Research HighlightsRetaining Flowers to Boost Soybean Yield
Highlights:
- A significant percentage of soybean flowers fall off before developing into pods. Improving flower retention increases yield.
- Collaborative research supported by multiple soybean research organizations funded work to tackle the complex challenge of tracking flower and pod development and mapping it to genetics.
- The research has resulted in a tool soybean breeders can use to track flowers converting to pods, as well as a list of candidate genes likely linked to flower and pod retention.
By Laura Temple
Under normal growing conditions, soybeans drop 25 to 30% of their flowers before they develop into pods. Under stress like drought or heat, that flower drop can increase as high as 80%.
“Flowers demand the most energy compared to any other organ in a soybean plant,” says Krishna Jagadish, a professor and crop physiologist at Texas Tech University. “Developing male and female reproductive organs and other floral structures uses a lot of energy and resources.”
Soybean plants attempt to balance the number of flowers that set pods with the resources and energy available to fill those pods. However, many soybean varieties are indeterminate, meaning they continue to add leaves and other vegetative growth during reproductive growth stages. That means their capacity to create energy to store in pods can increase during the season.
Jagadish calculates that reducing flower drop just 20 to 30%, or keeping roughly one-fourth of the flowers usually released, has the potential to enhance soybean yields by 10 to 15%. He believes this increase in pod set would remain within genetic yield potential.
To explore ways to improve soybean flower and pod retention, Jagadish is leading collaborative research funded by Soy Checkoff investments from the Atlantic Soybean Council, Mid-South Soybean Board, North Central Soybean Research Program, Southern Soybean Research Program and United Soybean Board. Key collaborators include William Schapaugh, professor and soybean breeder at Kansas State University, Henry Nguyen, professor of plant genetics at the University of Missouri, and Avat Shekoofa, associate professor and crop physiologist at the University of Tennessee Institute of Agriculture.
“Soybean flowers are complex, and they haven’t been studied in-depth for decades,” Jagadish explains. “Our team aims to capture the large genetic variation in flower abortion to discover ways to enhance flower and pod retention.”
Counting Soybean Flowers
The team started by studying a range of soybean genotypes to characterize observable differences, or phenotypes, of flowering and pod set. This included counting flowers converted to pods and flowers dropped.
“We discovered that it is very hard to phenotype soybean flowers,” Jagadish says. “They are small and hide under plant leaves.”
The team focused on figuring out how to use images and machine learning to count soybean flowers accurately, without double-counting or missing flowers that bloom continuously for a few weeks. Through a series of trials, they determined how to use GoPro cameras to capture video and train machine learning modules to count accurately. They also developed a process to ground-truth that work.
Then, they figured out how to track flowers to pod set and development, providing the percentage of flowers that drop.
“It is easier to see and count pods after leaf drop,” he notes. “But they grow in bunches and overlap. We had to train our machine algorithms to correctly count pods, as well.”
Thanks to intense teamwork across contributing universities, the team has developed a tool that any soybean breeding team can use to track flowering and pod development. This information supports improvements and breeders select crosses and lines to move forward.
Linking to Genetics
The flower and pod tracking tool allowed the team to identify soybean genotypes with above- and below-average flower retention rates under a variety of environmental conditions. They started crossing these varieties to begin the process of mapping flower retention to specific soybean genes.
“Through this work, we have an initial list of candidate genes that we plan to study further to improve flower retention,” Jagadish says. “Both gene editing and traditional breeding will allow us to identify genes that we believe can improve soybean yield by keeping more flowers and pods on each plant.”
The genetics, candidate gene validation and developing resilient germplasm will be the focus of the second phase of this multi-regional, multi-discipline research.
He believes that collaboration between universities and funding organizations made it possible to tackle this complex issue and make real progress. While all universities work closely together, Texas Tech led flower tracking efforts, while Kansas State led pod tracking work. The University of Missouri managed seed multiplication for field trials, and the University of Tennessee gathered contrasting flowering germplasm from the field into their greenhouse for in-depth study of the physiology of lines.
“Our deep collaboration included regular communication, sharing ideas for each challenge, finding solutions that worked for our team, and reporting the results to the full group,” Jagadish reports. “Although the initial work was much more complex than we anticipated, we are now confident that we can make a significant contribution to advancing soybean yields, providing a strong return on farmers’ investment in this research.”
Additional Resources
Multi-Regional Research Collaboration Focuses on Soybean Flower and Pod Retention for Improved Yield – SRIN article
Meet the Lead Researchers:
• Krishna Jagadish SRIN profile | University profile
• William Schapaugh SRIN profile | University profile
• Henry Nguyen University profile
• Avat Shekoofa SRIN profile | University profile
Published: Oct 27, 2025
The materials on SRIN were funded with checkoff dollars from United Soybean Board and the North Central Soybean Research Program. To find checkoff funded research related to this research highlight or to see other checkoff research projects, please visit the National Soybean Checkoff Research Database.
