Sprayed flowers still lure bumblebees – but one sip can steal their drive

A bumblebee crouches on a violet petal, its black eyes glinting and its fur dusted with golden pollen. The bloom beneath it seems fresh and full of life – but it may have been sprayed with herbicide a day or two earlier. Flowers can still look healthy and inviting for up to five days after being treated.

As the bee drinks, invisible traces of glyphosate slip into its body. It will fly again, yet its rhythm may falter – fewer flowers visited and less food returned to the nest.

“I wanted to know whether bumblebees really visit plants that have been sprayed with glyphosate,” says Kimmo Kaakinen from the University of Turku, Finland. “Everyone assumed that exposure happens, but no one had actually proven it.”

Since then, a careful field study with bumblebees has shown that they really can pick up glyphosate on and around flowers – but that study used pure glyphosate under controlled conditions. Kaakinen wanted to know what happens with bumblebees on real plants sprayed with the kinds of herbicide mixtures used in practice.

That question – will bees land, and what happens after they drink? – became the starting-point for Kaakinen’s PhD project. In field plots and flight cages, his team found that bumblebees showed no hesitation: they visited sprayed blooms as eagerly as clean ones. The flowers’ colour and scent still signalled food, but just one field-realistic sip of glyphosate-laced nectar cut their motivation to forage by half.

“Herbicides are meant for plants, not insects, so we have long assumed pollinators are safe,” Kaakinen says. “Our results challenge that assumption – not because glyphosate kills bees but because it changes their behaviour.”

Recently sprayed flowers, he concludes, may look harmless – yet they can quietly drain the strength of an entire colony, weakening one of nature’s most tireless workforces.

An invisible story between plants and pollinators

Every weed sprayed writes an invisible story between plants and their pollinators – and this study adds a troubling new chapter. It speaks to one of the greatest questions in ecology today: how the world’s most common chemicals are reshaping the creatures that keep it blooming.

Even far from farmland, exposure is easy to imagine. In Europe, glyphosate is not sprayed on flowering genetically modified crops, and yet traces linger across ordinary landscapes – from gardens and roadsides to forest edges.

In many countries, glyphosate has also been widely used to clear weeds in public areas and along railway lines, although current practices can differ from place to place. What matters for insects is that treated flowers and weeds in these everyday landscapes can still carry residues within hours or days after spraying.

The issue is anything but theoretical. Glyphosate’s licence in Europe was renewed in 2023 for a further 10 years, so it is now authorised well into the 2030s. And the stakes reach far beyond a few unlucky bees: insect pollination sustains about one third of global crops and nearly all wild flowering plants, and glyphosate-based herbicides are sprayed over hundreds of millions of hectares each year.

Understanding how such a widespread chemical might dull pollinator activity is not just about bee health – it touches food security itself.

Why a supposedly safe weed killer may not be safe after all

Insecticides are often blamed for the global decline of pollinators – but herbicides are used in far greater amounts. The most common of all is glyphosate, the active ingredient in Roundup, found in hundreds of weed killers worldwide.

Glyphosate works by blocking a chemical route in plants – the shikimate pathway – which they need to make essential amino acids. Animals do not have this pathway, so glyphosate was long considered harmless to insects and other wildlife.

“Safety by pathway is an appealing idea,” Kaakinen says. “But it does not rule out ecological side-effects – such as changes in floral scents or subtle shifts in bee behaviour.”

Bees, he explains, navigate the world through sight and smell. Floral scents – made up of dozens of volatile organic compounds – act like olfactory road signs that guide them toward nectar and pollen. Agrochemicals can blur those signals, changing how pollinators perceive their surroundings.

“We wanted to separate two things,” Kaakinen says. “Are bees exposed because they cannot tell the difference – or because the difference does not matter to them? And what if exposure happens inside the bee – does that change how keen they are to forage?”

How bees really meet glyphosate

Bees can encounter glyphosate in many ways – from drifting spray to traces left on petals, pollen or nectar soon after application. Sprayed plants do not die immediately; their blossoms often stay bright and fragrant for several days while still looking like perfectly normal flowers to an insect.

