Uncovering the community structure hidden in networks is crucial for understanding the function of networks. In this paper, an algorithm of biogeography-based optimization with consensus dynamics for community detection is proposed. The problems of seeking community structures in networks are exquisitely embedded into the framework of biogeography-based optimization. Hence the community structure unveiled in such an evolutionary and global manner is corresponding to the habitat with maximum modularity. We present a dynamical framework for generating initial distribution of solutions for the evolutionary process using consensus dynamics, which gives a reasonably good estimate of the community structure based on the topological information. Thereof, the proposed dynamical method of initialization promotes the efficiency of optimal solution search significantly, compared with the traditional random initialization. Then, the obtained partition is refined using biogeography-based optimization. In addition, a preferential selection strategy for generating the new solutions is developed based on local network topology. Furthermore, we also proposed an adaptive mutation operator that enhances the exploration ability of our evolutionary algorithm. The experimental results on both artificial random and real-world networks indicate the effectiveness and reliability of our algorithm. These findings shed new light on the role played by topological knowledge of networks extracted from consensus dynamics in the evolving optimization processes when finding complex mesoscale structures in networks such as community structure.