Bees as environmental sentinels: what honey bee colonies and wild bees reveal about their environment

1. Key points in brief

• The phrase attributed to Einstein about the disappearance of bees is most likely apocryphal, but it raises a real question.
• The honey bee can serve as a sentinel for certain environmental contaminations, especially if observations are standardised.
• Wild bees inform mainly about habitat quality, floral diversity and the intensity of land use.
• A problem observed in a colony is not automatically a pollution signal: varroa, viruses, nutrition, weather, queen and beekeeping management can produce similar signs.
• Since 2013, Belgium has developed monitoring programmes, but mainly around colony health and residues; the broader eco-epidemiological vision of the report has been only partially realised.

2. What the study shows

This chapter summarises the 2013 Belgian report: it distinguishes wild bees as ecosystem indicators from the honey bee as a potential sentinel of environmental contamination.

Research question. The report L'abeille, sentinelle de la santé et de l'environnement, indicateur des écosystèmes starts from a simple question: can bees really tell us about the state of the environment, or is the expression "sentinel bee" merely a slogan?

The question is introduced through a very familiar image: the phrase often attributed to Einstein according to which humanity would have only four years left to live if bees disappeared. The report points out that this phrase was probably never uttered by Einstein, but that its rapid spread reveals something important: in the collective imagination, the bee has become the symbol of an ecological boundary not to be crossed (L'abeille, sentinelle de la santé et de l'environnement, 2013). Quote-checking work points in the same direction: the formulation is most likely apocryphal (Calaprice, 2010; O'Toole, 2013).

Method. The document is not a single experiment but a scientific and institutional synthesis produced for the Belgian Federal Public Service Public Health, Food Chain Safety and Environment. It draws on the scientific literature, on a colloquium held in 2013, and on examples of environmental surveillance projects using bees.

The report first builds an essential conceptual distinction. An indicator species informs about the composition or state of an ecosystem. A sentinel species more specifically signals exposure to an environmental hazard, such as chemical, biological or particulate contamination. This distinction prevents the conflation of two different uses of living organisms: reading the state of a habitat versus detecting a risk.

Findings. The report concludes that non-Apis bees, that is wild bees in the broad sense, have strong potential as ecosystem indicators. Their species diversity, nesting requirements, floral specialisation and life cycles allow the quality of environments to be read. The presence or absence of certain species can provide information about floral resources, available habitats, agricultural practices or the state of semi-natural environments.

The honey bee, by contrast, is not a good ecological indicator simply by its presence. In our regions, it is largely housed, fed, cared for and propagated by beekeepers. Its presence in a landscape therefore reflects human activity as much as the ecological quality of the environment. On the other hand, the honey bee colony fits the criteria of a sentinel species well: it is exposed to air, water, nectar, pollen, dust and contaminants present on flowers; it forms numerous populations; it always returns to a fixed point, the hive; and its biology is well enough known for certain signals to be interpreted.

The report also emphasises the available matrices: adult bees, pollen, honey, wax, propolis. These materials can retain different substances depending on their chemical nature. Wax, for instance, can preserve lipophilic compounds; pollen informs about floral resources and certain associated contaminants; adult bees can carry or accumulate particles and residues. The hive thus becomes a biological collection point — not perfect, but potentially very useful.

Interpretation. The strength of the report is to reformulate the popular intuition scientifically: bees do not predict the end of the world, but they can provide warning signals. For these signals to become useful, however, they must be lifted out of anecdote. Mortality in front of a hive, a drop in activity or the weakening of a colony are not enough to prove contamination. These observations gain value only when integrated into a broader protocol: location, season, agricultural practices, health status, varroa level, weather, residue analyses and comparison between sites.

3. Critical assessment

This chapter clarifies what the report contributes, but also what the Belgian follow-up shows: surveillance has progressed without reaching the fully integrated eco-epidemiological system envisaged in 2013.

Strengths. The report has high pedagogical quality: it clearly separates three levels that are often confused. The first is symbolic: the bee as a flagship species capable of mobilising the public. The second is ecological: wild bees as indicators of habitat quality. The third is eco-epidemiological: the honey bee as a potential tool for monitoring contamination or environmental stress.

