During the past decade, the agricultural sector generated massive amounts of data, which included every aspect of crop production ranging from multiple management practices and biotic yield limiting factors, to final crop yield. When these data sources are coupled with environmental data (e.g., field soil properties, weather conditions), large, robust, and valuable datasets can be created. Recently, there have been several attempts by universities (e.g., Rattalino Edreira et al., 2017; Mourtzinis et al., 2018; Andrade et al., 2019) and private sector (e.g., The Climate Corporation) to gather, aggregate, and analyze data from diverse sources to identify practices and factors that can be manipulated by farmers to increase soybean yield. Most of the approaches successfully generalized the results of the analysis across large regions; however, they were not able to identify and advise optimum cropping systems at the field level. Being able to down-scale such information to each individual farmer would increase the usefulness as applied to specific farms and cropping systems.
An optimal pre-plant pest management strategy is difficult to define with confidence. This is due to the lack of knowledge of the anticipated pest-mitigated yield pressure that may arise during the growing season, primarily because of annually variable weather conditions. As a result, many farmers apply inputs prophylactically in the absence of pest pressure, which often leads to a negative return of investment (ROI) and increases risk of pest resistance and other negative environmental impacts. Currently, there are no attempts to develop systems that will evaluate the pre- and post-plant risk of pest pressure at the field level and provide advance notification to farmers for potential pest management action. A tool capable of better predicting when and where sustainable pest management options can reliably increase ROI and thus would be useful to soybean farmers.
Another important issue with current research approaches is that the highly variable and farm-specific management costs (e.g., different farmers often pay variable prices for similar seed) is not considered. Therefore, the effect of a management practice (e.g., seeding rate) on yield is disconnected from its associated cost (e.g., $/seed bag). Although this element does not negatively affect the recommended rate for optimum yield, it makes fine-tuning of profit optimization at the farm level impossible. Consequently, input over-application, or simply input application when there is no need, reduces net returns in the short-term and can have negative economic and environmental consequences in the long-term.
This proposal will address these issues by developing large databases that incorporate farmer management decisions (rates and costs) with the presence of yield-limiting biotic factors, yield data, soil properties, in-season weather, and remote sensing data. By amassing three years of data (2022-2024) on important variables of a field’s productivity, variable data can be related to farm profitability outcomes. Specifically, the objective of this proposal is to develop a tool to optimize soybean management practices that will allow for profitability optimization at the field level. Additionally, within field variability will be quantified to further increase farmer profit. The proposed approach is an opportunity for U.S. soybean farmers to increase profitability because it will allow for efficient input/management application at the field level. Currently, a similar approach using farmer data does not exist.

We propose a 3-year interdisciplinary and collaborative regional project, that will be Co-directed by PIs at the University of Wisconsin-Madison and The Pennsylvania State University, with collaborators in OH, MI, IA, MN, NE, and ND. The primary goal of the project is to analyze farmer data (yield, management practices and associated costs) in fields located across the NCSRP region. Our primary focus will be on increased profitability through improved knowledge of yield-limiting factors and best management practices. We believe that the proposed project fits well within the following three NCSRP targeted research areas: (i) “Basic and applied research that addresses soybean response to water, nutrients, climate, soil and environmental conditions specific to the North Central Region”, (ii) “Management of weeds and weed resistance to herbicides for species of common occurrence and threat across the North Central Region” and (iii) “Soybean production practices & crop management for increased yields and profitability in an environmentally sustainable manner”.

