Are Plants Smart? The Evidence for Plant Intelligence
Let’s go a little further, and talk cognition — do plants have any semblance of intelligence? Are plants smart?
We’re at a lucky time in history (at least in the narrow sense of plant cognitive research), for the question of plant intelligence is just now tentatively dipping its toe back into the bubbling hot tub of mainstream science. Or wait, backwards metaphor — maybe mainstream science should be the one dipping its toe. Regardless, we’re seeing over the last few years a welcoming of this field back into academia after many years in the wilderness. Instead of new-agey pseudoscientists, the question can now safely be considered by tenured professors, albeit ones that skew a bit new-agey. For example, our friend Dr. Monica Gagliano, whom we learned about in the last section — on her lab website, she mentions that she’s interested in “Plant Communication, Behavior & Congitive Processes (including Consciousness)” ! Now that’s bold.
And it looks like Dr. Gagliano can back up the boldness — aside from her stunning paper on plant communication, she’s also showed perhaps the first example that plants can learn — learning in the sense of Pavlov’s dog, in the sense of observing and remembering a lesson from experience. In a recent paper, Dr. Gagliano demonstrated the studying skills of the Mimosa pudica plant. Mimosa pudica is one of the very few plants that can move as quickly as an animal — like the Venus Flytrap, it rapidly moves its leaves in response to being touched. Whereas the flytrap is looking for a meal, the Mimosa pudica plant is just trying to protect itself. If you run your finger along its fern-like leaves, the leaves will recoil and close up on themselves:
A plant with cat-like reflexes sure is more convenient for reinforcement learning studies — imagine waiting around for an oak tree to react to a stimulus. And indeed, Gagliano showed that these plants are able to learn when to not react to something touching their leaves. The plants eventually stopped reacting when they were subjected to a harmless dripping water source, defiantly leaving their leaves open. The effect sure looked like learning — the plants were able to “remember” not to recoil from this harmless stimulus for several weeks, and they learned to squelch their reflex more readily when they were in a stressful environment (low light levels), where presumably it pays to pay attention — no sense wasting energy recoiling to harmless touches when precious sunlight is sparse.
Dr. Gagliano is a rock star among a small group of researchers who are pushing forward plant intelligence as a valid topic for research. Not in the same terms of an animal’s nervous system with a central brain, but perhaps more like the collective intelligence of a bee colony or an anthill — think more along the lines of distributed intelligence that emerges from the combined action of many dumb individuals, rather than a single, conscious, deliberating brain. Sometimes such a system can act surprisingly intelligent.
Another leader of the field, Dr. Stefano Mancuso, heads up the International Laboratory of Plant Neurobiology in Italy. Calling the research area “Neurobiology” is a bit of a stretch, as plants quite definitively don’t have neurons, but a stretch that was chosen intentionally. Mancuso and his collaborators (whom I’m now going to call the Mancuso Mafia, because I like the ring of it) aren’t suggesting that plants have as-yet-undiscovered neural anatomy, but their choice of term is meant to convey idea that plants might have a “nervous” system in the sense of some internal communication pipeline. Plants may not have a brain, but they might have a “brain” — a rudimentary sensory and decision-making system made from we-don’t-yet-know-what.
The Mancuso Mafia suggest that the “brain” of plants, if the metaphor fits at all, is probably down in the roots, who have perhaps the hardest cognitive responsibilities in the plant’s entire body. It’s not as though plant roots are defusing bombs or defeating chess masters, but they do have to decide where to grow in search of water and nutrients. Remember a plant can’t simply walk down to the 7-11 — if those roots make poor decisions about where to grow next, the plant starves. And befitting their heavy responsibility, the end tips of roots have a certain set of skills you find nowhere else in the plant body. Check out all the stuff that root tips can do:
- Sense gravity, of course — so they know to grow “down”
- Exquisite sensors of the underground chemical environment (to head toward nutrients, away from hazards, etc.)
- Root tips can distinguish between “self” (other root tips from the same plant) and “other” (from other plants)
- Root tips grow in concert with each other, akin to the “swarming” motion of a flock of birds or school of fish — they tend to align, especially when they grow near each other
- Avoid growing into an inert obstacle, even before touching the obstacle
- Can sense light (all the better to grow away from the sun down into the dirt, of course)
As we’ll learn, the Mancuso Mafia have come to see these root tips as the closest plants get to having a brain.
A Plant “Nervous” system?
Take my word for it, plants just don’t have nerves like we do — their anatomy just doesn’t contain neurons. So is there any evidence of a nervous system in plants? Anything that sends signals around, analogous to our nerves?
