This paper presents a novel optimization algorithm that we call the particle swarms swarm optimizer ((PSO)-O-2), which based on a hierarchical coevolution model (HCO model) of symbiotic species. HCO model introduced a number of M species each possesses a number of N individuals to represent the "biological community". Both the heterogeneous coevolution and the homogeneous coevolution aspects are simulated in this model to maintain the community biodiversity. This strategy enable the symbiotic species find the optima faster and discourage premature convergence effectively. The experiments compare the performance of (PSO)-O-2 with the canonical PSO, the fully informed particle swarm (FIPS), the unified particle swarm (UPSO) and the Fitness-Distance-Ratio based PSO (FDR-PSO) on a set of 6 benchmark functions. The simulation results show the (PSO)-O-2 algorithm markedly outperforms the four mentioned algorithms on all benchmark functions and has the potential to solve the complex problems with high dimensionality.