Connectivity between MPAs assessed by the temporal Lagrangian flow networks: A case study in a semi-enclosed sea

• Dynamic LFN method efficiently assesses connectivity between Bohai Sea MPAs. • Only 9% of MPAs connect within three months reveals low connectivity. • Four MPA clusters were identified, with key hubs driving regional connectivity. The connectivity between Marine Protected Areas (MPA), defined as the migration and exchange of materials such as individuals or organisms, is critical for species dispersal and ecosystem recovery. In this study, the complex network framework is introduced to measure the connectivity between 22 MPAs in the Bohai Sea. A time-varying Lagrangian flow network is constructed based on particle transport. Subsequently, Dijkstra’s algorithm is utilized to identify the most probable paths (MPP) within the MPAs. The results indicate that the overall connectivity among the MPAs in the Bohai Sea is weak, with specific areas forming tightly connected clusters. The connection time between MPAs ranges from 10 to 70 weeks, and only 9 % of MPAs achieve connection within three months, which exceeds the survival time of most species’ larval stages. Louvain network clustering analysis reveals four main MPA communities in the Bohai Sea: Liaodong Bay, Qinhuangdao Coast, Bohai Bay, and Laizhou Bay, which are closely connected internally but sparsely connected to other communities. Four MPAs in Daling River estuary, Beidaihe marine park, Yellow River estuary, and Laizhou Bay shoal national marine ecological special protected area are identified as hub nodes, connecting the most of other MPAs in the community, acting as critical nodes for maintaining connectivity within the protected areas. Overall, the connectivity between MPAs is largely influenced by the Lagrangian residual currents in the Bohai Sea, with the central anticyclonic eddy and the western coastal current playing significant roles in facilitating MPA connectivity. This study offers a transferable framework for assessing MPA connectivity and provides guidance for optimizing MPA network design.