Impact of changes in river network structure on hydrological connectivity of watersheds

River network connectivity is a key factor affecting watershed ecosystems and an important criterion for evaluating the health of river. The river network structure of the basin has changed dramatically in recent years due to human activity, but the response of connectivity to changes in physical structure is still poorly understood. In this paper, a dynamic index (connectivity index) based on daily-scale flow and probability density function is proposed to characterize the connectivity capacity of river networks. The Haihe River basin in China is selected as the study area, and a graphic model is constructed to assess the evolutionary structure of the river network. The 54 sections were classified into Low human impact (LHI) and High human impact (HHI), and the trends and causes of connectivity changes of the sections with the evolution of the network were analyzed. The results reveal that the river network degree and clustering coefficient keep increasing, the path length decreases and becomes more dense downstream, and the river network gradually evolves into a “small-world” network. The “small-world” characteristics improve the potential efficiency and specific functionality of the river network, but make the physical structure mismatch with the hydraulic gradient, and the network flow disperses severely, leading to a significant decrease in connectivity (HHI connectivity index is lower than LHI 52.1%). Moreover, the structural changes enhance the spatial and temporal differences in hydrological connectivity, which further restricts the function of the river system. “Small-world” and connectivity indicators (duration, start time, and intermittency) show strong correlations. Therefore, physical structure is an important factor influencing connectivity capacity of river networks. The planning and management of the watershed should consider the changes in hydrological connectivity caused by the structural modifications of the river network.