Shallow lakes and ponds represent a high proportion of aquatic ecosystems and, despite being highly managed, have been demonstrated to be hot spots of biodiversity. However, aquatic plants are potentially vulnerable to human activities and global change through several factors like eutrophication, landscape fragmentation, fish farming and global changes. We present in this paper a starting project, linking an ecology laboratory and a computer sciences laboratory, and aiming at the conservation of biodiversity in the context of ponds. The objective of species safeguarding must be achieved by considering the efficiency of the conservation actions peculiar to the concerned species, but also by taking into account the economical functions of ecosystems and the limited dedicated ressources. We propose to address this key issue by implementing an approach merging the “ecosystem-based” constraints extracted from a low-level individual-based model, and the “human-based” constraints as expressed by conservation instances. These unified constraints, linking parameters, variables and functions to be optimized, will then be solved by a dedicated engine using existing constraint programming languages.