Revealing the sneaky genius of bees

In a mountain valley in northern New Mexico, both humans and bees were preparing for winter. At Addelina Lucero’s adobe home in the Native American community of Taos Pueblo, a huge pile of firewood lay beside two stacks of rectangular beehives caked with red propolis, a resinous mixture that bees use to seal their homes for the coming months of cold and scarcity.

Indigenous beekeeper and National Geographic Explorer Melanie Kirby donned a lightweight mesh veil and pried open one of Lucero’s hives, pulling out the hanging frames and examining them to locate the colony’s queen, which was twice the size of the workers. “This is our mother breeder,” she said as Lucero looked on. “She’s survived many seasons.”

That’s sadly uncommon these days. Bees around the globe have been dying off in massive numbers, suffering an onslaught of woes: parasites, pathogens, pesticides, dwindling habitat. We’ve learned of these declines mostly from honeybees, whose lives are intertwined with human agriculture because beekeepers haul millions of hives to pollinate everything from almonds to apples, as well as dozens of other fruits and vegetables. Along with the 20,000-plus other bee species, they help pollinate roughly a third of the planet’s food supply and ensure the reproduction of more than three-quarters of all flowering plants. While we know less about wild bees, population surveys indicate that bees worldwide are in decline, from solitary species like leaf-cutter bees that nest alone and pollinate a few preferred plants to social ones like native bumblebees that live in colonies and collect nectar and pollen from nearly any flower they encounter.

In April 2025, American beekeepers reported that 55 percent of their colonies had perished over the previous year—their worst losses ever. Kirby’s honeybees, however, are thriving. She’s spent the past 20 years working with her farming partner Mark Spitzig to breed and raise colonies, cleverly branded “LongeviBEES,” that are well adapted to New Mexico’s high desert and the Rocky Mountain region. Kirby and Spitzig don’t treat their hives with the synthetic chemicals most commercial beekeepers use to fend off the varroa mite, an invasive parasite that feeds on bees, weakening colonies and carrying viruses, and is thought to be the primary culprit in recent die-offs. Instead, they wait to breed their queens until after they’ve lived for at least two years—long enough to prove the bees are hardy and wily enough to survive on their own.

Ten years ago, Kirby brought some of those bees to Lucero, who creates candles, balms, and salves from the beeswax she harvests. Kirby’s instructions were simple: Leave the bees mostly alone, and let them learn to solve the challenges of their volatile world. And they have. “They really can figure things out,” said Lucero.

Over the past few decades, scientists have been learning more and more about the ways that bees figure things out. They’ve studied how honeybee foragers fan out across miles of unfamiliar terrain in their six-week adult lifespan, navigating by sunlight and memory as they visit thousands of flowers to retrieve nectar for their colonies. They’ve followed bees back to their nests and seen how they dance to tell others where the best flowers are, and how they make collective decisions to swarm and relocate their homes.

Now researchers are uncovering remarkable new insights into how these industrious insects think. The breakthroughs have arrived thanks to a series of creative experiments designed to test how bees perceive the world, solve problems, and respond to unexpected situations. And the results have found that a single bee is much smarter than almost anyone imagined. These tiny creatures can make the sort of intelligent decisions that scientists previously believed were possible only in vertebrates.

Kirby’s sturdy and resistant bees also hint at a tantalizing possibility for people who are concerned about the insects’ long-term survival. As she sees every day in her apiaries, bees can adapt to their increasingly challenging surroundings—if we let them.

There was a time not too long ago when scientists believed that bees were automatons—mindless robots whose actions were hardwired into their genes. Even the most eminent scholars of animal behavior believed that bees’ actions were guided purely by instinct—“inherited through countless generations,” wrote German scientist Karl von Frisch, who won a Nobel Prize in 1973 for discovering how bees communicate. “The brain of a bee is the size of a grass seed and is not made for thinking.”

But in recent decades, researchers like behavioral ecologist Lars Chittka have designed a series of increasingly ambitious experiments that reveal the many ways that bees’ brains are indeed made for thinking. In 1990, when Chittka was studying bee neurobiology as a Ph.D. candidate in Berlin, he led a group of undergraduates to a gigantic agricultural field to explore how bees estimate distance and direction in a featureless landscape with no trees, bushes, or hills.

Over a bottle of Irish whiskey one night, they decided, as a joke, to design an experiment to see whether bees could count. “We initially laughed. It was a ridiculous idea,” he recalls.

But the next day, they constructed a row of identical, tent-shaped objects to act as landmarks for a colony of honeybees going back and forth to their hive. They placed a feeder filled with sugar water, to mimic nectar, between the third and fourth landmarks. Then, once the bees were familiar with the location of the feeder, the students varied the tents’ positions. The bees, however, still searched for the feeder after the third tent. They seemed to count the number of landmarks they passed on their way to the feeder. “At the time, that raised some eyebrows,” Chittka says, but other labs replicated the results and confirmed his findings. The success of that experiment, he says, “triggered me to probe even deeper into how much intelligence you could squeeze into a micro-brain like that.”

