What “bee decline” actually measures

Three signals, not one

When a newspaper headline says “bees are declining”, the underlying data is almost always one of three things:

1. Annual managed-colony loss surveys. A beekeeper-reported figure of how many managed hives died over the past year. Cadence: yearly. Geography: countries with formal apiary monitoring (US, EU, UK, Canada, Australia, and most COLOSS members). What it is: a stress signal on beekeepers and on managed Apis mellifera. What it is not: a wild-bee number.
2. IUCN Red List species assessments. A formal classification of how threatened each individual wild bee species is, on the standard IUCN scale (Least Concern → Near Threatened → Vulnerable → Endangered → Critically Endangered → Extinct in the Wild → Extinct). Cadence: every five to ten years per region. Geography: best for Europe and selected national / regional assessments. What it is: the gold-standard species-level status. What it is not: a population-trend number for a year-on-year news story.
3. Decadal range contraction / abundance trends. Peer-reviewed studies that compare modern and historic bee distributions over decades, often using museum specimens, long-running monitoring schemes, or aggregated citizen-science data. Cadence: published when a study comes out, not yearly. Geography: well-monitored regions only (UK, US, parts of continental Europe). What it is: the longest-baseline evidence we have on wild bees. What it is not: a global figure.

Each signal is real. Each is measured carefully by serious people. Each says something different. A story that picks one and labels it “bee decline” is not lying; it is just under-specifying. Once you can identify which signal a story is using, you can interpret it correctly.

Signal 1: Annual managed-colony loss

The canonical source in the United States is the Bee Informed Partnership (BIP) annual colony-loss survey, conducted with the Apiary Inspectors of America^[1]. It asks US beekeepers, each spring, how many managed colonies died over the prior summer (April–October), winter (October–April), and full year. The 12-month total US loss has hovered in the 30–50% range for most of the past 15 years.

That number is alarmingly high. It is also not a species number. It says that beekeepers in the United States are losing somewhere between a third and a half of their managed colonies every year — and replacing them, through hive splits, queen rearing, and package-bee purchases — in order to keep their headcount stable. The managed population persists because the beekeepers actively rebuild it.

The international equivalent of BIP is the COLOSS network, which runs the same kind of survey across more than 30 countries using a harmonised questionnaire^[2]. COLOSS findings are broadly consistent: managed colonies are under significant annual stress across the well-monitored beekeeping world, the loss rate varies year to year and region to region, and no single proximate cause explains it — Varroa destructor mites, queen failure, pesticide exposure, nutritional stress, and colony management all contribute.

If a news article cites a percent-loss figure and the source is a beekeeping or apiary-inspector organisation, you are reading Signal 1. The right interpretation is: this is bad news for beekeepers, but the managed-honeybee total may still be steady because they replace what dies.

Signal 2: IUCN Red List assessments

The European Red List of Bees, published in 2014 by Nieto et al. on behalf of the IUCN^[3], is the most-cited regional Red List for bees. It assessed 1,965 European wild bee species. Of the species that could be evaluated, roughly 9% were classified as threatened (Critically Endangered, Endangered, or Vulnerable), and another 5% as Near Threatened. Roughly 56% were Data Deficient — meaning the assessors had insufficient data to evaluate them at all.

That second number is at least as important as the first. The threatened-share figure is built on the subset of species for which the IUCN had enough data to assess; it almost certainly underestimates the true threatened share, because many of the most-data-deficient species are also the most likely to be in trouble (because they are rare, range-restricted, or both).

National Red Lists exist for several countries: the UK has Section 41 species of principal importance under the NERC Act, the US Endangered Species Act has the rusty patched bumblebee (Bombus affinis, listed 2017^[4]), Canada has SARA listings for several Bombus species, and Switzerland, Germany, Italy, and the Nordic countries each have their own national lists. The global IUCN Red List has assessed bumblebees globally but coverage of the other ~19,500 bee species is far from complete.

If a news article says “X% of bee species in [region] are threatened”, you are reading Signal 2. The right interpretation is: this is a species-level conservation status, valid for that region, almost certainly an underestimate of the threatened share because so many species are Data Deficient, and it has nothing to do with whether managed honeybees are doing well or badly.

