Fluorescence noise analysis as a sensitive indicator of dehydration stress in plants

Abstract

• Fluorescence noise enables early, sensitive, scalable detection of plant water stress. • Dehydration produces distinct, quantifiable fluorescence noise signal changes. • Histogram broadening indicates disrupted energy transfer and pigment–protein disorder. • Noise metrics distinguish hydration states more reliably than Φ PSII measurements. • Portable, low-cost method compatible with drones or towers for crop monitoring. Climate change is increasing the frequency and severity of droughts, necessitating rapid and cost-effective methods for detecting water stress in crops. Current technologies, such as soil humidity sensors and image processing, have limitations in accuracy, scalability, and cost. Here, we introduce a novel approach utilizing noise analysis of ambient chlorophyll fluorescence from plants to assess dehydration status. By analyzing the statistical properties of fluorescence fluctuations (“noise”), we identify characteristic changes that correlate with water deficiency. This constitutes a fundamentally new signal-processing framework, extending beyond conventional fluorescence intensity methods and enabling integration with automated sensor networks or drone-based platforms for large-scale agricultural monitoring. Our findings, validated in three different plant species, reveal that noise amplitude distributions and power spectral density (PSD) patterns shift predictably with dehydration, outperforming standard fluorescence measurement techniques in precision and reliability. Transmission electron microscopy further confirms structural changes in thylakoid membranes associated with these noise patterns. This scalable, non-invasive method holds significant potential for early drought detection in agricultural settings, paving the way for real-time, automated irrigation decision-making systems, while also offering a powerful tool for probing the energetic landscape in bio-physical systems through advanced signal-processing techniques.

Abstract from the paper’s OpenAlex record; open copy licensed cc-by. Read at the source ↗

The text the model read, with the source record’s inline markup (italics, sub- and superscripts) rendered and its character entities decoded; quote offsets refer to the raw text.