Now a professor at Queen Mary University of London, Chittka has become well-known for designing studies in custom-made testing arenas, confronting bees with unusual problems that they would never see in nature. (He has switched from using honeybees as his primary subjects to bumblebees, which live in smaller colonies that are easier to observe inside a lab.) By introducing bees to challenges “that none of their ancestors ever came across in their evolutionary past,” he says, his team explores the limits of the insects’ cognitive flexibility: their ability to alter their behavior as situations change. His pioneering work has spurred a growing body of research demonstrating that bees can recognize patterns, differentiate between symbols, identify human faces, cooperate on novel tasks, and plan for the future.

These unexpected abilities may stem from the unpredictable world that bees must negotiate. Foraging insects have to process more complex information than other insects. To retrieve nectar for their colonies, for instance, bees must search across miles to locate the sweetest and rowdiest blooms, remember where they are, and, on the fly—literally—perform cost-benefit analyses to determine whether the energy required to reach a sweeter flower is worth traveling the extra distance from the hive. They also must find water, dodge predators, and navigate a shifting compass of sun and sky. Social bees communicate that information to their sisters; solitary bees bear the additional burden of foraging, nestbuilding, and tending their young all alone.

“There are an awful lot of fairly nifty tricks that animals use to survive that are not the result of individual intelligence,” Chittka says, such as building hexagonal honeycombs or, in the case of a human toddler, learning to walk. But “intelligence, in the broadest sense, has to come about by individual learning” that goes beyond genetic programming. “No human being innately knows how to ride a bicycle,” he adds. “That’s something you have to learn.” So too must a bee learn to find and remember a patch of flowers.

One of the most radical leaps forward in understanding bee intelligence arose from another experiment that Chittka created half in jest. At a department meeting a decade ago, one of his colleagues lamented that the parrots in his lab had failed a string-pulling test—a classic experiment used to evaluate cognitive prowess in primates, dogs, and various bird species, in which an animal learns to pull a string to retrieve an otherwise inaccessible reward. “I sort of flippantly commented, ‘I bet our bumblebees could do that.’ Everyone laughed and said, ‘Oh, Lars has gone completely crazy.’ ” But Chittka thought about it some more and brought the idea up with his students. “I said, ‘Why don’t we just try?’ ”

After tinkering with a range of designs, his team placed an artificial flower—a blue dot filled with a sucrose solution—under a low transparent plastic barrier so the bees could see but not reach the reward. Instead, they were shown step-by-step how to pull the flower closer using the attached string. Shortly after the experiment’s design was finalized, one of Chittka’s students called him in to the lab. “I couldn’t believe what I was seeing,” he remembered. The bumblebees had quickly figured out how to pull it to them.

Then Chittka and his team decided to expand the classic framework by bringing untrained bees to watch demonstrator bees pull the string. The observers learned the task as well, and soon those bees were showing other foragers in the colony how to do it too. “The fact that a completely nonnatural behavior spread rapidly through an entire population was pretty spectacular,” he said. The experiment suggested that bees can recognize cause and effect—how pulling a string can produce a reward—and engage in social learning.

Bees’ surprising cognitive abilities also made Chittka begin to consider another radical idea: “If they’re that clever, maybe they can also feel things,” and are sentient, like humans and other animals, possessing the capacity to experience positive and negative emotions. In order to test for this sort of consciousness, Chittka and his team designed a set of robotic crab spiders—simulating dreaded predators—to briefly capture test bees as if in an unsuccessful attack. Afterward, the bees’ “whole demeanor changed,” he said. They’d inspect every flower carefully before landing and reject safe ones for fear of another spider. “That did look a bit like a post-traumatic stress disorder, where you’re conjuring up threats from the past,” he explained. In a different experiment, when bees encountered unexpected rewards, they engaged in more optimistic behaviors, like approaching unfamiliar flowers more quickly.

This doesn’t mean, as Chittka wrote in his 2022 book, The Mind of a Bee, that the insects “ponder the arc of their life from youth to death.” But he believes it may show that they feel pain, fear, joy, and perhaps even hope. “In addition to the well-justified motivation to conserve bees because they’re useful animals,” Chittka says, “I think the insight that they are intelligent and quite possibly sentient beings lends additional motivation to do something to look after them.”

Around the globe, a growing cadre of bee researchers is now trying to measure and understand how bees change their behaviors to solve and survive new challenges. And that means dreaming up ever more outlandish obstacle courses.

