Location: Genetics and Sustainable Agriculture Research
2025 Annual Report
Objectives
1. Develop dynamic, robust, and resilient cropping systems that integrate conservation and science-based solutions to improve short- and long-term crop production systems, increase input efficiencies, and provide adaptability for changing climate and supply chain shocks.
1.A. Evaluate cover crop management and soil amendment effects on diverse ecosystem services in dryland crop productions.
1.B. Develop new and modern cropping systems that reduce or eliminate production inputs, increase yield and profit, and enhance environmental health.
1.C. Developing plant science-based solutions to improve crop resilience to climate change.
2. Develop, expand, and deploy high throughput data acquisition and analytics systems and platforms for multi-faceted data streams to improve the sustainability and relevancy of agricultural production systems and ecosystem services with an emphasis on soil health, production inputs, water conservation, water quality, and greenhouse gas (GHG) emissions.
2.A. Identify region-specific environmental health indicators for emissions, soil biology, and nutrient uptake by utilizing high throughput sequencing, infrared greenhouse gas analyzers, and unmanned remote sensing systems.
2.B. Identify secondary, unintended effects on ecosystem services from agricultural practices, utilizing high throughput data acquisition and analyses.
2.C. Develop and evaluate a multi-sensor platform technology for within-the-canopy data collection and machine learning models for high throughput approaches to soybean, pea, and dry bean crop development.
2.D. Soil carbon and GHG monitoring at the farm scale using unmanned aerial vehicle (UAV) and unmanned ground vehicle (UGV) systems.
3. Advance engineering and computational technologies for cropping system best management practices and ecosystem services through innovations in precision agriculture, digital transformations, advanced hardware and software technologies, autonomous systems, computer vision, and artificial intelligence (AI).
3.A. Develop and evaluate AI-enabled techniques and systems for in-field monitoring of crop growth status by multisource remote/proximal sensing and meteorological observations to provide data and information to regulate the performance of the cropping and animal production systems.
3.B. Determine the ‘essential data resolution’ needed to develop effective models to quantitatively estimate crop resiliency at the genotype, environment, management, and its interactions by utilizing statistical analytics, machine learning and artificial intelligence methods.
3.C. To develop a variety of microwave (radio frequency - RF) sensors from small Unmanned Aircraft Systems (UAS) platforms, evaluate their use for water utilization and yield estimation in irrigated and rainfed farming, and create artificial intelligence-enabled algorithms for UAS-based precision agriculture.
3.D. Create the next-generation predictive and prescriptive tools for selection and deployment of climate-resilient cultivars adapted to the region.
Approach
Big data, artificial intelligence, and machine learning are powerful tools that rely on high quality data input, particularly when working with complex, interconnected datasets. Agroecosystems, and their sustainable production and maintenance of environmental health, are known for their interconnected complexities. The role that agronomic management plays in these systems is key towards feeding an ever-growing population, thus ensuring production for decades to come, particularly with increasingly volatile weather systems. Agroecosystems in the southeastern and northern U.S. are the economic platform for a largely agriculture-based society, with a key focus on corn, wheat, soybean, cotton, and animal agriculture. To maintain ecosystem health and to extract all yield potential requires an understanding of all, or many, of the various systems at work (e.g. biology, chemistry, and physics). To maximize the potential of these systems, we must employ every relevant tool, such as fertilizers, cover crops, industrial byproducts, and the confluence of these inputs. No single agronomic plan is a fit for every region, thus this project plan aims to study disparate regional systems to develop best management plans addressing multiple regional conditions such as soil structure, weather, and availability of agronomic inputs. This “systems based” plan addresses the problems, solutions, and impacts of modern agroecosystems. We address these problems via dynamic, data-driven acquisition of large amounts of multi-faceted data streams. Farm, field, experiment station, and laboratory/greenhouse-based experiments will be employed to address these issues. The project datasets comprise biology, chemistry, physics, emissions, remote sensing via unmanned ground and aerial vehicles utilizing the latest data acquisition technologies. Big data analysis and management provide readily accessible data to the public at large, which facilitates transparency and further collaboration. This project plan brings together a large team with varied expertise to develop novel, flexible, and targeted best management practices for sustainable agricultural systems.
Progress Report
Under Objective 1,cotton fields planted were subjected to unpredictable weather and unusual rainfall which prevented on time planting. To a lesser extent, corn planting was also negatively affected by rain events. Long-term studies and their related measurements relying on these fields will be curtailed for this year.
