The Drugs (Neonicotinoids) Don’t Work 3

Friday 25th July 2014

The controversy over the cost effectiveness of neonicotinoid pesticides continues to grow.  In response to a request from Members of Parliament to provide peer reviewed evidence that neonicotinoids are cost effective Bayer is only able to provide one paper.  Their flimsy and selective evidence to the Environmental Audit Committee raises more questions than it provides answers and further underlines how broken the pesticides approval process has become.


Neonicotinoids have featured regularly in my blogs. In March I blogged about the growing evidence that neonicotinoids don’t actually work and continued the theme with additional evidence in a second blog in June.

The Environment Audit Committee has been undertaking an Inquiry into the National Pollinator Strategy and their report is due on the 29th of July.  Buglife’s written evidence raised our concerns about the cost effectiveness of neonicotinoids.

As a result there was discussion of this issue at the oral evidence session on the 18th of June.  This included Caroline Lucas MP posing a question to the Bayer representative “can we ask Dr Little if he can name any peer-reviewed scientific publication that is in the public domain that does show that neonic seed dressings increase yields of arable crops”.  Dr Little responded “not off the top of my head”.

Bayer submitted supplementary written evidence, although it appears to only address the question of cost effectiveness from the point of view of Oil seed rape, despite the original question relating to all uses of neonicotinoids. Westminster pollinator event bee waitrose (c) James Bailey

Bayer’s evidence highlights just one peer reviewed paper, although they also have a very impressive photo showing a germinating field of oil seed rape apparently doing very well in one half (neonicotinoid treated) and very badly in the other half (not neonicotinoid treated).

The single paper selectively presented by Bayer is Soroka et al. (2008), this study is considered in the context of other evidence by ‘Heavy Costs’ a 2014 review of 19 published scientific studies by the Centre for Food Safety in the USA that concluded that neonicotinoids only rarely increase the yields of treated crops and even when they do, the increase in yield does not usually compensate the cost of the pesticide.

Soroka et al. compared various percentages of acetamiprid and clothianidin treated Oil seed rape seeds in the planting mix to assess whether farmers could reduce the percentage of treated seed they are planting and still maintain yields.  Decreasing treated seeds by one-third (67% treated) had no consistent effect on damage, yield, or cash return.  The authors concluded “reducing the proportion of treated seed sown by one third can be an effective means of reducing pesticide load to the environment while maintaining efficacy, especially in situations of low-to-medium flea beetle feeding pressure.”

If this is the most convincing scientific evidence of cost effectiveness for neonicotinoids from all the trials and tests done by the pesticide industry then they are in some difficulty.

Photos like the one presented in the Bayer evidence can seem compelling, which is why they are used by pesticide companies, but this is just anecdote.  Establishment varies across fields for many reasons.  There is a wealth of alternative theories that could explain the pattern of growth in the photo that are not addressed by Bayer’s evidence.  Were the crop varieties and genetics identical in both halves?  Were they planted on the same date?  Did they germinate on the same date?  What was the crop history in each area?  Seed treatments often include fungicides, pyrethroids and stickers, so the apparent effect could be due to these chemicals.  A replicated trial is needed, one farmer planting one field and taking a photograph does not prove anything at all!

Statements in the Bayer evidence such as Oil seed rape “will not yield a product if left untreated” need careful interpretation.  While HGCA Research Review No. 70 (Clarke et al. 2009) does suggest that there would be significant losses to the Oil seed rape crop without herbicides, the effect of not using any insecticides would be more moderate.  Without any insecticides the HGCA estimates that the loss to insects would be less than 9% and the loss to Turnip yellow virus would be less than 5%.

Aphids, specifically the peach–potato aphid, can infect OSR with the Turnip yellow virus.  Bayer claims that “Neonicotinoid seed treatments are very effective” and that alternatives such “pyrethroids are almost totally ineffective on this aphid as they have become resistant to this treatment”.  It is a great shame that Bayer do not actually provide any evidence to back up these claims as, if there is an economic benefit from Neonicotinoids it is likely to be in terms of control of Turnip yellow virus.

Ironically about half of the treated OSR seed planted was co-formulated with the pyrethroid beta-cyfluthrin, this prophylactic use on a vast scale probably contributed to the development of pyrethroid resistance.

The picture with Turnip yellow virus is more complicated than Bayer lets on.  1) Pymetrozine is an effective aphid control agent on the autumn crop, 2) Turnip yellow virus is a more significant issue in some areas of the UK than others and 3) resistant strains of OSR have been developed that may negate the virus as a future concern.

Note that in relation to control of viruses in sugar beet I quoted Dewar (1992) as saying “field trial results did not suggest differential control of the two viruses causing yellows” I should clarify that the study did report some control of the viruses, but could not differentiate control effectiveness between the two viruses. I hope to return to the issue of effectiveness against Turnip yellows in the not too distant future.

Bayer claim that when bad weather prevents the spraying of insecticides “clearly those that were seed-treated have a better level of protection against both flea beetle and aphid-mediated disease” (which implies that they think that sprays can still control Turnip yellow virus!) – but there is not a scrap of evidence presented to support this claim.

