Tetrahymena thermophila is an ideal species for experiments with microcosms on the determinants of viability, for several reasons in addition to the numerous advantages due to its ease of culture. Even if few studies have focused on ecological and evolutionary aspects, this species is well known from a genetic point of view because it has been used as a model organisms in microbiology for a long time. The genetic information available is relatively important and, because genetically distinct strains can be maintained -sexual reproduction is impossible between individuals of the same strain because they carry the same mating type-, studies of the genetic basis of determinants of viability are possible. Furthermore, T. thermophila allows the direct experimental study of dispersal processes (and not only patterns), a phenomenon playing a major role in (meta)population dynamics and viability. Owing to the species ability to produce morphologically distinct dispersal morphs, it is indeed possible to study dispersal propensity, and not only the result of a sucessful dispersal as in the vast majority of the species. Finally, this species is also appropriate for the study of the social aspects underlying the determinants of viability, because it shows a form of social aggregation et secretion of factors improving the survival of congeners. Selection pressures and constraints acting on life history traits are numerous, and this creates an interdependcy between life history traits, particularly as there is often a trade-off between them, e.g. between investiment in dispersal and in reproduction. The determinants of viability therefore show synergistic effects, very difficult to study in the «real world». The objective of our research is to use T. thermophila to jointly study the two determinants of viability studied on butterfly systems: dispersal and habitat quality. This species indeed shows differeces between strains (genetic clones) at the level of dispersal strategies and their trade-off with demographic performances: some strains are specialised in long-distance dispersal whereas other strains disperse at shorter distances but present better performances in terms of colonization ability and growth rate. Metapopulation systems, built with tubes (habitat patches) linked by pipes (dispersal corridors), allow us to study in detail how the viability depends on: (1) demographic and dispersal strategies by comparing the strains in similar conditions; (2) landscape structure by varying number and spatial configuration of habitat patches and dispersal corridors; (3) habitat quality and its temporal variability (environmental stochasticity) by altering the abundance of food.