Update:
In 2021, ISU experimental lines were entered in the Uniform Soybean tests.
In UTIIA (Uniform test, maturity group 2 early): A14011-77 (i.e., IAS19C3) yielded 1% less than the mean of the checks with data coming from 13 locations across the mid-west US. IAS19C3 compared to the only yellow hilum seed color, IA2102, yielded about 2% more than this check with similar days to maturity. Seed protein of IAS19C3 was 41.2% (on dry weight basis) while mean of checks was 38.6%.
In UTIIA (Uniform test, maturity group 2 early): A14004-126 (i.e., IAS25C1) yielded 1% more than the mean of the checks with data coming from 13 locations across the mid-west US. IAS25C1 compared to the only yellow hilum seed color, IA2102, yielded about 3% more than this check and two more days to maturity. Seed protein of IAS25C1 was 38.7% (on dry weight basis) while mean of checks was 38.6%. IAS25C1 has intermediate resistance against soybean aphid.
In UTIIB (Uniform test, maturity group 2 late): A14004-58 (i.e., IAS31C1) yielded 1.5% more than the mean of the checks with data coming from 13 locations across the mid-west US. IAS31C1 compared to the only yellow hilum seed color, IA2102, yielded about 4% more than this check and three more days to maturity. Seed protein of IAS31C1was 38.7% (on dry weight basis) while mean of checks was also 38.7%. IAS31C1has intermediate resistance against soybean aphid.
In 2021 regional testing, several lines with high protein and good yield were identified in clear hilum seed. In addition to IAS19C3, two other lines in MGII had seed protein higher than 40% (dry weight basis) and seed yield higher than comparable check (yellow seeded) IA2102. One line was large seeded, with seed protein ~42% and meal protein >48%; however, seed yield was 11% less than the mean of all check and 9% less than comparable check IA2102, although this line is also 5 days later than IA2102. In MGIII, one line had seed protein ~42% (dry weight) and meal protein ~42%, however, seed yield was 15 lower than mean of the check. Therefore, this line is only suitable for food grade market that require yellow hilum and high protein. For high yielding conventional lines several experimental lines were higher than the mean of the checks. For example, one line was ~9% higher yield than the mean of the checks and ~14% higher yield than IA2102, and seed protein was 39.3% (compared to 38.5% for the mean of the checks). Thirteen experimental lines were higher yielding than the mean of the check (10 locations of data across multiple states) in this test.
We identified one SCN resistant experimental line in MGII and three SCN resistant experimental lines in MGIII tests with Rhg1-Peking + Rhg4 genetic resistance against SCN.
Several accessions with exotic genetic background were tested in multi-state testing, and several high yielding and good protein lines were identified. This includes an experimental line (with 50% PI background), which was the highest yielding lines in the test and with >48% meal protein.
In addition to variety development efforts, several research experiments were performed. Examples of ISA supported graduate student projects are provided below. These projects are supported by multiple agencies, but ISA funding provided a critical infrastructure that allows us to complete the project.
Drone based phenotyping, and statistical analysis of field trials: Matt Carroll published two papers. The first paper was a review article for the use of aerial drones in small plot research and breeding trials. This paper was published in collaboration with multiple other research groups, from several institutions. This review provides a beginner in the field of drone-based phenotyping to have a good starting point to begin their research. The paper covers multiple important areas such as sensors, trait acquisition, best practices, and an overview of the previous work that has been done. Interestingly, this paper was motivated from discussion with Iowa farmers who had indicated a desire to see drone review that can help farmers to understand and utilize the drone based phenotyping on their farms. The second paper that was published on yield prediction in soybean by utilizing videos that were captured using a ground robotics system. In this paper we showed the viability of utilizing this type of system in a breeding program to estimate end season yield by identifying the pods in a plot and estimating the final yield utilizing deep learning techniques for both pod identification, and estimation of yield. This paper has led to several invited presentations, as it a very promising approach for automating breeding programs saving time cost and increasing efficiency in multiple steps.
