In this thesis we tested the predictions from the CSR theory for the community membership. Predictions are that, in the absence of competition, species from all strategies (Competitors, Ruderals and Stress-tolerants) will persist in low-stress habitats, whereas in high-stress habitats, only species with Stress-tolerant traits will survive. CSR recognizes that species evolved similar traits to one universal stress. For this reason, we were interested in testing which strategies will survive in different sources of stress. Our results from field and greenhouse experiments suggest that CSR theory does not predict community membership from the initial stages of a plant life-cycle. Instead, we found that the habitat stress plays a major role in determining the species that are incorporated into a community. In this thesis we also used a trait-based approach to evaluate: 1) the relationship between key traits using annuals species, and 2) the links between genome size and phenotypic variation within species. Firstly, we studied the triangular relationship reported for seed mass and leaf area in woody species. These traits are involved in the plant reproduction strategy and plant water and energy-use. We found a triangular relationship in annuals species, suggesting that is conserved across groups (woody and annuals). We also found that the driver of this relationship is related to soil fertility. Finally, for the relationship between genome size and phenotypic variation within species, we found that larger genome species display higher variation in traits than small genome species. This can be a potential advantage in heterogeneous environments where the amount of phenotypic variation would allow the species to adapt to them.