This distinction is valuable for beekeeping. It avoids asking the honey bee to deliver what it cannot. A hive placed by a beekeeper does not prove that an environment is rich in biodiversity. Conversely, a well-monitored colony, integrated into a network of observations and analyses, can contribute to documenting the contaminants to which bees are actually exposed.

The report is also strong because it does not reduce the hive to honey production. It treats it as a superorganism exposed to multiple environmental pathways. Foragers sample the landscape, resources flow into the colony, and certain effects can manifest at the level of individuals or of the colony. This approach gives a scientific basis to the idea of a "sentinel" without turning it into automatic certainty.

What followed in Belgium. The 2013 report did not stand alone. Belgium already had a first Federal Bee Plan 2012–2014 and then adopted a Federal Bee Plan 2017–2019. From 2016 onwards, the AFSCA launched the HealthyBee project to monitor bee health, obtain objective figures on mortality and examine certain risk factors. The 2016–2017 monitoring covered 193 beekeepers and 865 colonies, with three rounds of visits; the average winter mortality observed was 27.9 % (AFSCA, 2018).

This first HealthyBee cycle confirmed the multifactorial nature of the problem. The AFSCA notably found a significant link between varroa infestation and mortality. Bee bread analyses were also performed: 81 apiaries were sampled and 78 samples contained at least one residue, although none was classified as posing a high risk to bees according to the risk quotient used (AFSCA, 2018).

The 2017–2018 monitoring, conducted with epidemiological support from Sciensano, covered 174 apiaries visited three times. It included direct observations, interviews, measurements taken by qualified inspectors, as well as information on varroa and beekeeping management practices. The weighted winter mortality was estimated at 23.8 % for Belgium (Sciensano, 2022).

The draft Federal Pollinator Plan 2026–2028 indicates that this approach could continue and be broadened. It plans in particular to maintain HealthyBee as an official, active and continuous programme for monitoring bee health, with, from 2026 onwards, two visits to about 200 beekeepers by official AFSCA veterinarians: a first visit in autumn to establish the initial health status, a second in summer to collect data on winter and seasonal mortality. It also plans multi-residue analyses on Belgian honeys, in particular to detect residues of environmental pesticides and veterinary medicines, as well as work on indicators of agricultural pressures on pollinators (SPF Santé publique, 2026).

Main limitation. These follow-ups are important, but they realise only part of the 2013 vision. What has been put in place is mainly health and beekeeping surveillance: mortality, varroa, diseases, management practices, residues in certain matrices. That is already useful. But the most ambitious proposal of the report — systematically linking colony health, physiological biomarkers, residues, agricultural practices, industrial emissions, environmental data and human health — does not appear as a fully operational "One Health" device.

Possible biases and confusions. The main risk would be to turn the sentinel into an oracle. A declining colony does not automatically say: "the environment is contaminated". Varroa, viruses, Nosema, lack of pollen, a failing queen, mismanaged feeding, an incorrectly applied treatment, a cold spell, a heatwave or migratory beekeeping can produce similar signals. It is precisely for this reason that the report calls for a collective, multidisciplinary and standardised approach.

What cannot be concluded. One cannot conclude that an isolated hive is enough to measure the health of a landscape. Nor can wild-bee monitoring be replaced by honey-bee monitoring alone. Finally, colony mortality cannot be used as direct evidence of contamination without parallel data on parasites, viruses, nutrition, weather, treatments and the agricultural environment.

4. What related studies show

Recent work confirms the report's intuition but refines it: honey bees and wild bees do not provide the same type of information.

Direct support: the honey bee as a biomonitor of contamination. Recent reviews and studies confirm that honey bees and hive products can be used to track certain environmental contaminants: heavy metals, pesticides, polycyclic aromatic hydrocarbons, atmospheric particles or other chemical residues (Bargańska et al., 2016; Cunningham et al., 2022; Mair et al., 2023). This literature directly extends the idea put forward in the 2013 report: foragers cover a wide area, collect varied resources and bring back to the hive traces of the environments they visit.

