select search filters
roundups & rapid reactions
factsheets & briefing notes
before the headlines
Fiona fox's blog

expert reaction to two new papers on bees and neonicotinoids

Two papers published in the journal Nature have examined possible effects of neonicotinoid insecticides on the health and behaviour of bees. The authors of the first paper report that in laboratory tests bees preferred neonicotinoid-treated sugar water to an untreated solution, while in field studies the authors of the second paper report that bumble-bee colonies around neonicotinoid-treated fields had reduced growth rates and reproduction.


Dr Lynn Dicks, Biodiversity and Ecosystem Services Research Fellow at the University of Cambridge, said:

On the Rundlof study:

“This study provides startling evidence that the neonicotinoid clothianidin is a serious problem for wild bee populations in real farmed landscapes. Buff-tailed bumblebee colonies gained weight less quickly and produced only a third as many queens in the treated landscapes as in the control colonies. As it is only the bumblebee queens that survive over winter, this could have serious implications for the long term resilience of populations, even for this very common bumblebee species. The solitary-living red mason bees didn’t nest at all in the clothianidin-treated landscapes.

“The study builds on previous evidence that wild bees are exposed to neonicotinoids used in agriculture, at levels known to have substantial effects on their behaviour and reproduction, and that these effects translate into lower reproductive success. Before this, for wild bees (not managed honey bees) we had small number of laboratory studies, and a single field-based study of bumblebees, carried out in the UK at just three sites1. The new study in Sweden has 16 sites (8 treatment and control pairs), and is the first to measure effects on solitary bees, which are important pollinators and much more diverse than bumblebees.

“It is a well-designed and careful piece of research. The methods are entirely transparent and most of the fieldwork was carried out ‘blind’ – the field workers didn’t know whether they were working in control or treated fields. It is also independent of the neonicotinoid manufacturers – funded by the Swedish Government.

“Importantly, both control and treatment landscapes experienced normal agronomic management, including treatment of the oilseed rape fields with a range of other insecticides. The data clearly show that bees were exposed to higher levels of clothianidin in treated than control landscapes through nectar and pollen they had collected, but this was not true for some other insecticides tested. So it is scientifically very reasonable to assign the difference in bee reproductive success between control and treatment landscapes to an effect of clothianidin.

“For the amount of time and resources available, this is landscape-scale ecology at its best.  Of course, there are only a year’s worth of data so far and I would want to see the experiment repeated over five years, but still, the results are stark, implying that there is a relatively large effect here.

“It is very telling that the study does not find an effect of the clothianidin treatment on honey bees, although admittedly its statistical power would only pick up a relatively large effect of a 20% drop in colony size. In this, it accords with other field studies conducted by the neonicotinoid manufacturers (arguably less well designed2), which have looked for and not found effects on honey bees.

“It is also worrying that the study finds clothianidin residues in wild plants growing in the field margins around the treated fields. This raises a question about whether it’s a good idea to attract wild bees to field margins by sowing flowering plants there for them to feed on, if these systemic insecticides are being widely used.

“Rundlof’s team’s study and its implications must be considered very carefully by European politicians who will soon need to decide whether to retain or end a moratorium that currently restricts the use of clothianidin in Europe. For me, it’s pretty convincing evidence that the moratorium was a good decision, and should continue.

“We need a lot more studies like this, to fully understand the impacts of normal agronomic practices on wild bees and other beneficial insects. After all, while some insects are pests, many wild insects – bees, hoverflies, beetles and wasps – provide important and economically valuable services to agriculture, through crop pollination3 and consumption of the pest insects4.”


2 e.g.




Prof. Simon Potts, Director of the Centre for Agri-Environmental Research at the University of Reading, said:

“Together these two papers bridge three of the four major gaps in our understanding of the impacts of neonicotinoids on pollinators.

