The Light Eaters: How the Unseen World of Plant Intelligence Offers a New Understanding of Life on Earth

by Zoë Schlanger

In a paper titled “Broadening the Definition of a Nervous System to Better Understand the Evolution of Plants and Animals,” Llinás and Sergio Miguel-Tomé, a colleague at the University of Salamanca, basically argue that it makes no sense to define a nervous system as something only animals can have rather than defining it as a physiological system that could be present in other organisms in a different form. Defining it phylogenetically—meaning assigning it only to one portion of the tree of life—ignores the very real force of convergent evolution, where organisms separately evolved similar systems to deal with similar challenges. It happens all the time in evolution; a classic example is wings. Flight evolved separately in birds, bats, and insects, to very similar effect. Eyes are another example; the eye lens has evolved separately several times. It’s reasonable to imagine the nervous system as another case of convergent evolution, Llinás and Miguel-Tomé say. If a variety of nervous systems exist in nature, then what plants have is clearly one. If it walks like a duck and quacks like a duck, it’s probably a duck. Why not call that a nervous system already?

I was watching the plant becoming aware, in its own way, of my touch.

“Could the whole plant be something like a brain?” I asked. She smiled. We’d been talking for nearly three hours. She had fifteen minutes left. I’d saved the question for the last moment, in case she was entirely put off by it and ended our talk right then and there. And now I’d done it, and watched her smile, thinking I’d made a fool of myself. Then she leaned in a little and dropped her voice to a whisper. “I think you’re right,” she said. “I just don’t talk about it.”

It is nighttime in the rain forest of southeastern Cuba, and a long-tongued bat is sailing between the trees, plucking a clear path through the dense canopy at high speed and in total darkness. All gossamer wing membrane and stealth fluff, its whole body weighs hardly a third of an ounce. A paper airplane. The bat lets out a pulse of tones and listens for the echo they return to its oversize jackal ears. A cavalcade of clicks conjure a landscape of objects and air as the tiny mammal tilts its wings to slice between a tangle of vines. All of a sudden a tone comes back clear and crisp—again and again in the same way, despite the angle of incidence that changes over and over as the bat flies around and comes closer. The sound is irresistible in its clarity, a lure, a beacon in the night. The bat arrives to find a vine hung with a ring of sumptuous wine-colored flowers, their pollen-tufted faces tilted down toward red pitchers full of nectar. The bat unfurls its long tongue and squeezes its face between the flower and its pitcher. It begins to lap nectar, hovering in midair as it drinks. Its back takes a dusting of pollen in the process. Just above the ring of flowers grow a series of glossy leaves, conspicuously oblong and concave, like upright canoes. The deep, rounded shape reflects the same clear echo from many angles. To a moving bat in the acoustic mess of the forest, a loud, consistent echo coming from the same location stands out. And for a rare bat-pollinated vine sca...

Scientists watched as bats consistently landed only on the flowers that still had their hidden pollen keels intact, while avoiding the depleted ones. There were so many flowers—how did the bats find the right ones? A small concave appendage flanks the unopened flowers, like an extra petal on a hinge; researchers found that this acts as a perfect mirror for the bats’ sonar. The echo it sent back from multiple angles was of “astonishingly high amplitude,” they wrote, much like the echo from the leaves on the Marcgravia. Once the flower had disgorged its pollen on a bat butt, the mirror would lower itself, taking itself out of the acoustic arena. Bats could no longer find that flower and were directed instead to the ones with their mirrors still up. Plants have a particularly close relationship to sound.

Plants and insects interact all day long, and at every stage in both of their life cycles. It may be the most important relationship in either of their lives, if the insect is the type that drinks nectar or eats leaves, which is to say the great majority of them. Plants and insects together make up about half of all multicellular organisms on earth; it wouldn’t be an exaggeration to say theirs is one of the most important relationships on the planet.

Others will manufacture their own insect repellant, which in many cases is the bit of the plant humans most enjoy—it’s the rich oregano oil in oregano, the sharp spice in a horseradish root. Sometimes the approach is more sinister. One devilish case has been found in the humble tomato: the tomato plant will inject something into its leaves that makes the caterpillars look up from their chewing and turn to eye their fellow caterpillars. Soon, the leaf becomes irrelevant. The caterpillars begin to eat each other.

