Currently there is a multitude of bio-inspired approaches, from which, result complex systems qualified as biomorphic. Today, the advantage of using such approaches is well established: genetic algorithms, ant algorithms, artificial immune systems, etc. However, there is no research supporting the integration of these different approaches. In particular, hybrid approaches integrating several bio-inspired analogies at the same time are rare. In this paper, we investigate to which extent it is possible to propose a design approach which can integrate different bio-inspired paradigms. From this perspective, we show the interest of considering the multi-agent paradigm as the common denominator of bio-inspired approaches and we propose a generic agent-based model for bio-inspired systems.