摘要:Growing external pressures from human activities and climate change can exacerbate desertification, compromising the livelihoods of more than 25% of the world’s population. The dryland mosaic is defined by land covers that do not behave similarly, and the identification of their recurring or irregular changes over time is crucial, especially in areas susceptible to become desertified. To this aim, the methodological approach of this research is based on the integration of non-linear data analysis techniques, such as recurrence plots (RPs) and recurrence quantification analysis (RQA), applied to the Enhanced Vegetation Index (EVI), which is a functional ecological proxy of above ground net primary production. The research exploits the recurring change detected in vegetation cover over time to gauge the predictable (resilient) behavior of the EVI as well as its chaoticity in a semi-arid Mediterranean region (Apulia, Italy). Interestingly, the results have shown the spatial rendering of recurrence variables, confirming the well-known hot spots of soil degradation and desertification taking place in the region, which are characterized by greater EVI chaoticity, but they have also identified new potential candidate sites. As a result, the susceptibility to land degradation, as measured by the EVI-RQA approach, can help in measuring land desertification with evident operational benefits for landscape planning. The novelty of the research lies in the spatially explicit identification of resilient and less resilient areas to desertification that can support the definition of more targeted interventions and conservation priorities for better planning and sustainable management of Mediterranean drylands.