The studies also show that the choice of matrix is decisive. Adult bees and pollen can provide a sensitive environmental signal for certain metals or pesticides, while wax better preserves lipophilic substances and honey, although important for food safety, is not always the most direct reflection of environmental exposure (Calatayud-Vernich et al., 2018; Praus et al., 2023). Work on atmospheric particles also reinforces the interest of the bee as a biological collector of air pollution but requires rigorous protocols to distinguish external contamination, ingestion and effects on pollinator health (Papa et al., 2024).

Direct complement: wild bees as habitat indicators. Recent literature also confirms that wild bees are particularly useful for reading habitat quality. Their species richness, life-history traits, floral resource needs and nesting sites respond to the intensity of agricultural use, floral diversity, landscape structure and urbanisation (Ekroos et al., 2020; Papanikolaou et al., 2017). European studies show, for example, that flower richness, continuity of blooming, wildflower strips, semi-natural habitats and diverse urban gardens favour wild-bee communities (Ammann et al., 2024; Casanelles-Abella et al., 2022).

These findings confirm an important point: the honey bee should not be used as a substitute for wild bees. A European analysis concludes that managed honey bees are not a good "radar" for detecting wild-bee decline, because the two groups differ in life cycle, habitat dependence, human management and stress responses (Wood et al., 2020). To monitor biodiversity, wild-bee communities therefore have to be tracked directly.

Useful theoretical context: pollination is essential, but the four-year slogan is excessive. Studies on pollination services confirm that pollinators contribute strongly to fruit, vegetable, nut and seed crops, as well as to the reproduction of many wild plants. Wild pollinators often increase fruit set in crops independently of honey-bee abundance, showing their complementarity (Garibaldi et al., 2013). Globally, agricultural demand for pollination has grown faster than the stock of managed colonies, reinforcing the importance of pollinator conservation (Aizen & Harder, 2009). However, apocalyptic scenarios such as "four years before the end of humanity" are not supported by the data. Pollination losses would have serious effects on diet, human health and the economy, but not in the form of a rapid, mechanical extinction of humans (Smith et al., 2022).

Limits for practical interpretation at the apiary. The other studies also remind us that signals from colonies are heavily confounded. Winter losses and weakening depend strongly on varroa, viruses, nutrition, weather, queen quality and beekeeping practices (Goulson et al., 2015; Jacques et al., 2017; Requier, 2019). At broader scales, analyses often link colony losses, parasites, pesticides and extreme weather events, confirming that causes interact rather than simply adding up (Insolia et al., 2022).

Recent literature thus does not contradict the report: it refines it. Yes, bees can serve as sentinels or indicators. But honey bees, wild bees, beekeeping matrices and field observations do not answer the same questions. Sound scientific use means combining these signals, not conflating them.

5. What does this mean at the apiary?

For the beekeeper, the goal is not to make an environmental diagnosis alone, but to produce reliable observations that can contribute to a collective reading of the territory.

• Observe and document. Every beekeeper can record unusual mortalities, dates, weather, nearby crops, treatments carried out, varroa level, condition of stores, signs of disease and changes in behaviour. These simple data do not prove a cause, but they prevent signals from disappearing into anecdote.
• Rule out beekeeping causes before blaming the environment. Varroa, viruses, hunger, a failing queen, feeding error or mismanaged treatment must be checked first. A sentinel is useful only if the main confounding factors are documented.
• Embrace the network idea. The political and scientific potential emerges when many apiaries, associations, cantonal services, laboratories and researchers use compatible protocols. At the Swiss scale, and ideally cross-border, apiaries could form a watch network complementary to programmes monitoring biodiversity and environmental quality.
• Support wild bees as well. Hives do not replace wild bees. Promoting continuous flowering, reducing unnecessary mowing, preserving hedges, embankments, deadwood, bare soil and microhabitats helps maintain the most sensitive indicators of habitat quality.
• Call for monitoring frameworks. Beekeepers can contribute locally, but the overall picture requires harmonised protocols, analyses of beekeeping matrices, health data, agricultural information and coordination across regions. This is where the idea of the bee as sentinel becomes truly useful for public decision-making.