“First, the Rundlöf paper overcomes the widespread criticism that lab-based studies showing negative effects of neonicotinoids on pollinators are not representative of real world conditions. This well designed study, in a typical agricultural landscape in Sweden, clearly shows that neonicotinoid seed coatings can negatively affect the number of wild bees found on farmland. While this study tested just a single seed coating, clothianidin, the responses reported are so large that it raises major concerns for other neonicotinoids and other crops.

“Second, while stark impacts were recorded for a bumblebee and solitary bee, no such effect was found in managed honeybee colonies. Using honeybees as a model species in pesticide risk assessments is therefore inadequate as it may fail to capture the potential risks to the other 2,000 species of European bee and many other insect pollinators.

“Third, the Kessler paper shows that the criticism that bees in controlled feeding experiments may not have the choice to avoid pesticide-laced sugar solutions, even if they want to, doesn’t hold up. Both honeybees and bumblebees were shown not to be able to taste neonicotinoids and, far from being repelled by them, may even prefer to consume them. Under field conditions this could potentially mean that foraging bees may bring back more forage with neonicotinoids in it than previously thought and so increase the risk to growing colonies.

“The fourth, and still outstanding, question is whether the loss of bee numbers and fitness in real landscapes due to neonicotinoids, translates through to a loss of pollination services to our food crops. The likelihood is yes, since we know that in general fewer bees means less pollination, but we still need evidence to quantify the extent of the effect in terms of loss of production and economic value. This is particularly salient, since European agriculture receives annual benefits from insect pollinators in the region of 22 billion Euros, with the majority of contribution coming from wild bees (those most affected by neonicotinoids) and not honeybees (which seem to be less affected).

“The findings of these two studies, taken together with the previous body of evidence, now pose policy makers with an exceptional challenge. The strength of evidence for the negative impacts of neonicotinoids on pollinators has been significantly ratcheted up, yet a continuation of the restriction on the use of neonicotinoid seed coatings leaves farmers with few options. A return to the widespread use of older broad spectrum pesticides to replace neonicotinoids could also have highly detrimental impacts on pollinators and other wildlife – quantifying this risk is essential if policy is to weigh up the pros and cons of these different pest control measures. There are no new miracle pesticides on the horizon that will provide the same level of protection as neonicotinoids but without impacting on pollinators – and so industry doesn’t have a short-term solution. Alternatives which rely less on chemical pest control, such as integrated pest management (IPM), may represent a longer-term opportunity, but in Europe these practices are not well enough developed and effective enough to be widely adopted at present.

“The demand for a secure supply of cheap and nutritious food by the public will remain high and so farming has to find ways to keep meeting this demand, yet has to balance this with wider environmental protection required by society. The relevant policies on agriculture and pesticides have responded to a clearly characterised risk posed by neonicotinoids. However, the policy need to mitigate this problem is far ahead of scientists and industry to provide effective working solutions to reconcile food production and environmental concerns.

“We should consider this a large-scale policy “test case” with no easy quick fix – only through real innovations underpinned by robust evidence can the current mismatch be fully resolved.”


Dr Juergen Keppler, Ecotoxicology Group Leader at Bayer CropScience (which produces and sells neonicotinoids), said:

“We are pleased the Research Letter demonstrates again there is no effect of neonicotinoid seed treatments on honey bee colonies in realistic field conditions, consistent with previous published field studies.

“Based on own experiences from large-scale field studies in oilseed rape, which showed no adverse effects on monitored honey bees, bumble bees and solitary bees during and after oilseed rape flowering, our researchers and experts have questions concerning the methodology applied by the authors which has an influence on the robustness of the effect data on wild bees (e.g. nectar and pollen sampling and analytical verification of exposure, number of introduced solitary bees, consideration of patchiness in wild bee occurrence etc.).

“The method descriptions and results as summarized in the Research Letter do not – in our view – substantiate the conclusions that Spring-sown oilseed rape seed treatment with neonicotinoid insecticides affects wild bees. Bayer CropScience is however highly interested in this type of research and open for a dialogue with the authors of the study to discuss the Research Letter and the underlying methodology and results in more detail.”