Broiler litter was applied following the fall 2024 harvest of corn and cotton for two experiments. Subsequently, cool-season cover crops were planted. Leachate water was collected after each rain event. The cover crops were terminated using chemical and physical methods and above ground biomass recorded. Decomposition of cover crop residues was monitored in situ using litter bags. Soil moisture and temperature were continuously recorded. Plant traits were measured at four critical growth stages, as well as yield.
Edge-of-Field runoff and subsurface drained water were collected and analyzed for nutrient transported from the field. Litter and treatments were applied to the experimental units. The field strips were equipped with concrete approach sections, bubbler flow-measuring devices, and water sampling units. Runoff events were recorded and samples were collected and analyzed for water quality components.
Vetch was drill-planted in the fall and seven termination treatments were imposed before or after planting corn. Aboveground cover crop biomass samples were taken. Initial observation shows that planting corn into a living cover crop if terminated within 2 weeks of planting corn may not affect the corn stand.
Four legume and one grass species were planted to identify a legume that maximizes N fixation for corn. Corn was planted into the rows left unplanted with cover crops. Most cover crops and corn established an excellent stand in all plots. Preliminary observations show that vetch may be the best cover crop that fits the corn cropping system in the area. While winter pea and red clover had excellent stand, their slow growth in early spring may render them unsuitable as sources of N for corn.
A study was carried out at two locations with different soil types with low native soil K levels. The fields are no-till and conventionally tilled, respectively. Cotton was fertilized with variable K rates corresponding to 0 to 4X of the recommended rates. All plots received comparable nitrogen and phosphate fertilizations. Despite some indications that K may be taken up in greater quantities if supplied from litter, the overall 2024 results do not lead to a clear conclusion.
Integrated field and laboratory research was conducted to evaluate the impacts of various winter cover crop species and fertilizer treatments on soil aggregate stability, erosion resistance, and soil organic carbon distribution in rainfed no-tillage cropping systems in Mississippi. The winter cover crops included cereal rye (Secale cereale), hairy vetch (Vicia villosa), winter wheat (Triticum aestivum), and a mixture of mustard (Brassica rapa) plus cereal rye.
ARS researchers at Mississippi State, Mississippi identified cover and cash crops that reliably accumulate root length and biomass under drought, and sufficient, and excessive watering regimes in early stages of growth. Among crops tested, brassicas and corn demonstrated better potential to improve soil health and productivity of cropping systems. These results are based on replicated, controlled studies conducted under greenhouse and field setting where crops were subjected to three levels of water.
Under Objective 2, soil respiration measurements were taken from objective 1. Measurements included nitrous oxide (N2O) and carbon dioxide (CO2) soil flux. These compliment soil physical, biological, crop, and remote sensed data. Biweekly N2O and CO2 flux measurements investigated coal char and biochar amendments. Models are being explored to link UAV-CO2 flight data with ground respiration.
Rootzone samples collected from objective 1 experiments were collected and processed for gravimetric soil moisture analysis and extracted for microbial DNA from multiple time points throughout the experiments. DNA was assayed for the presence of various soil health related genes. DNA was sequenced and analyzed and enzyme activity to assess soil health in commercially grown turf. Wildlife fecal DNA was collected from multiple study sites throughout the southeastern US, in addition to the launching of a new experiment whereby the use of dung beetles to assess the antimicrobial resistance profile of a given environment is being tested. An experiment was launched in cotton fields to corroborate the findings of a previous study whereby machine learning was used to predict the abundance of soil health associated genes.
Soil samples were collected from objective 1 experiments. We assessed aggregate size distribution, mean weight diameter, geometric mean diameter, fractal dimension, and the carbon content. Multivariate statistical methods and random forest modeling were applied.
An cross-unit study was conducted, where whole corn ear detection was implemented using a novel corn ear hyperspectral imaging system. Detection experiments have been conducted with field ear samples inoculated with Aspergillus.
We developed and tested a leaf scanner to detect corn nutrient status; such data can be paired with unmanned aerial systems data and deep learning models to predict final corn yield. This development is in preliminary stages but can enable timely interventions and matching application rates to crop needs.
A modular platform was designed with dual-facing synchronized RGB-D cameras, enabling improved pod visibility within the crop canopy. The system is being expanded to mount on an autonomous rover for efficient field deployment. We transitioned from static image capture to continuous video acquisition to enhance machine learning. Real-time inference and pod localization are conducted on edge devices with a custom-built GUI enabling visualization of detections and system status.
