In an era marked by unprecedented climate challenges and escalating threats to global food security, a pioneering study out of Israel is redefining how scientists approach the resilience of wheat—one of the world’s most vital staple crops. Researchers at the Hebrew University of Jerusalem’s Faculty of Agriculture, Food and Environment, in collaboration with the Volcani Institute, have successfully leveraged cutting-edge drone technology combined with advanced spectral imaging to unlock the genetic secrets behind wheat’s ability to withstand drought and heat stress. This breakthrough not only accelerates the breeding of climate-resilient wheat varieties but also opens a new frontier in precision phenotyping driven by remote sensing and machine learning.
Harnessing the power of unmanned aerial vehicles (UAVs) equipped with hyperspectral and thermal cameras, the research team conducted extensive field experiments to monitor the physiological and biochemical traits of several hundred wheat genotypes grown under both well-watered and rain-out drought conditions. These drones captured highly detailed images that record variations in leaf thermal emission and light reflectance patterns, which are key indicators of critical plant traits such as stomatal conductance, leaf area index, and chlorophyll content. Such parameters directly relate to the plant’s water-use efficiency and photosynthetic capacity, offering a window into how different wheat lines manage water loss and carbon assimilation under environmental stress.
Traditionally, stomatal conductance — the rate at which CO₂ enters and water vapor exits the leaf through microscopic pores — has been measured by cumbersome, low-throughput instruments such as porometers, which require close contact with plants and considerable manual labor. This has severely limited large-scale genetic studies aimed at understanding plant physiological responses to stress. The revolutionary UAV-based approach developed by Ph.D. candidate Roy Sadeh, under the expert supervision of Dr. Ittai Herrmann and Prof. Zvi Peleg, circumvents these limitations by remotely acquiring thousands of data points rapidly across diverse germplasm collections, all without physically disturbing the plants.
.adsslot_cW2UdRtD0s{ width:728px !important; height:90px !important; }
@media (max-width:1199px) { .adsslot_cW2UdRtD0s{ width:468px !important; height:60px !important; } }
@media (max-width:767px) { .adsslot_cW2UdRtD0s{ width:320px !important; height:50px !important; } }
ADVERTISEMENT
Over two full growing seasons, the team’s drone flights at the Pheno-IL research facility involved capturing multiple spectral bands and thermal data, which were then fused into comprehensive phenotypic profiles through sophisticated computational models. Utilizing support vector machine algorithms—a subset of machine learning—the researchers translated raw imagery into quantitative estimates of water-use traits with an impressive 28% increase in accuracy over previous methods. This computational framework represents a significant advancement in transforming raw sensor data into biologically meaningful metrics at scale.
Crucially, integrating these precise phenotypic measurements with high-density wheat genotyping enabled a powerful genome-wide association study (GWAS). This analysis revealed 16 genetic loci significantly correlated with enhanced performance under both optimal and drought stress conditions, marking a pioneering stride in linking remotely sensed physiological traits with underlying genetic architectures. These genetic markers were subsequently validated in a follow-up field trial, cementing their potential utility as targets in wheat breeding programs.
The implications of this study extend far beyond academic curiosity. By enabling a high-throughput, non-invasive, and genetically informed phenotyping pipeline, the research team has effectively unlocked a fast track for breeders to select wheat lines that exhibit superior drought tolerance and carbon assimilation efficiency. This technology-driven breeding paradigm aligns perfectly with global efforts to build climate-resilient food systems, particularly as rising temperatures and erratic precipitation patterns increasingly jeopardize crop yields worldwide.
As Roy Sadeh aptly explains, “Our drone-based method fundamentally changes the pace and scale at which we can identify plants with desirable physiological traits. It empowers us to plant the seeds for future crop varieties that are better prepared to thrive in increasingly dry and hot environments, ultimately securing food production for generations ahead.” This statement underscores a paradigm shift in agricultural science where multidisciplinary innovations—from remote sensing and computational biology to genetics—converge to meet one of humanity’s most pressing challenges.
The research is particularly noteworthy for its utilization of hyperspectral imaging, which captures reflectance data across a broad spectrum of wavelengths invisible to the naked eye. This spectral richness enables decoding of subtle variations in leaf chemistry and structure, such as pigment concentration and canopy architecture, which are intimately tied to photosynthesis and water regulation. Coupled with thermal infrared imaging that detects leaf surface temperatures, the combined imaging modalities offer a multidimensional picture of plant health and stress responses in real time.
Furthermore, the experimental setting—a rain-out shelter at the Pheno-IL facility—allowed the team to precisely simulate drought stress conditions while maintaining uniform environmental controls. This setup ensured the reliability of trait measurements and the relevance of the findings to actual field situations where water scarcity is a prevailing concern. The rigorous validation approach, including the replication of genetic associations in independent trials, enhances confidence in the robustness and applicability of the results.
From a technological perspective, this study exemplifies how the fusion of airborne remote sensing platforms with machine learning analytics is revolutionizing precision agriculture. Support vector machines enabled pattern recognition and complex trait prediction beyond the capabilities of traditional statistical methods, highlighting the immense potential of artificial intelligence to decipher vast biological datasets. As computational power continues to grow and sensor technologies improve, such integrative approaches promise to become standard tools in crop improvement initiatives worldwide.
In summary, the integration of UAV-borne hyperspectral and thermal imaging with genome-wide genetic analyses opens a transformative new pathway in plant science and breeding. This innovative methodology offers a scalable, efficient, and precise means to dissect complex physiological processes like stomatal conductance at the genetic level, accelerating the development of wheat varieties that can endure the multifaceted stresses imposed by climate change. As global agriculture stands at a critical juncture, such forward-thinking research not only advances scientific understanding but also delivers concrete solutions critical for sustaining the food supply in a warming world.
The study, published on April 19, 2025, in the journal Computers and Electronics in Agriculture, represents a landmark achievement in crop phenomics and genetics. Supported by the Israeli Council for Higher Education’s Future Crops for Carbon Farming project, the Dutch Ministry of Foreign Affairs, and the Chief Scientist of the Israeli Ministry of Agriculture and Food Security, this work exemplifies the importance of international cooperation and innovation-driven funding in addressing global agricultural challenges.
Looking ahead, the research team envisions wider applications of UAV-based phenotyping across other crop species and stress conditions. By refining imaging and analytic technologies and expanding genetic databases, the precision breeding revolution sparked by this study promises to equip farmers and breeders with unprecedented tools to combat the uncertainties of climate change. As food security continues to dominate global priorities, this fusion of drone technology, spectroscopy, and genomics may well become one of agriculture’s most powerful weapons.
Subject of Research: Not applicable
Article Title: UAV-borne hyperspectral and thermal imagery integration empowers genetic dissection of wheat stomatal conductance
News Publication Date: 19-Apr-2025
Web References: http://dx.doi.org/10.1016/j.compag.2025.110411
Image Credits: Roy Sadeh, Ittai Herrmann, Prof. Zvi Peleg
Keywords: Agriculture, Crop domestication, Crop irrigation, Crop production, Crop science, Food science, Environmental sciences
Tags: advanced imaging for plant traitsclimate change and food securitydrones in agriculturedrought-resistant wheat developmentgenetic research in crop resilienceHebrew University research initiativesmachine learning in agricultureprecision phenotyping techniquesremote sensing in agriculturesustainable agriculture practicesUAV technology in plant sciencewheat breeding innovations
