Cork oak (Quercus suber) as a West Mediterranean species is known for its ecological, economic and social values. Wildfires are one of the most serious problems threatening Quercus suber, endangering its occurrence in its area of distribution. Therefore, knowing the behavior of the species after fire and the factors influencing its responses are particularly important for forest management. In this study we assessed the post fire vegetative recovery in 730 trees affected by wildfires on 2014 in Kiadi cork oak forest, located in the Western side of Akfadou Mountains in Algeria. Few months after the fire, individual tree mortality was very low (7.53%), and nearly, all the trees sampled survived the fire since almost all trees resprouted from canopy and some of them showed basal resprouts. Moreover, those two modes of post fire vegetative recovery were not correlated to each other. The performed redundancy analyzes (RDA) revealed that the cork oak post-fire response was highly correlated with individual characteristics and with the environmental data. The main variables influencing the likelihood of good or poor vegetative recovery were the understory height and cover, soil characteristics, fire severity, tree status (alive/dead trees), tree diameter and tree exploitation. Our results confirmed the fire resistance of cork oak species; which is also the only Algerian tree to resprouts. Hence, this makes the species a good candidate for reforestation programs in fire prone ecosystems.

Obtaining accurate forest cover information and dynamics of land occupation, through time, such as the spatial extent and pattern of disturbance and recovery is essential knowledge and assistance for forest managers and a crucial basis for the protection and conservation of current forest resources. Because most recent researches have focused on forest field survey and monitoring, a land classification containing information on forest cover dynamics is critically needed. Over the last decades, advances in remote sensing technology have enabled an accurate classification of different land covers from several sensors and remotely sensed data. We presently retained Tikjda forest (Djurdjura southerner, Algeria) as a case study to investigate the possibility of aerial photos classification and to analyze the historical dynamics of the area using a change detection analysis of multi-temporal data. To classify the study area’s main cover types, we used photographs collected over a period of 34 years (i.e., from 1983 to 2017). The results revealed that in 2017, Tikjda forest was composed of forest areas (24.1%), degraded areas (49.7%), and barren areas (26.2%). Throughout the investigated period, the analysis revealed a notable increase in barren areas (+9.8%), and degraded areas (+14.4%), While forest areas experienced a significant decrease (−24.2%). Moreover, the results confirm the potential of aerial photographs for an accurate classification of forests.