Taking the sting out of the honeybee controversy

Moving this kind of situation forward requires a procedure similar to that used in law courts, which gives voice to all the evidence and allows unfounded statements to be highlighted. Now for the first time, Laura Maxim from the Institute for Communication Sciences, CNRS in France and Jeroen van der Sluijs from Utrecht University in the Netherlands have developed such a structure. They put it to the test on the controversy over French honeybee losses.

Major declines of honeybee colonies have been reported in France since 1994. Some years later other parts of Europe and America experienced the phenomenon. Aside from the loss of revenue in honey sales this is worrying news because honeybees are one of the world's most important pollinators – 35% of agricultural crops rely on pollinators. France was one of the countries worst hit, with a reduction in sunflower honey yield of up to 70% in the apiaries affected.

During the crisis many French beekeepers and scientists involved in public research pointed the finger at Gaucho, an insecticide that was first used in 1994 to treat or "dress" sunflower seeds before planting. Meanwhile, representatives from Bayer Cropscience – the manufacturers of Gaucho – and the French Food Safety Authority (AFSSA) argued that there was no obvious causal relationship between Gaucho and the honeybee decline, and said that there were many other factors that could also be to blame. The French Ministry of Agriculture said that it considered Gaucho as one of a number of possible causes, although the Minister of Agriculture did ban its use in sunflower and maize seed-dressing as a precaution.

Maxim and van der Sluijs drew up the following procedure to try to understand why there is a difference between those arguing that Gaucho was the key causal factor in Gaucho-treated sunflower and maize areas, and those arguing for "many causes". First they introduced all the participants to an uncertainty scale from zero to ten, in which a score of ten is given to evidence that is convincing beyond any doubt, a one equates to "no reasonable grounds for suspicion", and a zero corresponds to impossible.

Then the pair drew up a list of factors that were considered to be potential causes for the honeybee decline. As well as the effect of Gaucho, these included unfavourable climatic conditions, diseases and viruses, insufficient pollen because of a rise in intensive agricultural practices, illegal use of pesticides, imported queens, which may not have adapted to local conditions, and changes in sunflower varieties.

Next the researchers drew up a list of precisely described signs of decline, including where and when they occurred. Finally they drafted a list of eight key questions designed to tease out the causality between the damage to the honeybee colonies and the signs seen in the field.

Stakeholders from both sides of the debate were asked the eight questions and to estimate their degree of uncertainty, using the prescribed scale, with their answer. Afterwards Maxim and van der Sluijs measured the statistical significance of the difference in answers given by different stakeholders. The findings are published in Environmental Research Letters.

By following this process Maxim and van der Sluijs revealed that the arguments of beekeepers and scientists involved in public research were referring to particular geographical areas and crops (Gaucho-treated sunflower and maize areas) and very specific signs, such as disappearance of honeybees during the sunflower and maize flowering season, whereas AFSSA and Bayer were generally referring to "honeybee losses" without being precise in their descriptions. In the view of Maxim: "This impreciseness made the debate appear much more confusing than the facts actually were".

In this case Maxim and van der Sluijs say that the heat was taken out of the argument and common ground found. They believe that science and fact were clearly separated from opinion, and rationality returned, helping people to take a broader view of the issue and move forward.

Crucially the findings should help to prevent a similar stalemate arising in the US. "We believe that the framework presented in our paper can help people from the US in addressing honeybee losses in their country," said Maxim.

It isn't just honeybees that stand to benefit from this process. The framework could be useful when applied to other debates in the environmental sciences. "Our method is generic and can be used for other similar cases where multiple potential causes have to be assessed regarding their relative contribution to a particular effect, such as climate change," said Maxim.

But the scientists warn that the timing of such an intervention is key. "For our analysis to play a fruitful role it is essential that constructive discussion among stakeholders having divergent views is still possible or that the analysis is carried out during a calm period in the controversy (for long controversies taking place over several years)," said Maxim. The earlier that this process is applied in a controversy, the better the likely outcome.