Currently we have collected flight data under a range of crop growth stages, flight altitude, and wind conditions. Analysis of the data reveals that the most significant challenges to the system that must be addressed to identify best practices include understanding the effect of the interaction between background wind condition and rotorwash from the UAV on air movement, estimating the footprint represented by the air sample, and translation of concentration (a static measurement) to flux (a dynamic measurement).
Two robotic platforms for weed management were upgraded with enhanced capabilities for targeted spraying. Updated mechanical weed removal heads were designed and installed. A laser weed control approach was also successfully tested in a controlled environment. Design modifications were started to integrate with a commercially acquired robotic platform.
Computer simulation of variable rate precision spraying has been implemented for cost-effective analysis. The research worked on the influence of different design factors and spraying methods on the performance of variable rate precision spraying.
Data integration has been designed and prototyped for multisource remote sensing big data on the national geospatial mapping system to create and manage projects in the Mississippi Delta, the Humid Southeast, and North Dakota. Accordingly, machine learning algorithm integration has been designed and prototyped as standalone and Internet-based.
All breeding and agronomy programs were successfully transitioned to the Genovix commercial platform for unified data management, including yield trials using this tool for experimental design and field plot mapping. A new version of ExLibris was launched to integrate, transform, and retrieve historical and current breeding data. For the analytics portion of ExLibris an improved Head-to-Head comparison analysis is now provided to our users for variety release.
Field data were collected into the fall season 2024 to assess various levels of plant stress and to test the developed software processing tool.
We established correlations between surface reflectivity and in-situ soil moisture while accounting for ancillary data like Normalized Difference Vegetation Index, crop height, surface roughness, and crop type. As a result, we have developed a physical-based surface reflectivity calculation model that uses water body measurements. This model is informed by a three-year fusion dataset covering the entire growing period for corn and cotton, with data including GNSS-R, multispectral-based NDVI, and LiDAR point cloud-based features such as plant canopy height.
The beta version of PredictPro was deployed on a test server, enabling genomic prediction through a range of statistical and machine learning models, and was utilized by two breeding programs for forward selection. The tool informed 2025 winter nursery selections in barley and hard red spring wheat breeding programs. A beta version of AgSkySight, a UAV image processing pipeline for RGB and multispectral sensors was completed. This tool is now being adapted to operate on the SCINet HPC system through a separate ARS NACA.
Data collection across the project plan remains central, with raw data collection continuing as our primary focus. We have encountered three key deployment obstacles: 1) historical data downloads don't meet research unit requirements; 2) drone image processing software selection constrained by vendor limitations; and 3) database platform establishment delay. Microsoft Dataverse has now been selected as our relational database platform. Raw data collection is nearly complete for 2019-2023, while 2024 collection is approximately 50% complete.
Accomplishments
1. Organic and inorganic fertilizer was integrated with cover crops to enhance dryland corn sustainability.. Current row crop production practices in the southeastern United States often rely on traditional methods of tillage and leaving fields fallow during winter. Additionally, they are heavily dependent on inorganic nitrogen (N) fertilizer sources. ARS scientists at the Mississippi State, Mississippi location evaluated integrating the effects of organic and inorganic fertilizer with cover crop species on dryland corn. Results indicated that broiler litter increased corn grain yield and grain nitrogen content compared to inorganic fertilizer. The mixture of crimson clover and cereal rye may serve as an ideal treatment to sustain crop N uptake while reducing the risk of N leaching and enhancing corn grain yield compared to using either species alone, particularly in the drier year. A mixture of crimson clover and cereal rye, combined with broiler litter, may serve as an optimal cover crop strategy for sustaining corn grain yield, supporting growers’ incomes while also promoting ecosystem quality.
2. Precision weed management is essential for reducing chemical use and labor in modern agriculture.. Weeds can potentially cause up to $13 billion in reduced income, if left unchecked. However, weed management is a costly operation; thus, reducing the amount of herbicides used in a season is a path towards agriculture sustainability. Scientists at North Dakota State University with a cooperative agreement with ARS at Mississippi State successfully developed and tested a laser-based weed control system in a controlled environment. This innovative system targets and neutralizes weeds using focused laser beams, minimizing damage to surrounding crops and eliminating the need for herbicides. The technology was evaluated across multiple weed species and demonstrated high accuracy in species-specific control. This system has potential to reduce a substantial on-farm cost, herbicide treatments, which will help farmers and producers across the country.
