Texas Tech agriculture researchers continue to lead efforts in developing a field-based imaging systems and innovative machine learning tools to address the challenge of lowering flower drops in soybeans.
Retaining even a portion of the 30% to 80% of soybean flowers lost under stressful conditions will allow for a 10% to 20% yield increase for U.S. soybean producers. This advantage can be extended to different soils, water availability conditions, temperatures, and is the major rationale for testing this assumption across different soybean growing states, allowing soybean producers to gain additional economic return at the same level of investment including seed costs, fertilizer levels and management.
Krishna Jagadish, team leader and professor of crop-forage-livestock systems within Texas Tech’s Department of Plant & Soil Science, noted that he is excited to continue working on this complex multi-regional and multi-institutional project, having Kansas State University, University of Missouri and University of Tennessee as collaborators with Texas Tech’s Davis College.
“This is a particularly interesting project as we have two different computer sciences teams from Texas Tech and Kansas State helping develop machine learning and artificial intelligence tools for achieving the project’s targets,” he said.
The on-going interdisciplinary project, titled, “Field phenotyping using machine learning tools integrated with genetic mapping to address heat and drought induced flower abortion in soybean,” is supported in 2025 by a $400,000 grant from regional soybean research collaborative programs, including the Atlantic Soybean Council, Mid-South Soybean Board, North Central Soybean Research Program, and Southern Soybean Research Program.
“The project utilizes cutting edge technology and scientific techniques and highlights a strategic partnership between the Whitacre College of Engineering and Davis College, along with our sister institutions across the country,” said Darren Hudson, Davis College Associate Dean for Strategic Initiatives & Assessment and the Larry Combest Endowed Chair for Agricultural Competitiveness.
A key element of the project is that it has a heavy imaging-based – greenhouse and field – machine learning component through Professor Hamed Sari-Sarraf and his research team which is based within Texas Tech’s Department of Electrical & Computer Engineering.
In addition to Jagadish and Sari-Sarraf, the transdisciplinary research team at Texas Tech includes Department of Plant & Soil Science Post-doctoral Fellow Juliana Espindola, Assistant Professor Gunvant Patil, Research Assistant Professor Impa Somayanda, Research Assistant Professor Christopher Turner, and Chair and Professor of Crop Physiology Glen Ritchie.
Atlantic Soybean Council Executive Director Danielle Bauer Farace noted in an email announcement that the project’s results will be submitted as a semi-annual report and final year-end report to the National Soybean Checkoff Research Database and may be published on the Soybean Research & Information Network as well as other sites.
According to the USDA Economic Research Service, the U.S. is the world's leading soybean producer and the second-leading exporter. Soybeans comprise about 90 percent of U.S. oilseed production, while other oilseeds—including peanuts, sunflower seed, canola, and flax—make up the remainder. According to the Agricultural Marketing Resource Center, dry soybeans are grown in more than 30 states in the U.S. with commercial-scale production in 18 states. North Dakota, Michigan, Minnesota, Nebraska, and Idaho are the top five producing states.
CONTACT: Krishna Jagadish, Professor, Department of Plant & Soil Science, Davis College of Agricultural Sciences & Natural Resources, Texas Tech University at (806) 834-7953 or kjagadish.sv@ttu.edu
0113NM25