A structure-process hydrological connectivity framework for estuarine wetland management: Case study of a typical...

Globally, sustaining hydrological connectivity is essential for estuarine wetland health and restoration, yet increasingly challenged by climate change and human impacts. Most existing assessments focus on a single dimension, thereby obscuring the dynamic interplay between physical structure and ecological processes during restoration. To address this gap, we developed a widely applicable dual structure-process framework. Structural connectivity was quantified using deep learning and multi-dimensional geomorphological and hydrological parameters to characterize tidal creek networks, while process-based connectivity was evaluated using a probability-based "source-sink" model of water and material exchange. Applying this framework to the Liaohe Estuary of Northern China revealed a marked asynchrony between the two connectivity types: restoration projects led to a steady 39 % increase in structural connectivity, yet process-based connectivity showed only an 11.2 % gain. Mechanistic analysis clarified that this decoupling arose from the inhibitory effect of geometric expansion-which lengthened flow paths and reduced exchange efficiency-contrasted with the non-linear benefits of topological optimization. Based on the combined structure-process states, we identified priority zones for protection and restoration. This transferable framework provides a robust basis for adaptive management in estuarine wetland restoration worldwide.