
IN 401 BC, 10,000 Greek mercenaries were marching home from war against the king of Persia. Passing through Colchis, on the eastern side of the Black Sea, they feasted on honey stolen from beehives dotted around the countryside. Soon thousands had fallen into a stupor. The historian and soldier Xenophon describes them acting like intoxicated madmen, as if under a spell. As strangers to the region, they knew nothing of the effects of the āmad honeyā made by local bees.
Growing skills: A scientific guide to the best gardening practice
Which gardening tips are worth the effort? We prune back some of horticultureās hardiest myths
The men recovered within a few days, but others were not so lucky. In 65 BC, troops led by the great Roman general Pompey against King Mithridates VI of Pontus, also on the Black Sea coast, lay weak and dazed after eating honeycombs strategically placed along their route by the enemy ā and were slaughtered. Mad honey had become a chemical weapon, and one that was to be used to terrible effect in this much-disputed part of the world. In 946, allies of the Empress (later Saint) Olga of Kiev tricked her Russian enemies into drinking mead made from mad honey, and then reportedly massacred 5000 stupefied men. Ivan the Greatās forces adopted a similar strategy in 1489. They left vast vats of toxin-spiked mead in their camp, returning later to wipe out the 10,000 Tatar troops who had stopped to drink it. still surface regularly, although it is rarely fatal.
Mad honey owes its awful powers to toxin-laden nectar. This is a far cry from the ancient Greek notion of nectar as the food of the gods, a golden potion that brought immortality to mere mortals who dared to drink it. But even in ancient times, some knew that nectar had a sinister side. In the first century AD, the great Roman naturalist Pliny identified the nectar of the common rhododendron (Rhododendron ponticum) as the source of mad honey, delicious but dangerous in the hands of those who knew its powers. Today, many dozens of plant species are known to have toxins in their nectar. Plants produce nectar to reward pollinators, so why risk repelling or even killing the animals that ensure their survival? Only now are we beginning to understand the paradox and discover just how cleverly plants manipulate pollinators for their own ends.
Advertisement
Nectar used to be thought of as little more than sugar-water, but closer scrutiny has revealed many more ingredients. These include an array of potentially harmful compounds: alkaloids, terpenes, iridoid glycosides, phenol-based molecules and more. Such chemicals are generally linked to plant defences, and are typically found in roots, leaves, stems and seeds. Their presence in low doses in nectar poses one of botanyās strangest puzzles. Why are they there? Do they simply leak into nectar as they travel around the plant, or do they have a purpose?
From the start, ecologists suspected that the toxins somehow improve a plantās reproductive prospects, and it looks as if they might be right. āEven if the toxins got there through āleakageā, once they were there it seemed likely plants would press them into service,ā says Phil Stevenson, a plant chemist at the Royal Botanic Gardens at Kew, London. āThereās now increasing evidence that plants control what goes into their nectar ā and how much. That suggests these compounds are there for good reasons.ā
Take caffeine, one of the most familiar of plant alkaloids. Coffee-drinkers are well aware of its effects: a morning cup sharpens us up, a night-time cup keeps us awake ā and too much caffeine causes shakes and palpitations. What might caffeine-spiked nectar do to an insect?
At least 15 genera of plants contain caffeine, primarily as a chemical defence against insects, or so it was thought. Caffeine tastes bitter to insects, and bees are no exception; they find it repellent. But this doesnāt deter them from visiting the flowers of coffee and citrus plants, the nectar of which has been found to contain caffeine. It turns out that the caffeine is present at levels too low for bees to taste, as Stevenson and neurophysiologist Jeri Wright at the University of Newcastle, UK, . Yet when offered a choice of nectar with or without a dash of caffeine, bees prefer it with. Wright suspected that caffeine acts as a drug, influencing a beeās mind in much the same way it does ours. She was right.
Mind-bending nectar
In the lab, Wright trained bees to associate nectar with a particular floral scent. When she added caffeine to a drop of nectar, twice as many bees remembered the scent three days later. Closer investigation suggests that caffeine produces this dramatic improvement in long-term memory by intensifying how the beeās brain reacts to information from its antennae, where smells are detected ().
