Abstract

Water stress periods are commonly used in olive orchards, specifically in regulated deficit irrigation (RDI) to improve oil quality. However, unplanned changes in the plant water status could alter a number of physiological processes, inducing reductions in yields, oil accumulation, vegetative growth and flower induction. In commercial orchards it is difficult to monitor physiological processes for short periods and at specific moments. The aim of this research was to explore a fast method to improve the assessment of plant water status and gas exchange of olive trees through an increase in the amount of samples per plot. In this work absolute reflectance (between 350-2500 nm) was correlated with net CO2 assimilation rate (A) stomatal conductance (gs) and transpiration (E) in order to find a reliable mathematical model to represent these variables. The research was carried out in a commercial orchard Quepu S.A., Pencahue, Maule Region, Chile. Four levels of water were applied in a completely randomized design (CRD) with 4 replications per treatment (T0: farmer's standard irrigation; T1: 30-40 days without irrigation; T2: 50-60 days without irrigation; T3: 70-80 days without irrigation). Model calibration (canopy and branch) was done using partial least squares (PLS) regression with independent validation, performed by Unscramble software. The results showed that the calibration process had a high coefficient of determination (R-2) (0.64-0.79), however, the validation with data measurements on the canopy presented a lower R-2 (0.33-0.42) compared with branches (0.77-0.81). When data were divided into low and high water stress levels only A in branches showed a strong association in calibration and validation analysis.