levels of both canopy closure and shrub cover, which in contrast hampered pine recruitment and the established oaks (particularly Q. ilex) to reach more advanced developmental stages. We also show that the occurrence of large stand-replacing fires triggers important vegetation changes from pinewoods to other types of woody vegetation (shrublands or forests dominated by resprouting tree species). At landscape level, the nature of these changes does not only depend on fire severity and topographic variables, but also on fire legacies and the characteristics of the pre-fire vegetation. Finally, our results support the feasibility of planting or sowing late-successional species for diversifying sub-Mediterranean pine forests when the process does not occur naturally. In this regard, our findings advise assisted migration programs to manage risks by thoroughly considering species and provenances selection, thermal migration distances and the occurrence of extreme cold events in the planting site. We also show that negative responses associated to these events could be partially buffered by sowing and planting under relatively closed canopy.