If bees remember which scent or colour indicates good food, they are more likely to return to the same sort of flowers, bringing other members of the colony with them. Plants that use caffeine to manipulate beesā minds in this way should benefit from the greater loyalty of their pollinators. The jury is still out on whether caffeine-laced nectar ultimately increases a plantās reproductive success, but itās looking increasingly likely. James Thomson of the University of Toronto, Canada, recently reported the results of trials with free-flying bees foraging from artificial flowers endowed with fake pollen and fake nectar. Flowers with caffeine-laced nectar received significantly more pollen ().
Messing with a beeās mind to encourage fidelity isnāt the only way in which plants harness their toxic assets to improve the odds of pollination. Tobacco plants enhance their prospects by altering the foraging behaviour of pollinators ā with the help of nicotine.

Hummingbirds are soon repelled by nicotine in nectar (Image: Danny Kessler)
Nicotine is a highly toxic alkaloid made by tobacco plants to ward off caterpillars, leafhoppers and other enemies. Danny Kessler and Ian Baldwin at the Max Planck Institute for Chemical Ecology in Jena, Germany, have spent many years studying Nicotiana attenuata, a wild tobacco from the Mojave desert in the US Southwest. They found that bitter-tasting nicotine repels carpenter bees, visitors that would otherwise consume nectar without pollinating the flowers. They also found that the hawkmoths and hummingbirds that pollinate this species find the taste equally repellent. That, however, turns out to work in the plantās favour ().
Along with nicotine, this wild tobacco spikes its nectar with benzylacetone, a powerfully attractive scent molecule. The scent lures pollinators; the nicotine cuts their visits short, leaving more nectar in the flower to attract new visitors. This chemical carrot-and-stick strategy results not just in the production of more seeds but also in greater movement of pollen between flowers on different plants. The result is a healthy mixing of genes and a greater capacity to survive challenges such as a changing environment ().
Flowering plants have famously evolved ingenious, often extravagant, features to enlist the services of specialist pollinators. Could this explain the nectar of the common rhododendron, the source of mad honey?
Rhododendron ponticum is native to southern Spain and Portugal, and across Turkey to Georgia, but itās in Ireland that researchers are unravelling the role of the nectar toxins that knocked out ancient armies. These belong to a group of nerve poisons called grayanotoxins. R. ponticum was introduced to Ireland in the late 18th century, probably from Spain, and is now naturalised over large swathes of the countryside, where it is a serious pest. Its flowers are visited almost exclusively by bumblebees, with occasional visits from solitary bees, flies, ants and wasps. āYou almost never see honeybees in badly infested areas,ā says ecologist Jane Stout at Trinity College Dublin in Ireland. She teamed up with Stevenson and Wright to find out why.

Bumblebees thrive on the nerve toxins in rhododendron nectar (Image: J. C. Stout)
Tests with captive bees showed that grayanotoxins have dramatically different effects on different types of bee. āThey have no apparent effect on worker bumblebees,ā says Stout. āMining bees show short-term symptoms of malaise. They lie on their backs with their legs in the air but recover later. But honeybees die within hours.ā In honeybees ā and humans ā grayanotoxins hold open the sodium channels present in all nerve and muscle cells, so that neurons keep firing until they are fatigued. In the bees, this leads to palpitations, paralysis and death. āWe donāt know why this doesnāt happen in bumblebees,ā says Wright. āLike any other drug, some animals are more susceptible than others.ā
Poisonous partnership
Stout suspects that grayanotoxins have helped to forge a partnership between rhododendrons and bumblebees. Rhododendron flowers are large, and conceal their nectar in a tubular fold in the back petal. āBumblebees make better pollinators because they are big and hairy and pick up pollen easily. They tend not to groom off the pollen, so it gets transferred to the next flower they visit,ā she says. But if R. ponticum selectively courts bumblebees by offering near-exclusive access to its nectar, how to explain mad honey? Stout suggests that honeybees living within the rhododendronās native range have had time to adapt to its toxic nectar. āThe subspecies of honeybee that makes mad honey in Turkey has probably evolved to resist the toxins in a similar way to the bumblebees.ā
Pollinators visit flowers for food, but what if nectar offers them something besides nutrients? New research suggests that some flowers can be pharmacies as well as food stores, and that some insects might seek out toxic nectar for its medicinal properties.