The proposed project has five major components:
(1) Data collection via surveys. Data collection is a key component of the proposed project. Based on our previous experience working with farmer self-reported data, a large number of reporting fields, which are well distributed across soybean production regions, are needed to account for large management/soil/weather variability. Collaborators in each state working with graduate and undergraduate students and technicians will be responsible for collecting the field level data. Requested information will include yield (captured by yield monitor), field location, and detailed information on crop/field/input management, such as planting date, soybean variety, tillage method. A primary focus, which is novel from previous studies, will be on pest management decisions and costs of major inputs (seed, fertilizer, foliar product + application cost etc.).
(2) Data collection via in-season scouting. Boots on the ground field scouting in a subset of soybean fields (minimum of 8-10 different farmers; can have multiple fields per farm if they are willing) selected from diverse environmental regions (TEDs) will take place on a weekly basis. Selected fields will be chosen specifically to allow for large environmental variability and expected pest pressure. Efforts will occur on an annual basis and continue for all three years of the project in order to examine year-to-year weather and pest presence variation. Individual field data and farmer contact information will be kept strictly confidential.
(3) Data assimilation. Collected data will be standardized into a single, consistent format, error-checked, and then inputted into a digital database. We will also retrieve soil data for each individual field (using its GPS coordinates) from readily available websites (e.g., USDA-NRCS SSURGO database). Daily weather data and satellite imagery will be collected. Ultimately, for each field-year data point supplied by an individual farmer, we will have a detailed description of weather, soil, management, and health progress (via the use of in-season reflectance data and scouting), which, when taken together, will help us to identify management factors that affect yield. Ultimately, by accounting for input costs, analyzing the data will allow us to identify opportunities for improving profitability regionally and more importantly, at and across the field level.
(4) Data analysis. We will use well-known methods to analyze the data and new techniques for information extraction. We will base our approach on the use of already developed robust protocols that have resulted in a new online decision tool (see screenshot of the user-friendly interface below) which was developed by Agstat (https://www.agstat.com/cropping-system-optimizer/), a private company that we will collaborate with in the proposed work. However, according to the developer, it uses algorithms trained on “synthetic and simulated datasets” which may not align with what farmer’s experience in their fields. At the moment, no in-season crop health information is considered nor does above model provide pre-or post-plant in-season pest pressure risk alerts. Therefore, a similar tool will be developed using farmer’s data and will be providing in-season recommendations, which we believe can result in increased accuracy and profit optimization compared to tools and products that currently exist. Given that the collected data will not have the attributes of traditional replicated field trials, where we can control potential confounding factors by replicated blocking, a combination of novel statistical approaches (e.g., multilevel models, repeated measures for spatial and temporal correlation) and machine learning algorithms (e.g., convolutional and recursive neural networks) will be used to analyze the data. We have worked with these methods to explore complex data sources, in particular to examine interactions among different variables which cannot be easily explored in traditional statistical methods (Mourtzinis et al., 2021-in review; Shah et al., 2021-in preparation). The methods we propose to use are also being applied by the industry as part of their digital agriculture programs, but our approach and analysis will differ because it will be based on the aggregated farmer database. The product of this work will be a tool (combination of sequential models and algorithms) that will have the potential to generate field-specific recommendations and early-season pest risk alerts using farmer’s data. The outcomes will provide information for optimal cropping systems and management practices that can increase farm profitability at the field level.
(5) Communication and dissemination of results. Due to the large variability in management practices and costs among different farms, results will be communicated with farmers following a 2-stage approach. In the first stage (at the end of the second year), we will perform a large-scale profit optimization across all fields and years in the database. The objective will be to identify the optimum management practices for increased profitability and report the estimated profit difference across the NC US. In the second stage (during the last growing season), a subset of farms will be selected to demonstrate the potential of the developed tool to increase farm profitability compared to what farmers typically use. We will work with our collaborators in every state to identify the demonstration farms. All results from this project will be disseminated to producers and public via peer-reviewed scientific and extension publications, presentations at scientific conferences and extension events sponsored by universities, natural resources districts, grower associations, and proprietary organizations that market their products to soybean producers.

Updated April 6, 2022:
We held a project kick-off meeting with collaborators on November 7, 2021. We discussed project objectives, outputs, and logistics. Additionally, we discussed the structure of the between-season survey that will be disseminated to farmers across the NC US region and the protocol to follow for data collection for the in-season field scouting part of the project. During the subsequent three months, weekly virtual meetings were held to discuss and finalize the between-season survey published online on March 1, 2022 (can be accessed here: https://arcg.is/1OD5an2). All state collaborators are contacting farmers to contribute to our project with their local field data. Concurrently with the between-season survey data collection efforts, the protocol for in-season data collection were developed, and the final field scouting protocol will be ready on April 15th. The goal is to scout a minimum of 10 fields per state every two weeks starting on May 15th throughout the growing season for various yield stressors (e.g., weeds, diseases, etc.). These data will serve as ground truth for subsequent analysis based on remote sensing data.

