Spatio-temporal analysis and identification of deforestation hotspots in the Moroccan western Rif

In the north of Morocco, forest areas are under threat mainly due to the expansion of agricultural land, forest fires and socio-demographic, economic and urban dynamics. Despite the magnitude of deforestation in this area, quantitative studies and investigations of its evolution in time and space are rare or non-existent. The present study aims to assess and quantify gross forest losses in the Western Rif between 2001 and 2020 through the exploitation of Global Forest Change (GFC) data developed by Hansen et al. (version 1.8) based on high resolution Landsat TM satellite images. We used Local Indicators of Spatial Association (LISA) methods to identify statistically significant spatio-temporal trends in forest losses with ArcGis 10.8 and Geoda software. The main results showed that over the last two decades, the Western Rif has lost 8.39% of its forest heritage, a total area of 272 km2. The greatest forest losses were recorded in the provinces of Chefchaouen, Tetouan and Larache (70% of total losses), while the highest annual deforestation rates are reported in the prefectures of M'diq-Fnideq and Tangier-Assilah (0.9% and 0.89% respectively). Moreover, we observed that 81% of the deforested areas were recorded in the formations of the following types: matorral, cork oak stands and artificial pine forest. A first analysis of local Moran's spatial autocorrelation based on the communal division and on the variables relating to forest losses and gross deforestation rates revealed the existence of 3 significant clusters of the high-high (HH) type. A second optimized analysis of deforestation hotspot (Getis-Ord statistical test) based on a 0.5 km × 0.5 km grid identified 26 deforestation fronts in the study area. Finally, we conducted an emerging hotspot analysis to take into account the time dimension and we concluded that the oscillating deforestation model predominates in the Western Rif. These results constitute valuable decision elements to better guide, prioritize and rationalize the monitoring, conservation and restoration efforts of forest ecosystems.