Work

Early detection of nitrogen status in cucumber plants using bioimpedance spectroscopy

Year: 2025

Type: article

Abstract: • Bioimpedance spectroscopy detects nitrogen deficiency in cucumber at the third leaf stage. • Nitrogen limitation increases extracellular fluid resistance and reduces cell membrane capacitance. • Nitrogen status was classified with up to 91% accuracy using PSO-optimized double-shell modelling. • Bioimpedance parameters reveal genotype-dependent nitrogen stress responses. Bioimpedance spectroscopy (BIS) analysis provides rapid, non-destructive insights into water and nutrient dynamics. However, its application for diagnosing nitrogen (N) deficiency in plants remains limited in the current literature. In this study, we evaluated whether BIS can reliably detect early N deficiency in cucumber plants and serve as a fast, non-destructive alternative to traditional diagnostic methodsTwo commercial hybrids (Nairobi F1 and ES2217 F1) were grown under high and low N conditions in a controlled light chamber, with optimal water and environmental conditions, where N was the only limiting factor. In addition to BIS, plant physiological parameters were evaluated. A custom four-channel BIS device recorded impedance spectra from the third true leaf of seedlings, and a double-shell electrical model was used to extract the impedance parameters. Parameter fitting was performed using a particle swarm optimization (PSO) algorithm with a custom cost fitness function. In this study, we found that plants under low-N supply showed significantly higher extracellular fluid resistance and lower cell membrane capacitance compared to high-N plants at an early phenological stage. In contrast, changes in vacuole fluid resistance were genotype-dependent, suggesting differential internal nitrate (NO 3 ⁻) storage or vacuole transport capacity between the hybrids. These changes were consistent with plant physiological indicators, including reduced total leaf N and NO 3 ⁻, lower membrane stability index. These two electrical parameters enabled reliable classification of N status, with ROC-AUC values of 0.86–0.92 and overall accuracy up to 91%. Our findings suggest that BIS, combined with PSO-based modelling, can detect early N deficiency quickly and non-destructively.

Source: Smart Agricultural Technology

DOI: 10.1016/j.atech.2025.101610