Do Plants Scream When They Are Hurt? The Ultrasonic Experiments

A greenhouse you cannot hear

A thirsty tomato plant stands under artificial light. Its leaves are beginning to lose water. There is no voice, no movement dramatic enough to catch the eye, no obvious sign of distress that a person on the other side of the room could hear.

Place an ultrasonic microphone nearby, and the greenhouse becomes a different place.

Researchers have recorded sharp airborne clicks from certain stressed plants. The sounds are mostly beyond normal human hearing. They are not an invented soundtrack added to a documentary. They are measurable acoustic events, captured with equipment designed to register frequencies that ordinary human ears miss. [S014]

The headline practically writes itself: plants scream when they are hurt.

But what does the experiment actually show? What is emitting the sound? Can any other living creature hear it? And does the word scream reveal a discovery or smuggle an explanation into an observation?

The experiment inside the chamber

In 2023, Itzhak Khait and colleagues published a study in Cell describing sounds emitted by stressed plants. They worked with tomato and tobacco plants and recorded them under different conditions, including drought stress and cutting. The researchers used an acoustic chamber and then tested recordings in a greenhouse setting. [S014]

Their measurements revealed that stressed plants could emit ultrasonic airborne sounds detectable from a distance. The team trained machine-learning models to classify some plant conditions from recorded acoustic patterns. In other words, under the experimental conditions, the sound contained information about what was happening to the plants.

This is the decisive observation. It does not depend on the listener believing that plants have minds. A microphone detected pressure waves. Those waves varied with experimental conditions, and statistical classification showed that the signals were not meaningless uniform noise.

The researchers did not establish that every cut plant emits the same sound. Nor did the study prove that an injured plant was consciously attempting to attract assistance. The word informative describes a property of measurable signals. It does not establish an intentional sender.

Why ultrasound changes the story

Human beings build most of their everyday world around frequencies we can hear. A sound above that range may be present in the air without registering as a sound in our experience.

That is why the discovery feels so cinematic. A garden can look silent to us while containing vibrations, movements and chemical changes outside our normal perception.

The researchers' acoustic recorders were important because they could separate plant-related emissions from background noise and measure them objectively. Their controlled conditions also helped address a familiar objection: perhaps the sound came from equipment, passing insects or the building rather than the plant.

The results supported acoustic emissions associated with plant stress. The broader ecology still requires further experiments. How far do different sounds travel outdoors? How do rain, wind, leaf motion and competing ultrasonic sources affect detection? Do species in natural habitats emit useful patterns often enough for another organism to learn them? [S014; S015]

A laboratory finding can open that door without yet walking through it.

What might be making the clicks?

One candidate is cavitation, a physical process associated with water transport. Water travels through plant vascular structures under tension. When that system is stressed, the formation and movement of gas bubbles can produce mechanical events. These may contribute to acoustic emissions. [S014]

Notice what this explanation does and does not do. A physical mechanism for making a click does not make the click uninteresting. Nor does it prove the plant intended to send a message.

A snapped twig also makes a sound. Its sound can reveal that a branch has broken even though the branch did not consciously choose to warn anyone. Similarly, a physiological sound can be informative about plant condition whether or not it evolved as communication.

The word scream suggests a state of suffering. The measurement confirms an acoustic event. An emotion or pain experience would require a different type of evidence.

Pain is not the same as damage

Plants respond powerfully to injury. A damaged leaf can trigger changes elsewhere in the organism. Defensive chemistry can rise. Electrical and calcium signals can travel through plant tissues. These are measurable physiological reactions. [S021]

But injury, nociception-like signaling, stress responses and consciously felt pain are different concepts. In animals, pain research involves nervous-system and behavioral evidence alongside reports of experience when reports are possible. Plants do not have animal brains or neurons. Whether there could be other forms of experience is a philosophical and scientific controversy, not something an ultrasonic click can independently settle. [S019]

It would be as careless to claim that all questions about plant experience are metaphysically closed as it would be to present recorded clicks as a direct recording of agony.

A good title can ask whether plants scream. The article must still explain the distinction between a vivid metaphor and what microphones demonstrated.

The unexpected listener: a moth

The acoustic story became more interesting with later work. Researchers investigated whether insects could use sounds associated with drought-stressed plants when choosing where to lay eggs. A peer-reviewed version of this study appeared in eLife in January 2026. [S016]

The experiments focused on female moths. In controlled tests, the researchers played sounds recorded from stressed tomato plants and compared the insects' egg-laying decisions under different arrangements.

