© Unsplash

Researchers at Tel Aviv University have recorded and analyzed sounds distinctly emitted by plants, using AI. The click-like sounds, similar to the popping of popcorn, are emitted at a volume similar to human speech, but at high frequencies, beyond the hearing range of the human ear. The researchers say they found that plants usually emit sounds when they are under stress, and that each plant and each type of stress is associated with a specific identifiable sound. While imperceptible to the human ear, the sounds emitted by plants can probably be heard by various animals, such as bats, mice, and insects.

The study was led by Prof. Lilach Hadany, together with Prof. Yossi Yovel and a team of researchers from Tel Aviv University and was announced in a press release.

How they did it

Professor Hadany says: “From previous studies we know that vibrometers attached to plants record vibrations. But do these vibrations also become airborne soundwaves – namely sounds that can be recorded from a distance? Our study addressed this question, which researchers have been debating for many years.”

At the first stage of the study the researchers placed plants in an acoustic box in a quiet, isolated basement with no background noise. Ultrasonic microphones recording sounds at frequencies of 20-250 kilohertz (the maximum frequency detected by a human adult is about 16 kilohertz) were set up at a distance of about 10cm from each plant. The study focused mainly on tomato and tobacco plants, but wheat, corn, cactus and henbit were also recorded.

“Before placing the plants in the acoustic box we subjected them to various treatments: some plants had not been watered for five days, in some the stem had been cut, and some were untouched,” prof. Hadany comments, “Our intention was to test whether the plants emit sounds, and whether these sounds are affected in any way by the plant’s condition. Our recordings indicated that the plants in our experiment emitted sounds at frequencies of 40-80 kilohertz. Unstressed plants emitted less than one sound per hour, on average, while the stressed plants — both dehydrated and injured — emitted dozens of sounds every hour.”

AI algorithms

The recordings collected in this way were analyzed by specially developed machine learning (AI) algorithms. The algorithms learned how to distinguish between different plants and different types of sounds, and were ultimately able to identify the plant and determine the type and level of stress from the recordings.

Moreover, the algorithms identified and classified plant sounds even when the plants were placed in a greenhouse with a great deal of background noise. In the greenhouse the researchers monitored plants subjected to a process of dehydration over time and found that the quantity of sounds they emitted increased up to a certain peak, and then diminished.

A resolved controversy

“In this study we resolved a very old scientific controversy: we proved that plants do emit sounds!” Prof. Hadany announces, “Our findings suggest that the world around us is full of plant sounds, and that these sounds contain information – for example about water scarcity or injury. We assume that in nature the sounds emitted by plants are detected by creatures nearby, such as bats, rodents, various insects, and possibly also other plants, that can hear the high frequencies and derive relevant information. We believe that humans can also utilize this information, given the right tools – such as sensors that tell growers when plants need watering. Apparently, an idyllic field of flowers can be a rather noisy place. It’s just that we can’t hear the sounds!”

Selected for you!

Innovation Origins is the European platform for innovation news. In addition to the many reports from our own editors in 15 European countries, we select the most important press releases from reliable sources. This way you can stay up to date on what is happening in the world of innovation. Are you or do you know an organization that should not be missing from our list of selected sources? Then report to our editorial team.

CurrencyAmount