So how would different parts of the plant communicate with each other? For example, all the little root tips — how do they share information? One thought is from releasing volatile chemicals. If plants don’t have any kind of electrical-powered communication system, then the fastest way to get information from leaf A to leaf B is for leaf A to release some smell that will waft over to leaf B. That’s thought to be a way that plants can quickly mount a plant-wide defense against predators. But that isn’t going to work underground, now is it? How would these oh-so-amazing root tips talk to each other?
This was surprising to me, but there’s actually quite a bit of research showing that plants do have neuron-esque electrical and chemical signaling systems. It’s nowhere near worked out yet, but the evidence is piling up. In retrospect it seems pretty obvious that certain fast-acting plants like the Venus Flytrap need something quicker than leaf farts to snap shut on a fly — turns out they have an electrical response quite like the action potential in animal neurons that controls their quick response. Plant cells can conduct their own action potential “spikes”, which in animals are the key unit of information transfer.
In fact the Mancuso Mafia point to a particular region in the tips of the roots, a spot that seems to control where the root grows next. Specifically they point out an area called the “transition zone”, just a teensy bit behind the very tip of the root. This area contains excitable cells (like a neuron) that they suggest is the source of the root’s “intelligence” and decisionmaking. Recently, they published evidence for the same kind of synchronized behavior in the transition zone that you’d normally see in neuroscience papers (i.e animal brains) — Masi et al from the Mancuso Mafia (Spatiotemporal dynamics of the electrical network activity in the root apex) were able to record a spreading action potential spike in the transition zone as it moved from cell to cell. This is the first time any such pattern has been seen in plants — plant cells spiking in concert, rather than just randomly. What’s more, these plant cells were spiking on their own, not in response to some stimulus. Conceivably electrical spiking would be used by just a handful of plants that need to respond in a hurry to some stimulus (like a fly landing), but now we’re seeing the plant cells using electrical communication without any external signal — e.g. we might be eavesdropping on internal communication. (A transcript from this internal communication reads: “that Mancuso guy is sticking electrodes in us again. Quite painful. Recommend we hurry up and evolve some feet.”)
So at minimum, this root tip might act as a sophisticated sensor, which has to make some sort of decision (e.g. keep moving forward, turn right, hit a pocket of fresh cow poo, etc.) based on what it senses (water, nutrients, other roots, competitors, pipes and other obstacles), and possibly pass back the news to the rest of the plant. This counts as a sensorimotor computation just like what you find in animals. Since a plant can’t get up and move, the only way for a plant to search out new food and water is to grow — specifically, shoot out little root tendrils to ever-new spots underground. So in a sense, these roots are at the plant’s “face” — the portion of its body closest to the nutrients it needs to find. This spot in the root is apparently the control center for its root growth and communication to the rest of the plant. Even if it doesn’t indicate complex thought, the work by Masi & others is pretty exciting. Plus they got to deploy the term “spatiotemporal dynamics”, which is always a good thing.
But this nervous system sans nerves isn’t just limited to the roots. Recently Dr. Ted Farmer showed that electrical signals are conveying information even up in the leaves. His team attached electrodes at various points on an Arabidopsis plant, and watched as an electrical signal propagated away from a spot being nibbled on by a cotton leafworm. Sounds pretty similar to our own nervous system, even thought the plant doesn’t explicitly use neurons. And the information was received — soon after the electrical signal was sent, the rest of the plant released a chemical called jasmonic acid in an attempt to drive away the munching worm.
And there’s more — Stanislaw Karpinski‘s research seems to show that plants can signal information from leaf to leaf, using more than just the wafting vapors I talked about earlier. They appear to use a signalling cell called a “bundle sheath cell” that transmits information using electrical signaling, just like neurons in animals. Karpinski cleverly showed the leaf-to-leaf communication by shining a particular color of light on only one leaf, then watching as the other leaves reacted to that color. It appears that the plant is actively monitoring the quality of the light, perhaps to know when the sunlight is particularly tasty, or to know when it’s in shade, or to know what season it is in the year from the particular spectral content of the sunlight during a particular season. And the plants are able to remember information about the exposure to certain kinds of light. According to Karpinski, plants need to know roughly what time of year it is, to mount up a defense against pathogens that crop up in specific seasons. Sort of like getting yourself ready for flu season. He hypothesizes that plants infer what season it is from the spectral quality of the light, then adjust their immune systems accordingly. In his experiments, he sees differences in how well his plants are able to resist a pathogen depending on what kinds of light he shone on them before the exposure.
