long run. In the end, this survey gave us a map. It’s a map crowded with names—fluvalinate, coumaphos, chlorothalonil; permethrin and fipronil in the margins—and it shows bees navigating a chemical landscape that’s layered, persistent, and shared across much of the continent. If we want healthier colonies, we have to manage the landscape they actually inhabit, not the simplified one our lab tests imagine. That starts with recognizing that in a hive, nothing is alone: not a bee, not a brood cell, and certainly not a pesticide.