3. Crops reliably accumulate root length and biomass under various water stresses are key to sustainable Southeastern systems. Farming under unpredictable precipitation makes incorporating cover crop into a cropping system more difficult. Scientists at Mississippi State University in conjunction with a cooperative agreement with ARS at Mississippi State, Mississippi tested brassicas (cover) and corn (cash) which demonstrated better potential to improve soil health and productivity of cropping systems. Results are based on controlled studies conducted under greenhouse and field setting where crops were subjected to three levels of water, and evaluation was made on physiological and morphological plant parameters. The implication of these results is that farming systems which include brassica cover crop species in corn production may be more resilient to a range of weather conditions, when compared with competing options, thus providing MS farmers with key sustainable agriculture systems.
4. Poultry litter is a profitable fertilizer for dryland soybeans. Research has shown that fertilizing soybeans with poultry litter increases yield but whether this practice is profitable has not been investigated. ARS Scientists at the Mississippi State, Mississippi location and Mississippi State University investigated the profitability of fertilizing dryland soybeans with poultry litter relative to fertilizing with conventional synthetic fertilizers in a marginal upland soil. Partial budgeting was used to determine revenue, costs, and net returns of cover crops and fertilizer sources. Poultry litter outyielded synthetic fertilizers by 12.5%. Cover crops did not increase soybean yield in any of the 5 yr. Soybean yield, regardless of cover crop or fertilizer, was highly dependent on the amount rainfall received in July. The economic analysis indicated that despite its high cost, poultry litter was about 20% more profitable than synthetic fertilizers due to greater yield gains. The results show that poultry litter fertilization is profitable for dryland soybean in marginal soils, but use of cover crops may not be justified on a short-term basis. This provides MS farmers with practical information that some sustainable management systems may be situationally dependent.
5. A whole corn ear hyperspectral imaging system has been developed for aflatoxin detection. Aflatoxin contamination detection is important throughout corn and other industries. Rapid detection of aflatoxin in the field can swiftly quell problems in the field, potentially saving substantial losses. ARS scientists at the Mississippi State, Mississippi location developed a novel hyperspectral imaging system for whole corn ear imaging and aflatoxin detection. The instrument records a unique fluorescence signal linked to fungus infected and aflatoxin contaminated corn kernels. This research is the first to report on the use of the kernel crown fluorescence for aflatoxin detection and highlights the potential of whole-ear hyperspectral imaging for non-invasive contamination screening and phenotypical trait analysis in corn breeding, which can reduce farmer losses throughout the US and world.
6. Soil moisture assessment utilizing surface reflectivity can improve field scale soil moisture assessment. Measuring soil moisture via traditional methods can be labor intensive and is often only associated with small sample sizes, not representative of the entire field. Understanding field soil moisture variability is a key to healthy crop production throughout the growing season. Scientists at Mississippi State university through a cooperative agreement with ARS scientists at the Mississippi State, Mississippi location have established correlations between surface reflectivity and in-situ soil moisture readings. A model was developed to estimate soil moisture across a range of field conditions. This model has the potential in conjunction with further refinement to provide US farmers with on the go sensing of moisture conditions across entire fields assessed via UAS.
7. Precision weed spraying can be improved through the use of vision-based applicators. Spraying simulation models have been developed. Variable rate herbicide applications can improve weed control efficiency. The applications become more feasible with the development of deep learning and artificial intelligence in weed detection and management. ARS scientists at the Mississippi State, Mississipi location developed spraying simulation models to simulate real-time machine vision-based variable rate precision spraying. This system utilizes weed data from simulated weed density distributions in crop fields. The results provide valuable suggestions for determining variable rate application strategies for precision weed management in crop production.
8. A new analytical infrastructure for smart agriculture has been designed and developed to integrate agricultural remote sensing, big data, and algorithms. To facilitate agricultural remote sensing with artificial intelligence applications, a big data hub to process, organize, visualize, and interpret multisource remote sensing data is needed. Scientists at the Mississippi State, Mississippi location developed a machine learning algorithm which can be synchronized to provide a user-friendly, high-performance crop growth process modeler, which incorporates remote sensing, weather, and ground-truth data. The data and algorithm integrated system has been designed and tested with various machine learning algorithms through benchmark statistical analysis in conjunction with remotely sensed images and in-situ data collected from various field experiments. This development will deliver an analytical infrastructure that significantly improves agricultural remote sensed data management for timely interpretation, which greatly facilitates the use of AI techniques to accurately characterize and analyze crop growth processes.