This past year Sam Blair continued our project on the use of drones to estimate the relative maturity of soybean genotypes within a testing field. Selecting for maturity among genotypes is critical for a breeding program. The current process of rating for maturity involves many hours of human-labor. Performing this same rating process with a UAS and associated data analytics, significant saving of time and labor is observed. In his second project, we are deploying ground robot-based yield estimation and use in the breeding program.
Root related research: Throughout the last year, Clayton Carley continued his work on enhancing the learning about the biology, growth, and development of nodulation in early soybean growth stages. Him and other colleagues published a paper “Using Machine Learning to Develop a Fully Automated Soybean Nodule Acquisition Pipeline (SNAP)” and have subsequently used the SNAP method to evaluate the spatial and temporal placement of nodules in diverse soybean genotypes. We have also been exploring the relationships between nodules and final seed nitrogen and have found that the most significant correlations come from the total volume of nodules on the plant taproot as opposed to the secondary roots. We have also observed that genotypes have different modes of nodule control, including adjusting the quantity of nodules present and the size of nodules at different rates leading to changes in total nodule volume overall. He is continuing the work on genomics studies to identify candidate genes for nodule development and placement. This, in conjunction with a fine mapping project for root system architecture traits, will enable a deeper understanding of root growth and traits which breeders can utilize and maximize in their programs. In the past year, Clayton was intricately involved in mentoring one of our lab’s undergraduate students, Melinda Zubrod. Clayton guided her through protocol development, writing her own research questions, executing the seed nitrogen study, and helped her through the data analysis and presentation. Melinda successfully completed the project, and presented her work at the annual SASES meeting in Salt Lake City where she won the Darrel Metcalfe Student Journalism Contest for this work.
Heat stress research: The focus of this work is to screen for soybean heat tolerance, as well as understand the genetic mechanisms controlling the trait. She worked on developing heat stress screening protocols in indoor and field conditions. The initial genetic screening for soybean heat tolerance (preliminary results) was completed and analysis is on-going. These experiments and other planned experiments will gave a greater insight to the available genetic diversity of these traits for soybeans with maturity suitable for production in the US Midwest region.
Drought stress research: With the support of ISA funds breeding efforts on drought stress were made, and experimental lines were developed which will be evaluated for variety release.
Our students won several awards including: Clayton won first runner up in the graduate student oral & poster competition with his talk, “Breeding Below Ground, Which Nodules Count?” Clayton also received the university-level Research Excellence Award for his research and academic merit. He also won the Agronomy Department’s Outstanding Graduate Student Award. Matt received the C.R. Weber award for excellence in Plant Breeding.
Papers published:
1. Singh AK, A Singh, S Sarkar, B Ganapathysubramanian, W Schapaugh, FE Miguez, CN Carley, ME Carroll, MV Chiozza, KO Chiteri, KG Falk, SE Jones, TZ Jubery, SV Mirnezami, K Nagasubramanian, KA Parmley, AM Rairdin, JM Shook, L Van der Laan, TJ Young, J Zhang (2021). High-throughput phenotyping in soybean. In Concepts and Strategies in Plant Sciences High-Throughput Crop Phenotyping.
2. Shook JM, D Lourenco, AK Singh (2021). PATRIOT: a pipeline for tracing identity-by-descent for chromosome segments to improve genomic prediction in self-pollinating crop species. Front Plant Sci, 12: 676269.
3. Jubery TZ, CN Carley, A Singh, S Sarkar, B Ganapathysubramanian, AK Singh (2021). Using machine learning to develop a fully automated soybean nodule acquisition pipeline (SNAP). Plant Phenomics, 2021: 9834746.
4. Riera-Garcia L, ME Carroll, Z Zhang, JM Shook, S Ghosal, T Gao, A Singh, S Bhattacharya, B Ganapathysubramanian, AK Singh, S Sarkar (2021). Deep multiview image fusion for soybean yield estimation in breeding applications. Plant Phenomics, 2021: 9846470.
5. Shook JM, T Gangopadhyay, L Wu, B Ganapathysubramanian, S Sarkar, AK Singh (2021). Crop yield prediction integrating genotype and weather variables using deep learning. PLoS One, 16(6).
