Imaging-based seed phenotyping through imbibition kinetics for quality assessment and breeding

Overview

This solution offers an automated, low-cost imaging platform for evaluating seed quality through imbibition kinetics — the initial stage of seed germination in which seeds absorb water. By replacing traditional gravimetric methods with digital image analysis, the approach delivers seed quality assessments in less than 24 hours, enabling rapid single-seed selection for plant breeding programs.

The technology is particularly relevant for bean breeding, where seed quality directly impacts germination performance and field establishment. Beyond quality assessment, imbibition kinetics phenotyping can reveal genetic variation associated with water uptake behavior, supporting the identification of genes and quantitative trait loci (QTL) linked to tolerance to water-limiting conditions. This makes the platform a valuable tool for breeding programs focused on climate resilience and drought adaptation.

Technical specifications

• Camera-based imaging system that captures seed images every 15 minutes over a 24-hour period, generating approximately 17,000 data points per experiment
• Gompertz model fitting applied to imbibition curves to extract quantitative parameters describing water uptake dynamics
• Validated in bean recombinant inbred lines (n = 128), demonstrating a relationship between imbibition kinetics and seed quality
• Scalable to large germplasm collections — currently being tested on 1,100 bean accessions from a worldwide germplasm collection
• Low-cost and high-throughput compared to traditional gravimetric or scanner-based methods
• Single-seed resolution enabling individual seed selection rather than bulk population assessment
• Genetic analysis pipeline including genotyping-by-sequencing for SNP identification, genetic linkage map construction, and QTL analysis
• Transcriptome profiling via nanopore RNA-seq of parental lines to identify differentially expressed genes associated with imbibition behavior

Technology readiness level

The platform has been experimentally validated using a population of 128 bean recombinant inbred lines, confirming a relationship between imbibition kinetics and seed quality. Current efforts are scaling validation to 1,100 bean accessions from a worldwide germplasm collection to assess broader applicability across genetic diversity.

Future validation plans include crossing accessions with contrasting imbibition kinetics, evaluating F3 progeny under water-limiting conditions, performing QTL analysis and RNA-seq transcriptomics, and selecting plants based on imbibition response. The method builds on published precedent in maize seeds (Miller et al., 2018) and is positioned at a mid-to-late experimental validation stage, with ongoing development toward breeding-ready applications.