Non-invasive methods to assess seed quality based on ultra-weak photon emission and delayed luminescence

Abstract

Abstract Seed quality is the set of physical, genetic, and physiological characteristics, reflecting the overall germination potential. Maintaining an optimal seed quality is essential for agriculture and seed banks to preserve genetic diversity. Compared to conventional methods (e.g., germination tests), non-invasive approaches allow a more sustainable and rapid evaluation of seed quality but this is limited by high costs. The measurement of ultra-weak photon emission (UPE) and delayed fluorescence (DL), defined as biological phenomena potentially related to the physiological status of living systems, may represent a suitable approach to estimate seed quality. To test this hypothesis, seeds of five agriculturally relevant legume species (Phaseolus vulgaris L., Lathyrus sativus L., Cicer arietinum L., Pisum sativum L., and Vicia faba L.), stored at different conditions (room temperature or -18 °C) for several years, were analysed using a LIANA© prototype to collect data regarding DL and UPE occurring after UV excitation. The obtained data were integrated with germination parameters which underline species-specific behaviours in response to storage conditions. The prediction models show variable efficiency in classifying seeds based on germination which underline species-dependent links between photon emission and seed quality. Therefore, these measurements represent novel, non-invasive, and rapid approaches to evaluate seed quality.

Keywords

• Delayed luminescence
• Leguminosae
• Machine learning
• Non-invasive assessment
• Seed quality
• Ultra-weak photon emission