Location: Plant Germplasm Introduction and Testing Research
Project Number: 2090-21000-037-020-S
Project Type: Non-Assistance Cooperative Agreement
Start Date: Jul 1, 2024
End Date: Aug 30, 2025
Objective:
The overall objective of this NPGS Horticultural Germplasm Evaluation Proposal is to characterize 291 accessions of lima beans (Phaseolus lunatus) in the GRIN collection for traits relevant to succulent bean production and consumption to inform breeding efforts and improve current greenhouse-limited characterization methods while regenerating and characterizing 101 not currently available accessions. The specific objectives are as follows:
1. Characterize 291 accessions for seed size, shape, and color using image analysis.
2. Evaluate 291 accessions for plant architecture, growing degree days to flowering, leaf shape, and lima bean anthracnose symptoms in a replicated field trial.
3. Run PACE markers for photoperiod sensitivity, plant architecture, and seed coat color to validate potential markers for community use and further characterize the accessions.
4. Develop and test a protocol to track phenological relative to growing degree days during genebank regenerations
5. Integrate phenotypic data into GRIN and publish a data note to maximize accessibility.
Approach:
Objective 1. Twenty seeds from each accession will be photographed with color and size standards using a flatbed scanner. A blue background will be used to facilitate the segmentation of seeds during image analysis. Images will be color-corrected and segmented from the background. The length, width, area, and circumference of each seed will be measured using the python library OpenCV (Bradski et al. 2000).
Objective 2. The trial will be planted spring 2024 and will use standard management practices for the region. The trial will be phenotyped throughout the growing season using Field Book (Rife and Poland 2014). Phenotypes to be collected include stand count, growth habit, leaf shape, flowering time, and lima bean anthracnose symptoms. All phenotypes will be collected using the common bean ontology as appropriate, as a lima bean ontology is not yet available.
Leaf shape will be evaluated by image analysis following the protocol used by Ernest (2020). One trifoliate leaf per plant from each of three plants per plot (9 leaflets) will be cut from the plant and immediately transported to the lab for imaging. The total diseased leaf area, number of lesions, and average lesion size will be measured using a custom python script. Due to the lack of protocol or isolates for field-based inoculation with Colletotrichum truncatum, these observations will rely on natural infection. Given the high rate and severity of infection with this disease in recent seasons, it is likely that adequate natural infection will occur. In the event that infection does not occur, this trait will not be phenotyped.
Objective 3. Leaf tissue will be collected from three plants in each plot in the field trial. PCR-quality DNA will be extracted from the samples using a pulse-tested sodium hydroxide method from KnowPulse Knowledgebase (Sanderson et al. 2019), and each sample will be evaluated using allele-specific PACE markers for three traits: photoperiod sensitivity, plant architecture, and seed coat color. Marker results will be compared with observed phenotypes from the field trial (Objective 2) and seed images (Objective 1).
Objective 4. Collaborator will perform standard greenhouse regenerations with a focus on P. lunatus accessions which are putatively photoperiod-neutral, not currently available or backed up, and vulnerable to loss. Phenology, first flower, first shelling bean, and first dry bean phenotypes will be collected during these regenerations. Each of these traits is a standard bean trait listed on GRIN. Data loggers will be used to associate plant development with day length and growing degree days experienced in the greenhouse environment.
Objective 5. All phenotypic data and raw images will be curated, labeled with metadata, and made freely available via upload to GRIN Global within six months of data collection. All project-related code will be made publicly available via GitHub.
