Skip to content
NatureHQ

Claim check

Do all plants need bees?

Exaggerated

Something true sits underneath, stretched well past what was shown.

Around 88% of flowering plant species depend on animals — but the animals are tens of thousands of species of bee, fly, beetle, moth, butterfly, bird and bat. And the crops supplying most human calories are wind-pollinated: wheat, rice and maize need no animal at all. The finding underneath is more alarming, not less: the pollination that matters is largely done by wild insects nobody counts.

The claim as it circulates

“Without bees, plants cannot reproduce and a third of our food would disappear — bees pollinate everything.”

Where you may have met it: Conservation campaigning and "save the bees" fundraising; Food-security journalism, usually with a striking percentage attached; Product marketing, where buying something is linked to bee survival

Taking it apart

What was claimed
That bees are necessary for plant reproduction generally and for a large fraction of the human food supply, such that losing bees would remove both.
What was actually observed
A survey of community-level data puts animal pollination at roughly 88% of flowering plant species, higher in the tropics and lower in temperate regions. The staple cereals are wind-pollinated. Across some 600 fields in 41 crop systems worldwide, wild insect visitation raised fruit set in every system, while honey bee visitation did so in 14% — and adding honey bees did not substitute for the wild insects.
What the evidence supports
That animal pollination is widespread and important, and that its loss would fall heavily on fruits, nuts, oilseeds and vegetables — the nutritionally richest part of the diet. That wild pollinator diversity has value managed colonies cannot supply.
What it does not support
That bees means honey bees: the honey bee is one managed species among roughly twenty thousand bee species, and is not the most effective pollinator of many crops. That plants generally need bees — most flowers are visited by many kinds of animal, and some need vibration a honey bee cannot produce. That staple calories are at risk from pollinator decline.

The awkward part of this is that correcting it can sound like a reason to care less, and it is the opposite. Honey bee colony numbers globally have risen over recent decades, because honey bees are livestock and people keep more of them. The species in genuine trouble are wild — solitary bees, hoverflies, moths — and they are the ones doing the work the campaigns are describing. A story about honey bees can be resolved by installing more hives, and installing more hives does not help, and can compete with the wild insects that would.

The other detail usually lost is that some plants cannot use a honey bee at all. Tomatoes, aubergines and blueberries hold their pollen in tubes that open only to vibration, and a honey bee cannot produce it. On those flowers a honey bee is not a substitute for a bumblebee; it is an animal that cannot open the door.

Go deeper

  • Pollination

    Who actually pollinates what, and how it is measured

  • Bumblebees

    The buzz pollination honey bees cannot do

The claims underneath

Each one carries its own evidence, scope and caveats. Expand any of them to reach the studies.

About 88% of flowering plant species need animal pollinators — and wheat, rice and maize do not

Well supported

Good evidence backs this, though some details remain open.

Community-level surveys across latitudes indicate that approximately 87.5% of flowering plant species are animal-pollinated, rising to about 94% in tropical communities and falling to around 78% in temperate ones. The wind-pollinated remainder includes the grasses, and therefore the cereals that supply most human calories.

Who this applies to
flowering plant species globally
Studied in
Angiospermae

You may have heard

“Without bees we would starve”

Two errors stacked. The cereals supplying most human calories are wind-pollinated and would be unaffected; what animal pollination supplies is much of the variety, nutrition and interest in a diet, which is a serious loss and not starvation. And "bees" undersells it — the pollinators concerned are tens of thousands of species of bee, fly, beetle, moth, butterfly, bird and bat, and the honey bee is one managed species among them.

Why we rate it this way, and what the caveats are
Well supportedHigh confidence

A systematic estimate from community-level data with an explicit method and stated uncertainty, and the figure most cited in the pollination literature since.

How far it can be extended

Estimated from community surveys spanning tropical and temperate regions, with a consistent latitudinal gradient.

Caveats

  • A count of species, not of abundance, biomass or crop yield, which behave differently.
  • "Animal-pollinated" covers everything from obligate dependence to marginal benefit.
  • Community surveys are unevenly distributed geographically.

Still unanswered

  • How much does the species-level figure translate into yield dependence for real agricultural systems?

Last reviewed 2026-08-10

The evidence (2 studies)

An insect on a flower may be pollinating it, robbing it, or doing nothing at all

Established

Specialists would state this without hedging. Multiple independent lines of evidence agree.

Pollination requires viable conspecific pollen to reach a receptive stigma. Flower visitors vary from effective pollinators to nectar thieves and robbers that bypass the reproductive structures entirely. In buzz-pollinated species — roughly 6% of flowering plants, including tomato and blueberry — pollen is released only by vibration, which honey bees essentially cannot produce, so a honey bee visit transfers nothing.

Who this applies to
flower visitors generally; buzz pollination demonstrated in Solanum and relatives
Studied in
Solanum, Bombus, Apis mellifera

You may have heard

“Bees pollinate everything”

A honey bee can sit on a tomato flower all day and pollinate nothing, because the pollen only comes out under vibration and it cannot vibrate. Commercial tomato growing depends on bumblebees for exactly this reason. Counting insects on flowers measures visits, and visits are not pollination.

Why we rate it this way, and what the caveats are
EstablishedHigh confidence

Buzz pollination provides a mechanical demonstration that cannot be argued around: a bee physically unable to vibrate at the required frequency cannot extract the pollen, however often it lands.

How far it can be extended

The visitor-versus-pollinator distinction is general across pollination ecology; the buzz-pollination case is the cleanest documented instance.

Caveats

  • Effectiveness varies by flower, visitor, weather and time of day, so single-visit measures generalise poorly.
  • Nectar robbing is not always costly to the plant, and can be neutral.
  • Honey bees are effective pollinators of many crops; the point is that visits are not interchangeable.

Still unanswered

  • How much of measured "pollinator visitation" in survey data represents actual pollen transfer?

Last reviewed 2026-08-10

The evidence (4 studies)

Go deeper

Pollination

A honey bee can sit on a tomato flower all day and pollinate nothing. It cannot buzz.

Last reviewed 2026-08-10