It was evidence that plants could really hear, in their own earless way. Sound, to them, is pure vibration.

Once one opens the door to plants registering the sounds of a caterpillar chewing, other considerations come into play. The world is a noisy place. What else might plants hear?

Evolution, ever scanning for a benefit, will give the organism ways to use its awareness to further its project of survival.

If plants could be made to produce pesticides through simply playing sounds to them, it could reduce or eliminate the need for synthetic pesticides on farms, and in some cases increase the levels of compounds that the crop in question is grown for. In a crop like mustard, for example, the plants’ own pesticide is the very thing it is farmed for—mustard oil. Putting a lavender bush on high alert by playing the right sounds would cause it to make more of the defensive compounds we prize in lavender oil.

In 2019 researchers at Tel Aviv University found that the beach evening primrose—a lemon-yellow teacup-shaped flower that grows low to the ground—would increase the sweetness of its nectar within three minutes of being exposed to an audio recording of honeybee flight. The primrose would completely ignore sounds that fell outside of the frequency of the hum of bee wings. The team, led by evolutionary biologist Lilach Hadany, theorized that the sweeter nectar—it had a higher sugar content than flowers not exposed to bee sounds—would better entice pollinators, and increase the chance for cross-pollination.

The flower, in this case, was definitely the part of the plant responsible for “hearing”—and it suggests that it had taken on the bowl shape for exactly the same reason satellite dishes are concave. “We found a potential hearing organ, which is the flower itself,” she said. When she looks at flowers now, she sees ears everywhere. Roots, it seems, can be just as acoustically sensitive. Why have ears above ground only, when half your body is under the dirt? There’s plenty to hear down there too.

The seedlings appeared to be able to parse various sensory cues, ranking the inputs in terms of priority to their own health, Gagliano thought. But more pressingly, they could hear—and move toward—the sound of real running water. This would likely not surprise a plumber. Plumbers are accustomed to the frustrating phenomenon of tree roots bursting through sealed water pipes. Cities spend millions each year repairing municipal pipes punctured by “root intrusion.” Germany, for example, spends an estimated thirty-seven million euros per year repairing root-burst pipes. The U.S. Forest Service points to root intrusion as the cause of more than half of all sewage pipe blockages.

Over a century after the first evidence emerged, science refused to believe that bats could be using sound to orient themselves in space. It seemed too far outside assumptions of what animals could possibly do. Scientific disbelief hampered the discovery of echolocation in bats;

Already, scientists have found compelling evidence that language is not entirely confined to the human realm; prairie dogs appear to use adjectives, specific repeated sounds they use to describe the size, shape, color, and speed of predators. Japanese great tits have syntax; they use distinct strings of chirps to instruct their comrades to scan for danger, or tell them to move closer. We’ve heard about songbirds using backchannels for alarm calls, and risk-averse chipmunks screaming at the slightest spook. Perhaps it would be small-minded of us to foreclose on the possibility of a sound-based plant language emerging too.

I’ve come here because it occurred to me that memory must be the basis of all complex behavior.

Henning returns to my first question, about where Nasa poissoniana’s memories could possibly be stored. That is, of course, still a mystery. But, Henning says, “maybe we are just not able to see these structures. Maybe they are so spread all over the body of the plant that there isn’t a single structure. Maybe that’s their trick. Maybe it’s the whole organism.” Memory, even in humans, is still mostly shrouded in mystery. Neurobiologists have found ways to “see” certain human memories on brain scans, as particular connections of neurons, but many more are yet invisible to science. And then there are the memories the human body keeps but which have nothing to do with neurons at all. Our immune cells remember pathogens, and draw on those memories to respond the next time they appear. Epigenetic memories in cells can be passed down through human generations; we now know that the toll of stress and trauma—as well as exposure to things like air pollution—travels down the bloodline to children and grandchildren, potentially impacting things like inflammation markers. The body, it has been said, keeps the score. But these aren’t the type of memories we include in the landscape of our consciousness. The memories our body keeps for us are only silent until they emerge as a change in our health. Then they become very tangible. But epigenetics itself is a field we are only beginning to peel back the curtain on.

