Résumé : Ecological restoration of degraded metal-rich habitats and bare metal-rich soils created by mining activities has nowadays become a major environmental issue to reduce mining impacts on the erosion of biodiversity and the degradation of landscape, soil and water.Studies based on species identification for restoration purposes without reference to plant trait are limited only to local pool of species which makes the interpretation and applications in various ecological contexts rather difficult. This study aims at trait analysis of metal-rich habitats in order to test the differences in plant traits in degraded primary habitat and along a successional gradient in secondary habitats. In this context, investigation of the existence of resilience or the need for ecological restoration is attempted. A second aim of this work is also the identification of traits candidates for appropriate species selection for revegetation of metal-rich bare soils purposes.Results showed that the cover of xylopodia trait values decreases according to the intensity of degradation among habitats, and can be considered an indicator of habitat degradation. Differences of traits values among primary and old secondary habitats highlighted that the absence of resilience of degraded primary steppic savanna was not explained by 14 soil factors measured.The main plant traits for revegetation of metal-rich bare soils are the following: annual life cycle, growth phenology in wet season, depth 0-10. cm of underground system. , bud bank by seeds, dispersule size <2. mm. ×. 2. mm and dispersal mode by adhesion.Future challenge would be to develop new functional ecosystem on metal-rich bare soils promoting biodiversity conservation and ecosystem services related to water and soil protection. This would require the association of species selected for revegetation and phytostabilization from plant traits occurring in secondary metalliferous habitats and species selected from plant traits specific to primary habitats.