Dr Peter Campbell, Senior Environmental Risk Assessor at Syngenta (which manufactures and sells the neonicotinoid Thiamethoxam), said:

On the Kessler paper:

“This is another artificial laboratory experiment exposing either honey bees or bumble bees to neonicotinoid spiked sucrose.

“Whilst for thiamethoxam the range of concentrations used varied from 1 to 1000 nM solutions, only the 1 and 10 nM solutions are environmentally relevant to pollen and nectar residues measured in the field from thiamethoxam seed treated crops like OSR.

“When you look at the actual amounts of sucrose consumed in these choice tests, bumble bees exposed to the environmentally relevant doses of 1 & 10 nM thiamethoxam were reported to have a 30% and 15% increase in consumption respectively, compared to controls. Honey bees had only a 12 % increase in consumption compared to control. Although statistically significant, the ecological significance of these findings is questionable.

“Even under these worst case laboratory conditions, there was no statistically significant effect on bee survival after 24 hours for either species at the environmentally relevant thiamethoxam concentrations of 1 & 10 nM. Furthermore, no significant effect on survival was reported even at 100 nM which is 10 times higher than thiamethoxam concentrations we would normally expect in pollen and nectar under field conditions. This is consistent with the finding of our published field study on thiamethoxam-treated oil seed rape with honeybees, which reported no effects on bee mortality, foraging behaviour, colony development and overwintering success, following 4 consecutive years exposure (Pilling et al, 2013).”


Dr Christopher N Connolly, Reader in the Medical Research Institute at the University of Dundee, said:

On the paper by Kessler et al.:

“This is a very interesting paper that identifies that bees actually choose to consume certain neonicotinoids when present at environmentally relevant (1-10 nM) levels.  Interestingly, they cannot taste the presence of these neonicotinoids, which suggests that the stimulus being sought may be similar to nicotine seeking behaviour in humans. It will be interesting to see if insects become addicted to neonicotinoids over time as humans become addicted to nicotine. Given that the neonicotinoids are commonly found in our farmed environment at these levels, this may have already occurred.

“It is unclear why the bees do not also prefer the neonicotinoid clothianidin, especially as thiamethoxam (that they do prefer) is a pro-pesticide that is metabolised to clothianidin.  Nevertheless, the actual exposure to neonicotinoids may be higher than expected.”

On the paper by Rundolf et al.:

“This paper demonstrates further field evidence of the delivery of neonicotinoids to honeybees and bumblebees (10-14 ppb) and their detrimental effect on bumblebee colony growth. The authors extend the relevance to include an impact on solitary bee nesting behaviour. However, it is unclear what the impact of this is on solitary bees as nests could be established away from the experimental site provided.

“The density of wild bees is increased in line with the amount of oilseed rape provided (regardless of treatment or not). However, the density of bees is decreased when clothianidin-coated seeds (compared to uncoated seeds) are used. It is hard to reconcile this with the results of Kessler, where bees show a preference for some neonicotinoids (although no preference or avoidance is reported for clothianidin).

“In the case of the honeybees, no effect on adult bee numbers were found. This is perhaps unsurprising given that the levels of neonicotinoids are sublethal and bees develop over 21-24 days before emerging as adults. Much longer periods would be required to detect deficits in honeybee colonies and the impact of the rest of the environment would be confounding.

“In summary, the major advance of this work is to extend the evidence on the impact to bumblebees and demonstrate that it can result from field-relevant use of neonicotinoids as a seed coat protection of oilseed rape.”


Prof. Lin Field, Head of Biological Chemistry and Crop Protection at Rothamsted Research, said:

“Although the study by Rundlof et al does show an effect of neonicotinoids on wild bees I don’t personally think that it alone is enough to ‘put an end’ to the argument. We simply need more data before we can really say what the risks are. We also have to consider the reason why we use these compounds: can we afford not to control pest insects? Is it acceptable that yields would be reduced as a result? Are the alternative insecticides any safer to bees? These are questions that a two-year moratorium on neonics is unable to answer.

