Location: Plant Science Research
2024 Annual Report
Objectives
Objective 1: Develop genomic tools for alfalfa to accelerate breeding and facilitate identification and validation of genes for important agronomic traits.
Sub-objective 1.A: Compare alfalfa genomes to improve understanding of genome architecture and complexity.
Sub-objective 1B: Identify DNA markers associated with biotic stresses to facilitate germplasm development.
Subobjective 1C: Improve transformation and gene editing in elite genotypes to accelerate field testing of novel edited plants.
Objective 2: Develop breeding methods and understanding of genetic control of important agronomic traits in alfalfa for plant improvement.
Sub-objective 2.A: Improve environmental resiliency and abiotic stress tolerance in alfalfa using genomic selection and machine learning.
Sub-objective 2.B: Develop alfalfa germplasm with enhanced forage quality and digestibility.
Sub-objective 2.C: Develop germplasm with novel root traits that enhance herbage biomass and utilization of alfalfa in agroecosystems.
Sub-objective 2.D: Develop novel germplasm with protein and nutritional profiles desired in human food products.
Objective 3: Establish innovative methods and new standards for assessing and evaluating alfalfa quality for multiple uses.
Sub-objective 3.A: Quantify and characterize variability of non-structural carbohydrates in alfalfa for improved nutritive value.
Sub-objective 3.B: Increase alfalfa fiber digestion by the investigation of cell wall lignification and digestion in alfalfa with reduced lignin concentrations.
Objective 4: Increase understanding of the interactions among forage crops, soils, and microbiomes to reduce risk, improve agronomic outcomes, and build resilience.
Sub-objective 4.A: Evaluate the impacts of alfalfa cultivar, cutting frequency, and fall dormancy on root production, root turnover, root litter quality, C inputs to soil, and interactions with microbial communities.
Sub-objective 4.B: Develop methods to isolate, identify, and characterize emerging plant pathogens.
Sub-objective 4.C: Evaluate alfalfa establishment and termination strategies, diverse crop rotations, and winter annual cover crops for improving soil C balances and reducing greenhouse gas emissions.
Sub-objective 4.D: Measure total root biomass, root:shoot ratios, root responses to management, and C fractions in forage systems to improve C accounting and model parametrizations of alfalfa and other forage crops.
Approach
Alfalfa is the engine that drives dairy and beef production and is unparalleled for providing environmental services. However, the slow progress in increasing forage yield and re-establishment costs after winter injury have discouraged greater utilization of alfalfa. Modern breeding methods and -omics technologies provide the opportunity to break the yield bottleneck, improve plant persistence and forage nutritive value, and develop novel products and environmental services. In support of these goals, we will assemble, annotate, and carry out an in-depth characterization of the alfalfa genome for structural and repeat number variants using bioinformatic tools. DaRTag SNP markers will be used to identify superior germplasm with improved winter survival, greater forage nutritive value, and root architecture to increase yield potential, and to identify markers associated with resistance to major yield-limiting diseases. Methods will be developed to improve alfalfa gene editing in diverse germplasm. Compositional analyses of alfalfa herbage using biochemical and chromatographic methods will lay the groundwork for utilization of alfalfa in human food products. In-depth analyses of stem cell wall development and ruminal degradation will be done to gain a better understanding of developmental and structural changes that improve forage quality. High throughput sequencing and metagenomic analyses will help fill knowledge gaps in the composition and function of key microbial communities associated with alfalfa in diverse soils and with biotic and abiotic stress. Conventional and novel crop rotations utilizing forages will be evaluated for their effect on greenhouse gas emissions and carbon sequestration using field-scale eddy covariance flux measurements and measurements of total root biomass, root:shoot ratios, root responses to management, and C fractions in forage systems to improve C accounting and model parametrizations of alfalfa and other forage crops.
