Throughout the Mediterranean region, vegetation dynamics are affected by human activities which are either ‘stresses’ or ‘disturbances’, depending on their frequency, intensity and spatial distribution. To minimize or reduce anthropogenic degradation caused by land use and other disturbances, it is necessary to understand and predict the various responses of plant communities to disturbances. In particular, detailed but integrative approaches are required to assimilate large databases on vegetation and to make them directly useful for managers and restorers.

We describe two case studies undertaken to evaluate the effects of logging or overgrazing on plant species diversity in pine forests of southern France and steppe ecosystems of southern Tunisia. Both studies employed the same methodology to identify plant functional traits (morphological, life history and regeneration traits) associated with community response to disturbance. The results of these analyses allowed us to develop state and transition models that could be used to plan and predict ecosystem trajectories, assess ongoing degradation processes and monitor community and ecosystem responses to management and restoration practices. We discuss the relevance and the use of plant functional types (PFTs) as tools for ecosystem management and planning and for monitoring restoration in southern Europe, northern Africa and elsewhere. Using this approach it is possible to improve management strategies for the conservation, restoration and sustainable exploitation of biodiversity and of ecosystems.

Nomenclature: Le Floc'h & Boulos (In press); Tutin et al. (1964–1980).

Abbreviations: GLM = Generalized linear model; LDMC = Leaf dry matter content; PPC = Perennial plant cover; SLA = Specific leaf area.