The most widely used insecticides in the world—neonicotinoids—turn bumblebees into more bumbling pollinators, leading to lower yields for apples and perhaps other plants that they visit. That’s the message from a new study by Dara Stanley at the Royal Holloway University in London, the first to show that neonicotinoids affect the plants that rely on bees, as well as the insects themselves. It’s the latest in a damning and growing body of evidence about the negative consequences of these chemicals.

Bees and other pollinators help to fertilize around seven in eight species of flowering plants, including many crops. They play invaluable roles in nature, and their contributions to human agriculture are enormous. “One in three bites of food we eat comes from crops pollinated by bees,” says Stanley.

But “pollination services,” as they are often called, might be suffering because of neonicotinoids. These chemicals are systemic pesticides: Rather than just sitting on the surface of plants, they are absorbed and transported to all kinds of tissues, including nectar and pollen. That’s where bees and other pollinators can encounter them. And even if the neonicotinoids aren’t present at anywhere near a lethal dose, they can still affect insect behavior.

They impair a bee’s memory and its ability to navigate, reducing the foraging skills of workers, and creating a vicious cycle that hobbles the colony’s workforce. Ineffective workers means a shortfall of food, which means that workers are sent on more foraging trips at the expense of caring for youngsters, which means that the colony produces fewer new workers, which means even less food, and so on. This slows the growth of existing colonies and, by reducing the production of queens, prevents the creation of new ones.

“Much of this previous work looked at things from the bees' perspectives,” says Stanley. But she wanted to know if the plants are also affected.

For two weeks, her team fed bumblebees on nectar spiked with extremely low levels of a common neonicotinoid called thiamethoxam. They then released the bees in outdoor cages containing virgin apple trees, which had never been visited by pollinators before.

The team found that bees that drank from the spiked nectar actually spent more time foraging, and visited more flowers. That ought to make them better pollinators, but not in this case. They didn’t succeed in collecting any more pollen and the flowers they visited didn’t produce more apples. Perhaps they were behaving differently on the actual flowers, or failing to learn from experience as bees normally can. Either way, the neonicotinoid-sipping bees became busier, but no more effective.

And even though each individual was more active, the insecticide-treated colonies deployed fewer foragers in total. As a result, the apple trees received fewer visits and produced fruit with 36 percent fewer seeds—a sign of poorer quality.

These changes were obvious when the bees’ nectar contained 10 parts per billion of thiamethoxam, but not when it contained 2.4 parts per billion. This suggests that dose matters, and that there could be a level where bees provide the same level of pollination services. But that’s academic, says Stanley, because “these are levels that have been actually measured in the nectar and pollen of treated crops.”

Apples aren’t just pollinated by bumblebees, but by many species of solitary bees that seem to be even more sensitive to neonicotinoids. It's possible, then, that Stanley’s results underestimate the costs that these insecticides exert upon apple orchards, especially since one study estimated that poor pollination already costs U.K. apple growers around £5.7 million a year.

“The obvious conclusion is that farmers using these chemicals could potentially experience reduced yields, as could their neighbors who may not be using the chemicals,” says Dave Goulson from the University of Sussex, who was not involved in the study. “There may also be knock-on effects for pollination of wildflowers growing on or near farms.” Indeed, if bumblebees are providing poorer pollination services to apples, they could also be short-changing others crops like beans, berries, tomatoes, and oilseed rape, not to mention legions of wild species.

Maj Rundlöf from Lund University says the study has two major limitations. It was “conducted in cages and on potted apple trees, so it is uncertain if the results can be transferred to more realistic situations.” Also, it used “a very artificial route of exposure where the bees are forced to feed from a contaminated source.” Still, Rundlöf praises the work and says it provides “a strong indication that farmers should consider both the gains and the costs of pesticide use when selecting pest-management strategy.”

Both the gains and costs have been hotly contested. Pollinators aside, neonicotinoids have been recently linked to declining populations of insect-eating birds and aquatic invertebrates. Meanwhile, the U.S. Environmental Protection Agency released a report in 2014 claiming that the insecticides provide “little or no overall benefits to soybean production in most situations.” And an independent assessment in the U.K. suggested that crops of oilseed rape have experienced a bumper yield despite a recent moratorium on the use of neonicotinoids. These findings have been opposed by farmers unions and pesticide manufacturers, and both the U.S. and U.K. governments have been accused of gagging entomologists and pesticide experts.

Still, in a few places, the pendulum has swung towards greater regulation. In December 2013, the European Union voted for a two-year ban on three neonicotinoids on crops that are attractive to bees. In July 2015, Ontario, where nearly all corn and most soybean seeds are treated with these pesticides, enacted a new law aiming to reduce usage by 80 percent within two years.

It's hard to say what effect this will have on bees. We should view the impact of neonicotinoids in the context of the broader pollinator declines that we’ve been concerned about for some time,” says Nigel Raine from the University of Guelph, who led Stanley's new study. “There’s not a single issue that’s affecting them. Are there enough flowers? And nesting sites? Are those sites well enough connected in the landscape? Are the pollinators dealing with nutritional stress, or parasites?”

“These factors are all quite complex,” he adds. “But you can see why pesticides are being looked at because that’s something we are adding to the system. We might be more able to easily regulate that.”