How Felicity Muth is decoding bee intelligence

In the rural mountains of the western United States during what researchers call “queen season,” Felicity Muth looks for solitary queen bumblebees. The “mega bumblebees,” as she calls them, are giants compared to their worker counterparts — and they navigate the world alone before they establish their colonies, demonstrating impressive intelligence in the process.

“They’re a bit more elusive, they’re fewer in number and people haven’t really worked on their cognition before,” explains Muth, a National Geographic Explorer whose research is focused on the cognitive abilities of wild bees. “During the pandemic I had a bit more time on my hands, and I went out and caught some of these bees and started working with them.”

Her findings add to what is known about bee intelligence.

“I found that they’re [queen bumblebees] really good at learning,” says Muth. “It’s been known for decades that the workers were good at learning, but these queens seem to be even better at learning than the workers.”

From her base at the University of California, Davis, Muth has documented behaviors signaling that queen bumblebees possess learning abilities that surpass not only famously intelligent worker bees, but also larger-brained animals in some cognitive traits and tests. 

“Now we’re trying to work out more about differences in how they make decisions, using this as a model for thinking more broadly about how differences in animals’ environments can lead to different cognitive abilities.” 

Muth’s research suggests that bees express decision-making biases — demonstrating that less-preferred options can influence choices between better alternatives. “Much like humans at a supermarket,” she says.

“Some of the biases we have are the same between humans and bees,” she highlights. “We all have neurons. The very basic building blocks are the same.”

From Muth’s perspective, “by understanding bees, we can also better understand other animals, including humans.” 

Building her path

Muth’s path to becoming a bee cognition expert began with birds. Her doctoral studies focused on nest-building behavior in weaver birds, taking her to Botswana to observe their constructions. She also spent time observing zebra finches from a laboratory, studying the cognition involved in their nest making. 

Her childhood fascination with animal behavior, sparked by David Attenborough documentaries and her father’s gift of classic animal behavior literature, kickstarted her professional direction. And eventually, a chance conversation with a friend about pursuing animal behavior set her on her academic trail.

“My trajectory has been very predictable,” she laughs. She pursued her animal behavior doctorate, followed by animal behavior postdoctoral studies and became a professor in animal behavior six years ago.

Recently, she’s circled back to investigating nest construction, now in bumblebees. “Bumblebees make these really cool complex wax nests that look chaotic, but what we’ve been finding is they’re really flexible in their building behavior.” 

Bumblebee colony nests differ from the perfect hexagons typical of honeybees.

“People are now finding that even the honeybee colony, which looks so stereotyped, has some flexibility in it. And I think that with the bumblebees, there’s even more flexibility in how they build in these seemingly chaotic spaces.” 

To investigate further, Muth is conducting a photographic experiment in collaboration with fellow National Geographic Explorer and photographer Anand Varma. Using 3D-printed structures, Muth’s team discovered that bumblebee nest building can involve incorporating foreign objects and completing partial structures.

“We can give them half a honey pot and they will complete it with their own wax,” she explains, showing images of bees adapting plastic structures into their nests.” The bees will also fill plastic honey pots full when they’re provided. 

Muth and Varma are experimenting with how to document the bumblebee nest construction process.

It’s part of Muth’s mission to shift public attention beyond honeybees to the thousands of native bee species that exist, and are at risk. One in four native species is in danger of extinction.

“So much of the focus is on honeybees, but they’re not a native bee in North America. They’re a species that we use in pollination that were brought over from Europe,” she explains. “Really, all the native bees are the ones we should be worried about, and yet the honeybee gets all the attention.”

There are around 20,000 species of wild bees in the world, with over 4,000 species of native bees in North America. Of those native species, around 25 different kinds of bumblebees are in California — comprising about half of North America’s total bumblebee varieties, explains Muth. The diversity in this region made it an ideal location for her bee cognition studies, which are funded by the National Geographic Society and the Templeton World Charity Foundation as part of the Wildlife Intelligence Project — an initiative that supports investigations related to animal intelligence. 

Species vary in many aspects of their ecology, but even intra-specifically — within a species — “bumblebees have a number of discrete roles and castes with different ecological requirements,” Muth emphasizes. “For example, foraging alone rather than part of the colony structure means that queens can't afford to make mistakes and therefore this could be one reason that they are better at learning associations.” Bumblebees are “generalists,” they can thrive on many species of flower, which requires sorting out which ones are useful and recalling them by qualities like color and smell. Compared to queen bumblebees, “workers may be able to afford to take more risks — which might not pay off, but could lead to finding new nectar locations,” she adds. 

Down the line, Muth plans on applying her investigation across species. 

What bees do in the wild

What sets Muth’s work apart is the emphasis on bees’ natural environment as the experimental ground. While most bee research has been conducted in laboratories, and has been focused primarily on two commercially available species, Muth spends significant time in the field to understand how wild bees deploy their intelligence in their native environments.

“People have been working on bee behavior for over a hundred years, and there are thousands of studies out there,” she cites.

Karl von Frisch’s experiments resulted in the groundbreaking discovery that honeybees communicate through a “waggle dance,” and demonstrated that bees have color vision. Bees demonstrate knowledge acquisition and sharing behaviors, and research shows bees are trainable. One recent study showed how a bee can be trained to feel shapes in the dark — like a ball or a cube — and learn that either is tied to a reward. Later, when shown the shape in a photo in the light, they’d recognize and gravitate toward the shape associated with a reward. Scientists call it cross-modal object recognition.

