130 bee colonies lost: Northern Ireland’s better winter still took a heavy toll

130 honey bee colonies were lost in a single winter among the Northern Ireland beekeepers who answered AFBI’s latest husbandry survey. They had started winter 2025/26 with 570 living colonies: roughly one colony in four did not make it through. That is an improvement on the previous winter, but still a substantial loss—not evidence, on its own, of a recovery. ¹

The Agri-Food and Biosciences Institute’s annual survey drew responses from more than 100 beekeepers, its highest participation in five years. Those answers provide a picture of their colonies and practices, not a census of every hive in Northern Ireland. The distinction matters: 130 is the loss reported by participants, not a regional total. ¹

There is also an unresolved problem with the published percentage. AFBI’s announcement on 27 October 2026 gives a loss rate of 23%. BBC Northern Ireland’s account on 28 October gives 24%, alongside the counts of 130 colonies lost from 570. Dividing those counts produces approximately 22.8%, which rounds to 23%. ^{2 1}

That arithmetic does not establish how the survey calculated its rate, nor explain the BBC figure. A methodological explanation or a correction could resolve the difference; the available evidence establishes neither. The sound conclusion is narrower but clear: either published rate is below the 33% loss average reported for winter 2024/25. ^{1 2}

The recent history makes a single year’s improvement less reassuring than it first sounds. Reported losses were 11% in 2021/22 and around 14% in 2022/23, before rising to 24% in 2023/24 and 33% in 2024/25. The latest result, whether expressed as 23% or 24%, is close to the level two winters earlier—not a return to the lower losses before that. The highest historical figure identified in the BBC account was 43%, in winter 2012/13. ¹

The survey has run annually for 18 years, but a long-running exercise does not automatically yield a reliable long-term trend. The available evidence does not establish whether the same beekeepers took part across recent winters, how recruitment changed or how consistently the surveys covered the region’s colonies. More responses add information; they do not, by themselves, prove that the sample represents all beekeepers. ¹

AFBI entomologist Joanna Kirbas described the survey’s purpose as monitoring honey bee health, tracking winter losses and examining disease prevention. Its value is in keeping that record and improving the evidence—not in making every fluctuation carry a verdict about the future. ¹

Archie Murchie, head of AFBI’s Plant Health and Integrated Pest Management branch, said weather influences bees, their food sources and their susceptibility to infection. But his assessment also carried an important limit. ¹

That is a plausible contribution, not a measured explanation of this particular winter. The findings available here do not quantify the effect of a named storm, a temperature pattern or another specific weather condition on the 2025/26 losses.

The largest specified loss category concerned the queen bee. About 51% of losses were linked to queen problems—roughly half the colonies lost, not half of all 570 colonies that entered winter. Other reported outcomes included dead bees or an empty hive. These findings describe reported problems and outcomes; they should not be turned into a complete, independently verified explanation of every loss. ¹

A lost colony does not necessarily mean a hive full of dead bees. The BBC account explains that the bees may have died or left, sometimes leaving a queen and some nurse bees behind. The unit being counted is a colony, rather than an individual bee or the physical hive. ¹

Another 10% of losses were grouped under storms, flooding and theft. Although the account labels these “natural hazards”, theft is not a natural hazard, and the combined figure cannot honestly be described as the weather’s share. Without a breakdown, it tells us neither how many losses were attributed to storms nor how many to flooding or theft. ¹

For a beekeeper approaching the next winter, a lower survey percentage cannot substitute for a functioning colony with food it can reach. National Bee Unit guidance published in September 2026 recommends beginning preparations in late summer, aiming for colonies with a functioning queen, healthy bees, adequate food stores and weatherproof hives. ¹³

Food needs to be accessible as well as present. The unit’s starvation guidance calls for monitoring both the quantity and accessibility of stores; its winter guidance recommends regular checks through winter, especially in early spring when starvation risk is high. These are general husbandry measures for the coming winter, not proof that inadequate feeding or a particular management decision caused the losses just reported. ^{12 13}

The survey also examines disease prevention and management of Varroa, a parasitic mite that feeds on honey bees and can contribute to colony collapse and death. AFBI records its first UK detection in 1992 and its first Northern Ireland detection in 2002; it is no longer a statutory pest. ^{1 17}

That established pest must be kept distinct from Tropilaelaps, an Asian-origin parasite that attacks developing bees. Official National Bee Unit guidance treats Tropilaelaps as an exotic, notifiable pest and recognises accidental introduction into the UK as a serious risk. Nothing in the survey account establishes it as a cause of Northern Ireland’s winter 2025/26 losses. ^{27 1}

There is a concrete development behind the warning. The National Bee Unit says Turkish authorities confirmed detection on 31 August 2026 in Borçka, bordering Georgia, where the pest had previously been reported. That is a confirmation date, not necessarily the date the mite arrived. Murchie placed the Georgian detection in 2024; the official notice confirms previous Georgian reporting without independently establishing that year. ^{26 1}

Murchie warned that the mites breed rapidly and can quickly overwhelm a hive. But the available material supplies no numerical likelihood or timetable for wider European spread, and does not settle the pest’s current presence or absence in Northern Ireland. A prospective threat should not be mistaken for an explanation of losses already counted. ^{1 27}

The immediate finding remains substantial enough: participating beekeepers lost 130 colonies despite a better winter. The next tests are to reconcile the published percentage with the reported counts and establish whether improvement lasts. Until then, this is relief after a worse year—not yet a demonstrated recovery.