“I think everyone would agree that direct exposure of neonics to bees will be damaging – and indeed potentially lethal – as some neonics are very toxic to bees. What is debated is whether the relatively small amounts that bees encounter in pollen/nectar are enough to have ‘sub-lethal’ effects.

“Many of the tests to show this have used doses that are higher than likely to be encountered.  There is also significant variation between tests where some have shown effects and others have not.  One problem is the difficulties in doing work in the field and differences between different compounds and different pollinators.  Indeed the study by Rundlof et al shows effects on wild bees but not on honey bees – a good example of different pollinators responding in different ways.

“The study by Kessler et al suggests that bees can’t detect neonics in sucrose solution, which would imply that they would ingest them in the same way as the sucrose control.  But why they might take more is not obvious.  The authors suggest it is via effects on the receptor that neonics bind to, which lead to changes in ‘learning’.  This is very interesting but I think this needs more work to see if it is really true and if so how it works.”


Prof. David Goulson, Professor of Biology at the University of Sussex, said:

“The new paper by Rundlof is a major step forwards in clarifying the neonicotinoid debate.  There was already a large body of evidence which very strongly suggested that exposure of bees to neonicotinoids at field-realistic doses did them substantial harm.  The evidence from bumblebees was particularly clear.  In contrast, the evidence from honeybees was conflicting – while there is abundant evidence that neonicotinoids harm individual honeybees, for example impairing their navigation and learning and making them more susceptible to disease, some large scale field studies had failed to find significant effects on whole colonies.

“These field studies were either conducted by or funded by the agrochemical companies that manufacture the chemicals, leading to some distrust of their results.  They have also been widely criticised for other reasons – for example their control bees were contaminated with neonicotinoids, and the bees were exposed to small experimental plots of treated crop (1-2 ha) rather than realistic-sized fields, so that most of the bees probably foraged elsewhere.  Nonetheless, this contrast had led many of us to speculate that honeybees may be resilient to some neonic exposure – their large colony size buffering them against loss of workers (up to a point).

“This new study used full-sized fields, either treated or untreated with neonics. They were far enough apart that there was minimal contamination of controls.  In short, this is the first fully field-realistic, well-replicated trial so far, an impressive piece of work.  It is also the first to simultaneously examine effects on honeybees, bumblebees and solitary bees.  The results: solitary bees failed to breed at all near treated fields, while bumblebee colonies grew slowly and produced far fewer queens (exactly replicating the results of Whitehorn et al. 2012, Science 336: 351-352).  Honeybees appeared little affected, at least within the short time-frame of this study.

“At this point in time it is no longer credible to argue that agricultural use of neonicotinoids does not harm wild bees.”


‘Seed coating with a neonicotinoid insecticide negatively affects wild bees’ by Maj Rundlof et al. and “Bees prefer foods containing neonicotinoid pesticides” by Sebastien Kessler et al. published in Nature on Wednesday 22 April. 


All our previous output on this subject can be seen at this weblink:


Declared interests

Dr Connolly: “I am lead applicant on a Biotechnology and Biological Sciences Research Council, the Department for Environment, Food and Rural Affairs, the Natural Environment Research Council, the Scottish Government, and The Wellcome Trust, under the Insect Pollinators Initiative (United Kingdom) Grant BB/1000313/1​. Geraldine Wright [co-author of the Kessler paper] is a co-applicant on this programme.”

Prof. Field: “Some of the work in my department at Rothamsted is supported financially by Agro-Chemical companies, including Bayer and Syngenta, but I personally receive no remuneration from any companies.”

Dr Dicks: “I am an independent scientist, with no conflicts of interest.  I am funded entirely by the Natural Environment Research Council.  I am a Coordinating Lead Author of the forthcoming IPBES thematic assessment on pollinators, pollination and food production:

Dr Campbell’s employer, Syngenta, manufactures and sells the neonicotinoid Thiamethoxam.

Dr Keppler’s employer, Bayer, produces and sells neonicotinoids and products containing them.

in this section

filter RoundUps by year

search by tag