Read the original study

L'abeille, sentinelle de la santé et de l'environnement, indicateur des écosystèmes. Report produced for the Federal Public Service Public Health, Food Chain Safety and Environment, September 2013. Official download link: [À VÉRIFIER].

Further reading on ApiSavoir

• The bee, sentinel of health and the environment
• Wild bees in Switzerland: way of life, importance, threats and protection
• Safeguarding pollinators
• Wildflower strips support bees
• Practical Guide: 3.1.2 Bee poisoning
• Wax hygiene: a central lever for colony health

References

L'abeille, sentinelle de la santé et de l'environnement, indicateur des écosystèmes. (2013). Report produced for the Federal Public Service Public Health, Food Chain Safety and Environment.

Agence fédérale pour la sécurité de la chaîne alimentaire. (2018). Santé des abeilles : l'AFSCA communique les résultats du projet HealthyBee 2016–2017. https://favv-afsca.be/fr/publication/sante-des-abeilles-lafsca-communique-les-resultats-du-projet-healthybee-2016-2017

Aizen, M. A., & Harder, L. D. (2009). The global stock of domesticated honey bees is growing slower than agricultural demand for pollination. Current Biology, 19(11), 915–918. https://doi.org/10.1016/j.cub.2009.03.071

Ammann, L., Bosem-Baillod, A., Herzog, F., Frey, D., Entling, M. H., & Albrecht, M. (2024). Spatio-temporal complementarity of floral resources sustains wild bee pollinators in agricultural landscapes. Agriculture, Ecosystems & Environment. https://doi.org/10.1016/j.agee.2023.108754

Bargańska, Ż., Ślebioda, M., & Namieśnik, J. (2016). Honey bees and their products: Bioindicators of environmental contamination. Critical Reviews in Environmental Science and Technology, 46, 235–248. https://doi.org/10.1080/10643389.2015.1078220

Calaprice, A. (2010). The Ultimate Quotable Einstein. Princeton University Press.

Calatayud-Vernich, P., Calatayud, F., Simó, E., & Picó, Y. (2018). Pesticide residues in honey bees, pollen and beeswax: Assessing beehive exposure. Environmental Pollution, 241, 106–114. https://doi.org/10.1016/j.envpol.2018.05.062

Casanelles-Abella, J., Fontana, S., Fournier, B., Frey, D., & Moretti, M. (2022). Low resource availability drives feeding niche partitioning between wild bees and honeybees in a European city. Ecological Applications, 33. https://doi.org/10.1002/eap.2727

Cunningham, M. M., Tran, L., McKee, C. G., Polo, O., Newman, T., Lansing, L., Griffiths, J., Bilodeau, G. J., Rott, M., & Guarna, M. M. (2022). Honey bees as biomonitors of environmental contaminants, pathogens, and climate change. Ecological Indicators, 134, 108457. https://doi.org/10.1016/j.ecolind.2021.108457

Ekroos, J., Kleijn, D., Batáry, P., Albrecht, M., Báldi, A., Blüthgen, N., Knop, E., Kovács-Hostyánszki, A., & Smith, H. G. (2020). High land-use intensity in grasslands constrains wild bee species richness in Europe. Biological Conservation. https://doi.org/10.1016/j.biocon.2019.108255

Garibaldi, L. A., Steffan-Dewenter, I., Winfree, R., Aizen, M. A., Bommarco, R., Cunningham, S. A., Kremen, C., Carvalheiro, L. G., Harder, L. D., Afik, O., Bartomeus, I., Benjamin, F., Boreux, V., Cariveau, D., Chacoff, N. P., Dudenhöffer, J. H., Freitas, B. M., Ghazoul, J., Greenleaf, S., Hipólito, J., et al. (2013). Wild pollinators enhance fruit set of crops regardless of honey bee abundance. Science, 339, 1608–1611. https://doi.org/10.1126/science.1230200

