Project : USDA ARS

ARS Home » Midwest Area » Columbia, Missouri » Cropping Systems and Water Quality Research » Research » Research Project #435625

Research Project: Sustainable Intensification of Cropping Systems on Spatially Variable Landscapes and Soils

Location: Cropping Systems and Water Quality Research

2023 Annual Report

Objectives
Objective 1: Use a GxExM research approach to develop decision support tools for on-farm implementation of sustainable and resilient cropping systems. 1a: Develop knowledge to aid planting-time decision support for optimizing corn emergence on variable soils and landscapes. 1b: Improve decision support for variable-rate grain crop nitrogen management. 1c: Develop and evaluate new and improved soil health assessments. 1d: Develop and evaluate proximal sensing approaches to provide spatially-dense information important in soil management and soil health applications. Objective 2: Develop and evaluate sustainable and resilient cropping systems using a site-specific GxExM framework. 2a: Evaluate production and soil health of grain and perennial grass cropping systems on degraded claypan soil landscapes. 2b: Evaluate effects of cover crops and reduced tillage on soil health and crop productivity. 2c: Evaluate spatial aspects of sustainability in site-specific management systems.

Approach
In this project, our interdisciplinary team will address key knowledge and technology gaps limiting the development of site-specific management systems using a genetics by environment by management (GxExM) research approach. In the first objective we focus on developing new decision support tools and the underlying knowledge needed to facilitate improved, targeted crop management systems. Here we will conduct field studies to understand how to vary planting depth to optimize corn emergence and yield and investigate the effect of emergence date on crop modeling (1a). We will conduct multiple analyses of a previously collected dataset to develop decision support guidelines for in-season variable-rate nitrogen management in corn (1b) We will collaborate with ARS colleagues in Oregon in developing decision support technology for variable-rate nitrogen management in wheat (1b). We will develop new laboratory-based soil health assessments and evaluate them in field experiments (1c). We will develop and evaluate in the field new proximal soil sensing techniques to support soil health and other management decisions (1d). In the second objective we develop, apply, and evaluate innovative management systems that incorporate information about spatially variable soil resources. Many of the studies incorporate application and evaluation of the decision tools described above. In long-term field experiments, we will investigate the effect of cropping systems and landscape variability on soil health and crop production and profitability (2a). We will quantify differences in energy yield of bioenergy crops grown across variable landscapes (2a). Also in field experiments, we will investigate the effects of cover crops and reduced tillage on soil health and crop productivity (2b). We will use a model-based approach to spatially compare production between site-specific and whole field management and validate model results with measured field data (1c). We will conduct field research that uses crop sensor technology to evaluate soybean drought and flood tolerance (1c). Much of the research in the second objective supports, and is coordinated with the Central Mississippi River Basin Long-Term Agroecosystem Research (CMRB LTAR) project, which is part of another research project within this ARS unit. Specifically, decision tools and knowledge from this project will inform possible future changes to the aspirational cropping system design for the CMRB LTAR common experiment.