Garlic can go into the ground in October, or November, if you’re procrastinating, but much longer and you’re pushing your luck; the ground still needs a little heat in it to convince a clove it’s safe to put out roots.

What the garlic needs, in order to sprout, is the memory of winter. That the spring eventually comes is not enough to make life emerge—a good long cold is crucial. This memory of winter is called “vernalization.” Apples and peach trees won’t flower or fruit without it. Tulips, crocuses, daffodils, and hyacinth, often the first blooms of spring, need a good strong vernalization too. If you live in a warm climate and buy tulip bulbs, the garden supply store clerk might wisely advise you to put your bulbs in the refrigerator for a few weeks before planting, or you’ll never see a flower.

Perhaps most instructive of all is that plants know how to wait, how to endure the inhospitable, knowing their time has not yet come but will, and that their flourishing is not a question of whether but when.

The remarkable thing about vernalization is that it means plants remember. The term undisputedly applies; plants use information stored about the past to make decisions for the future. This isn’t a singular example. Plants take note of the length of a day and the position of the sun. Cornish mallow, a pink-flowered plant, will turn its leaves hours before sunrise to face the horizon in exactly the direction it expects the sun to rise.

Researchers have messed with the mallow by simulating a more chaotic “sun,” switching the direction of its light source. The mallow learns the new location. This response is “extraordinarily complex—yet extremely elegant,” in the words of a research team eager to learn from the mallow to make smarter solar panels.

Plants that rely on vernalization must have some way to record the passage of time to be sure the elapsed period of cold—and warmth—is enough. Emerging in a two-day warm spell in February could be disastrous. So they seem to count the days. That’s why many plants wait until the warmth is sustained for four days or more. It’s less likely to be a fluke.

Memory clearly has deep roots in biology. This makes sense; if the trajectory of all evolution is toward survival, then the ability to remember has a natural evolutionary advantage. It’s incredibly useful for staying alive.

But the flytrap keeps counting after it closes. If the trigger hairs are disturbed five times in quick succession—eliminating all doubt that it has caught a living, wriggling creature—the plant injects digestive juices into the trap, and the meat meal commences. Digestion takes many days, so it’s important to be sure. But if the trap is triggered twice, snaps shut, and the triggering stops, the trap will open again within one day. Clearly whatever is inside is too small to bother with, or not a living creature at all, but perhaps a bit of twig or stone—or, in the case of all the flytraps that have told us anything about their kind, the cold tip of a botanist’s probe. The flytrap corrects for its error.

The ability to choose wisely is one hallmark of intelligence. The Latin root of “intelligence,” interlegere, means “to choose between.” Dodders are wonderfully fun for watching a plant make a choice: they prefer tomatoes to wheat, for example. Wheat is hard to climb, and not particularly juicy. When a dodder seedling is grown between a wheat and a tomato, it begins to circle the air almost as soon as it pops out of the soil. After a few perambulations it turns, with determination: it has noted its neighbors from afar. Now like a baby snake it crosses the air, aimed directly at the tomato, shunning the wheat.

In Trewavas’s assessment, the brain is just one strategy for building intelligence and consciousness. Plants simply took a different evolutionary route, according to their needs: their attention and awareness is localized in each of their parts, but each of their parts communicates and strategizes across the whole, producing consciousness all the same. “The individual plant containing many millions of cells is a self-organizing, complex system with distributed control permitting local environmental exploitation but in the context of the whole plant system,” he writes. “Consciousness is thus not localized but is shared throughout the plant in contrast to the more centralized location in the animal brain.”

Memory likely propelled our most distant ancestors toward more complex lifestyles that called for yet more complex decision-making. We may not know where plants store their memories—somewhere, it could be said, in their brainless mind—but the knowledge that they evidently have them is enough to change our world.