This seems like a little bit of a stretch to me, but I’m a random schmoe running a science education site, not an NIH grant reviewer, so I gleefully report on his ideas, no matter how fanciful. And boy are they fanciful — Karpinski even manages to bring quantum physics into it. I quote:
Our recent results allow us to suggest that plants actually work as a biological quantum computing device that is capable to process quantum information encrypted in light intensity and in its energy with help of NPQ and quenched singlet stages, transmute it into analogous (PEPS) information and finally is capable to physiologically memorize it with help of qp , O2 .- and H2 O2 .
I don’t know who’s more of a hoot, Karpinski or the Mancuso Mafia. I love it when a scientist has the cojones to connect their research to quantum mechanics — the more preposterous the stretch the better. Is it a stretch in this case? He suggests that plants are doing “biological quantum computation”. Is it really quantum computing? Or “quantum” in the sense that every biochemical reaction involves chemistry, which is explained by quantum mechanics? From my novice point of view, I think it’s more toward the latter, but I’d love to be proved wrong. To read more on Karpinski and his quantum houseplants, check out his homepage.
Thinking without Thinking
Whereas we’re focusing on rummaging around plant anatomy for something neuron-esque, we could be missing intelligent problem-solving behavior conducted by run-of-the-mill parts of the plant we already know about. Much like the collective intelligence of an ant colony, parts of the plant might be making sophisticated decisions without the help of any sort of nervous system. In a paper with the fantabulous title Evidence for complex, collective dynamics and emergent, distributed computation in plants, David Peak and colleagues contend that the leaves of plants are capable of calculating the answer to a complex question relating to breathing. As we all know from grade school, plants breathe CO2 — turns out they do this through tiny openings on their leaves called stomata. But a plant can’t afford to just let its stomata hang wide open all day, lest it dehydrate from water evaporation through the very same stomata. So the plant needs to decide when to open its pores, and by how much — too much and you lose all your fluids, too little and you don’t get enough precious CO2 to photosynthesize sunlight. It’s a trade off of two competing drives, and plants have a strong evolutionary incentive to get that tradeoff just right.
So how do they do it? Park et al suggest something so cool, I wound up devoting an entire other article to it here on timeblimp.com which I encourage you to check out. Park says the plant’s leaf is set up as a Cellular Automaton, and that setup allows the leaf to solve the stomata tradeoff problem without having a central intelligence to make a decision. If true, this is one of the very first discoveries of Cellular Automata actually being employed by a living system to solve some problem. What’s a Cellular Automaton, you ask? Why, read my other article on this paper to find out! Too lazy? Yeah, I feel you — in short a Cellular Automaton is a very simple computational model used for any system that has a lot of simple interacting parts. It’s super simple — essentially just a set of dots that move around a fixed checkerboard grid interacting with each other. They’re so simple, yet lead to such complex-looking patterns, that they’ve been enthralling people for decades. They’re so enthralling that Stephen Wolfram built a theory of the universe based on them. Until lately I was pretty skeptical that he could be right, given the lack of evidence that anything in the natural world uses these things. But holy crap, here we are looking at a potential Cellular Automaton built right into in the leaf of a plant. But the saga of Wolfram and the Cellular Automata is a story for another time (that you should go read immediately after this) — let’s for now admire the possibility that plant leaves might be using pretty sophisticated computational techniques to come to a decision about opening their stomata — and they do it with no help from any centralized “brain”.
Who’s doing the thinking?
This proves little either way regarding plant intelligence, but the Mancuso Mafia also point out the astounding fact that plants can be anesthetized. Their awareness and reaction to the environment can be dimmed to a state somewhat resembling sleep. And from the same chemicals that anesthetize animals. Claude Bernard showed this almost two hundred years ago, and the Mancuso Mafia bring it up again to illuminate the similarities between plants and animals. Being anesthetizable doesn’t of course imply that you have a central brain, but it’s another suggestive fact that leads us to ponder whether plants have more going on inside than it first appears.
So could plants really be intelligent? Could they be conscious? My take — they probably aren’t conscious like we are, but if plants can solve sophisticated problems, then they probably have to be called intelligent. If an ant colony persistently finds ways into your house, that’s relatively smart problem-solving, even though no individual ant is smart. If swarm intelligence can solve sophisticated problems and adapt to a changing environment, then we pretty much need to call them intelligent. One might argue that even though such “emergent” intelligence might appear surprisingly smart, it may not actually be all that smart — ever hear of an ant mill?
In the next (and final) section, we’ll relax from the hard-nosed scientific facts discussed so far, and allow ourselves to speculate — what would it feel like to be a plant? Read on for more.
An Excellent New Yorker article on plant intelligence and consciousness — this is a good one for further reading. In fact you should probably read it instead of this article. Sorry I didn’t tell you earlier.
Dr. Monica Gagliano ’s research page
Stefano Mancuso’s International Laboratory of Plant Neurobiology