How Groundbreaking Tree Sensor Can Save Our Groundwater

Startup invents tree-whispering smart irrigation system that can tell when the plant is thirsty, not just when the ground is dry.

Anat Solomon Halgoa

Fresh water is growing scarcer in our warming world, not least in California and the Middle East. There is a crying need not only for crops that can survive in harsh conditions, but for better irrigation technology.

Saturas, an Israeli startup founded in 2013 under the auspices of the Trendlines Agtech incubator, is developing a sensor that measures a plant's actual thirst. Its aim is to make irrigation maximally effective, resulting in healthier plants with better yields. The sensor will also help preserve the groundwater, says company CEO Anat Solomon Halgoa.

With today's wasteful farming methods, farmers normally over-irrigate and the excess water seeps into the groundwater together with plant nutrients – which cause, in a word, contamination.

Most "smart" irrigation techniques today measure the water level in the ground, explains Solomon Halgoa. But what counts is whether the plant is thirsty, not whether the soil is. Thus, she explains, even smart irrigation is a crude beast. There is no feedback on how much water the plant is actually taking up.

A better measure, she says, is the water potential in the plant's stem. Think of the water potential as the "need for water".

Water rises from the roots because of the water pressure gradient created as water evaporates from the leaves during sunlight hours, Solomon Halgoa explains. Accurately measuring that water potential is presently labor-intensive and takes hours. Leaves near the trunk are wrapped in foil to halt water evaporation. After waiting about an hour for the leaves to reach equilibrium, they are picked and put into a device that measures the pressure required to extract the first drop of water, which is observable using a magnifying glass.

"A person can check maybe 40-60 leaves a day. It's not efficient," says Solomon Halgoa.

Not to mention, it's hardly real-time data. The Saturas sensor, which gets embedded in the tree trunk forever more, can measure the water potential in real time.

Right now the prototype sensor is a tad large, 2.5 centimeters in diameter, suitable only for trees and mature vines. Saturas is working on miniaturization and envisions a final diameter of 5-8 millimeters, which would also be suitable for tomato plants, cotton and so on, Solomon Halgoa says.

Meanwhile the larger version has been undergoing months of successful testing in peach and citrus orchards, she adds. The company hopes to start selling in about a year and a half, by which time management hopes to have an international collaboration agreement in place.

How it works

The sensor consists of a small cell that contains a solution at high osmotic pressure. Embedded by drilling into the tree trunk, the sensor is nestled against the water-containing tissue of the tree.

Between the sensor cell and the water tissue is a membrane permeable only to water. When the tree is saturated (not thirsty), water seeps through the membrane into the sensor cell by osmosis.

Since the sensor cell is hermetically sealed, even a few molecules of water passing through will raise the pressure inside it. That difference can be measured, claims Solomon Halgoa.

When the tree is thirsty, the opposite happens. Water seeps through the membrane from the cell into the tree, and the pressure inside the cell drops. "When we know the osmotic pressure exchange rate we started with in the cell, the rest is a simple algorithm," says Solomon Halgoa.

Theoretically, the company plans for the sensors to remain operable for three years, after which they stay in the tree doing nothing, including doing it no harm.

Naturally one doesn't ram a sensor into each plant. One or two per hectare should do the trick. The data gets transmitted to a central automated irrigation control system.

Right now the company is thinking of $300 per sensor. On flat ground and uniform crops ("a reasonably homogenous area"), a farmer doesn't need a sensor per tree in an orchard, but only 1-2 representative trees per 10 dunam. If he happens to grow loquats on a hillside, for instance, he'll need more. Leaving aside farmers who cultivate exotic landscapes, the return on the investment for the farmer is estimated to be less than a year.