My thesis aims to design and evaluate an in-silico computer method to select one or more combinations of fish species that are compatible and complementary. This method aims to promote an integrated multi-trophic aquaculture production model that is environmentally friendly, and fostering biodiversity conservation while improving its eficiency. The identification of the selected combination(s) of species will be conducted in Nancy in collaboration with the L2A team (formerly UR-AFPA), and then tested in ponds in Cambodia.