āSome insects might seek out toxic nectar for its medicinal propertiesā
Many plants produce antimicrobial compounds. In nectar, these probably help suppress microorganisms that consume sugar, so preserving it for pollinators, but they could also stave off disease. āFlowers are like airports, a hub of intense activity. As insects fly from flower to flower, they pick up and deposit pathogens, which can then infect the plant,ā says Lynn Adler at the University of Massachusetts, Amherst. Pollinators, too, are exposed to these parasites and pathogens, and Adler suspects some nectar toxins could reduce the spread of insect diseases.
Until recently, the idea that chemicals in nectar could ātreatā insect diseases was little more than speculation. showed that gelsemine, an alkaloid in the nectar of an American vine called yellow jessamine, has a therapeutic effect on bumblebees. In worker bumblebees infected with the common gut parasite Crithidia bombi, gelsemine reduced the number of parasites by as much as 65 per cent. The parasite, spread via faeces deposited on flowers and in the nest, shortens the life of both individual bees and colonies.
Adler and her colleagues Becky Irwin and Leif Richardson at Dartmouth College in Hanover, New Hampshire, decided to see whether other nectar toxins might have a similar medicinal effect. Testing eight of these on bumblebees, they found that all reduced levels of infection. Four produced dramatic results, reducing the number of parasites by as much as 81 per cent (). āBy reducing parasite loads, these toxins could play an important role in reducing transmission between bees and between colonies,ā says Adler.
But although these studies suggest toxic nectar benefits some plants and pollinators, it could have a darker side. Some pollinators feed on nasty nectar with no ill effects, but other insects may be harmed or even killed. āWhatās one bee speciesā medicine may be another beeās toxin,ā says Adler. As the list of species with toxic nectar grows, ecologists are wondering whether it could be contributing to the global decline in pollinators (see āNatureās poisoned chaliceā).
The ancients may have known enough about nasty nectar to gain an edge over their enemies, but we are still a long way from understanding its role in nature. One thing seems clear, however. If plants harness their noxious assets to medicate, manipulate and mess with the minds of the animals they depend on, then they probably have plenty more surprises in store.
Natureās poisoned chalice
Almost 90 per cent of flowering plants are pollinated by animals, which are rewarded for their services with sugary nectar. With the worldās pollinators in alarming decline, how worried should we be to discover that a wide variety of plants lace their nectar with poisons?
The consensus is that pollinators are disappearing as a result of a combination of stresses ā diseases, parasites, inadequate nutrition and synthetic pesticides such as neonicotinoids, which end up in nectar and pollen. āUntil now, the role of nectar toxins has been overlooked,ā says Phil Stevenson at the Royal Botanic Gardens at Kew, London. āBut added to other sublethal effects, such as poor nutrition, they could be the tipping point.ā
There have been many anecdotal reports of pollinators dying after feeding from certain flowers, and lab studies have revealed that bees fed nectar toxins can suffer a range of harmful effects, from reduced mobility to reproductive failure and lower rates of survival. In natural landscapes, however, nasty nectar probably poses less of a hazard. āJust as bees learn to prefer flowers that offer the best reward, they can also learn to avoid flowers with toxic nectar,ā says Jeri Wright at Newcastle University, UK. If they can taste toxins, they reject the food. āBut they donāt just have to rely on taste,ā she says. āWeāve found they are able to learn what made them feel unwell, and use associated cues ā such as a flowerās odour ā to avoid it next time.ā
Toxic nectar is more of a threat in landscapes where a single crop or invasive plant that produces nasty nectar covers vast areas. Then, a pollinatorās options are limited. āIf bees feed on toxic nectar, they might be poisoned. If they avoid it but thereās little else around, then theyāll suffer from inadequate nutrition. Either way, that could have a severe impact on their colony,ā says Stevenson.
So, how to avoid adding to pollinatorsā many problems? āWe can try to manage the landscape in ways that make it better for pollinators,ā says Stevenson. āThe most obvious way is to ensure there is a wide enough choice of flowers to satisfy the needs of a diversity of bees and other pollinators.ā
This article appeared in print under the headline āSweet deceitā