Agro-hydrological models and field measurements to assess the water status of a citrus orchard irrigated with micro-sprinkler and subsurface drip systems

Compared to the micro-sprinkler irrigation, traditionally used in citrus orchards, subsurface drip systems (SDS) allow increasing the water use efficiency (WUE); when coupled with water-saving strategies, like regulated deficit irrigation (RDI), further increase of WUE are possible. Combining measurements of soil water content (SWC) and weather data with measurements of midday stem water potential (MSWP) makes it possible to identify irrigation scheduling parameters for the RDI. However, measurements of MSWP are destructive and time-consuming, and also require skilled operators. For all these reasons, the use of the agro-hydrological models, such as the FAO-56 model, can be considered a surrogate aimed at the indirect evaluation of soil and crop water status. Objective of this work was to assess the potential of the FAO-56 model to predict soil and crop water status of a citrus orchard under subsurface drip and traditional micro-sprinklers and different irrigation strategies, as well as to identify the model outputs that can be used for irrigation scheduling purposes. Experiments were carried out in a citrus orchard during 2018 and 2019. The field was irrigated with micro-sprinkler and SDS. In the latter, two different irrigation strategies, i.e. full and regulated deficit irrigation, were applied. A standard weather station and probes to measure SWC were installed in the orchard. Additionally, Predawn Leaf Water Potential and Midday Stem Water Potential were monitored. After a site-specific calibration, the predicted average SWC fitted fairly well with the corresponding measured in the layer 0-50cm, with root mean square errors (RMSE) lower than 0.028 m3/m3. The performance of the model to identify the crop response to soil water deficit was also assessed by considering the observed similarity between the temporal dynamic of simulated crop water stress coefficient, Ks, with the measured MSWP. Even extending the analysis, the model was able to estimate the seasonal water stress integrals characterizing the examined treatments, demonstrating that even this variable can be used for irrigation scheduling purpose.