Progress Report
Since its initiation in January 2024, significant progress was made in support of Objective 1. Whole genome resequencing and chromosome level assembly based on PacBio HiFi sequencing was done for eight genotypes varying in fall dormancy. To aid in gene annotation, RNA was extracted from roots, root nodules, stems, leaves, flowers, and seed pods of each of the genotypes and used for RNA-Seq and long-read Iso-Seq analysis. RNA-Seq data was aligned to the corresponding genome sequence and assembled into transcripts to analyze expression in each tissue. To aid in further assembly of the four haplotypes of each genome sequence, F1 genetic mapping populations were made, and plants established for DNA extraction to be used in genotyping and phenotyping disease resistances. Also, DNA was submitted for mapping chromosome scale interactions using Omni-C technology. A previously generated F1 mapping population was evaluated for resistance to Aphanomyces root rot and a second novel locus for high resistance was identified. A publication describing this research is in preparation.
In support of Objective 2A1, an analysis of genetic diversity and population structure was conducted on alfalfa germplasm established to evaluate winter survival and biomass yield using a panel of 3,000 single nucleotide polymorphic markers from Diversity Array Technologies (DArTag) and Breeding Insight. Approximately 12,000 microhaplotypes were found within the 3,000 markers. Principal Component Analysis and Discriminant Analysis of Principal Components revealed five distinct population structures based on geographical origin, with the check cultivars forming a central cluster. Inbreeding coefficients (FIS) ranged from -0.1 to 0.006, with 27 out of 28 populations showing negative FIS values, indicating an excess of heterozygotes. The most significant variation among populations was found in the check cultivars at 10.6% of total molecular variation, while the lowest variation was observed in BASE populations at 7.3%. The study identified 16 quantitative trait loci for winter survival, and genomic selection and machine learning models were developed from these data. A manuscript reporting these results has been submitted. In addition, potato leaf hopper and canopy structure notes were taken in 2023. About 400 individual plants were selected in early spring 2024 for fast early spring regrowth, high winter survival, large leaf size, and high branch numbers. For Objective 2A2, two field experiments were established with transgenic phosphate hyperaccumulating plants but the plots suffered from severe winter injury and will need to be replanted. A greenhouse experiment is underway to identify plants that have improved agronomic traits for the field study. In support of Objective 2B to develop alfalfa germplasm with enhanced forage quality and digestibility, over 6,000 samples have been separated into leaf and stem fractions, ground, and are being tested for forage quality traits by near infrared spectroscopy. A genomic selection and two machine learning models were developed from 500 genotyped and phenotyped lines. From 1,499 genotypes, 200 were identified based on genomic selection and machine learning, 100 with the highest and 100 with the lowest stem fiber digestibility for the next breeding cycle. For objective 2C, four populations differing in root architecture were selected and cycle 5 seeds produced using caged bees and by hand-pollination. The seeds were planted in the field in Rosemount and Becker, MN at two densities, 3 inches between plants and a solid seeded plot. Twenty minirhizotron tubes were installed to track the dynamics of root system development. For objective 2D, samples collected in 2023 for the evaluation of functional proteins and nutritional components in a diverse alfalfa population were ground and a subset have been analyzed for water-soluble and fat-soluble components as well as for saponin concentrations. Variation for these compounds was found. The 2024 harvest of alfalfa samples were collected, and flash frozen for later analyses.
In support of Objective 3, for analysis of nonstructural carbohydrates in fine roots of alfalfa under different management regimes, samples from two seasons have been collected with the third and final season of collection currently in progress (25% completed). Optimization of an ion chromatograph methodology and the extraction protocol for water-soluble carbohydrates from leaf and root material was completed. For the comparison of stem structure in plants with diverse in vitro neutral detergent fiber digestibility (IVNDFD), the majority of the first season (2023) stem degradation assays have been completed. The material for the composition analysis have been ground and weighed out. Currently, the stem samples for the 2024 season are being harvested at two locations according to physiological time points (50% completed).