Progress Report
This is the final report for the project which terminated in October 2023. The Progress Report summarizes research over the life of the project. In support of Objective 1, “Use a genetics by environment by management (GxExM) research approach to develop decision support tools for on-farm implementation of sustainable and resilient cropping systems”, 1) three years of field studies examining the effects of corn seeding depth and within-field soil variability of claypan and alluvial soil fields on emergence and yield showed that variable seeding depth resulted in more uniform stands and yields across fields. Related studies developed unmanned aerial vehicle approaches to early corn stand evaluation, developing methods applicable across a range of soil and crop conditions. Overall, this research resulted in five journal manuscripts focused on corn emergence and two related manuscripts on cotton emergence; 2) a sensitivity analysis found that the Agricultural Policy/Environmental eXtender (APEX) model was not sensitive to planting depth. These results will be combined with related results on parameterizing the APEX model for buffer versus no buffer conditions and published in a future manuscript; 3) a large, multi-state project focused on variable rate grain crop nitrogen management was completed and the associated dataset representing 49 site-years across eight states from 2014–2016 was published. Analysis showed that no single corn nitrogen fertilizer rate recommendation tool was helpful across all growing environments in the U.S. Midwest. However, tools were improved when site-specific soil and weather information were used as part of the fertilizer rate determination process. Overall, this project and related work on tools to optimize nitrogen for corn grain production resulted in 16 publications; 4) the successful “Yield Editor” software (over 12,000 downloads to date) was updated and enhanced to accommodate wheat protein mapping and development of variable-rate nitrogen fertilizer prescription maps. The software was demonstrated to producers and other stakeholders to get their suggestions for improvements. Although improvements were implemented, the software was not released due to the resignation of the developer prior to final software packaging. Functional components of the software are substantially complete, and plans are to contract with a software developer to finish the steps needed before product release; 5) across multiple related studies, a suite of popular soil health measurements were compared and evaluated for cost-effectiveness and sensitivity to management practices at the field, regional, and national scale. A set of five soil health indicators: soil organic carbon, active carbon, aggregate stability, total protein, and soil respiration were selected. A continental scale Soil Health Assessment Protocol and Evaluation (SHAPE) interpretation tool was developed that provides a score for each of the five indicators based on climate and edaphic conditions. The SHAPE tool is publicly available online as a Shiny application and as a GitHub accession. A total of 18 publications resulted from this and related work on soil health indicator methods and protocols; 6) an inversion analysis of soil electrical conductivity sensor data from diverse Missouri field sites showed promise for generation of maps of depth-specific estimates of soil texture, as might be useful for model inputs, in many soil types. Coupling these depth-specific data with in-situ soil moisture sensor data allowed mapping within-season soil water content dynamics using repeated soil electrical conductivity sensing campaigns, for both alluvial and claypan soils. Data collection and soil sample analysis for a three-dimensional soil carbon mapping project has been completed, along with a preliminary analysis. Next, more complex machine learning modeling techniques will be applied and evaluated. The first phase of a project evaluating commercial sensors for estimating soil organic matter and other soil properties has been completed. Estimates of soil organic matter by these sensors are considerably less accurate than those from more complex and expensive full-spectrum instruments but may provide useful qualitative estimates. Publications include nine journal articles and two book chapters. Progress under Objective 2, “Develop and evaluate sustainable and resilient cropping systems using a site-specific GxExM framework” included: 1) the field work, data collection, and validation were completed annually for a 10-year analysis comparing the productivity of grain and perennial grass cropping systems on claypan soils. The scientist leading the project retired before data analysis and publication were completed; however, a replacement scientist has been hired and a publication is expected within the next 12 months; 2) a study of microbial community biomass and diversity in corn production systems in three Missouri soils showed a significant but minor influence of nitrogen management on microbial abundance and biomass, while soil order and corn growth stage exerted a stronger influence on the bacterial and fungal communities. However, the primary factor influencing corn growth was nitrogen management, suggesting more research is necessary to incorporate soil microbial information into corn management decisions. The project was delayed due to resignation of the cooperator’s bioinformatician, but data analysis was concluded this year and a journal manuscript has been submitted. Related studies resulted in five publications demonstrating the soil health and microbial community benefits of perennial systems including switchgrass, agroforestry, and reconstructed prairie, over annual grain crop production systems; 3) an initial analysis comparing energy yield between switchgrass and corn was conducted using previous yield components (2009 to 2013) but additional years (2009 to 2022) of data were deemed necessary to include a wider range of weather scenarios. The expanded research plan includes analysis of water use efficiency, nitrogen use efficiency, and associated energy yield calculations. The scientist leading the project retired before data analysis and publication were completed; however, a replacement scientist has been hired and a publication is in development and expected within the next 12 months; 4) a study of soil health in irrigated cotton production was initiated in 2019 to assess the soil health and production benefits of cover crops and reduced tillage. No difference in yield was observed between the treatments in 2020, but in 2021, a yield drag was observed in the cover crop treatment, similar to what many producers report when first adopting cover crops. Yield data from 2022 has been collected and will be analyzed soon for comparison with 2020 and 2021 results. Soil samples were collected in spring 2023, laboratory analysis is underway, and a publication evaluating the soil health benefits of cover crops and reduced tillage in cotton systems is expected in 2024; 5) a study examining the productivity and soil health benefits of reduced tillage and extended rotations by comparing a continuum from traditional 2-year corn-soybean rotations up to 5-year extended rotations was initiated in 2015. Yield data has been collected annually. Soil sampling was delayed by two years until 2020 to allow for better discrimination of treatments. Laboratory analysis was completed in 2023 and a manuscript is expected in fiscal year 2024. Multiple related studies demonstrated the effects of cover crops and reduced tillage on soil health and productivity in grain cropping systems at the field, regional, and national scale, resulting in 16 publications; 6) the study evaluating the spatial aspects of sustainability in site-specific management systems with MAIZSIM and GLYCIM crop models was not completed due to a critical vacancy. The cooperator has hired a scientist for the modeling position and the project is moving forward as part of the Long-Term Agroecosystem Research (LTAR) network regionalization project under the new 211 project (5070-12000-001-000D). Work is proceeding with installation of 48 soil moisture sensors covering four depths at 12 locations in the LTAR aspirational management field; 7) to demonstrate that crop canopy sensors can distinguish between tolerant and sensitive soybean genotypes for both drought and flooding stress, drought tolerant soybean genotypes and high-yielding commercial varieties were planted under a variable-rate irrigation center pivot system. Crop growth and yield were evaluated under varying levels of irrigation. A high-clearance sprayer tractor was obtained to serve as a sensor platform, allowing mobile sensor-based monitoring of crop development later in the growing season with minimal impact on the crop. Unfortunately, the genotypes used were extremely sensitive to dicamba herbicide. Even after waiting until late June to plant the study in hopes of encountering less herbicide injury, the damage to the plants made the planned measurements meaningless. Given the high levels of dicamba use in the region, the planned approach was deemed unfeasible, and the primary experiment was abandoned. Related research evaluating the spatial aspects of site-specific management systems resulted in nine publications.

