Estimation of stem water potential in drip-irrigated peach trees

Identification of the plant water stress is essential when applying deficit irrigation strategies. Stem water potential (Psi(stem)) is accepted as one of the most accurate plant water status indicators. However, because its monitoring is tedious and labor consuming, alternative methods for Psi(stem) estimation are desirable. In this respect, the midday Psi(stem) of adult peach trees (Prunus persica Batsch 'Flordastar') submitted to different drip irrigation treatments in a long-term trial under Mediterranean conditions (Murcia, SE Spain) was fitted by linear regression analysis as a function of soil water content (SWC) and agro-meteorological variables, with the soil water content being the main contributor to estimate Psi(stem). A multiple linear regression equation, based on SWC in the soil profile (0-0.8 m), mean daily air vapour pressure deficit (VPDm) and growing degree hours (GDH) explained 72% of the variance in Psi(stem). A non-linear approach (hybrid soft computing techniques) allowed selection of the relevant inputs to Psi(stem) (SWC at 0.3 m, the day of the year, and mean daily air temperature) and identification of five fuzzy rules, yielding an accurate model (86% variance explained). The suitability of these approaches as an alternative to the field measurements of Psi(stem) is studied.