One result was surprising: when there was no real plant in the arena, moths sometimes favored the side where plant-like sounds were played over silence. But when actual plants were available, the preference shifted toward silent plants, in line with a choice that could favor better-hydrated hosts. Experimental disruption of hearing removed the relevant sound-based preference. [S016]

That combination is more informative than a sensational headline. It suggests the animals were responding to acoustic information, while the direction of their behavior depended on the larger context. A cue that says there may be a plant nearby is not the same cue as one that says this specific plant is the best nursery.

The researchers interpreted the findings as evidence for an insect-plant acoustic interaction. The work does not mean a tomato knowingly called a moth. It suggests that, at least in these conditions, another organism can exploit plant-produced sound.

What the reviewers questioned

Peer review matters most when a story is irresistible.

The eLife assessment highlighted the potential importance of the findings, while noting limitations including the relationship between playback frequencies and typical plant sounds, inconsistent responses depending on whether real plants were present, and challenges in generalizing experimental effects to the field. [S017]

These limitations do not erase the experiment. They frame what it can support. The obvious next steps include natural field recordings, better estimates of real-world acoustic exposure, and tests of whether additional insects respond across habitats and plant species.

This is how a new area of biology matures: a laboratory observation is followed by a proposed receiver, then by more exact questions about ecological context and reproducibility.

Is a sound automatically a message?

Imagine a night insect flying through a garden. In one direction it detects a patterned series of ultrasonic clicks. In another it hears relative quiet.

If the insect adjusts its movement or egg-laying behavior in response, the sound functions as information for the listener. Does that make it communication?

There are two ways to answer, and neither should be hidden.

In everyday language, a system in which one organism produces an identifiable cue and another reacts may be described as communication. In evolutionary biology, researchers often distinguish a cue, which another organism can use, from a signal shaped through selection to affect a receiver. Demonstrating a benefit to the emitter can matter to that distinction. [S011; S016]

The receiver can learn from a sound even if the producer never benefited from emitting it. A moth could eavesdrop on the physical consequences of drought. That would still be a remarkable sensory relationship.

The same discipline helps with the claim that plants talk to one another. To establish plant-to-plant acoustic signaling, scientists would need to show that a neighboring plant detects the emitted sound and changes its physiology because of that sound rather than vibration through the ground, wind, or other chemicals.

The experiments discussed here do not yet show that complete chain.

The older dream of listening to plants

Long before modern ultrasonic studies, writers and investigators speculated about the hidden sensory world of plants. Peter Tompkins and Christopher Bird's The Secret Life of Plants popularized dramatic claims about awareness, emotional reactions and extraordinary sensitivity. A copy is listed in the Avalon Library archive. [S002]

The historical influence of that book is worth understanding. It encouraged people to imagine plants as active beings rather than passive scenery. But it also illustrates why a compelling narrative can outpace reliable experimental evidence.

The later acoustic findings should not be retroactively treated as confirming every claim in older popular literature. Measuring a sound does not validate telepathy, mind reading, or plant pain.

Nor should old exaggerations prevent us from appreciating an authentic discovery when better methods reveal one.

What a garden might eventually teach us

There are practical possibilities here. If drought stress produces an acoustic pattern under relevant conditions, future monitoring technology might help detect water stress in cultivated plants. If insects can hear some plant sounds, agricultural ecologists may discover interactions that currently pass beneath our hearing threshold.

But it would be premature to promise farms that can diagnose every sick plant by sound. Species, environmental noise, timing and calibration all matter. Researchers would need reliable field performance before transforming a proof of concept into a general tool.

A scientifically honest story can remain startling without pretending the future technology already exists.

The silence was ours

When a person walks through a greenhouse, its silence may say more about the limits of human hearing than about the absence of biological events. The plants are moving water, responding to light, modifying chemistry and sometimes producing ultrasonic clicks.

The new research invites a larger question: how many relationships in nature exist in channels we rarely measure?

That is not proof of a secret language of suffering. It is an invitation to build better microphones, sharper experiments and more precise concepts.

A tomato plant under stress can make a sound. Some moths, in experimental conditions, can respond to such sounds. Between those findings and the image of a plant crying for help lies a fascinating frontier of biology. We do not need to settle it with a metaphor. We can keep listening. [S014; S016; S017]