Accomplishments
1. Conservation practices improve soil health and provide economic benefits to farmers. Adoption of conservation practices has been hampered by the lack of regionally relevant soil health information and lack of a demonstrated relationship between soil health measurements and crop yield. ARS scientists at Columbia, Missouri, in cooperation with the University of Missouri, collected data from over 5,300 production fields in Missouri, and showed that biological and physical soil health status was greater in systems with increased rotational diversity (three or more crops) and reduced tillage. In related work with public and private partners, they linked soil health scores to corn and soybean yield using data from 96 farms across nine Midwestern states over five years, demonstrating an economic benefit from improved soil health. Furthermore, specifically in Missouri, subfield variability in soil health was linked to corn yield and a threshold level of active carbon, a popular soil health indicator, was identified for optimal corn grain productivity. These results benefit farmers by providing them with science-based information on the economic and soil health benefits of conservation management practices.

2. Accuracy of implement-mounted soil sensing systems improves with local calibration. Collecting information about soil properties during seeding or tillage operations has potential to improve machinery management and crop performance. Implement-mounted sensing systems are commercially available but independent evaluations of their performance are lacking. ARS researchers at Columbia, Missouri, evaluated two commercial systems for their ability to estimate soil organic matter (SOM). Both systems were able to detect relative differences in SOM, but more accurate SOM estimates required local calibration using laboratory SOM analysis of four to ten soil samples. A follow-up study compared SOM estimation by commercial sensing and advanced analytical techniques using multiple combinations of soil reflectance bands in the visible and near-infrared spectrum across selected soils and soil water contents. The commercial system detected relative differences between low and high SOM, but estimates were impacted by soil water content. The more complex and costly system reduced SOM estimation errors by 85% across a range of soil water contents. This information will help producers determine if measurements from commercial sensors are useful for cropping system management and will contribute to the development of improved sensor-based control systems for precision agriculture applications.

3. Optimized corn nitrogen deficiency detection to maximize farmer profits. Knowing when and where corn is nitrogen deficient is essential to optimize productivity, but current methods to assess nitrogen stress, such as collection and analysis of plant tissue samples, are expensive and time consuming. ARS scientists in Columbia, Missouri, in collaboration with researchers from industry and eight universities, combined data from active canopy reflectance sensors with soil, weather, plant genetics, and management information from 49 sites across the Midwest. A variety of advanced data analysis techniques were evaluated and compared to measured plant tissue nitrogen content and plant biomass. Machine learning algorithms successfully integrated the multi-source data with an 84-93% accuracy in predicting corn nitrogen status and outperformed any of the individual measurements alone. Using machine learning algorithms to integrate weather (precipitation and corn heat units), soil (texture and nitrate concentration), and on-the-go canopy reflectance measurements (normalized difference red edge) could help farmers assess corn nitrogen status on a plant-by-plant basis and determine if additional fertilizer is needed for improved yield and profitability.

4. Soil health information did not improve potassium or phosphorus fertilizer recommendations for corn in Missouri. Incorporating soil health tests into fertility recommendations may reduce farmer inputs, optimize crop yield, and mitigate negative environmental outcomes. However, studies linking soil health tests with fertility recommendations are lacking. ARS researchers at Columbia, Missouri, in cooperation with the University of Missouri, collected soil health, fertility, and yield data from monitoring sites (with and without added potassium and phosphorus based on traditional fertility recommendations) on 101 producer corn fields across northern Missouri. In this study, traditional soil fertility measurements alone provided the most effective fertilizer recommendations for optimal corn yield. These results help farmers understand how to use soil tests to make the best management decisions and optimize crop productivity and profitability.

U.S. DEPARTMENT OF AGRICULTURE

Cropping Systems and Water Quality Research: Columbia, MO