This is the authors' abstract. We don't add key points for this paper.
High-throughput and non-destructive assessment of seed quality is essential for accelerating genetic improvement, ensuring seed quality assurance and supporting long-term germplasm conservation. Although conventional viability and vigour tests are reliable, they remain destructive, labour-intensive and time-consuming. In this study, near-infrared reflectance spectroscopy (NIRS) combined with modified partial least squares (MPLS) regression was evaluated for predicting key physiological seed quality parameters across 122 genetically diverse soybean ( Glycine max ) accessions. These accessions represented wide variability in seed coat colour, size, shape and storage history with freshly harvested and one- and two-year old naturally-aged seed lots. Spectral pre- processing involving standard normal variate (SNV) correction, detrending and derivative smoothing effectively minimised scattering effects and enhanced biochemical signal clarity. The developed calibration models exhibited high predictive accuracy (RSQ external = 0.71-0.84) and strong discrimination ability (RPD = 1.63-2.25), with germination showing the highest prediction performance (RSQ = 0.84, RPD = 2.25). External validation confirmed model robustness and statistical reliability. These findings demonstrate the potential of NIRS as a rapid, accurate and non-invasive approach for assessing seed viability and vigour in soybean, reducing dependence on destructive laboratory tests and enabling large-scale, real-time phenotyping for breeding, seed gene bank management and routine seed quality evaluation.
Answers are grounded in this paper and cite their sources.
No takes yet. Share an insight, caveat, or question.
Patil et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: