Bees and neonicotinoids – another twist in the tale.

To the agrochemical companies and to many farmers they are essential tools ensuring efficient crop production. To environmentalists and to many bee scientists they are dangerous chemicals contributing to declining bee populations. I am talking, of course, about the neonicotinoid insecticides widely used in this country to control insect pests.

Last week two papers were published on-line in the journal Nature emphasising the dangers posed by these chemicals.

One paper, from a team at Newcastle University, investigated whether honeybees and bumblebees showed any preference (positive or negative) for food containing neonicotinoids; there had been suggestions that bees might avoid neonicotinoid-treated crops in the field. The new lab-based work showed that, when offered a choice, bees preferred to eat sucrose solution (nectar) containing neonicotinoids (imidacloprid or thiamethoxam) rather than control sucrose solution. If this occurs in the field then bees may forage preferentially on crops containing chemicals toxic to their health and inadvertently bring back these toxins to their nests.

The second paper, from scientists at Lund University in Sweden, showed that the neonicotinoid clothianidin, when used in the field, damaged wild bee (bumblebee and solitary bee) populations but was without effect on honeybees. I want to focus on this paper as it is the first controlled study of the effects of a neonicotinoid on honeybees, bumblebees and solitary bees under field conditions. Let’s start by looking at the background.

Three years is a long time in science

Back in 2012, two studies were published showing that neonicotinoids, even when they didn’t kill bees, could affect bee behaviour in a way that impaired survival of honeybee and bumblebee colonies. The findings showed that these chemicals could contribute to a decline in the numbers of bees and other pollinators. These observations had a big effect on policy and indirectly contributed to the current partial ban on the use of three neonicotinoids in Europe.

One of the papers was from Dave Goulson’s lab, then at Stirling, and found that even at the low doses typically encountered on treated crops in the field, neonicotinoids substantially reduced the number of queens produced by bumblebee colonies, so impairing survival. In Goulson’s study, bumblebees were treated with neonicotinoids in the lab before being allowed to fly freely. This is called a semi-field design and some have suggested that the findings cannot be extrapolated to the real world. The levels of neonicotinoids have also been criticised although these were very carefully thought through.

There was considerable media interest in this work, nicely described in Goulson’s latest book (A Buzz in the Meadow) and the UK Food and Environment Agency (FERA) labs in York hastily set up a preliminary field trial to examine the problem. During the spring and summer of 2012 they studied three fields of oil seed rape, one treated with the neonicotinoid, imidacloprid, another treated with the neonicotinoid, clothianidin and a third untreated field to act as a control. Bumblebee (Bombus terrestris) nests were placed by each field and the occupants were left to fly freely and build their colonies. Extensive analyses were performed including colony growth, the pollen collected by the bees and pesticide residues detected in pollen and nectar.

Superficially this sounds like a rigorous study (free flying bees, compare neonicotinoid-treated with control, plenty of analyses and so on) and it should have given an indication of the effect of neonicotinoids under real-world field conditions. The success of the trial depended on the bees foraging on the crop near their nests so there should have been clothianidin-exposed colonies, imidacloprid-exposed colonies and neonicotinoid-free controls. Unfortunately the bees had other ideas; they largely ignored the flowering oil seed rape by their nests, foraging further away and returning with pollen from crops treated with another neonicotinoid, thiamethoxam. This completely negated the original design so that, in my view, no valid conclusions can be drawn from the study, despite official pronouncements.

The new Swedish study

The FERA work was a valiant but flawed attempt to study the effects of neonicotinoids on bumblebees under field conditions and it was clear that a properly conceived field trial was required. The ambitious new study from Lund University mostly fills that void.

The Lund team selected 16 landscapes surrounding spring-sown oil seed rape fields in different parts of southern Sweden in 2013. The landscapes were divided in to 8 pairs on the basis of the nature of the surrounding countryside and for each pair one was sown with seed treated with the neonicotinoid, clothianidin and the other was sown without neonicotinoid, to act as a control. All seed also contained a fungicide. The neonicotinoid-treated seed also contained a pyrethroid to protect plant roots but there was no evidence that the bees came in to contact with this chemical.

Honeybee hives, bumblebee (Bombus terrestris) colonies and solitary bee (Osmia bicornis) cocoons were placed by each field during the flowering season and their progress was followed. Additionally, the numbers of free flying wild bees (bumblebee and solitary bee) were assessed at each field during the flowering season. Pollen brought back by bees was analysed for flower type and for pesticide residues. The study was large enough to allow statistical analyses to be performed so that valid conclusions could be drawn although the design was unable to detect effects smaller than 20%.

Unlike the FERA study, bees did forage on the oil seed rape adjacent to their colony based on pollen analysis, although this was not an exclusive choice. Bees near treated fields brought back pollen and nectar containing clothianidin whereas those near untreated fields did not. The study design seemed to have been successful and the results were clear:

1. Honeybee hives behaved similarly by treated and untreated fields; there was no effect of the pesticide on colony strength. This is good news for honeybees and for beekeepers and agrees with a 2014 field study from Canada also using clothianidin.

2. Bumblebee colony growth and reproduction were reduced near treated fields. This agrees with earlier semi-field studies so that it is now difficult to avoid the conclusion that agricultural use of neonicotinoids has damaging effects on bumblebees.

3. Whereas solitary bees placed by untreated fields emerged from cocoons and built new nests, this did not occur for cocoons situated near treated fields, possibly because of navigational problems caused by insecticide exposure.

4. The number of free-flying wild bees was reduced by about 50% at the treated fields.

The implications of the new Swedish study

Honeybees were not affected by foraging from a neonicotinoid-treated crop whereas wild bees (bumblebees and solitary bees) suffered reductions in numbers and reproductive ability. Wild bees are, therefore, more sensitive to neonicotinoids than honeybees. This may be because, as James Cresswell at Exeter has shown, honeybees break down neonicotinoids more quickly than bumblebees so that honeybees experience lower doses.

The effect of the neonicotinoid on wild bees is an important result for several reasons. Wild bees are important pollinators, contributing more than half of the “pollination service” required for crops. Crop yield and quality will suffer if these insects are lost so we need to look after them and that may need to include rethinking use of pesticides.

The differential sensitivity of bee types to neonicotinoids shows that environmental risk assessment of new and existing insecticides, typically performed in short term lethal studies on honeybees, is inadequate to determine long term effects on different bee types in the field.

These new data will add to the pressure to extend the European moratorium on the use of three neonicotinoids as seed dressings after the initial two year period ends in December. There is, however, concern that if neonicotinoids are not available then farmers will use older insecticides that may be more dangerous for bees. Perhaps farmers should rethink their use of pesticides and return to a “treat when required” policy rather than using neonicotinoids prophylactically.

When the three neonicotinoids were temporarily banned in 2013, farmers claimed that the yields of crops would be drastically reduced. It is interesting that in the first season when crops were grown from untreated seed in the EU, the yields actually increased. Perhaps we should rethink the use of chemicals more generally in modern farming.