Topological structure of the reservoir significantly impacts the network performance of Echo State Network (ESN). Synaptic plasticity learning rules influence the network performance of ESN by altering the topological structure of the reservoir, but this effect is unclear. To investigate this problem, five synaptic plasticity learning rules are applied to the ESN to change the topology of the reservoir. Through the experimental analysis, we explore the relationship between the change of the topology structure of the reservoir and the network performance. The experimental results show that in terms of predictive performance, anti-Oja rules perform well on nonlinear autoregressive moving average (NARMA) system and Oja rule perform well on Mackey-Glass system.