Are plant functional types relevant to describe degradation in arid, southern Tunisian steppes?

Abstract. In the Tunisian arid zone disturbances (e.g. overgrazing and agriculture) and stresses (e.g. aridity, low fertility) drive changes in the structure and functioning of rangelands, with a decrease in perennial plant cover, changes in floristic composition and erosion. Long-term monitoring requires (1) an understanding of the dynamics of vegetation change and associated ecological processes and (2) identification of relevant indicators. Using data from the arid zone of southern Tunisia we tested the hypothesis that plant functional response types could be used to address these two goals. We identified plant functional response types in response to a gradient of soil and vegetation types characterized by changes in perennial plant cover, dominant species and associated soil types. Vegetation samples were stratified by contrasted vegetation patch types with varying perennial plant cover (1.6 to 22%). We focused our analysis of trait responses within dwarf–shrubs, which are the dominants in typical steppe ecosystems of south Tunisia. Available trait data concerned morphology (plant height, leaf type), regeneration (dispersal mode, phenology and regeneration mode) and grazing value. Although we found it difficult to recognize ‘indicator response types’ that could be used directly to monitor changes in community composition, we were able to identify plant response syndromes that are relevant to long-term vegetation changes, and in particular degradation processes, in the region. Two main response types were identified: the decreaser type, made up of small or medium chamaephytes with high grazing palatability and the increaser type with medium to tall chamaephytes and low grazing palatability. These response types are proposed as key elements in a state-and-transition model of vegetation dynamics in the context of agropastoral disturbances and climatic and edaphic stresses.