“That creates a short but realistic exposure window,” says Kaakinen. “The flowers still look and smell inviting, so bees keep visiting – and each visit means a mouthful of glyphosate.”

Earlier studies with honeybees had hinted at similar effects – reduced learning and altered feeding after glyphosate exposure – but none had tracked how it unfolds in real foraging. “What is new here,” Kaakinen explains, “is that we tested real field doses and real flowers with bumblebees. This is the first time anyone has followed exposure as it actually happens in nature.”

Even tiny amounts can ripple through a colony. When just a few workers slow down, brood feeding and temperature control begin to falter.

“The big ecological question is not whether glyphosate kills bees outright,” Kaakinen says. “It is whether it dulls their motivation just enough to slow food collection, stunt brood growth or weaken colonies when resources are tight.”

Will bees land? The field test

To trace how bumblebees meet glyphosate, Kaakinen and his colleagues built a series of experiments – from open meadows to flight cages and scent chambers – each asking a deeper question: Will bees land? Can they smell the difference? And what happens once they drink?

“First, we asked the simplest question: will bees land on treated flowers?” Kaakinen says. “Everyone assumes they would not, but there were no data.”

In a flowering meadow, the team marked small one-square-metre plots. Half were sprayed with a commercial glyphosate-based herbicide at the normal farm rate; the rest were sprayed with water.

“One hour later and over the next three days, we counted every flying insect pollinator that landed on those plots,” he explains. “We then looked at the bee numbers separately, because they were the main focus of our study. The result was clear: treated squares drew just as many visitors as controls. Temperature mattered – treatment did not.”

When scent and sight deceive

To remove background distractions such as wind, competing flowers and wild scents, the researchers moved into semi-natural flight cages filled with potted oilseed-rape – a crop bees cannot resist. Two plants were sprayed, two were left clean and the bees were tested about 24 and 48 hours later.

“We tracked which plants each bee chose and how many flowers it visited,” Kaakinen says. “Again, spraying made no difference. The only thing that predicted visits was how many flowers were open – not whether they had been treated.”

Next, the team tested scent alone. Using a transparent Y-shaped maze, they channelled air from bagged treated and untreated plants to see which aroma bees preferred.

“They split almost perfectly fifty-fifty,” Kaakinen says. “So even if the floral scent changed, it was not enough for them to care.”

The exhaustion test

Chemical analysis showed that the flowers’ scent did change – but only slightly. Using gas chromatography–mass spectrometry, a technique that separates and identifies scent molecules, the team found that three compounds – alpha-farnesene, beta-ocimene and myrcene – dropped after 24 hours, while the overall bouquet stayed the same.

“It is a subtle modulation of the bouquet, not a switch-off,” Kaakinen says. “The flowers still smell like food.”

Then came the crucial step: oral exposure. Individually marked bumblebees were fed sugar water containing a realistic field dose of glyphosate and released into cages filled with oilseed-rape – either alone or alongside blue lupine, a large, irresistibly fragrant plant.

“This is when we saw the strongest effect,” Kaakinen says. “Exposed bees were roughly half as likely to forage at all.”

The difference was striking: normal bees darted between blossoms, while treated bees barely moved. “It is not poisoning,” he says quietly. “It is more like exhaustion.”

A silent slowdown

The results confirmed what Kaakinen had suspected all along: bumblebees do not avoid glyphosate-treated flowers – and once they ingest the herbicide, their behaviour changes dramatically.

“In every set-up, the bees kept visiting sprayed flowers,” he says. “Even when untreated plants were right beside them, they did not discriminate.”

Field data told the same story: treated and control plots attracted equal numbers of bees. Temperature influenced how active they were – but treatment did not. Inside the flight cages, the pattern held. Bees visited sprayed and clean plants alike. “The two kinds of plants seemed equally rewarding,” Kaakinen says. “Their colour and scent cues just had not changed enough to matter.”

And when the bees did feed, the effect was unmistakable. “They were not dead – just slow and unwilling to work,” Kaakinen says. “More like exhaustion than poisoning.”

