ALIAS: A Remote Sensing Approach to Monitor Ailanthus altissima Invasion and its Ecohydrological Impacts

Biodiversity loss is a growing threat to natural ecosystems, driven by a combination of anthropogenic and natural factors (e.g., urbanization, deforestation, and, notably, climate change). Such factors can alter wildfire regimes, with the possible consequent creation of more available space for the establishment of invasive alien species. These, often highly adaptable and with rapid growth capabilities, can profoundly alter local ecosystems, disrupt hydrological processes, and reduce native biodiversity. Among the most concerning invasive species, Ailanthus altissima has rapidly spread across the globe. Ailanthus is distinguished by its ability to adapt to a wide range of environmental conditions. It is resilient to extreme temperatures, able to grow on various soil types, and tolerant of high levels of air pollution, making it adapted also to disturbed/degraded environments. The species can regenerate even when it is cut or burned. Seed dispersal occurs through wind, but also via water, animals, and humans. Ailanthus demonstrated a strong dependence on water availability, employing deep root systems and efficient water uptake strategies to thrive in waterlimited environments. This exacerbates competition with native species, particularly in regions under hydric stress, where Ailanthus can monopolize water resources and disrupt local ecohydrological balance, such in the case of Mediterranean ecosystems. This work presents ALIAS (Ailanthus Locator and Identification Algorithm Suite), a machine learning-based classifier based on the Support Vector Machine (SVM) model, that uses highresolution PlanetScope satellite imagery, designed to enable accurate remote detection of Ailanthus in specific areas of interest. ALIAS was calibrated by focusing on the presence of Ailanthus along transportation corridors, where species frequently establishes itself due to the wind generated by vehicles facilitating seed dispersal, and in hydrologically connected areas, such as riparian zones. Validation was conducted in an area with a confirmed invasion, i.e., the “Vallone Piano della Corte” Nature Reserve (Sicily, Italy). It represents a sensitive site where local biodiversity and water resources are threatened by dense clusters of Ailanthus. Over the past four decades, the species has progressively expanded, creating populations that competed with native plants and disrupted the natural ecosystem balance. Particularly, four distinct clusters of Biodiversity loss is a growing threat to natural ecosystems, driven by a combination of anthropogenic and natural factors (e.g., urbanization, deforestation, and, notably, climate change). Such factors can alter wildfire regimes, with the possible consequent creation of more available space for the establishment of invasive alien species. These, often highly adaptable and with rapid growth capabilities, can profoundly alter local ecosystems, disrupt hydrological processes, and reduce native biodiversity. Among the most concerning invasive species, Ailanthus altissima has rapidly spread across the globe. Ailanthus is distinguished by its ability to adapt to a wide range of environmental conditions. It is resilient to extreme temperatures, able to grow on various soil types, and tolerant of high levels of air pollution, making it adapted also to disturbed/degraded environments. The species can regenerate even when it is cut or burned. Seed dispersal occurs through wind, but also via water, animals, and humans. Ailanthus demonstrated a strong dependence on water availability, employing deep root systems and efficient water uptake strategies to thrive in waterlimited environments. This exacerbates competition with native species, particularly in regions under hydric stress, where Ailanthus can monopolize water resources and disrupt local ecohydrological balance, such in the case of Mediterranean ecosystems. This work presents ALIAS (Ailanthus Locator and Identification Algorithm Suite), a machine learning-based classifier based on the Support Vector Machine (SVM) model, that uses highresolution PlanetScope satellite imagery, designed to enable accurate remote detection of Ailanthus in specific areas of interest. ALIAS was calibrated by focusing on the presence of Ailanthus along transportation corridors, where species frequently establishes itself due to the wind generated by vehicles facilitating seed dispersal, and in hydrologically connected areas, such as riparian zones. Validation was conducted in an area with a confirmed invasion, i.e., the “Vallone Piano della Corte” Nature Reserve (Sicily, Italy). It represents a sensitive site where local biodiversity and water resources are threatened by dense clusters of Ailanthus. Over the past four decades, the species has progressively expanded, creating populations that competed with native plants and disrupted the natural ecosystem balance. Particularly, four distinct clusters of Ailanthus were identified on the south-facing slope of the site. In contrast, the north-facing slope hosts native flora, i.e., Quercus pubescens forest stands. A diachronic analysis was also performed, reconstructing the invasion of Ailanthus from the late 1980s and performing field surveys with drone acquisitions to obtain the current distribution and the area of invasion. These historical insights were critical for validating ALIAS and demonstrating its reliability. The results obtained highlight classifier’s potential as a predictive tool for identifying regions at high risk of invasion, particularly in hydrologically sensitive areas. By enabling efficient monitoring of Ailanthus in both confirmed and potentially at-risk zones, ALIAS provides critical insights into the ecohydrological dynamics of invasive species. Furthermore, the classified images resulting from the use of the classifier form the basis for validating vegetation dynamics models (e.g., CATGraSS model), able to reconstruct invasion dynamics and to predict the future expansion of Ailanthus under different climate change and hydrological scenarios.