Why does a colony replace its queen?

► Read for you

A colony may replace its queen for several reasons: advanced age, insufficient egg-laying, poor fertilization, injury, or general weakness. But a recent study suggests that another factor may sometimes be at play: a severe viral infection in the queen could impair her reproductive function and alter her pheromonal signal, to the point of triggering supersedure.

1. Key Points

• The central question is whether high viral infection in a queen can alter her pheromone signal to the point of prompting workers to initiate her replacement.
• Among the measured compounds of the queen's pheromone bouquet, methyl oleate is the only one showing a robust association with viral infection and with ovarian mass. [strong empirical basis]
• In a queenless colony trial, adding methyl oleate to a queen pheromone blend suppressed queen cell rearing more than the blend without this compound. [empirical basis, but in an artificial experimental setup]
• The study makes a chain plausible — high viral infection → reduced ovarian investment → altered pheromone signal → supersedure — but does not yet demonstrate this full sequence in ordinary colonies at the apiary. [theoretical, supported by several empirical findings]
• For the beekeeper, the main value is interpretive: supersedure is not necessarily just a matter of queen age or a vaguely «poor queen», but may also reflect a physiological or health problem.

2. What the Study Shows

The study combines controlled trials, field observations, and biochemical analyses to investigate a supersedure mechanism.

Research question. The authors seek to understand how high viral infection in the queen can destabilise the social organisation of the colony. Their hypothesis is that severe viral stress alters a key pheromone signal of the queen, prompting workers to initiate her replacement. The study focuses primarily on infections involving Deformed Wing Virus B (DWV-B) and Black Queen Cell Virus (BQCV), with a broader interpretation in terms of total viral load.

Methods. The work comprises several components. A cage trial involved 27 young queens, divided into three groups of 9, microinjected with saline solution, live viral inoculum, or inactivated viral inoculum. A second dataset comprised 32 nucleus colonies in five-frame hives with queens of the same age; however, inoculation did not produce a clear difference between groups, and this component was ultimately analysed observationally, with 29 queens retained in the final analysis. The authors also conducted a pheromone trial on 30 queenless colonies divided into three groups of 10, as well as a comparison between 10 queens with small ovaries obtained by caging and 10 queens with large ovaries allowed to lay freely.

Results. Of the seven measured components of the queen retinue pheromone, only methyl oleate emerged robustly: it declined as viral load increased in the experimental trial, and was positively correlated with ovarian mass in the field data. In parallel, the most heavily infected queens also showed broader lipid changes, notably a decline in numerous triacylglycerols, interpreted as major energy reserves. [strong empirical basis]

Interpretation. The most noteworthy finding is that simply reducing ovarian investment experimentally — by caging and restricting laying — also reduces methyl oleate. In other words, workers probably do not react primarily to the virus as such, but rather to a deterioration in the queen's reproductive condition that the virus contributes to causing. Moreover, in the queenless colony trial, the pheromone blend containing methyl oleate inhibited queen cell rearing more than the blend without this compound. This supports the biological role of methyl oleate in maintaining queen acceptance, without on its own proving the entire causal chain under natural conditions. [theoretical, supported by several empirical findings]

A further result nuances the immunological interpretation: when ovaries are reduced without infection, the «canonical» immune proteins do not increase, even though ApoLP-III declines. A reproduction–immunity trade-off reversible in both directions is therefore not demonstrated here. This point is of greater relevance to the biological understanding of the mechanism than to immediate apiary practice.

3. Critical Assessment

The findings are coherent and well constructed, but several limitations call for cautious reading.

Strengths. A key strength is the convergence of multiple approaches: experimental infection, field observations, independent manipulation of ovarian mass, and then a functional test of a pheromone compound. This architecture strengthens the credibility of the proposed scenario, especially because it does not rest on a single correlation. The fact that only methyl oleate stands out clearly among several measured compounds also reinforces the specificity of the observed signal. [strong empirical basis]

Methodological limitations. The principal limitation is that the field component did not remain truly experimental: inoculation did not produce a difference in viral load between groups, and the authors had to treat this component as an observational dataset. This does not invalidate the results, but reduces the causal strength of this part. Furthermore, the pheromone trial was conducted on queenless colonies receiving a synthetic blend at a dose chosen by reasonable approximation; biologically this is informative, but it is not the exact equivalent of a colony with a living queen who is infected and gradually declining.

Possible biases and confounders. The study is set in a North American context, with trials in Canada and complementary analyses on queens imported from Northern California. Direct transposition to Swiss apiaries should therefore remain cautious. In addition, the authors did not test other non-viral pathogens across all experimental components, and the exact site of methyl oleate production remains unknown. Finally, lipid measurements were taken from queen heads, which is relevant for pheromones but less than ideal for assessing overall energy reserves across the organism.

What cannot be concluded. This study does not show that all supersedure is of viral origin. Nor does it demonstrate that there is a simple viral threshold at the apiary beyond which workers replace the queen, or that a beekeeper could today use methyl oleate as a practical diagnostic tool. Finally, it does not directly prove that a given varroa treatment strategy will reduce this specific type of supersedure, even if this hypothesis becomes more plausible within a virus–varroa framework. [cautious assessment]

4. What the other related studies show

Source: Gauthier et al. (2011)

The related work supports several links in the proposed mechanism, but does not yet demonstrate the entire chain under ordinary apiary conditions.