9. Terminated cover crops (CC) generate many positive effects during the cash crop season like increased soil moisture, decreased soil temperature. Valuation of cover crop usage is needed to encourage farmer adoption for soil improvements. Scientists at the Mississippi State, Mississippi location, investigated graded levels of rye CC on soil respiration and nutrients, finding that greatest biomass coverage increased carbon dioxide flux by 40% but decreased nitrous oxide flux 65% relative to no CC. This demonstrates a valuable tradeoff for soil respiration. This is the first study to quantify soil respiration relative to CC in the southeastern United States, and will be the basis for prescribing CC recommendations to MS farmers by providing optimized systems preventing valuable losses of nitrogen.
10. Integrating winter cover crops and poultry litter significantly improve soil structural stability, increases carbon stocks, and reduces erosion risk in rainfed no-tillage systems. Soil degradation and erosion remain critical challenges in the Southeastern United States, where conventional fertilization practices often fail to enhance soil structure or promote long-term soil health. Scientists at the Mississippi State, Mississippi location in conjunction with a postdoctoral researcher utilized multiple years of field experiments and laboratory evaluations to determine that cereal rye and mustard-rye mixtures were the most effective cover crops for improving soil aggregation. Poultry litter consistently outperformed inorganic fertilizers in promoting the formation of larger, more stable aggregates and reducing soil erodibility. The highest concentration of soil organic carbon was found within the microaggregate fraction (0.25-0.053 mm), particularly under the combined use of cover crops and poultry litter. These results directly support federal conservation and sustainable agriculture programs, and provide practical, research-based guidance for farmers, land managers, and agricultural advisors seeking to improve soil health and reduce erosion through locally adaptable practices.
Review Publications
Chen, D., Huang, Y. 2025. Integrating reinforcement learning and large language models for crop production process management optimization and control through a new knowledge-based deep learning paradigm. Computers and Electronics in Agriculture. 232(110028): 1-12. https://doi.org/10.1016/j.compag.2025.110028.
Balderas, J., Chen, D., Huang, Y., Wang, L., Li, R. 2025. A comparative study of deep reinforcement learning for crop 1 production management. Smart Agricultural Technology. 10(2025):2772-3755. https://doi.org/10.1016/j.atech.2025.100853.
Signh, K., Huang, Y., Young, W., Harvey, L., Hall, M., Zhang, X., Lobaton, E., Jenkins, J.N., Shankle, M. 2025. Sweet potato yield prediction using machine learning based on multispectral images acquired from a small unmanned aerial vehicle. Agriculture. 15(420):1-23. https://doi.org/10.3390/agriculture15040420.
Chandan, K., Jagman, D., Huang, Y., Reddy, K.N. 2025. Field-scale corn yield prediction using UAV multispectral data and explainable machine learning models. Computers and Electronics in Agriculture. 231 (2025) 109990. https://doi.org/10.1016/j.compag.2025.109990.
Kharel, T.P., Tyler, H.L., Mubvumba, P., Huang, Y., Bhandari, A.B., Fletcher, R.S., Anapalli, S.S., Joshi, D.R., Mengistu, A., Birru, G.A., Adhikari, K., Dhakal, M., Maskey, M.L., Reddy, K.N., Clay, D.E. 2025. Machine learning on multi-spectral imagery to estimate nutrient yield of mixed-species cover crops. Agricultural & Environmental Letters. https://doi.org/10.1002/ael2.70009.
Dai, W., Feng, G.G., Huang, Y., Adeli, A., Jenkins, J.N. 2024. Impact of cover crop management in a corn-cotton cropping system on soil aggregate stability and related factors. Soil & Tillage Research. 244(2024):106197. https://doi.org/10.1016/j.still.2024.106197.
Dai, W., Feng, G.G., Huang, Y., Tewolde, H., Shankle, M.W., Jenkins, J.N. 2024. Water-stable soil aggregation and associated carbon in a no-till soil with cover crops and poultry litter. Soil & Tillage Research. 248(2025):106399. https://doi.org/10.1016/j.still.2024.106399.
Adeli, A., Brooks, J.P., Miles, D.M., Read, Q.D., Huang, Y., Feng, G.G., Jenkins, J.N. 2025. Integrated effects of tillage and fertilizer sources with cover crop on dryland cotton. Agronomy Journal. 2025(117):e70019. https://doi.org/10.1002/agj2.70019.