Which crops need pollinators, and which bees do the work

Dependence is a sliding scale

Every major assessment of crop pollination sorts crops into the same five classes, first set out by Klein and colleagues in 2007. They measure how far a crop's production falls when animal pollinators are excluded: essential means by 90 per cent or more, great by 40 to 90, modest by 10 to 40, little by less than 10, and none means no increase from animal visits at all.

Two things stand out. The four largest crops in the table by tonnage — maize, rice, wheat and potatoes — sit in the 'none' class, which is why a world without pollinators would not run out of calories. And the crops that do depend on pollinators are not all small: soybean, grown on more than a hundred million hectares, is classed as modestly dependent, and apples, cucumbers and mangoes as greatly dependent. Where the aggregate picture of pollinators and food comes from — and why the famous 'one bite in three' overstates it — is covered in the guide on why bees matter.

Needing bees is not the same as renting them

Whether a crop needs bee pollination is a question about its flowers. Whether a grower pays for managed colonies is a question about a farming system in a place. The USDA's crop-by-crop record for the United States keeps the two apart, and they disagree more often than the usual story suggests.

• Oranges and strawberries are recorded as not requiring bee pollination, and managed pollinators are used on both — for some orange cultivars and for honey, and on strawberries to complement wind pollination.
• Tomatoes are recorded as requiring bee pollination and are unattractive to honey bees. In glasshouses the managed pollinator is the bumble bee, which can vibrate the pollen out of the flower.
• Alfalfa seed is pollinated commercially by alfalfa leafcutting bees and alkali bees as well as honey bees, and the forage crop is cut before bees matter at all.
• Almonds depend on cross-pollination for most cultivars. The IPBES assessment cites around 1.7 million hives moved into California for a six-week bloom — the largest managed pollination event there is, and a reason the almond crop is exposed to anything that affects honey bee supply.

Why the managed honey bee is not the whole answer

The honey bee is the only pollinator most farmers can order, and that makes it tempting to treat pollination as a logistics problem. The field evidence synthesised in the IPBES assessment does not support that. In a worldwide study of 41 crops in 600 fields, visits by wild bees increased fruit and seed set, on average, about twice as much per visit as visits by honey bees, and stocking fields with more hives did not make up for fewer wild bees — though honey bees added to what the wild bees achieved.

Wild pollinators come from somewhere, and distance matters. Across studies, the rate at which bees visited crop flowers fell by half at about 670 metres from natural vegetation, and the number of bee species by half at about 1,500 metres. A larger analysis of 20 crops in about 1,400 fields found that most of the wild-bee contribution came from a subset of common species — which argues for keeping ordinary farmland bees abundant rather than for treating crop pollination as the reason to conserve every rare one.