Assessing Nitrogen Fertilization Impacts on Avocado Ecophysiology: Insights from UAV-Derived Remote Sensing and Nutritional Experiments Responses

Uniform and excessive nitrogen (N) fertilization practices in agriculture often lead to spatial and temporal crop variability, resulting in inefficient nutrient distribution, environmental harm, reduced orchard yields, and fruit quality. This study investigates the effects of N fertilization levels on avocado ecophysiology using advanced remote sensing (RS) techniques, focusing on the estimation of Nitrogen Use Efficiency (NUE), Leaf Nitrogen Content (LNC [% dry weight]), Plant Area Index (PAI), and Canopy Nitrogen Content (CNC [kg per tree]) to estimate crop responses to over and under fertilization. NUE was calculated as Partial Factor Productivity (PFP) of N; kg of N in harvested avocado per kg N applied. This research includes three experimental systems: (1) a lysimetric experiment at the Gilat with five continuous N fertigation treatments ranging from 115 to 1400 kg-N ha; (2) Kfar Menachem Commercial Orchard, a controlled orchard with three continuous N fertigation treatments ranging from 70 – 570 kg-N ha; and (3) Kabri Orchard, a mature commercial avocado orchard (~10 ha) with standard N application (270 kg-N Ha). Key eco-physiological parameters were measured on the ground bi-monthly in tandem with UAVs during 2022-2023. The ground measurements included nutrient uptake (in the lysimetric experiment only), LNC, stomatal conductance, chlorophyll fluorescence, and growth rates. Yield, fruit quality, and shelf-life were assessed at harvest and after cold storage and shelf life. UAV flights integrated multispectral, thermal, and LiDAR sensors. Canopy metrics, such as height, volume, PAIwere estimated on the based segmentation model, and spectral indices related to vegetative growth, were extracted and incorporated into a CNC prediction model and used to evaluate spatial-temporal variability. Random forest models demonstrated high predictive accuracy on CNC (R² = 0.94, RMSE = 0.4 kg per tree). Results revealed significant site-specific leaf nitrogen responses, with notable NUE differences (p-value < 0.05). A strong negative correlation was observed between NUE and PAI (p-value < 0.05), underscoring the importance of precise nitrogen management. This research highlights the potential of advanced remote sensing and site-specific fertilization strategies to optimize precise N use, enhance avocado yield and quality, and mitigate environmental impacts. By addressing spatial variability and leveraging predictive RS modelling, this study contributes to sustainable avocado farming practices.