For example, in the 1990s, she was studying the triangle of drama between corn, caterpillars, and wasps. First a caterpillar chews a corn plant. The plant notices this, and samples the combination of saliva and regurgitant the caterpillar leaves behind on its leaves; now the plant knows the caterpillar’s species—or at least which species of wasp it needs to come parasitize it. The plant then releases a finely tuned chemical gas. Within an hour, the correct wasps arrive. The wasps, no doubt appreciative of the ideal scene before them, insert their needle-like appendages into the caterpillars’ bodies, injecting their eggs inside them. When the eggs hatch, the wasp larvae use their especially oversize mandibles to eat the caterpillars from the inside out. They then spin their Tic Tac-shaped cocoons, which they glue to the emptied husks of their caterpillar. The result is a green wormish form bristling with white silky spines, like hedgehog quills made of felt. And thus the plant attempts to save itself. De Moraes discovered this behavior in corn, tobacco, and cotton in 1998.

The marks looked like tiny crescent moons, always a tell for a bumblebee mouth. But why were bumblebees biting plants? She kept watching. The bumblebees, she realized, were starving. They’d been flying around the closed buds of mustard flowers for days, but no bloom had yet opened. If they didn’t slip their tongues in a pool of sugary flower water soon, they would begin to slow down, their bodies desperately trying to conserve calories. Eventually they would land on the dirt, crawl a while, and die. The poor bees; their timing was all wrong. The flowers weren’t due to open for another month. It wouldn’t do. The bees, she saw, began to bite the plants’ leaves. The next day the flowers bloomed. The bees drank the nectar and survived. Interesting, De Moraes thought. Bees don’t eat leaves. There seemed to be no good reason to waste precious energy on something that wouldn’t feed them. But here they were, biting them regardless. There were moons all over the place. Someone must have noticed this before, she thought. “And then you do this literature search, and it’s like, how did they miss this?” She set up an experiment with all the proper controls and found that bees biting plants made their flowers bloom as much as thirty days earlier than they would otherwise. Obviously, the bees benefit here—but, Consuelo found, so do the plants; they bloom in time to have bees around to pollinate them. Timing in nature is like that: all species rely on other species in one way or another. ...

De Moraes talks about the “arms race” of insects, plants, and viruses, each outsmarting the other and being outsmarted at turns. “Everybody is trying to survive. All of them.”

Interspecies chatter of this sort is constant and entirely invisible to human perception. The communication is sophisticated, dynamic, multilayered, and quick—all of this happens in a matter of mere moments. Only a sliver of it is so far understood, De Moraes says. “I’m always surprised by how much we don’t know.”

I THOUGHT BACK to the corn and tomatoes summoning wasps. They were in essence enlisting collaborators. Or, seen in a slightly different way, they were using the wasps as tools. The line between collaboration and coercion is sometimes blurry. The wasps certainly benefited from the relationship too. But either way, it was the plant that proposed the arrangement in the first place. From the plant’s perspective, it had found the right tool for the job. I thought about how an ability to use tools is a classic test of animal intelligence.

Bittersweet nightshade, a plant in the same family as tomatoes, potatoes, and tobacco, secretes sugary nectar in order to recruit ants as bodyguards. The ants, hooked on the sticky syrup the plant oozes for them, dutifully pluck off the larvae of the bittersweet’s mortal enemy, the flea beetle, which are clinging to the plant’s stem. They must be quick, before the wriggling flea beetle babies have a chance to bore themselves into the bittersweet’s body and wreak havoc. The ants march the larvae deep into their ant nest. The larvae are never seen again. Several other plants appear to hire ants in this way, enough for the botanical community to give them their own informal name: ant plants. (And a formal one: myrmecophytes.) Some species of ants can no longer survive without their ant plant, as is the case for the symbiotic ants of the tropical tree genus Macaranga, who quickly die out when separated from it. Acacias have a similar relationship; they feed their ants and provide them with specialized places to nest in the hollow thorns on their limbs. In turn, the ants aggressively attack anything that disturbs their ant-plant nest tree.