Inside a small laboratory at the University of Oulu in Finland last October, behavioral ecologist Olli Loukola and doctoral researcher Akshaye Bhambore watched as a large buff-tailed bumblebee tried to negotiate a rectangular testing arena. Bhambore had placed a fake flower—a blue dot the bee had been taught to associate with a sugary reward—on the ceiling, hidden from the bee’s view as it entered the mazelike area from its hive. He then put a foam ball about the size of a shooter marble in the testing ground. Without any prior training, the bee had to roll the ball to the dot, then clamber onto it like a step stool. Bumblebees typically scramble into flowers, rather than hovering around them. If the bee could roll the ball under the flower, climb up, and probe the dot for nectar using its proboscis within a prescribed time limit, it passed the test.

The experiment is an extension of work that Loukola and Chittka pioneered when Loukola was a postdoctoral researcher in Chittka’s London lab. Building on Chittka’s string-pulling experiments, Loukola had taught bees to “play soccer,” training them to roll a small yellow ball to a floral “goal” to reach a nectar reward, another remarkable demonstration of bees’ ability to learn. But this new trial is even more challenging. It could demonstrate that they can, without previous experience, establish an objective—the sugary reward—remember it when it’s out of sight, and then use the ball as a tool to reach their objective. If the trial is successful, it will prove that bees are capable of something similar to “insight learning,” solving problems through sudden “aha” moments, rather than trial and error. This is “true intelligence, or complex cognition,” said Loukola. “The animal needs to really understand what they’re doing.”

The bee they were observing did seem to understand the task at hand, though she was clumsy. After exploring the arena for a while, she nudged the ball, then crawled up onto it and flipped it over on herself. Many bees roll the ball backward this way, moving their bodies in the direction they want to go. She somersaulted awkwardly, tumbled with the ball and abandoned it a few times, then finally rolled herself and the ball to the reward’s location, climbed up on the ball, and plunged her proboscis into the dot. She had passed the test with time to spare.

Loukola and his team encountered some healthy scientific trial and error of their own as they designed the experiment, reworking the testing arena a number of times. Though their work is ongoing, the results look promising so far, and if they prove out, it will be, Loukola said, “the first evidence in any insects where we show that they can solve problems spontaneously.”

While the highly controlled environment in the lab has allowed Loukola and Chittka to investigate the outer bounds of bees’ cognitive abilities, other colleagues have found ways to witness and corroborate bee intelligence in nature. In California, UC Davis biologist and National Geographic Explorer Felicity Muth has demonstrated that lone bumblebee queens are even faster and better learners than the foraging worker bees used by Chittka and Loukola, perhaps because these queens have to forage, build, and provision their nests all alone in the first weeks of their lives. In one experiment, Muth placed strips of colored paper dipped in sugar water or plain water in a meadow. The queens learned to visit the colors that contained sugar water more quickly than the workers did.

In France, University of Toulouse behavioral ecologist Mathieu Lihoreau has glued tiny, magnetic transponders onto the backs of bumblebees to track how bees constantly alter their routes as plants, flowers, and weather patterns change—with an aim of developing a predictive model for how bees visit flowers. Understanding the way bees think and adjust could not only help farmers plan agricultural pollination to increase crop yield, it could also be used in conservation efforts such as aiding in pollination of endangered plant populations.

Like the bees they study, Lihoreau, Muth, Loukola, and Chittka have had their share of revelations as they’ve come to understand how bees negotiate a radically shifting landscape with agility and vigor—and just how vastly we’ve underestimated the capabilities of their tiny brains.

Perhaps, bee advocates suggest, we’ve also misjudged the insects’ ability to adapt to the raft of new challenges, like the parasitic varroa mites that continue to devastate the beekeeping industry. Kirby, for one, has come to believe this is the case, at least where honeybees are concerned. Beekeepers have for too long been stuck on a treadmill of chemical treatments to fend off the mites. But eventually, those treatments stop working and we see huge die-offs like the one last year. Working with a group of like-minded beekeepers, researchers, and scientists, Kirby founded the Adaptive Bee Breeders Alliance, a national network that promotes beekeeping practices that align with the insects’ instincts and intelligence, encouraging the breeding of bees that have been shown to be resilient in their climate and environments. More beekeepers are beginning to embrace the notion that the best and strongest bees are the ones that survive on their own. Zia Queenbees, the company Kirby owns with Spitzig, sells as many as 300 queens each week during spring and summer to beekeepers across the country, and bigger, more commercial operations are also beginning to come around to this point of view. “Don’t tell the bees how to do the job,” says beekeeper and breeder Randy Oliver, who keeps 1,500 colonies—around 60 million bees—in California. “Just promote the ones that do.”

A bee’s job, we’ve come to understand, is to learn its world: to navigate and remember its ever-changing surroundings to support its colony. And perhaps our job as humans is to support bees in their process of learning and adaptation, being mindful of the habitat they need and the poisons and stresses they don’t. Humans need to learn to listen to the bees around them, says Kirby. “I always feel like the bees still have so much more to teach us.”