The thesis research concerned evolutionary ecology and community ecology of amphibians in two different areas of a Mediterranean region. The goal of the research was to test how anuran species coexist along a lentic freshwater gradient from ephemeral to permanent ponds. The community structure across the gradient has been explained by different ecological models, based on different trade-offs and inherent properties of the species. To test the different models he used four-year field surveys of communities to characterise pond-breeding habitats (with respect to temporariness, predator abundance and competitor abundance) and to evaluate the dynamics and spatial structures of the metacommunity. He also designed lab experiments to quantify species traits and phenotypic plasticity (life-history, morphological and behavioural traits) in response to different circumstances found in nature: pond drying, presence of invertebrate predators and intra- and interspecific competition. Comparative analysis of phenotypic plasticity traits was made in relation to species ecological breadths (quantified from field surveys) and phylogenetic relationships. In general, species that use a wide variety of habitats or unpredictable environments showed a greater plasticity of responses than those occurring in predictable habitats. At the two extremes of the hydroperiod (ephemeral and permanent ponds) there were specialists with limited plasticity, whereas species from intermediate temporary ponds showed higher levels of plasticity. Results therefore supported the hypothesis that interspecific differences in plasticity are adaptive and are related to ecological breadth and unpredictability of habitat. The correlations among traits of the different species reflected trade-offs suggested by the models (colonisation-competition; predator-permanence gradient; and competition ability-permanence gradient), but correlation coefficients did not favour any single trade-off model over the others. These results suggest that the community studied can be interpreted as a metacommunity in which local interactions and regional processes (colonisation-extinction rates) are related, and they emphasise the importance of habitat heterogeneity for both: local and regional diversity maintenance.