He knew that certain orchids only attracted certain species of wasps, so the chemistry must be rather specific. He figured they had to be using some combination of the more than 1,700 floral scent compounds already known. “We could not have been more wrong,” Peakall said, in a keynote to the global community of botanists at the discipline’s annual conference in 2020. Almost all the semiochemicals he and his team analyzed were entirely new to plant science. And these were just a handful of orchids. How many more compounds were out there, at work in the air, manipulating the environment in subtle ways, as of yet invisible to us? Not only that, but thanks to recent advances in gas-sensing technology, Peakall was able to see that the wasp attraction depended on the orchid concocting an exact ratio of two or more of these compounds. In each orchid, the recipe was different. One species of orchid synthesized a gas that was an exact 10:1 ratio of two compounds. Another orchid used a 4:1 ratio of two entirely distinct compounds. All of them were new to science. The specificity was mind-boggling and threatened to exceed even the most advanced modern tools to detect it.

Natasha Myers, an anthropologist of science at York University, and Carla Hustak, who studies the history of science at the University of Toronto, offer a different way of seeing the relationship between orchid and wasp. The orchid and insect bodies, as even Charles Darwin noticed in 1862, were exactly articulated to one another, the most perfect adaptation in nature. But he still saw it as essentially deceptive; since the insect was not getting any reproductive advantage from the encounter, the orchid must be deceiving it. But, ask Myers and Hustak, could this be something else? Perhaps something other than Darwinian survival-of-the-fittest mechanics? They suggest a sort of flirtation between insect and orchid, an interspecies dance in which both parties consent to the arrangement and move toward it with pleasure. Perhaps these encounters could be proof of a different sort of ecological arrangement, where “pleasure, play, or improvisation within or among species” is the norm.

It must also be said that research has found that these orchids tend to slightly underperfect their chemical mimicry, subtly changing some aspects of the chemical concoction to be convincing but not absolutely indistinguishable from the real thing. This makes sense: if they outperformed the female wasps, luring males too well, perhaps male wasps would be so entranced as to not ever copulate with a real female wasp at all. The orchids would risk losing their pollinator. One wonders what the wasp, who might notice the discrepancy, thinks of all this when he decides to copulate with a flower anyway.

Some plants almost exclusively clone themselves, like aspens or dandelions, and still others clone themselves sometimes and have sex other times, like the strawberry. Many plants are bisexual, with male and female genitalia occurring together on the same flower (in plant anatomy these are called, intriguingly, “perfect” flowers). The ancient ginkgo tree can spontaneously switch the sex of a section of its body, producing a female branch on an otherwise male tree. Ginkgo are one of the oldest lineages of trees we live alongside, having persisted for hundreds of millions of years, stubbornly surviving since the time of the dinosaurs. Their sexual fluidity may make them remarkably resilient to whatever the eons have thrown at them.

There is something decidedly queer in all of this—the orchids and aspens and strawberries and antplants and ginkgoes—a sense of sensual entanglement that disregards binaries, runs across the species boundary, and almost gleefully defies heteronormative modes of reproduction.

Basically, air pollution makes an absolute mess of plant communication. And it’s only getting worse. The cascade effects are daunting. Plant defense mechanisms—like the ability to turn bitter when warned of an oncoming insect attack, for example—are often the main way those pests are kept at manageable levels. This sort of natural pest control doesn’t work as well if plants can’t pass messages. “Some species of pests that are normally controlled might reach outbreak level,” Holopainen said.

On top of this, some evidence is trickling in that the way we grow food appears to be getting in the way of plant communication, though this is far from a generalizable situation. Some domesticated plants actually produce more volatiles than their wild varieties. But research has found that commercial varieties of corn are much less able to produce volatile signals than local cultivars when they notice herbivores laying their eggs on them. They are entirely unable to summon beneficial predators. There are, it seems, silent fields of corn, mute in their moment of danger. This raises questions about whether fields planted with highly engineered plants grown in vast identical plots for food, like the corn, have had communication unknowingly bred out of them. Or perhaps it has been rendered unnecessary by the hand of selection for a plant that is given everything it needs to survive without endlessly defending itself. This is, in part, why so many pesticides are necessary in modern industrial-scale farming: some plants seem no longer so able to do things like warn each other of invading pests, or summon beneficial predators at will.

