Sensors to Enhance Agricultural Productivity

With an increasing world population, there is concomitant increase in food demand. The COVID-19 pandemic that disrupted global supply chains has also highlighted the importance of ensuring food security. But with extreme and erratic weather under climate change, agricultural yields are expected to reduce due to crop damage and poorer plant growth. One way to overcome these challenges and produce more food for the world is to boost agricultural productivity through the use of sensors.

A sensor can be broadly defined as a device or tool that detects changes in its environment and relays the information to other electronics for further processing to trigger counter-responses to the changes. They are ubiquitous in our everyday lives. Some examples include the thermostat in our air conditioner that detects and regulates temperature changes, and micro-controllers on our smartphones that are sensitive to mechanical touch or pressure. Sensors are also now being developed to monitor plant health for enhancing agricultural yield.

Sensors for water potential

Water plays a crucial role in plant physiology, and studies have shown that the amount of water in plants correlates with plant growth, crop yield, quality, and vulnerability to diseases. Despite the great incentive to measure water potential in leaves, the available tools are limited. 

One current tool—the Scholander pressure chamber—makes use of pressurized gas (typically compressed nitrogen) to force water content out of a leaf placed within a sealed chamber. However, it is destructive to plants. An alternative tool is the psychometer. It makes use of thermocouple heating to measure the water vapor pressure of leaves and the pressure probe, which is a microcapillary that is directly introduced into a plant cell to measure volume changes. These sensors are also damaging to plant cells and their microenvironment. They are also expensive and labor intensive.

To overcome the limitations of existing techniques, researchers at Cornell University invented a new method named AquaDust in which hydrogel nanoparticles coated with fluorescent reporters are imaged to provide water potential based on the phenomenon of Förster resonance energy transfer (FRET). The hydrogel particles swell when they are wet, and collapse when they are dry. This causes fluorescence emission signals from the nanoparticles to change depending on their wet or dry states, and their relative FRET efficiency can be used as an indicator of water potential.

The team performed extensive material optimization to create the water potential sensor. First, they decided to use poly(acrylamide) as a material for their hydrogel nanoparticle as it is weakly dependent on pH, ionic strength, and temperature for swelling to ensure reproducibility. Second, they chose fluorophores with signals that minimally overlapped with leaf autofluorescence and reabsorbed emission from chlorophyll to provide signal specificity. Third, they synthesized nanoparticles of around 42 nm in diameter as particles smaller than 10 nm were likely to pass through cell wall pores, while they wanted to measure extracellular waterpotential.

AquaDust was able to coat the cell wall and minimally penetrate through it. There were also no significant physiological changes to the plants after AquaDust was introduced into the leaves, including carbon dioxide and water vapor exchange rates. The method also worked for different varieties of plants including maize, coffee, and pokeweed for at least five days in a greenhouse under fully illuminated conditions. Importantly, the experimental data for water potential also correlated well with theoretical models and predictions. 

The team next utilized AquaDust to characterize resistance to water flow along a leaf blade. “Major unknowns remain about flow of water through different tissues in the leaf, as it moves from the xylem vessels to tissue outside the xylem, while inevitably some of it is lost through stomata,” says Piyush Jain, the co-lead author of the paper and a graduate student at Cornell University. “Water potential is a measure of water stress and gradients in water potential is a measure of the driving force for the flow of water. Using AquaDust, we observed that tissue outside xylem became a dominant source of resistance to flow of water under moderate and high drought stress. These measurements allowed development of a mechanistic model to accurately predict water stress in field-grown plants with varying environmental conditions.” 

“Our aim is to make use of our portable Raman technology to enhance yields, minimize resources, and sustain Singapore’s efforts to increase local production…”

To demonstrate further translational application of AquaDust, the team also extended their technology to monitor water status in plants under fluctuating climate conditions. 

Compared to existing techniques, there are a few distinct advantages of AquaDust. This method is minimally disruptive and can be used to measure water potential on intact leaves. It is also highly localized, which means that water potential at different parts of the plant can be measured accurately. Finally, it has a suitable dynamic range for physiologically relevant water potential in plants while not being affected by parameters like temperature and pH. These advantages open doors for AquaDust to be deployed as a tool for phenotyping to quantify water-use efficiency in crops for crop selection and breeding. Nevertheless, as this technology still requires the use of manual labor for injection of AquaDust hydrogel nanoparticles, the throughput can be limited. Additionally, injection has been shown to cause damage to plants and this could also limit longitudinal study on the same plant.

“Our first priority with AquaDust will be to use it to answer basic questions about the physiological mechanisms by which plants respond to drought stress,” says Jain. “AquaDust enables unprecedented resolution for measuring water potential at cellular scales. With further development, we envision AquaDust will be embraced in the context of breeding of drought-resistant crops as it provides new opportunities to measure traits across genetic diversity to target improved water-use efficiency (WUE—ratio of carbon uptake/water loss). Improving WUE is crucial across crop management and breeding of various species.” 

Sensors to detect plant stress

Plant stress can be triggered by nutrient scarcity, leading to substantial loss of valuable crops. Shilpi Gupta, postdoctoral researcher at Singapore-MIT Alliance for Research & Technology Centre, and colleagues designed a leaf-clip Raman sensor to measure nitrogen deficiency, as well as to analyze plant metabolites such as carotenoids and nitrates, in maize, bok choy, and choy sum. 

The Raman sensor consists of a Raman fiber connected to a portable Raman instrument and makes use of inelastic scattering of light to measure molecular vibrations and provide chemical fingerprints of samples. It can simultaneously hold lead in place, and block ambient light and transmitted laser radiation to provide reproducible readings. 

“Deploying sensors to phenotypically profile and select plants and to monitor plant health is an excellent strategy to boost agricultural productivity.”

The team showed that their Raman spectra were distinctly different in plants that were grown with and without nitrogen. When the team tested their instrument on plants that were subject to drought and temperature stress, the leaf-clip Raman sensor was also able to provide disparate plots based on their principal component analysis plot. 

“Our aim is to make use of our portable Raman technology to enhance yields, minimize resources, and sustain Singapore’s efforts to increase local production, so that Singapore can produce 30 percent of nutritional needs locally by 2030,” says Gajendra Singh, scientific director of the Disruptive & Sustainable Technologies for Agricultural Precision (DiSTAP) research program. “The sensor can analyze biochemicals in plant leaves in a greenhouse setting and measure nutrient stress in plants in real time. Currently, there is no other sensor that can achieve this. As our system is portable, it can be easily carried to different sections of an urban farm by a single person. In the future, we hope to demonstrate the usefulness of this technology for other types of plant stress, like light stress or stress due to bacterial or fungal infection. Our plan is to further miniaturize the system and also make it more affordable to users like urban farms.” 

Sensors for the future

Food security is an important goal. Deploying sensors to phenotypically profile and select plants and to monitor plant health is an excellent strategy to boost agricultural productivity. 

There are still challenges associated with existing sensor technologies. First, sensors need to be species independent so that they can be broadly applicable to crops other than model organisms like maize and tobacco. Second, they should possess wider dynamic range of detection as this can assist in earlier and more sensitive detection of plant diseases and abnormal changes in plant health. Third, the sensors should be higher in throughput as the land for growing crops is vast, and it is also equally crucial that high throughput should not destroy the plant samples—i.e. the sensors should be minimally destructive. Lastly, sensors need to be more affordable and preferably automated, as this can further reduce manual interventions and relocate human labor for more meaningful work. 

Optimistically, through the use of sensors, we can progress toward precision agriculture to make crop production more sustainable, cheaper, and more secure.