Biological invasions, the second major threat to biodiversity, pose significant challenges to conservation management and eco-evolutionary research. Even though invasion processes have been studied for more than 150 years, our capacity to predict their presence today and in the future is still rudimentary. This deficiency stems mainly from the difficulty involved in reliably assessing the ecological niche of an invader, i.e. those environmental and biotic conditions that allow the species to maintain viable populations. In particular, disentangling the abiotic and biotic components of the ecological niche and accounting for their changing over space and time due to evolutionary dynamics is difficult, albeit crucial for the quality of predictions. The main objective of my PhD has been to address these challenges by improving methodological approaches of niche estimation, advancing our understanding of the role of biotic interactions for invasion processes and studying in greater detail how evolution may affect spatio-temporal niche dynamics. More precisely, (1) with a comprehensive literature review, I started by describing the limits of the different modelling approaches usually applied to predict invasive species distributions. (2) Then, I provided a modelling framework for improving regional environmental niche estimations. (3) Thirdly, I focused on the identification of biotic interactions, and the methods commonly used to identify patterns of symmetric competition in ecological communities. I also implemented a simulation model of community assembly to test the efficiency of these methods. (4) In a fourth part, I studied invaded alpine plant communities and showed that characteristics of the biotic environment in these communities (e.g. symmetric vs. asymmetric competition) were good predictors of invaders' presence. (5) Finally, I provided a first example of a genetic-based, climatic niche expansion of the invasive weed Ambrosia artemisiifolia L. in the French Alps by combining information on its environmental niche, genetic structure and functional traits. Taken together, the results of these studies highlight how tightly the different facets of invasion ecology and evolution are interrelated and open the way to an integrated modelling approach that would advance both eco-evolutionary research on invasion dynamics and applied tools for biodiversity protection.