The world uses about two million tons of conventional pesticides to control weeds and bugs each year. (The United States alone says it uses one billion pounds a year.) And these aren’t onetime applications; most crops need to be doused in pesticide several times a growing season to keep the pests away. As a matter of course, pests evolve to become resistant to pesticides, requiring higher and higher doses, until entirely new formulas must be developed. The consequences of all this to human health can be severe. In the United States alone, as many as 11,000 farmworkers are fatally poisoned by pesticides each year, and another 385 million are severely poisoned but don’t die, to say nothing of the birth defects, breathing disorders, and other long-term health impacts of constant exposure to regular doses of the stuff.* Meanwhile rainwater that runs over sprayed fields takes the pesticides off farms and into streams and rivers, which taint water supplies, extending the health impacts to the general population of humans along with fish and aquatic wildlife.

Some even point to returning to the old knowledge of companion planting, the practice of paying attention to which plants survive and grow better in the company of other plants—their natural companions. Strawberries offer a clear example of the advantages of companion planting. A strawberry flower is self-fertile; it can produce fruit using its own pollen, or essentially by having sex with itself. It also can cross-pollinate with other strawberry plants, though this requires the help of flying insects. Farmers know that strawberries will produce a third more fruit—and much of it higher quality—when planted beside borage, a medicinal herb that blooms in perfect blue stars. The borage attracts the strawberry’s pollinator; the better, more copious berries emerge when the sexually adaptive strawberry opts to copulate by way of insects rather than with itself. While home gardeners and Indigenous farmers have used companion planting techniques for ages, it is still a rare approach in conventional, large-scale agriculture.

At this point, Baluška was famous among botanists, or infamous, depending on your view, for being a founding member of the Society for Plant Neurobiology, and for experiments where he found it was possible to anesthetize plants. If plants can be knocked unconscious, does that make them conscious? Baluška says absolutely. “I think consciousness is a very basic phenomenon which started with the first cell,” he says.

Early wheat farmers, weeding by hand, pulled out and discarded the rye-weed to keep their planted crops healthy. So, to survive, a few rye plants took on a form more similar to wheat. Farmers still extricated the pesky rye, when they could spot it. This selective pressure molded the rye to evolve to trick a farmer’s discerning eye. In this case, only the best impressionists survived. Eventually, the rye became so excellent a mimic that it became a crop itself. “Vavilovian mimicry” is now a basic fact of agriculture.* Oats are a product of the same process; they also got their start mimicking wheat.

Crop science is typically seen as the domestication of scrawny wild species to turn them into plump, useful food machines, a testament to human will and ingenuity. But Baluška objects to it being true “domestication” at all. “Domestication would be when one partner has more influence than the other one. But there is no evidence for this,” he says. “A better word would be coevolution. We are changing them, but they are changing us.”

We know that some plants are hallucinogenic, some are addictive, and gardening is proven to reduce depression. What about the compounds that may be inside an apple, or an ear of corn? The question becomes, in what other ways are they influencing us? A fleet of humans carefully tending a field of crops can certainly start to look like an army of plant symbionts, diligently serving the plants’ needs. I think about Vavilovian mimicry: we didn’t domesticate oats; oats domesticated us. When I look at a field of cabbage or pumpkin or blueberries, I wonder: Have they conscripted a symbiont, and is that symbiont us?

“I think the plants are primary organisms, and we are the secondary ones. We are fully dependent on them. Without them, we would not be able to survive,” Baluška says. “The opposite situation would not be so drastic for them.”

It was titled, “Vision in Plants via Plant-Specific Ocelli?” That innocent question mark did nothing to soften the implications. Ocelli is a science term for simple eyes, and Baluška and Mancuso were asking if plants might have them. It included a mention of boquila, which was discovered by Gianoli two years prior to be able to mimic the shape and feel of the leaves of other plants, right down to their color, vein pattern, and texture. Recent research suggests that an ancient cyanobacteria, an early ancestor of plants, had (and still have) the smallest and oldest example of a camera-like eye. Baluška and Mancuso ventured to suggest that plants, which evolved from the union of that organism with an early alga, might never have dropped that useful evolutionary feature. In their letter, they note that the cells closest to the surface of plant leaves tend not to have chloroplasts—the cell type that enables photosynthesis—despite the fact that, logically, the very surface of the leaf is probably the best spot to use for photosynthesizing. “This phenomenon is not easy to rationalize,” they write. Could it be because those cells are used like ocelli? In other words, were they some sort of very simple eyes?