As mentioned above, preliminary results of this project have already been shared at a half dozen events including the international Agronomy meetings, Prairie Grains, Best of the Best, and MN Soybean Expo. This topic is in great demand by producers. An additional year of data with summarization, including an economic analysis, will make this a hot topic for discussion at many meeting within and beyond Minnesota. We anticipate writing a series of blog articles about this work and we will develop final summaries to be published on www.soybeans.edu. In addition, this work will be published in peer- reviewed publications.
Protocol Overview: Overall, we had an extremely successful research season in 2021. We were able to capture yield responses to common IDC management strategies at six unique field sites on 3 farms.
Results were covered in recent progress reports, at the Prairie Grains Conference, at the CPM Short Course, at the Best of the Best, and at the MN Ag Expo. We were able to carry out research as planed in 2021, and we will repeat these same protocols in 2022. Details can be found below.
We plan to evaluate all combinations of two varieties (tolerant and highly tolerant), two seeding rates, three rates of iron chelate (SoyGreen), and two levels of IDC intensity by supplemental N. The study (24 treatments x 4 replicates) will be planted in 10x30’ plots in two areas in each producer field. One area will be placed in the ‘hottest’ IDC part in the field, and the other area will be planted nearby in a representative part of the field. These two paired studies will be planted in three farm fields in western Minnesota.
Bayer Collaboration: The Bayer Crop Science Regional Technology team has an intense interest in this work. They have agreed to continue to cooperate on this project. Although, Bayer has agreed to cooperate with us on this project, funding from the MSR&PC for a portion of this work is a critical foundation for this project to go forward. Any collaboration will allow for two-way data sharing without any intellectual property restrictions.
As in 2021, the Naeve project will conduct small plot research trials in six field locations on the full set of treatments (2 x 2 x 3 x 2). Specific treatments are shown below. The Bayer Technology Development team has a complement of medium-scale research equipment to examine the same treatments outlined here; however, their locations are limited to the Benson area. This research plan leverages the strengths of each organization to test the same questions at additional locations.
Graduate Student: An MS student, Maykon da Silva, has assisted with managing this research trial and coordinating with Bayer. Maykon will continue to collect from each research site high-resolution, multispectral drone imagery on a weekly basis and to examine leaf area and NDVI accumulation over time. This work is critical for developing robust yield estimation models for IDC with remote-sensed data. While he did not have time to do so in 2021, Maykon may also contribute to remote sensing on Bayer research sites and at on-farm research trials in 2022. While this project will be Maykon’s primary responsibility and the focus of his MS degree, this proposed grant only supports 50% of his appointment. Other funds will be sought to supplement his graduate study funding from other sources.

Protocol Detail: This proposal covers the deployment of six small-plot research studies on three cooperating famers’ fields. To vary the intensity of IDC, small-plot studies will be located in a ‘hot-spot’ on each producer field as well as in an area of each field that is representative of the majority of the acres in that field or less chlorotic. We developed an efficient protocol whereby we were able to increase the intensity of IDC symptoms through the application of supplemental N. Therefore, in addition to the primary treatments, each study will be blocked with and without supplemental N application.
In total, we will replicate this study across 12 unique IDC environments (3 fields, 2 areas within each field, and 2 levels of IDC intensity within each study area). With four replications, we are planning for 768 individual plots.
• 2 Variety treatments
o Moderately tolerant
o Highly tolerant
• 2 Populations
o 125,000
o 175,000
• 4 Iron Chelate (SoyGreen) rates
o 0
o 2 lbs
o 4 lbs
• 2 IDC levels
o With N
o Without N
• 2 Locations within fields
• 3 Producers

Three cooperating producers will be selected to represent distinct IDC environments so that unique soil and climate factors are examined annually to allow a broad inference of research findings. The six field studies will be planted in production soybean fields by the Naeve project. A split-block design will be utilized, where iron chelate rate will be the main block and varieties, populations, and N will be randomized within the main block.
After emergence, plots will be evaluated weekly for IDC symptoms using visual scoring, ground-based NDVI as well as though high-resolution drone imagery. These data will be utilized to evaluate each treatment for timing and intensity of IDC symptomology relative to yield.

Upon completion of field activities, a complete economic analysis will be conducted. Yield data will be examined relative to a range of grain prices and appropriate input costs. Economists will be consulted on the most appropriate analyses based on the data.

Update:

View uploaded report

The present study evaluated the effectiveness of variety selection (highly tolerant [MT] vs. moderately tolerant [MT]), iron chelate application (0, 2, and 4 lbs. Fe-EDDHA A-1), and increased seeding rates (125,000 and 175,000 seeds A-1) as management strategies for soybean iron deficiency chlorosis. Further, profitability and risk analysis were conducted to assess the impact of each of the management practices on overall economic returns. Planting a highly tolerant variety, applying iron chelate in furrow at planting, and increasing the seeding density were all effective at mitigating IDC. As IDC severity increased by one-point in a 1 to 5 visual scale to rate symptoms, the HT variety yielded 5 bu A-1 more, on average, than the MT variety. A similar trend was found with the application of Fe-EDDHA in furrow at planting. Soybean yield increments averaging 6.5 and 7.5 bu A-1 were observed with the application of 2 and 4 lbs. Fe-EDDHA A-1 for every one-point increase in IDC intensity, respectively, relative to the untreated plots. When IDC was severe (average visual chlorosis score of 4 at R5.5 growth stage), iron chelate application at rates of 2 and 4 lbs. A-1 increased yield by 21 and 25 bu A-1, respectively, on average. A smaller but consistent effect was found for increased seeding rates. Compared to 125,000 seeds A-1, increasing the seeding density to 175,000 seeds A-1 significantly increased grain yield by 2.2 bu A-1, regardless of IDC severity or treatment combinations. Our risk efficient frontier analysis revealed that the best management alternative in terms of risk to reward for low-moderate IDC conditions was the combination of a HT variety with 2 lbs. Fe-EDDHA A-1 planted at 125,000 seeds A-1. Conversely, the single best option for severe IDC, both in terms of risk and expected returns, was the combination of a HT variety with 4 lbs. Fe-EDDHA A-1 planted at 175,000 seeds A-1. This study also investigated the utility of UAV-based vegetation indices for estimating grain yield of soybean grown under IDC stress conditions as a tool to aid growers, researchers, industry and policy makers with crop management, market planning, market research, and policy writing. In order to develop the yield estimation model, the most relevant VI’s for soybean yield estimation had to be determined. For that, a path analysis was performed, and indicated that NDRE at R5.5, OSAVI at R5.5, and NDVI at R1 had the strongest positive direct effect on soybean yield under IDC stress conditions. The three VI’s indicated by the path analysis as the most important for yield estimation were used as independent explanatory variables in a regression tree algorithm with grain yield as the principal dependent variable. The resulting regression tree algorithm was able to predict soybean yield with a relatively low RMSE (7.88 bu A-1) and MAE (6.7 bu Aa-1), while explaining more than 93% of the yield variability. Results from this study can help soybean growers increase productivity, improve economic returns while controlling economic risk, and provide an advantage when it comes to agricultural decision making.