In support of Objective 4A1, a second research site was established in 2023 in Waseca, Minnesota and manure was applied prior to establishment of the alfalfa varieties. Yield, stand density, crown health, and root biomass data were collected in the establishment year. The first site at Rosemount, Minnesota has been rotated to corn after 3 years of manure-applied alfalfa to evaluate the N-credits to corn with varying manure management strategies. In support of Objective 4A2 to identify root-associated microbial communities associated with alfalfa rotation effects, soil and plant samples have been collected from long-term rotation experiments, DNA was extracted, and microbiomes have been sequenced from first-year samples following different rotation crops. Preliminary analyses of bacterial, fungal, and arbuscular mycorrhizal marker genes indicate that preceding crops have a significant effect on subsequent plant-associated microbiomes, including numerous potentially beneficial taxa involved in nutrient acquisition. We anticipate acquiring at least 3 years of data to establish robust patterns in rotation effects on plant microbiomes. In efforts to determine a core alfalfa microbiome (Goal 4A3), alfalfa was grown in distinct soils, DNA was extracted, and sequencing is underway. Experiments to characterize the dynamics of microbiomes during plant development are underway. Host-mediated microbial engineering experiments (Goal 4A4) have been initiated to select for microbiomes conferring suppression to Aphanomyces root rot and the first round of selection has been conducted. However, this experiment may require modification to achieve sufficient disease pressure to differentiate between effective and ineffective microbiomes. In efforts to characterize metagenomes on dryland alfalfa (Goal 4A5), initial sampling of paired dryland and irrigated alfalfa fields has been conducted. Additional timepoints over the growing season are planned. In support of Objective 4C, eddy covariance, soil, and crop yield data collection continued in two production fields, one in a perennial continuous living cover system (intermediate wheatgrass, 2023 - 2025) and the other an annual continuous living cover system (soybean - winter barley, 2023). Data analysis is underway and preliminary findings were presented at the ASA, SSSA, and CSA annual meeting in fall 2023. In support of Objective 4D, third year data collection of root production, root turnover, root litter quality, and C inputs continued for alfalfa of varying fall dormancy and cutting frequency.
Accomplishments
1. Alfalfa root system architecture image analysis using artificial intelligence. Roots are critical for nutrient and water acquisition in plants and breeding plants with improvements of plant root structure are important for increasing aboveground traits such as biomass yield and increased nutritive value. However, categorizing root structure manually is time-consuming and is prone to error by individual plant breeders. A modern method for investigating plant root structure pairs labeled images with artificial intelligence (AI) to detect root properties such as total root length and the number of secondary and tertiary roots off the main tap root to categorize the entire root system architecture. ARS scientists in St. Paul, Minnesota analyzed 15,000 root systems of alfalfa plants using digital images as input into an AI model to test its ability to predict root types and compare its accuracy and speed to human predictions. Initial model results were approximately 64% accurate at predicting the root type and the model was improved an additional 11 to 13% by applying reactive machine learning and confident machine learning to achieve an overall prediction accuracy of 86%. This method can be used by non-specialists, requires only a phone camera, and reduces the time for identifying root types from 22 weeks to 2 weeks while reducing human errors caused during sampling and plant selection. The AI root analysis method has accelerated the breeding process and improved selection accuracy for developing alfalfa plants with diverse traits; more fibrous roots that have an increased the number of nitrogen-fixing root nodules, plants with a deep tap root with greater drought tolerance, and plants with highly branched roots for wet soil tolerance; to make the crop more productive and resilient under diverse climatic conditions.
Review Publications
Botkin, J., Farmer, A.D., Young, N.D., Curtin, S.J. 2024. Genome assembly of Medicago truncatula accession SA27063 provides insight into spring black stem and leaf spot disease resistance. BMC Genomics. 25. Article 204. https://doi.org/10.1186/s12864-024-10112-9.
Liu, H., Xu, Z. 2023. Editorial: Machine vision and machine learning for plant phenotyping and precision agriculture. Frontiers in Plant Science. 14. Article 1331918. https://doi.org/10.3389/fpls.2023.1331918.
Weihs, B.J., Heuschele, D.J., Tang, Z., York, L., Zhang, Z., Xu, Z. 2024. The state of the art in root system architecture image analysis using artificial intelligence: A review. Plant Phenomics. 6. Article 0178. https://doi.org/10.34133/plantphenomics.0178.