Pesticide Residues and Bees – A Risk Assessment

For decades, the standard way to decide whether a pesticide was safe for bees was to spray it directly on them and watch what happened over the next 48 hours. That was the test. That was the whole test. It had a name — the hazard quotient, calculated by dividing the predicted concentration a bee might encounter in a field by the LD50, which is the dose that kills half the test population. Simple, tidy, and, as it turned out, almost entirely blind to how bees actually live. Bees don't stand in fields being sprayed. They live inside hives, eating stored pollen and honey for weeks at a time, accumulating whatever residues those food stores carry. The 48-hour spray test missed that completely. It missed the slow build-up of chemicals in the hive, the months-long dietary exposure of nurse bees and larvae, and the quiet chemical partnerships between fungicides and insecticides that can multiply lethality tenfold, or in one case, five hundred sixty-fold. Sánchez-Bayo and Goka set out to ask the harder question. What they found in the pollen and honey first should give you pause. Across published residue surveys, one hundred sixty-one distinct pesticide compounds have been detected inside hives — one hundred twenty-four in pollen, ninety-five in wax, and seventy-seven in honey or nectar. That's not a few chemicals drifting in from nearby fields. That's a pharmacopeia. Average residue loads hit one hundred twenty-six parts per billion in wax and sixty-six parts per billion in pollen. Among the most commonly found compounds are neonicotinoids — a class of systemic insecticides, meaning they are absorbed into plant tissue and end up in pollen and nectar directly. Thiamethoxam, one of the most prevalent, showed up in over fifty percent of honey samples in some surveys, with average residues around twenty-nine parts per billion. Its oral LD50 — the dose that kills half a bee population when ingested — is just zero point zero zero five micrograms per bee. That is a vanishingly small number. Pyrethroids, which act primarily through contact, were also widespread, along with the organophosphate chlorpyrifos and, crucially, a class of fungicides that would turn out to matter enormously. The sheer diversity of what's in the hive is itself the first problem — because it means bees aren't being exposed to one chemical at a time. They're being exposed to mixtures, constantly. The methodological core of Sánchez-Bayo and Goka's work is a shift in how you think about dietary risk. Instead of a single lethal dose delivered at once, they ask: how long does a bee have to eat contaminated food before the accumulated intake reaches its LD50? The key metric is T50 — the time to accumulate the median lethal dose through daily ingestion. Put plainly, it moves the question from "is this dose lethal?" to "does this bee die before it stops eating?" Those two questions have very different answers. The team distinguished three exposure scenarios. Nurse bees eat pollen almost exclusively for about ten days inside the hive. Larvae are exposed for just five days during development. Nectar foragers drink nectar or honey for around twenty days in summer, and winter bees for up to one hundred days. The same residue concentration that barely registers for one group can be lethal for another, simply because of how much they consume and how long they're exposed. The model assumes no elimination of residues from the body — a conservative worst-case — and even so, the findings are striking. For imidacloprid, nurse bees reach their LD50 in roughly seven to nine days — just inside their ten-day pollen-feeding lifespan. For thiamethoxam, the calculation suggests large mortality can occur within about a day of cumulative exposure. The relationship between toxicity and time isn't linear, either: mortality accelerates with exposure duration, scaling with a power factor between roughly one point five and two. That acceleration matters, because it means the risk isn't just additive — it compounds. And that compounding gets dramatically worse when you introduce the finding that upended the conventional wisdom about fungicides. Ergosterol-inhibiting fungicides — compounds like propiconazole, penconazole, and myclobutanil, used widely to protect crops from mold — have generally been considered harmless to bees. They're not classified as insecticides. They're not designed to kill animals. But Sánchez-Bayo and Goka show that when these fungicides are present alongside certain insecticides, the combination is catastrophic in a way that neither compound would predict alone. Propiconazole increases the toxicity of cyhalothrin, a pyrethroid, by a factor of sixteen point two. It increases the toxicity of thiacloprid, a neonicotinoid, by a factor of five hundred sixty. Acetamiprid, another neonicotinoid, becomes roughly one hundred times more toxic in the presence of propiconazole. The likely mechanism involves the cytochrome P-450 detoxification system — the enzymatic pathway bees use to break down foreign chemicals. Ergosterol-inhibiting fungicides appear to disable it, leaving insecticides to accumulate unchecked. Other fungicide classes that don't interfere with P-450 don't show this synergism. The practical consequence is stark. A compound rated low-risk by every standard metric can become acutely dangerous the moment an ergosterol-inhibiting fungicide is present. These fungicides show up in pollen at a prevalence of roughly one point eight to five point five percent — not ubiquitous, but not rare. Where they co-occur with pyrethroids or certain neonicotinoids, the amplified toxicity is not a statistical artifact. It's a five hundred sixty-fold increase. That is the number conventional bee risk assessment has been ignoring. When Sánchez-Bayo and Goka pull together both contact and dietary risks into a full ranking, the picture clarifies. By contact exposure — contaminated pollen landing on a bee's body — pyrethroids and neonicotinoids are the dominant hazard. Five compounds showed risks above five percent: thiamethoxam, phosmet, chlorpyrifos, imidacloprid, and clothianidin. Imidacloprid stands out particularly for bumble bees, where estimated contact risk runs from thirty-one point eight to forty-nine percent — compared to ten point three to sixteen percent for honey bees. That gap matters. Bumble bees are wild pollinators, not managed livestock. There's no beekeeper tracking their colony health. By dietary exposure — ingestion of contaminated pollen and honey — the systemic neonicotinoids again lead the risk ranking. Thiamethoxam and imidacloprid produce the highest modeled mortality rates in nectar foragers and nurse bees. For bumble bee nectar foragers, dietary risk from imidacloprid reaches fourteen point five to fifty-seven point four percent. Chlorpyrifos appears in the moderate-to-high dietary risk category for both species. About sixty-five percent of individual residue compounds showed negligible risk in isolation — but the small set that didn't, combined with the synergistic mixtures, is where the real exposure burden sits. The policy implications are direct. Regulatory frameworks currently test pesticides one at a time, using acute contact exposures over twenty-four to forty-eight hours. Sánchez-Bayo and Goka argue this misses three things that demonstrably matter: the prevalence of residues in actual hive food, the cumulative dietary burden over a bee's lifespan, and the toxicity amplification that occurs in mixtures. They identify about eighteen compounds posing a threat through contact alone and flag five insecticides plus four insecticide-fungicide mixtures as carrying risks above five percent under real-world residue conditions. They're careful to note that pesticides aren't the whole story — Varroa mites, viral diseases, Nosema, and other biological stressors are also implicated in colony collapse. But the chemical burden is measurable, it's present in the food bees eat every day, and it's being underestimated by the very tests designed to catch it. The ask from Sánchez-Bayo and Goka is specific: residue-based exposure metrics rather than modeled concentrations, explicit evaluation of mixtures including fungicide-insecticide combinations, time-cumulative risk calculations that reflect how long bees actually eat contaminated food, and regulatory prioritization of the compounds that measured data — not theoretical drift estimates — show to be most dangerous. The old forty-eight-hour spray test answered a narrow question cleanly. It just wasn't the right question. The bees were always eating the answer. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.