As an emerging approach in biodiversity conservation, ecological conservation networks are increasingly used to represent spatial connectivity among habitats and to support ecosystem-level management and sustainability. However, current conservation networks studies rarely account for migratory marine fish species and how seasonal migration reshapes functional connectivity and ecological network structure remains poorly understood. Developing dynamic, cross-seasonal ecological conservation networks can facilitate more effective protection strategies aligned with species’ life-history characteristics. In this study, the small yellow croaker (Larimichthys polyactis), an economically and ecologically important species in China, was selected as a model organism. Based on offshore fishing data from 1971 to 1982, a random forest model was applied to predict its winter and spring distribution. Cross-seasonal resistance surfaces were constructed, and the least- cost path and circuit theory approaches were integrated to build a dynamic ecological conservation network. The resulting network identified 16 ecological corridors and 7 core conservation areas, revealing clear migration pathways and habitat use patterns between wintering and spawning grounds. The cross-seasonal ecological protection network captured both spatial and temporal dimensions of ecological connectivity, advancing traditional static network approaches. The proposed conservation network can support conservation planning for economically important fish species. It can also inform adaptive fisheries management and support sustainable marine ecosystem development.