Hidden risks in bloom

Chemical analysis backed up the observation. Although three scent compounds fell slightly after spraying, the overall floral bouquet stayed rich. “It is a mild detuning, not a warning signal,” Kaakinen says. “The flowers still smell like food.”

Then the picture shifted. Once the bees actually ingested glyphosate, their drive to forage collapsed.

“Those bumblebees were really, really passive after acute exposure,” Kaakinen recalls. “They were not dead – just slow and unwilling to work. And they were supposed to be hungry when we freed them, so it was not that they were full – they just lost motivation.”

In numbers, only about half as many exposed bees bothered to forage compared with controls. Even the presence of blue lupine – a large, vividly coloured plant that normally acts like a magnet – did not change that.

“Normally, lupine pulls them straight in,” he says. “But after glyphosate, they just ignored both options.”

Beyond survival: when motivation fades

The team also noted that sprayed oilseed-rape plants stayed alive and blooming for several days after treatment – a period when they still looked healthy and inviting.

“That is the risky phase,” Kaakinen says. “The plants appear fine, the bees visit freely and each sip delivers glyphosate.”

Together, the findings tell a clear story: bees cannot detect danger and keep foraging – until a small oral dose dulls their motivation.

“We have speculated everything from gut to brain effects,” Kaakinen says. “Something in the nervous system may be involved – but no one has studied that yet.”

“It is not that they get lost,” Kaakinen concludes. “They just lose the motivation to work. A few quiet bees today can mean an underfed colony tomorrow.”

Timing is everything

The message from the study is simple but sobering: recently sprayed flowers still attract bees, and even small doses of glyphosate can make them work less. For farmers and land managers, that creates a short but critical risk window after spraying.

“If herbicide use is unavoidable, timing is everything,” says Kimmo Kaakinen. “It should be done when pollinators are not flying – in early spring or late autumn – and never near flowering field edges or wildflower patches.”

He stresses that he is not against herbicides themselves, only their careless use. “I am not opposed to herbicides,” he says. “They are sometimes necessary – just not during pollinator flight.”

The findings also expose a blind spot in current safety testing: risk models focus on survival and not behaviour. “Even short pauses in foraging can ripple through a colony,” Kaakinen notes. “Fewer trips mean less food – and greater stress.”

Exposure beyond the fields

Kaakinen points out that most exposure does not happen in crop fields at all. “In the European Union, glyphosate use on flowering crops is already restricted,” he says, “but bees still visit roadside weeds, gardens and field edges. Those can easily become exposure hotspots.”

The same pattern appears elsewhere. In regions that allow pre-harvest spraying or grow glyphosate-tolerant crops, the contact window is far wider – and so are the ecological risks.

“Glyphosate might actually be the least harmful option among herbicides,” Kaakinen admits. “But its effectiveness is also its biggest danger – it kills all the weeds, and then there is nothing left for pollinators. It is ironic,” he adds. “The same quality that makes it a powerful tool for farmers can quietly erase the food base for bees.”

The quiet collapse of communication

The scent results also open a subtler line of inquiry into plant–pollinator communication. The floral bouquet changed slightly but not enough for bees to care. “It shows how resilient – and how vulnerable – this communication is,” he says. “Bees use colour, shape, smell and memory together – and sometimes the buffet wins over the bouquet.”

What remains uncertain is how long the behavioural effect lasts. “We saw changes within a day,” Kaakinen says. “But does the apathy fade after a few hours, or persist for the rest of the bee’s life? We do not yet know.”

For now, the takeaway is pragmatic: avoid spraying when pollinators are active, buffer margins against drift and reconsider herbicide use in public and natural spaces.

Kaakinen hopes that future studies will broaden the focus beyond bees. “It would be so interesting to study hoverflies,” he says. “They play a big role in pollination too, but they are not studied as much.”

“Pollination is one of nature’s quietest but most valuable services,” he concludes. And in a world in which one of every three bites we eat depends on bees, that lost motivation may become the costliest harvest of all.