Signal 3: Range contraction and abundance trends over decades

The third signal is the hardest to summarise because it lives in the peer-reviewed literature rather than in any single dataset. Representative landmarks:

• Cameron et al. (2011) compared modern bumblebee distributions across the contiguous US to museum-specimen records spanning the prior century, and found significant range contraction and relative abundance declines in several species (most starkly in Bombus affinis, B. occidentalis, B. terricola, and B. pensylvanicus)^[5].
• Soroye, Newbold & Kerr (2020) used long-running citizen-science and museum data across North America and Europe to estimate climate-driven occupancy declines in bumblebee species over roughly 50 years^[6].
• Powney et al. (2019) analysed the UK Bees, Wasps and Ants Recording Society database to estimate occupancy trends in 353 wild bee and hoverfly species across Great Britain from 1980 to 2013, finding net declines across the group but very uneven across individual species^[7].

These studies are the most direct evidence we have that wild bees are not doing well in the well-monitored parts of the world. They are also, by construction, restricted to the regions where someone has been collecting bee data for decades — almost entirely Europe and North America. The global situation outside those regions is much less well understood, which is one reason why citizen-science contributions to GBIF and iNaturalist are genuinely valuable: they expand the geographic baseline that future Signal 3 studies can use.

If a news article cites a long-term percentage decline in wild-bee abundance, you are reading Signal 3. The right interpretation is: this is a careful, peer-reviewed estimate of how some bee populations have changed over decades, valid for the regions studied, and the best evidence we currently have on wild-bee trajectories.

Why the three signals get confused

Each signal supports a different sentence and they are easy to mix up:

• “Bees are in trouble.” (Loosely true if you mean Signal 1 stress on beekeepers, Signal 2 species statuses for many wild bees, or Signal 3 declines in monitored regions. Vague enough to slide between them.)
• “Bee populations are crashing.” (Almost always Signal 1, often misread as “all bee species”.)
• “Bees are doing fine, the panic is overblown.” (Almost always a Signal-1 counter using rising FAOSTAT managed-hive totals, deployed against a Signal-2 or Signal-3 worry. Cross-talk in both directions.)
• “Save the bees.” (A campaign slogan with no clear referent. We do not use it. It does not direct action accurately.)

How to read a bee headline

A practical checklist when a bee story arrives in your feed:

1. Which bees? Honeybees (managed Apis mellifera), or wild bees, or both? If the article doesn’t say, treat its conclusion as soft.
2. Which signal? Annual colony loss (Signal 1), species-level conservation status (Signal 2), or decadal abundance change (Signal 3)?
3. Which region? Most authoritative numbers are regional. A US colony-loss figure does not generalise to Kenya; a European Red List percentage does not generalise to Indonesia.
4. What does the underlying source actually say? If the article cites a study, find the abstract. If it cites a Red List, find the assessment page. If it cites a survey, find the methods.

This is the same discipline applied to any scientific story; bees just happen to be a topic where the popular coverage is unusually muddled.

Sources

1. Bee Informed Partnership — annual US colony-loss surveys. https://beeinformed.org/
2. COLOSS — honey bee research association, multi-country annual surveys. https://coloss.org/
3. Nieto, A., Roberts, S. P. M., Kemp, J., Rasmont, P., Kuhlmann, M., et al. (2014). European Red List of Bees. Luxembourg: Publications Office of the European Union. https://www.iucnredlist.org/resources/european-red-list-of-bees
4. US Fish & Wildlife Service — Rusty Patched Bumble Bee (Bombus affinis) species page. https://www.fws.gov/species/rusty-patched-bumble-bee-bombus-affinis
5. Cameron, S. A., Lozier, J. D., Strange, J. P., et al. (2011). Patterns of widespread decline in North American bumble bees. PNAS 108(2): 662–667. https://doi.org/10.1073/pnas.1014743108
6. Soroye, P., Newbold, T., & Kerr, J. (2020). Climate change contributes to widespread declines among bumble bees across continents. Science 367: 685–688. https://doi.org/10.1126/science.aax8591
7. Powney, G. D., Carvell, C., Edwards, M., Morris, R. K. A., et al. (2019). Widespread losses of pollinating insects in Britain. Nature Communications 10: 1018. https://doi.org/10.1038/s41467-019-08974-9

This primer is published under CC BY-SA 4.0. Copy, translate, and republish freely; please keep the citations intact.