Goulson Lab

Neonicotinoids and bees; what is the fuss all about?

When I was first asked to support a ban on neonicotinoid insecticides, I was very sceptical. I’d previously been asked to sign up for campaigns claiming that bee declines were due to mobile phones, GM crops, overhead power lines, and various other eccentric or implausible notions. However, a review of the science suggested that there was something worth investigating, and for the last two years I seem to have had time for little else.

The issue has focussed largely on harm these chemicals might be doing to bees, both domestic honeybees and wild pollinators such as bumblebees. I’ll focus on bumblebees here, for they are the subject of my particular expertise. Neonics are widely applied as a seed dressing to crops such as oilseed rape, and being systemic they spread through the plant tissues into the nectar and pollen. They are highly toxic to insects; for example the “LD50” (the dose that kills half of test subjects) in honeybees is about 4 billionths of a gram. To put that in context, 1 gram – little more than the weight of a sachet of salt – would provide an LD50 to 250 million honeybees, or roughly 50 metric tonnes of bee. They are neurotoxins, binding to neural receptors in the brain and causing swift paralysis and death.

Until recently, there had been few studies of neonics and bumblebees, all small scale and nearly all performed in cages or glasshouses. Taken together, they seemed to suggest that exposure to very low levels of these chemicals, such as might occur when a bee fed on a treated oilseed rape crop, was not sufficient to kill many bees, but it seemed to affect their behaviour, particularly their ability to learn, gather food, and navigate. These are not impacts that are assessed by any of the safety tests by which pesticides are judged (although there are plans to introduce such tests in Europe in the very near future). They are also impacts that are likely to be much more important in the field, when bees naturally perform astonishing feats of navigation and learning when gathering food from flowers. It seemed to me, and to a number of other researchers around the world, that there was a need to study what impacts these sublethal affects might have on bee colonies in natural situations.

To investigate this, we exposed bumblebee nests to the concentrations of neonics found in the pollen and nectar of oilseed rape for two weeks, and then placed the nests out in the field to see how they fared compared to control nests. Over the following 6 weeks the treated nests grew more slowly, and ultimately produced 85% fewer new queens. Since our study, work from other labs has confirmed that field-realistic doses greatly reduce pollen collection in bumblebee workers, potentially explaining why our treated nests performed poorly. It has also been found that concentrations of neonicotinoid as low as 1 part per billion (ppb) in their food cause a drop in egg laying in bumblebees of 30%. Considerably higher concentrations than this have been found in pollen of treated crops (the typical range in pollen is about 1-10 ppb, although 50 ppb has been found in lucerne pollen, and over 100 ppb in melon pollen).

For bumblebees, the evidence so far is convincing and coherent; exposure to levels of neonic commonly found in crops have profound impacts on colony success. The only study which apparently contradicts this was recently placed online by Defra, a report describing a study they conducted in 2012 in which they attempted to repeat our work but with the pesticide exposure phase of the experiment occurring in the field. They placed bumblebee nests next to the only untreated field of oilseed rape they could find, or next to one of two other fields treated with two different neonics, and they followed the colonies over time. Unfortunately they had no replication – just one field per treatment – and, disastrously but interestingly, the control nests all became contaminated with neonics. The summary of this report claims the study shows no clear effect of neonicotinoids on bumblebee colonies; hardly surprising, given the absence of any controls – and hardly the sort of convincing evidence one would wish government to base its policies upon. It is also not really true – in fact there was a highly significant negative relationship between neonicotinoid levels recorded in each nest and how well the nest performed, but the authors removed this relationship by “taking out outliers” – by which they mean removing the nests that received the highest exposure (which also happened to be the nests that performed most poorly) from their data set and then re-ran the tests.

Perhaps the most valuable lesson to be learned from this work is that bumblebee nests placed into the English landscape become contaminated with multiple types of neonic, even when on farms where none are used. Many of the nests had concentration of neonic higher than 1 part per billion, and most had detectable levels of at least two different neonics.  Bizarrely, the results suggest that bumblebees have a particular predilection for the neonic thiamethoxam, for the control bees seemed to have ignored the untreated rape field right next to their nests and flown a long way to find a thiamethoxam-treated crop.

