Oxidative stress assessment by a spectroscopic approach in pomegranate plants under a gradient of ozone concentrations

Techniques to monitor oxidative stress pre-visually are essential to optimize plant management. Here, we investigated the capability of hyperspectral reflectance (350−2500 nm) to characterize responses of two pomegranate cultivars (Parfianka and Wonderful) under ozone (O3) episodes at a gradient of concentrations (50, 100 and 200 ppb for 5 h). Analyzing spectral signatures collected rapidly and non-destructively from asymptomatic leaves, we accurately discriminated the two cultivars, as well as controls from plants exposed to O3, in particular those under the higher oxidative stress (i.e. 200 ppb). These discriminations were especially accurate in Wonderful at the end of the exposure (5 h from the beginning of exposure; FBE), and at 24 h FBE. Furthermore, using a partial least squares regression (PLSR) approach, we constructed predictive spectral models to estimate from spectra an array of commonly used physiological and biochemical leaf traits related to plant/oxidative stress interaction (photosynthesis, lipid peroxidation, enzymatic and non-enzymatic antioxidants). Most traits were relatively well predicted by spectroscopic models (model goodness-of fit for validation, R2: 0.77−0.50). Finally, variations of spectra-derived vegetation indices and leaf traits derived from spectra confirmed the lower O3-tolerance of the Wonderful cultivar, when exposed to 200 ppb. Overall, the present study shows that the proposed spectroscopic approach can rapidly and non-destructively assess early oxidative stress conditions in plants, and consequently it can help in increasing plant yield and quality. Limitations of the approach are also presented and discussed.