In addition, bees can be trained to pull strings or roll balls to reach a reward.

“What we’re interested in is, what do wild bees actually do?” she says. “If we go out into a meadow, we’re studying that bee’s behavior and how it’s making decisions and how it’s learning about those flowers.”

Her field setup is simple: a small apparatus allows her to train wild-caught bees using colored paper strips and different scents. The bees learn quickly that blue means reward, and yellow is “bad, or different scents,” explains Muth. She’s also played with the anatomy of wild flowers, removing their nectar and adding artificial nectar. 

“The great thing about bumblebees is that they’re super motivated. At first they’re a bit freaked out, and then you give them some sucrose, and then they’re just psyched.”

The bees are so fond of her experimental setup that when some tagged individuals are released back to the wild they return. “I’ve had tagged bees try to get back in the apparatus because they’re like, ‘This is the best flower I’ve ever visited.’”

Muth has applied the experiment to bees in a lab setting as well — bees rescued from Sacramento homes (of the species Bombus melanopygus) that would normally be relocated to fields. “We’ve been able to experimentally overwinter bees in the fridge. So, normally, bees will be nesting underground, under the snow, for the winter. We can put them in the fridge and basically trick them into thinking that they have overwintered, and then, we can bring them out as spring queens in the lab and have them forage on flowers.”

Muth is also interested in bee life cycle transitions, specifically, queen bumblebees’ unique shift from forager to egg layer. Unlike queen honeybees, which rarely leave the hive, bumblebee queens go from being foragers who can navigate miles visiting flowers. They go from “this really visual factory world, to then being in the dark for the rest of their lives.”

“What happens, if anything, to their brains? What happens to their cognition in that transition?”

A spectrum of specialized abilities

Of the countless hours she’s spent observing bees — from their construction behavior to their general decision-making, she’s concluded that bees “seem really different from us.” However, “When you spend hours and hours staring at bees, you see all kinds of little quirks that feel very human, sometimes.”

Rather than viewing cognition as a hierarchy with humans at the top, Muth sees it as a spectrum of specialized abilities shaped by evolution and environment. 

“Every animal has the cognitive abilities that they have evolved for their environment,” she explains. “Some of those are better than what humans have, some are worse, some are just different — they’re deployed in different ways.”

She points to chickadees as an example: “They cache seeds, and they can remember the locations of thousands of seeds because that’s what they need to do. I can’t remember the locations of thousands of seeds in the world. So, they are better at spatial memory than I am, but I’m better at language than they are.”

“It seems like an unfair way to measure animals is to critique them on human skill when they’ve got all these other things going for them,” she presses. 

Underlying her work is an urgent conservation reality: bee populations are plummeting worldwide. The causes are multifaceted — habitat loss, pesticides, pathogens — but according to Muth, the solution requires understanding what is being lost.

She’s previously researched pesticides, finding that neonicotinoids specifically impair bees’ olfactory learning and scent discrimination. For creatures that “live in this really olfactory world,” using scent for navigation, nest recognition, and communication, such impairment could be devastating, she explains.

“If we’re going to understand how humans are affecting bees, we also need to have a baseline understanding of their behavior, to then understand when their behavior is changed and how it’s changed them,” she notes.

She adds that in “well-meaning” intention, people often buy honey bees to combat the reality of bee population declines. “Well, actually, that’s not a good thing for native bees, you’ve just introduced a non-native competitor that might spread a disease to them.”

In 2022, Muth authored a children’s book, “Am I Even a Bee?” as a way of challenging bee misinformation and introducing young readers to the diversity of bee species. 

“I felt if I could teach kids that there was more than one type of bee out there, then that might be a better contribution than all of my scientific career combined,” she reflects.

The book, published with Sundance Books Baobab Press publishing branch through a serendipitous encounter with a local shop in Reno, Nevada, has taken her to schools, storefronts and libraries for readings and other outreach events. In South Lake Tahoe she recently attracted the biggest crowd yet. “Normally I’m wrangling 15 children, and the other day it was over 300.”

Working intimately with bees has taught Muth lessons about more than just cognition. She’s noticed her own learning patterns — how getting stung makes her hypervigilant for weeks despite knowing the danger hasn’t necessarily increased. She has, of course, been stung, “but not as much as you might think.”

She’s also noticed how she’s ended up training herself while training the bees: “It’s all very meta,” she laughs. “You’re doing these weird things you’ve never done before, at first you’re not very good, and then you get better,” she says. “So you end up noticing how you learn things as you’re also working with the bees and they’re learning things.” 

Understanding bees’ capabilities is, for Muth, a part of a fully-lived experience. “One of the joys of being human is being able to ask questions and try and understand our world. And so, I think these are amazing animals, and I’m trying to just get a better understanding of how their minds work.”

ABOUT THE WRITER For the National Geographic Society: Natalie Hutchison is a Digital Content Producer for the Society. She believes authentic storytelling wields power to connect people over the shared human experience. In her free time she turns to her paintbrush to create visual snapshots she hopes will inspire hope and empathy.