Table 1. Anticipated number of fields to scout in each state.
State Number of fields to scout in 2022 season
Iowa 10-15
Michigan 10-15
Minnesota 10-15
Nebraska 10-15
North Dakota 10-15
Ohio 10-15
Wisconsin 18
Pennsylvania 10-15
Total: Minimum of 80 fields that will be scouted

Project PIs Dr. Shawn Conley (WI) and Dr. Paul Esker (PA), along with Dr. David Kramar (ND), Dr. Cucak Mladen (PA), John Gaska (WI), and Dr. Spyridon Mourtzinis (WI), will supervise data collection and will be responsible for quality control of the data. The WI-PA-ND core team will retrieve soil, satellite, and weather information for each field each year from existing databases. The WI-PA-ND core team holds bi-weekly virtual meetings to discuss and monitor project progress.
At the end of this 3-year project, we will have validated a novel tool (combination of models and algorithms) that utilizes self-reported on-farm production practices with associated costs to identify management practices (single and cropping systems) that can result in increased profit. A “beta version” of the online tool will be ready for in-depth, extensive validation in the growing season following project completion. We will also strengthen state-to-state research collaboration through the managed coordination of the on-farm research. The potential impact of the outcomes derived from this study is significant, including helping farmers adopt new approaches and understand the effect of management decisions on soybean yield and profitability.

The development of the tool proposed in this project should prove of significant value to producers as they will better understand the impact of decisions made throughout the production cycle in terms of outcomes at the end of the year. Additionally, as a bonus, we presume that they will recognize the value of participating in such efforts and further promote the importance of this research by showcasing its benefits.

Updated January 31, 2023:

Updated January 31, 2023:
We held a project meeting with many collaborators in Baltimore, MD, on November 6, 2022. We discussed project objectives, outputs, and logistics. Additionally, we discussed and refined the approach for the between-season data collection and the protocol to follow for data collection for the in-season field scouting part of the project. The between-season survey can be accessed here: https://survey123.arcgis.com/share/7a5e999e1832473a8b5fd2a2d21d234f. During the subsequent three months, all state collaborators contacted farmers to contribute to our project with their local field data. We aim to collect data from at least 10 scouted fields per state. In 2022 we scouted the following number of fields per state: IA (25), MI (13), MN (3), NE (10), ND (12), OH (11), WI (25), PA (36) for a total of 135 fields. These data will serve as ground truth for subsequent analysis based on remote sensing data. Additionally, we aim to collect data from at least 100 fields with between-season data. This is ongoing.

Project PIs Dr. Shawn Conley (WI) and Dr. Paul Esker (PA), along with Dr. David Kramar (ND), Dr. Santosh Sanjel (PA), John Gaska (WI), and Dr. Spyridon Mourtzinis (WI), will supervise data collection and will be responsible for quality control of the data. The WI-PA-ND core team will retrieve soil, satellite, and weather information for each field each year from existing databases. The WI-PA-ND core team holds bi-monthly virtual meetings to discuss and monitor project progress.

At the end of this 3-year project, we will have validated a novel tool (combination of models and algorithms) that utilizes self-reported on-farm production practices with associated costs to identify management practices (single and cropping systems) that can result in increased profit. A “beta version” of the online tool will be ready for in-depth, extensive validation in the growing season following project completion. We will also strengthen state-to-state research collaboration through the managed coordination of the on-farm network. The potential impact of the outcomes derived from this study is significant, including helping farmers adopt new approaches and understand the effect of management decisions on soybean yield and profitability.

The development of the tool proposed in this project should prove of significant value to producers as they will better understand the impact of decisions made throughout the production cycle in terms of outcomes at the end of the year. Additionally, as a bonus, they will recognize the value of participating in such efforts and further promote the value of this research by showcasing its benefits.

In 2022, the data driven team built and populated a scouting app, actively scouted 135 field across the North Central region and uploaded the scouting results to our cloud based platform. The team is actively uploading yield monitor data from these fields and will layer, weather, soils and satellite imagery to the cloud. The team is also actively collecting farmer survey data using an on-line system that will also be layered into the system.

By the end of this 3-year project we will have validated a novel tool (combination of models and algorithms) that utilizes self-reported on-farm production practices with associated costs, to identify management practices (single and cropping systems) that can result in increased profit. At the end of the project, a “beta version” of the online tool will be ready for in-depth extensive validation in following growing seasons. We will also strengthen state-to-state research collaboration through the managed coordination of the on-farm network. The potential impact of the outcomes derived from this study are significant.