While the debate has focussed heavily on bees so far, there are broader issues to consider. Neonics are very persistent in soils, and evidence has recently emerged that they accumulate over time; a study in East Anglia performed by Bayer in the 1990s found concentrations up to 60 parts per billion in soil after 6 years of annual use. This is likely to be enough to have profound impacts on soil life. It has also emerged that they get into field margin vegetation, and into streams and ponds. A concentration of just 0.6 parts per billion is enough to kill mayfly nymphs. It may be that, while focussing on bees, we have missed the bigger picture. 

The economics

A common argument in favour of neonics is that they provide huge economic benefits, and that the alternatives are worse. A glossy document produced by the “Humbolt Form” (funded exclusively by the agrochemical industry) in early 2013 claimed that a ban on neonicotinoids would cost the EU 17 billion Euros and 50,000 jobs. However, the hard evidence for these claims seemed to be lacking. Indeed, the evidence that neonicotinoids are important for crop production is surprisingly difficult to find. Studies from USA show that yields of soya bean do not benefit at all from neonic seed dressing, despite their application being standard practice. Sadly, similar experiments in the UK have not been conducted. Since farmers get most of their agronomic advice from companies that supply pesticides, it is reasonable to suppose that a good proportion of UK pesticide use may be unnecessary.

Whatever happened to IPM?

When I was at University in the 1980’s, I read Rachel Carson’s Silent Spring, and was taught about the terrible mistakes made in agriculture in the 1950s and 1960s when indiscriminate use of persistent, broad-spectrum insecticides, and an abandonment of traditional cropping practices such as rotations led to huge pest outbreaks. The pest insects had all become resistant, while their natural enemies had largely been eradicated. As a result, an approach called Integrated Pest Management (IPM) had been developed, and we were taught that this was the future of pest control. IPM is predicated on minimising pesticide use: farmers monitor their crop pests, and only take action when necessary; they encourage natural enemies as far as possible, use crop rotations and other cultural controls to suppress pests, and only use the insecticides as a last resort. Even then, they avoid those that persist in the environment.

Whatever happened to this philosophy? Why are we now applying pesticides prophylactically to more or less all crops? Did we learn nothing from our past mistakes?  

The politics

The EFSA spent 6 months evaluating all the evidence, and concluded that current use of neonics poses unacceptable risks to bees. The UK’s Environmental Audit Committee (a group of MPs) agreed. Numerous NGO’s, including the RSPB, who are normally very cautious, also agree. Even the WI support a ban. What was once a radical stance is now where the vast majority of informed opinion lies; on 29 April the EU member states voted for a 2 year moratorium on use of the three most-commonly used neonics on crops visited by bees.

The UK voted against the ban, with Chief Scientific Advisor Sir Mark Walport saying that everyone else had misinterpreted the evidence. He also argued that applying the precautionary principle would lead us to continue using neonicotinoids, an argument that few of us could follow.

What Next?

The moratorium is only for two years, yet we have clear evidence that it will take far longer than this for neonicotinoids that have accumulated in soils to break down. In any case, they are still being widely used, on winter wheat for example, and some types of neonic are not included in the ban.  So it is unlikely that we will see the benefits of the moratorium in rebounding bee populations any time soon. In any case, there seems to be no plan to actually monitor bee populations, so even if they did recover it is not clear how we would know.

What will farmers do without neonics? It is hard to say. It would be nice if they returned to an IPM approach, but with their main / only source of advice being from people with a vested interest in selling chemicals, this seems unlikely. 

Have we solved the bees’ problems?

I’m afraid the answer is an emphatic NO. Bees have been declining for many decades, and much of their decline has been due to loss of flower-rich habitat, which has been exacerbated by the arrival of non-native diseases, and by widespread use of pesticides. We have taken a step to reduce (but not remove) their exposure to some brands of one type of pesticide, for two years. That is nowhere near enough. If we want to ensure healthy populations of honeybees, bumblebees, and other wild pollinating insects upon which we depend for our crop production, and more generally if we wish to support the healthy, diverse ecosystems upon which our future health and wellbeing depends, then we need to find ways to produce food in a sustainable way which incorporates the needs of biodiversity. At present we are failing to do this.


Thank you for sharing your very important research findings.. It makes for an incredibly scary read

By James on Wednesday, May 8, 2013