Mindless activity: Brittle stars demonstrate associative learning despite the fact that their neurons don’t converge into a central brain-like structure.
Penta Springs Limited / Alamy Stock Photo
When the COVID-19 pandemic forced Julia Notar to cancel a summer’s worth of international marine research, the Duke University doctoral student swapped sun-dappled tide pools for a dingy, socially distanced basement lab — and faced dwindling options to finish her dissertation. She took on what she calls a “risky proposition”: training brittle stars to anticipate food, much like Pavlov’s dogs.
Notar had already been working with the stars, which are closely related to sea stars (commonly known as starfish) and sea urchins but, unlike those creatures, have long, willowy arms that help them scramble over the ocean floor. So she knew well that the stars’ neurons, which help them move and see, don’t converge into any central structure resembling a brain — something long thought to be required for learning and memory.
But she had also noticed that her stars were particularly shy. They almost always hid behind their tank filters unless they were being fed. And that suggested a training opportunity: Every day, Notar turned off the light and fed eight brittle stars in the darkness; 30 minutes later she turned the lights back on, waited a few hours and then fed another eight stars in the light. To ensure the former group became conditioned to the temporary darkness and not to a specific time of day, she varied the time of day she fed both groups. Cameras set above each of the tanks monitored the stars’ movements.
After two weeks, the stars fed in darkness started crawling around their tanks soon after the lights dimmed. And even after Notar fed the stars at progressively later time points during the dark period and eventually in the light again, the stars continued to venture out in the darkness — associating it with a forthcoming meal.
Like Pavlov’s dogs (but presumably less slobbery), the brittle stars fed in the dark had formed a connection between two unrelated environmental cues — darkness and incoming food — demonstrating associative learning. Meanwhile, the stars fed in the light, having formed no such association, largely remained hidden during the half-hour periods of darkness.
In other words, Notar’s gamble had worked.
Historically, associative learning has been held up as somewhat of a “crowning achievement” of cognition, possible only in organisms with complex neural architectures, says Chris Reid, assistant professor of natural sciences at Macquarie University, who studies behavior in insects and slime molds and was not involved in the brittle star research.
But Notar’s results, published in November 2023, are just the latest to challenge that view. Findings across a range of brainless organisms are helping to turn the tide against the longstanding assumption that only higher-order animals can learn. So much so that within the past 15 years, Reid says, he has seen the field widen its perspective about who — and what — is capable of cognition.
“Learning is a change in how the nervous system works as a function of experience,” says Mariam Aly, assistant professor of psychology at Columbia University, who studies memory and learning in the human brain. In humans, a lot of the changes that enable learning happen in our brains, she adds. But “that doesn’t mean that animals with simpler nervous systems can’t learn.”
nd indeed, learning has cropped up in some surprising corners of the animal kingdom.
Despite their astral anatomy, brittle stars are actually among “some of our closest invertebrate cousins,” Notar says, even closer than insects. But just two months before she published her study, she points out, another group of researchers based in Europe published evidence of associative learning in a more distant evolutionary relative.
Only about half an inch in diameter at most, the box jelly Tripedalia cystophora lives in the shallow waters of tropical mangrove lagoons. These animals use their complex eyes to navigate the intricate latticework of underwater mangrove roots, foraging for prey sheltering among them. It’s critical the jellies gauge distance accurately: Too close to the roots and they crash into them, damaging their delicate bodies. Too far away, though, and they aren’t able to find food.
Box jellies can learn to judge those distances over time in a circular tank painted with black or gray stripes to mimic the silhouettes of mangrove roots, the European team showed. Jellies easily steered clear of tank walls painted with high-contrast black stripes, suggestive of dangerously close roots. But the lower-contrast gray stripes proved trickier. Initially, the jellies perceived them as farther away and frequently collided with the tank walls. But after a while, they learned that gray “roots” were closer than they appeared — a kind of learning they might need in murky water.
The paper makes the argument that even without a brain, “learning [is] an intrinsic property of nervous systems, period,” Notar says.
Other teams are exploring what kinds of learning may take place in creatures even further away from humans in the evolutionary family tree, such as single-celled organisms. Without any kind of nervous system at all, they rely on intracellular communication to integrate all the information they gather.
The slime mold that Reid studies — Physarum polycephalum — has yet to show any convincing signs of associative learning, he says. But these “surprisingly charismatic blobs of yellow goo” display other cognitive behaviors, including solving mazes, making complex choices and habituating to repetitive stimuli, he says.
The latter is also true for Stentor coeruleus, a large bluish-green protist that is usually shaped like a flowering Calla lily but instantly curls up into a tiny ball if disturbed. When S. coeruleus grows in petri dishes rigged to a device that periodically jiggles them — sending the protists into microscopic whiplash — the cells initially curl up after every tap, a November 2023 study showed. Over time, however, they increasingly remained unfurled, suggesting the microbes habituate to the tapping and can effectively ignore it.
Until recently, there was little compelling evidence of unicellular organisms learning anything besides habituation, Reid says. But in 2021, an international team of researchers reported seeing associative learning in three different species of amoeba.
The experimental design makes the results particularly convincing, he adds. The researchers conditioned a group of cells in an electrical field by placing appetizing peptides at the anode, or negative, end. Amoebas have a natural tendency to move toward the cathode, or positive electrode, but these cells started to migrate to the food-stocked anode end instead. Even in a peptide-free field, the conditioned cells still moved toward the anode, indicating that they had formed an association between the electrical current and the presence of food.
There have been bits of evidence of learning in single cells over the past 75 years, says Ken Cheng, professor of natural sciences at Macquarie University, who was not involved in the 2021 work. “[This study] is probably the best because it got the amoeba to go against their natural tendencies.”
ven in organisms that aren’t capable of learning, Cheng says, cognition (or acquiring, integrating and using information about the world) is universal.
“All life is cognitive,” he says, including single-celled organisms. “How can you climb a chemical gradient without using information about chemical concentrations?” A microbe needs to sense a particular chemical (amid countless others) and then compare the current concentration with the concentration moments before. Then it must act to continue on the same trajectory or try a different path.
It may be tempting to dismiss this kind of seemingly intelligent behavior as reflex — mere functions of biochemistry and thermodynamics, says Alison Hanson, a postdoctoral researcher at Columbia University who studies neural processing in Hydra. But, she adds, it’s not scientifically consistent to set human cognition apart as something special.
“That’s not how evolution works. [Intelligence] doesn’t just appear in a human being all of a sudden. Nature uses the same building blocks over and over again. Just because [simpler organisms] don’t look like humans and they’re not playing Tetris in an MRI scanner doesn’t mean they’re not doing intelligent and very complicated things,” Hanson says — or that by studying their cognition we cannot learn something about our own.
Unraveling the molecular mechanisms behind different cognitive behaviors in different creatures “will give us a much broader appreciation of how different things solve the same problems,” Reid says, “which is much more useful than just repeatedly showing that they can.”
And studying cognition in any animal that’s not usually studied is “really, really useful to understand where our brains are coming from,” says Björn Brembs, professor of neurogenetics at the University of Regensburg. “What were the driving forces that led to something like us?”
“After 150 years, even among the scientific community, the concept of evolution hasn’t really fully sunk in and been soaked up completely,” he says. “Because what does evolution really mean? It means we share things, and we’re different.”
The Spectrum team is excited to announce the launch of The Transmitter, a new publication for the neuroscience community. Spectrum is now a key section of The Transmitter and will continue to publish news and perspectives about autism research. Soon, you will be able to find all previous Spectrum articles at thetransmitter.org.