Congestion in a network is determined by the resource constraints and the number of deployed sensor nodes. Congestion can significantly degrade the quality of services (QoS) in wireless sensor networks (WSNs) regarding throughput and end-to-end delay. In this paper, a hybrid bio-inspired algorithm is proposed for congestion control in large-scale WSNs. First, a competitive Lotka-Volterra (C-LV) model to avoid congestion is employed, while fairness among sensor nodes is maintained. Second, particles swarm optimization (PSO) is employed to enhance C-LV by optimizing the parameter for minimizing end-to-end delay. PSO makes this scheme adaptive to change. Simulation results verify that the proposed scheme improves the QoS in WSNs.