This study does not stand in isolation: it extends earlier work by the same team. Chapman et al. (2024) had already shown, by combining laboratory and field approaches, that infected queens had smaller ovaries and resumed egg-laying less often, and that in the field, infection was a significant predictor of the presence of supersedure queen cells. The 2025 study thus adds a possible chemical link to this scenario: the decline in methyl oleate. This compound is nothing new in itself: Keeling et al. (2003) had already identified it as one of the components of the pheromone bouquet associated with the attraction of workers around the queen.

Where the other studies converge. Several works make the causal chain plausible without fully demonstrating it. Fiévet et al. (2006) showed that deformed wing virus (DWV) is not confined to the digestive tract but can be detected in several tissues of the reproductive castes, notably the ovaries and fat body of queens as well as the seminal vesicles of drones. In the queens analysed, the ovaries even showed the highest viral loads among the organs tested. This study therefore supports the idea that DWV can reach tissues directly involved in reproduction, while remaining cautious: the histological localisation of the virus in queen ovaries could not be confirmed by all the methods used.

Gauthier et al. (2011) introduce an important nuance. The authors described, in honeybee queens, an ovarian pathology characterised by yellowish discolouration and degenerative lesions of the follicles. In the most severe cases, these lesions were associated with egg-laying deficits, and the affected tissues contained numerous viral particles, attributed mainly to DWV and VDV-1. But the study also shows that high titres of these viruses could be observed in the ovaries or abdomen of functional queens. The authors therefore found no simple correlation between viral titre, ovarian degeneration and egg-laying deficit. This result is particularly useful for interpreting the McAfee et al. study: it confirms that the viruses can reach the queen's reproductive tissues, but it serves as a reminder that the pathological effect probably depends on other factors, such as the physiological context, the stage of infection, the tissue affected, or as yet poorly identified cofactors.

Chapman et al. (2021) point in the same direction, associating high natural viral loads with lower-quality queens, with lower sperm viability, smaller ovaries and altered ovarian protein profiles. In the experimental part of this study, infection with Israeli acute paralysis virus also reduced ovarian mass. Amiri et al. (2016) further showed that DWV can be transmitted during natural mating by infected drones and can infect certain queens. These results support the idea that a queen's reproductive function can be compromised by viral infections, including at sensitive stages of her life.

The link between chemical signals, brood and queen rearing is also long established. Pettis et al. (1997) showed that a signal associated with eggs and very young larvae acts together with queen mandibular pheromone to reduce queen rearing. This result does not concern methyl oleate, but it serves as a reminder that the workers' decision to rear queens or not depends on a set of signals linked to the queen, the brood and the state of the colony.

Where the other studies temper the picture. The step "high viral infection → queen replacement" must not be hardened into an automatism. Lang et al. (2023) observed that a single exposure to DWV-A, by oral or venereal route, did not significantly alter colony-strength indicators for two of the three queen genotypes tested, with little subsequent infection of the brood. One genetic source, by contrast, showed greater losses, suggesting that susceptibility also depends on genotype and context.

A still-open underlying debate. Whether queen pheromones are "honest" indicators of the queen's reproductive quality, or rather agents regulating worker reproduction, remains debated. Strauss et al. (2008) concluded that certain components of the mandibular pheromone function more as suppression agents than as reliable indicators of reproductive value. More recently, McAfee et al. (2024) likewise showed that certain pheromone components vary with the queen's age or status, but that ovarian mass does not necessarily explain acceptance by the workers. These results do not directly contradict the 2025 study, but they caution against reducing supersedure to a single pheromone compound.

Relevance for Europe. The viral context nonetheless makes this question relevant to European apiaries. Paxton et al. (2022) described the rapid expansion of the DWV-B genotype and its possible substitution for DWV-A. Sircoulomb et al. (2025), based on samples from fifteen European countries, confirm a clear dominance of DWV-B and of A/B recombinants. This reinforces the relevance of the studied mechanism for temperate Europe, including Switzerland, but does not yet prove that the sequence "high DWV-B → decline in methyl oleate → supersedure" occurs predictably in local apiaries.

Overall, the related literature lends good support to the general idea that a viral infection can compromise a queen's reproductive quality and that workers respond to chemical signals and to the brood context. The convergence is therefore real, but limited: it strengthens the plausibility of the proposed mechanism without turning it into a rule directly applicable at the apiary.

4. Practical Takeaways for the Apiary

At the apiary, this study primarily informs the interpretation of queen replacements, rather than prescribing new practices.

• Early, repeated, or apparently «unexplained» supersedure is not necessarily due solely to queen age; it may also reflect a physiological or health problem.
• In a temperate European context, this study indirectly reinforces attention to the varroa–virus pairing, even though this lever was not tested here as a practical intervention.
• The study does not yet justify any direct change at the apiary in the form of a new diagnostic tool, a pheromone application, or a specific technical recommendation.
• Nor does it support the conclusion that requeening alone will resolve a broader health problem within the colony.
• Its main contribution is therefore interpretive: it helps to read certain apiary signals more clearly, without in itself being sufficient to change management practices.

Read the original study

► Elevated virus infection of honey bee queens

Further reading

• Queen pheromones
• Queen cells
• Renewing colonies and queens
• Deformed wing disease
• Integrated varroa management through the seasons

Bibliography

McAfee, A., Chapman, A., Alcazar Magaña, A., Marshall, K. E., Hoover, S. E., Tarpy, D. R., & Foster, L. J. (2025). Elevated virus infection of honey bee queens reduces methyl oleate production and destabilizes colony-level social structure. Proceedings of the National Academy of Sciences, 122(42), e2518975122. https://doi.org/10.1073/pnas.2518975122