Experiment II

Phomopsis Stem Canker in sunflower plants 

(PSC sunflower 2024) 

The Experiment was conducted from May to August 2024 at the "Enrico Avanzi" Center for Agro-environmental Research in San Piero a Grado, Pisa (University of Pisa, Italy; coordinates: 43.674462 N, 10.318070, 10 m a.s.l.), to evaluate the effectiveness of the biostimulant BlueN in managing Phomopsis stem canker and developing an eco-friendly method for early disease detection using remote sensing measurements. Sunflower seeds from two cultivars (P64HE144, less susceptible, and P64HE167, more susceptible) were sown in 32 plots, applying four treatments: control, inoculated control (Diaporthe helianthi), BlueN-treated non-inoculated, and BlueN-treated inoculated (Figure 5). Each subplot contained eight plants, spaced 50 cm apart, resulting in a planting density of 8 plants per m². To ensure consistency and comparability, all leaf measurements focused on the youngest mature leaves. Throughout the experiment, field and remote sensing measurements were collected to monitor plant health. Two intensive survey campaigns were conducted between July (T1, July 11–12) and August (T2, (July 31–August 1), resulting in a comprehensive dataset, including hyperspectral imaging, thermal infrared (TIR), LiDAR, and solar-induced fluorescence (SIF). Field survey included continuous canopy reflectance measurements, leaf-level hyperspectral assessments, physiological response evaluations, and root microbiome analysis to better understand plant-pathogen interactions.

A preliminary investigation of hyperspectral data focused on selected plots based on plant health and vegetation cover homogeneity. The analysis aimed to extract spectral signatures and assess temporal variations, differences between healthy and infected plants, and disease progression between resistant and susceptible cultivars. Hyperspectral images were taken around 2 PM to minimize shadow effects, and a mask was applied to focus on sunlit leaves. Regions of interest (ROIs) were manually drawn for each plot, ensuring robust statistical significance.