Planificación de estrategias de riego deficitario en cultivos leñosos ante distintos escenarios de disponibilidad de agua de riego

Abstract

The Region of Murcia, like the rest of the regions located in the Mediterranean basin, is under a worrying scarcity and irregularity of rainfall, which, together with the high evaporative demand of the area, have caused a severe structural water deficit over the Segura River Basin. Likewise, short and long term climatic forecasts indicate that, due to climate change, this situation will become increasingly complex. In this context, the role played by irrigation communities becomes more relevant, given the enormous difficulty in adapting the water supply for irrigation, the quantity and quality, to the water needs of crops. In this way, farmers are forced to adopt irrigation strategies focused on water saving, with the least possible impact on crop yield, maximizing the efficiency of its use to ensure the economic and environmental sustainability of the sector. Therefore, the main objective of this PhD Thesis was the development and validation of regulated deficit irrigation (RDI) strategies in three of the most common woody crops in the Mediterranean area (table grape, flat peach and apricot), based on information from plant wáter status indicators and through predictive modeling, so that these sustainable protocols obtained at field scale can be extrapolated to a larger irrigable area. In table grapes, the RDI strategies applied during post-veraison caused a maximum water stress of 0.22 MPa of ΨS compared to well-irrigated vines, establishing an irrigation protocol for a range of water availability for irrigation between 4000 and 7000 m3 ha-1 that did not affect yield. In this way, water use efficiency (WUE) increased by 50% during the first two years and around 81% during the third, even improving berry quality. In addition, with the data collected during the study period, it was possible to develop a predictive model using the supervised learning algorithm of gaussian process regression (GPR), to predict the irrigation requirements of the crop for nonlimiting irrigation conditions and for RDI, according to two easily determined variables such as daily maximum temperature and day of the year. In the same way, using data from 7 consecutive crop cycles in which different RDI strategies were applied in table grapes during post-veraison, saving an average of 40% of irrigation water, it was possible to predict the effect of water stress intensity, based on ΨS values, and the number of clusters per vine on berry quality, determined as their firmness and color, using GPR. From these results, it was possible to determine that WUE increases linearly up to an accumulation of the wáter stress integral of up to 30 MPa day, improving the color and firmness of the berries. In flat peach, by applying RDI during the non-critical periods to water deficit: phase II of rapid fruit growth and late post-harvest, in which irrigation scheduling was carried out to satisfy 70% and 50% of crop evapotranspiration (ETC), it was possible to achieve weekly irrigation reduction of around 43 and 109 m3 ha-1 in phase II and late post-harvest, respectively. This represented a water saving of 33.6% in the RDI treatment compared to well-irrigated trees, and an increase of around 45% in WUE, reaching values of around 4.16 kg m-3. Likewise, models were developed for irrigation scheduling based on crop phenology and climatic parameters based on irrigation thresholds according to ΨS for well-irrigated trees and for trees with deficit irrigation. Finally, for trees under RDI and without water restrictions, the threshold value of ΨS during post-harvest was -1.4 and -1.1 MPa, respectively.In apricot, two RDI strategies were applied during the phenological periods not sensitive to irrigation reduction with different water stress intensities; during phases I and II of fruit growth, irrigation was reduced by 20% compared to well-irrigated trees, and during late post-harvest, two threshold values of ΨS were considered for the start of irrigation, the first, corresponding to a moderate water stress of -1.5 MPa (RDI1) and on the other hand, a severe stress of -2.0 MPa (RDI2). The scheduling of deficit irrigation in both treatments did not affect the yield or the quality of the fruit at harvest, reducing the contributions of irrigation water by 1124 and 2133 m3 ha-1 and increasing WUE by 13.2% and 25.6%, for RDI1 and RDI2, respectively. Likewise, a water stress integral of 30.2 MPa day during post-harvest could be considered optimal since, when reaching around 41 MPa day in the RDI2 treatment, vegetative growth was reduced by 35%. Therefore, given the limited availability of water for irrigation, the high uncertainty faced by farmers, and the need to reduce pressure on water resources, deficit irrigation protocols have been developed and validated in three representative crops of the agriculture in Murcia. These protocols have significantly increased the irrigation water use efficiency in semi-arid conditions, taking into account: i) the phenology of the crop, for the more specific delimitation of non-critical periods in each of them; ii) an adequate level of water stress that does not negatively affect crop yield or vegetative growth, as it could affect its future production; and iii) the obtaining and validation of predictive models in order to extrapolate the irrigation protocols to a larger irrigable area.