Goulson, D., Nicholls, E., Botías, C., & Rotheray, E. L. (2015). Bee declines driven by combined stress from parasites, pesticides, and lack of flowers. Science, 347. https://doi.org/10.1126/science.1255957

Insolia, L., Molinari, R., Rogers, S. R., Williams, G. R., Chiaromonte, F., & Calovi, M. (2022). Honey bee colony loss linked to parasites, pesticides and extreme weather across the United States. Scientific Reports, 12. https://doi.org/10.1038/s41598-022-24946-4

Jacques, A., Laurent, M., Ribière-Chabert, M., Saussac, M., Bougeard, S., Budge, G. E., Hendrikx, P., & Chauzat, M. P. (2017). A pan-European epidemiological study reveals honey bee colony survival depends on beekeeper education and disease control. PLoS ONE, 12, e0172591. https://doi.org/10.1371/journal.pone.0172591

Maeterlinck, M. (1901). La Vie des abeilles. Paris : Eugène Fasquelle.

Mair, K., Irrgeher, J., & Haluza, D. (2023). Elucidating the role of honey bees as biomonitors in environmental health research. Insects, 14, 874. https://doi.org/10.3390/insects14110874

O'Toole, G. (2013). If the bee disappeared off the face of the earth, man would only have four years left to live. Quote Investigator. https://quoteinvestigator.com/2013/08/27/einstein-bees/

Papanikolaou, A. D., Kühn, I., Frenzel, M., Kuhlmann, M., Poschlod, P., Potts, S. G., Roberts, S. P. M., & Schweiger, O. (2017). Wild bee and floral diversity co-vary in response to the direct and indirect impacts of land use. Ecosphere, 8. https://doi.org/10.1002/ecs2.2008

Papa, G., Pellecchia, M., Capitani, G., & Negri, I. (2024). The use of honey bees (Apis mellifera L.) to monitor airborne particulate matter and assess health effects on pollinators. Environmental Science and Pollution Research International. https://doi.org/10.1007/s11356-024-33170-8

Praus, L., Urbanová, S., & Száková, J. (2023). Honey bees and associated matrices as biomonitors of soil trace elements: Assessment of their sensitivity in a regional rural environment. Environmental Toxicology and Chemistry. https://doi.org/10.1002/etc.5793

Requier, F. (2019). Bee colony health indicators: Synthesis and future directions. CAB Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural Resources. https://doi.org/10.1079/PAVSNNR201914056

Sciensano. (2022). HealthyBee : Surveillance de la mortalité des colonies d'abeilles mellifères belges 2017–2018. https://www.sciensano.be/fr/projets/surveillance-de-la-mortalite-des-colonies-dabeilles-melliferes-belges-2017-2018

Service public fédéral Santé publique, Sécurité de la Chaîne alimentaire et Environnement. (2026). Projet de Plan fédéral pour les pollinisateurs 2026–2028. https://www.health.belgium.be/sites/default/files/media/files/2026-01/em_projet_plan_pollinisateur_federal_26-28_30012026.pdf

Service public fédéral Santé publique, Sécurité de la Chaîne alimentaire et Environnement. (2017). Le Plan fédéral Abeilles 2017–2019. https://www.health.belgium.be/fr/le-plan-federal-abeilles-2017-2019-0

Smith, M. R., Mueller, N. D., Springmann, M., Sulser, T. B., Garibaldi, L. A., Gerber, J., Wiebe, K., & Myers, S. S. (2022). Pollinator deficits, food consumption, and consequences for human health: A modeling study. Environmental Health Perspectives, 130. https://doi.org/10.1289/EHP10947

Wood, T. J., Michez, D., Paxton, R. J., Drossart, M., Neumann, P., Gérard, M., Vanderplanck, M., Barraud, A., Martinet, B., Leclercq, N., & Vereecken, N. J. (2020). Managed honey bees as a radar for wild bee decline? Apidologie, 51, 1100–1116. https://doi.org/10.1007/s13592-020-00788-9