From smartphone to gene: Automated image-based phenotyping captures flowering diversity of common buckwheat

Automated smartphone/camera imaging with Mask R-CNN quantified 22 flowering traits across 169 common buckwheat accessions, with heritability up to h2=0.93 and strong agreement to manual stages.

• Why it matters: Field phenotyping of buckwheat flowering is slow and expertise-dependent, limiting breeding for flowering patterns needed to reduce yield losses and improve mechanized harvesting.
• What they did: Over five growing seasons, 37,078 images were segmented and flowering dynamics were modeled to extract traits for flowering time (n=10), intensity (n=6), and abscission (n=6).
• The result: Genome-wide association studies identified 171 genomic regions linked to flowering traits, including marker-trait associations and candidate hormone-signaling genes such as “Auxin response factor” and “Ethylene-insensitive protein.”

AI summary of the abstract · may contain errors · not medical advice · read the paper.