Bayer say that Pollen beetle is a “damaging pest which left unchecked can cause serious damage to the flower buds” but we know that the plants just replace the buds and yield is not usually affected.  In any case Bayer then states that neonicotinoids at the pre flowering stage “will have long lost their efficacy” – so this appears to be just an advert for their Thiacloprid spray Biscaya.  Note that just because the neonicotinoid is no longer killing pests in big plants, that doesn’t mean that it can’t be reducing the fitness and reproductive rates of bees feeding on the flowers.

HGCA Research Reviews No. 70 and No. 77 (Nicholls 2013) estimate average yield loss as a result of Cabbage Stem Flea beetle (CSFB) in OSR crops not treated with any pesticides is about 1% of national yield.  Pre-neonic ban loss of OSR to CSFB was 0.5% (Clarke et al. 2009).  Therefore the difference between pre-neonic ban loss and average loss of yield with no pesticides, and hence the benefit provided by neonicotinoids, is 0.5%, not the 1% quoted in the Bayer evidence.

Bayer then go on to produce a series of excuses as to why they have been only able to present one peer reviewed paper to evidence the cost effectiveness of neonicotinoids.

First they claim “Academic journals have no interest in publishing papers on such matters”.  However, there are 19 published studies quoted by the Centre for Food Safety review and more in Goulson (2013).  There are journals such as the Journal of Economic Entomology and Pest Management Science established to publish exactly these types of studies.  It is not the fault of the academic journals that the evidence is absent.

Bayer then points at “the loss of government funding of independent field research, essentially requiring the likes of ADAS and the Scottish Colleges to generate funds by collaborating with commercial companies in near market research”.  I can whole heartedly agree with Bayer on this point and would add the Centre for Ecology and Hydrology to their list, the loss of government funding to independent research in this field is indeed a significant hurdle to understanding the environmental impacts of pesticides, providing independent advice on effectiveness and developing public confidence in their use.

Then comes the most concerning sentence of all – Bayer states “The primary focus of crop protection companies is on generating sufficient reliable data to justify the registration of a new product and then demonstrating the efficacy and cost effectiveness of the product to growers and advisers”.  The implication is that Bayer do not believe that it is necessary to establish efficacy before registering a new product.  However, the Plant Protection Products Directive (Recital (24) of Regulation 1107/2009) emphasises that it must be demonstrated that plant protection products “present a clear benefit for plant production” and for approval a product “shall be sufficiently effective”.  Treating efficacy and cost effectiveness as a marketing tool rather than a fundamental foundation of the approvals process is further evidence of how the regulation system is failing.

Authorisation of a product requires as a first step the establishment of efficacy – this data should be produced at a publishable quality and the publishing of the data should be a small final step.

Bizarrely Bayer then pose a question back to the MPs – “Does it matter?”.  The answer has to be an emphatic YES, it does matter if the pesticides work or not.  It is important because 1) fear and commission based sales are very effective at increasing sales without efficacy, 2) users are subject to experiences that can be very impactful, but actually misleading, perhaps as the photo in Bayer’s evidence demonstrates? 3) it’s the law.

Bayer finish off with a gloriously circular argument.

The sale of seed treatments demonstrates that they work, farmers think that they work so they buy them, they must work because companies are allowed to sell them.

Vast numbers of insurance products sold to individuals have subsequently been found to be superfluous and mis-sold.  Unfortunately sales are no proof of efficacy.


In conclusion, the single peer reviewed paper selectively presented by Bayer found that similar levels of cost effectiveness could be achieved using less neonicotinoid pesticide, the remaining content of Bayer’s evidence lacks any robust, scientifically credible basis.  The question – ‘Are neonicotinoids cost effective, an expensive insurance policy or ineffective?’ – remains open across all crop uses.

Government should ask the Competition and Markets Authority to undertake a review of the extent of commission based selling in the sector, the vulnerability of the market to distortion and restricted consumer choice, the prevalence of miss-selling and the effects on vulnerable farmers and the environment.  I hope that, amongst other things, this is a recommendation of the Environmental Audit Committee next week.


Clarke J, Wynn S, Twining S, Berry P, Cook S, Ellis S, Gladders P (2009) Pesticide availability for cereals and oilseeds following revision of Directive 91/414/EEC

Dewar, A. M. (1992). The effects of imidacloprid on aphids and virus yellows in sugar beet. Pflanzenschutz-Nachrichten Bayer (German edition) 45:423–44

Goulson, D. (2013) An overview of the environmental risks posed by neonicotinoid insecticides. Journal of Applied Ecology 50, 977-987.

Nicholls CJ, (2013) Implications of the restriction on the neonicotinoids: imidacloprid, clothianidin and thiamethoxam on crop protection in oilseeds and cereals in the UK. HGCA. Research Review No. 77.

Soroka, J. J., L. F. Grenkow, and R. B. Irvine. 2008. Impact of decreasing ratios of insecticide-treated seed on flea beetle (Coleoptera: Chrysomelidae, Phyllotreta spp.) feeding levels and canola seed yields. J. Econ. Entomol. 101 (6): 1811-1820.