6. Guo W, ME Carroll, A Singh, TL Swetnam, N Merchant, S Sarkar, AK Singh, B Ganapathysubramanian (2021). UAS-based plant phenotyping for research and breeding applications. Plant Phenomics, 2021: 9840192
7. Nagasubramanian K, TZ Jubery, F Fotouhi Ardakani, SV Mirnezami, AK Singh, A Singh, S Sarkar, B Ganapathysubramanian (2021). How useful is active learning for image-based plant phenotyping?. Plant Phenome J, 4(1): e20020
8. Shook JM, J Zhang, SE Jones, A Singh, BW Diers, AK Singh (2021). Meta-GWAS for quantitative trait loci identification in soybean. G3, 11(7).
9. Chiozza MV, KA Parmley, RH Higgins, AK Singh, FE Miguez (2021). Comparative prediction accuracy of hyperspectral bands for different soybean crop variables: from leaf area to seed composition. Field Crops Res, 271: 108260.
10. Natukunda MI, JD Hohenstein, CE McCabe, MA Graham, Y Qi, AK Singh, GC MacIntosh (2021). Interaction between Rag genes results in a unique synergistic transcriptional response that enhances soybean resistance to soybean aphids. BMC Genomics, 22: 887.
11. Kohlhase DR, CE McCabe, AK Singh, JA O'Rourke, MA Graham (2021). Comparing early transcriptomic responses of 18 soybean (Glycine max) genotypes to iron stress. Int J Mol Sci, 22(21): 11643.
Several other research papers were published with support from other projects.
Invited presentations made:
* Singh AK (2021). “Building Better Beans - use of HTP and AI for plant breeding.” Texas A and M Plant Breeding Symposium, (virtual), Feb 12, 2021.
* Singh AK (2021). “Plant Breeding and Computer Science Need to Work Together to Enhance Agriculture Productivity.” Department of Computer Sciences, Missouri Science and Technology University, (virtual), Feb 12, 2021.
* Singh AK (2021). “Expanding the breeding toolbox to develop soybean cultivars.” University of Nebraska. April 30, 2021.
* Singh AK (2021). “Soybean Protein Improvement.” IA Soybean Research Center. October 27, 2021.
View uploaded report 
This project lead to the development of competitive lines for Iowa soybean farmers. Three varieties were commercialized with foundation seed sales made in 2020 and in 2021: IAS19C3, IAS25C1, IAS31C1 were successfully launched with high market demand. These varieties are non-GM, and yellow hilum with an excellent package of traits, and suitability for the state of IA. Foundation seed increase was done in 2020 and 2021, and contracts were established with multiple companies for seed sale of these lines. These lines were tested in Iowa Crop Performance Tests in 2021.
Highlights (from the 2021 ICPT test):
• In northern Iowa testing (early season) based on 2 year mean results, IAS19C3 ranked 6th in the state competing against private company entries. IAS19C3 is the highest yielding conventional line, suitable for food applications. Only one variety significantly outyielded IAS19C3 based on 2020 and 2021 data. This variety also has high seed protein, therefore has an excellent combination of hard to achieve high yield and high protein.
• In Northern Iowa testing (full season) based on 2 year mean results, IAS25C1 was the highest yielding conventional line, suitable for food applications. In Central Iowa districts (2-year means), IAS25C1 was the second highest yielding conventional variety. IAS25C1 has improved aphid tolerance, which has been a high priority for organic food grade soybean growers.
• In Central Iowa testing (full season) based on 2 year mean results, IAS31C1 ranked 4th in the state competing against private company entries. IAS31C1 is the highest yielding conventional line, suitable for food applications. No variety significantly outyielded IAS31C1 based on 2020 and 2021 data from Central full-season districts. In 2021, this variety was the third highest yielding variety in Central Iowa full season district tests competing against private company varieties. IAS31C1 has improved aphid tolerance. This line did not perform as well in southern districts (2-year means).
Numerous other varieties are in the pipeline and have shown merit in IA multi-location tests, as well as in 2021 multi-state tests. These lines bring an attractive package of good yield with high protein, and suitability for food grade markets. In addition, lines with specialty traits are near commercialization.
Breeder seed production of most advanced stage lines was successfully completed, enabling foundation seed increase in future.
In the funding period, several graduate student projects were completed, and peer-reviewed scientific papers were published.