Sudden bee deaths: poisoning, disease, or something else

A pile of dead bees is a question

Finding many dead or dying bees in front of a hive is alarming, and the first explanation most people reach for is a pesticide. Sometimes that is right. But the scene a beekeeper sees — bees dead on the landing board, others trembling, crawling, unable to fly — is produced by several different things, and the sign alone does not say which. A virus can do it. So can a robbing fight, a cold spell after a dearth, or a colony running out of food.

What you see, and what could cause it

The table works from the sign. Each row gives what a beekeeper sees, the things that could produce it — in no order — the questions that make one more or less plausible, what would sit badly with the obvious reading, what settles it, and when an official report is the next step. Read across any row and the pattern holds: the sign narrows the possibilities, and something else decides between them.

Some of the context questions do a lot of work. Poisoning from a spray or from drilling treated seed usually reaches many colonies at once, often in several apiaries in one district, in good foraging weather and within days of the field work. The National Bee Unit's account of chronic bee paralysis contrasts it in exactly those terms: poisoning tends to leave dead bees across the apiary, while the virus piles them at one entrance, with trembling, flightless bees, some hairless, black and shiny. Crumpled wings point to deformed wing virus and a high mite load instead. None of these contrasts is a rule; each is a reason to ask the next question.

When it is not an infection

Three common causes of a dead or dying colony are not infections at all. Starvation leaves workers dead head-first in cells, with no stores near the cluster; in winter a cluster can also starve with food a few centimetres away, if the cold stops it moving. A failing queen leaves a colony without eggs, or with drone brood scattered in worker cells, and it dwindles. And acute poisoning, when it happens, kills foragers in the field and at the hive within days.

The Northern Ireland survey gives a sense of proportion, with its limits. For winter 2025-26, beekeepers classed 66 of 130 lost colonies as unsolvable queen problems, 50 as dead or empty, and 14 as natural hazards. Of the 50 dead colonies, 8 had workers dead in cells with no food left, and 20 had them dead in cells with food still present — the second picture is not simple starvation. Those are respondents' descriptions, from 101 volunteers; the colony losses guide sets out what they can and cannot show.

Poisoning: what a real incident looked like

The best-documented large incident in Europe is the one in Germany's Upper Rhine valley in spring 2008. In late April and May, more than 11,500 colonies of about 700 beekeepers showed dying bees with cramping, disorientation and abnormal wing movements, in front of and inside the hives, with up to several thousand dead a day per colony. Damage ranged from 10 to 90% of a colony, and total losses were few. The cause was dust abraded from maize seed treated with the insecticide clothianidin, blown off the fields by pneumatic drills during sowing.

What settled it was not the signs. Germany's federal examination centre for bee poisoning analysed dust, plants, dead bees, fresh pollen and bee bread and found the insecticide residue, and screened bee samples for varroa, Nosema and amoebae, finding no correlation with pathogens. Both halves mattered: a residue alone does not prove a residue killed the bees, and a pathogen screen alone cannot rule a poisoning in or out. The same report notes that the region had recorded above-average winter losses, of up to about 40%, before the incident — a reminder that colonies can carry more than one problem at once.

How a pesticide could reach a colony

'Pesticides kill bees' runs four questions together. Is the substance hazardous to bees at all — a laboratory question, answered product by product? Does it reach bees by a given route, and how much? Was an effect on bees measured, and was that in a cage or in a field? And was a consequence for colonies documented in the field? The European Food Safety Authority's 2023 bee guidance assesses contact and dietary exposure through the treated crop, weeds in the field, field margins and adjacent crops reached by spray and dust drift, and the crop sown next. It left contaminated water out, because there were not enough data on how much bees drink to quantify it.

Reporting a suspected poisoning

In England, a suspected pesticide poisoning of bees is investigated through the Wildlife Incident Investigation Scheme, and the Health and Safety Executive's page says bee notifications often go straight to the National Bee Unit. Liaise with the NBU and the bee inspector from the start and say why a pesticide is suspected. At least 50 bees are enough for the standard tests: put them in a cardboard box, not a plastic bag, and freeze them if they cannot be sent at once. The NBU screens them for disease within 48 hours before they pass on for residue analysis. Where fumigants are suspected, do not collect samples at all. Natural England can be reached on freephone 0800 321 600 in office hours.

The scheme does not take further an incident where the animals died of something other than a pesticide — disease, starvation or injury — which is why the disease screen comes first. Scotland and Wales run their own arrangements, and Ireland and Northern Ireland have their own authorities; HoneyHQ has not read their poisoning routes, so it gives none. Ask the national authority. Where the dead bees show signs of a notifiable brood disease instead, the notifiable-disease rules apply, and a suspicion is reported at once.