Applying machine learning to the study of environmental dynamics and sustainable management of the Argan grove in the El Guerdane region (Souss plain, Morocco)

The forest ecosystem of the Argan grove in the region of El Guerdane, located at the heart of the plain of Souss, has experienced significant natural and anthropogenic pressure in recent decades, leading to observable changes in the landscape of the Argan grove. However, the nature and intensity of these changes remains unclear. This study aimed to draw up an inventory of the dynamics of environmental changes between 1985 and 2017 through the diachronic exploitation of Landsat satellite images. It appears that the support vector machine (SVM) algorithm is an appropriate approach to highlight the evolution of this environment. It helped to identify the changes that occurred in 32 years and understand that the extension of the built environment and modern takes place at the expense of the argan tree stands. Thus, the loss of about 40.60% of the Argan forest area in this region during this period has negative impacts on biodiversity, the environment, the economy, and cultural and regional history; therefore, a quick and effective intervention could be a positive action for the sustainable development of this unique forest area.