Structural evolution, block differentiation, and driving factors of the spatial network of fishery economic resilience in China

China’s fisheries economy has long faced multiple external shocks, and enhancing its resilience is crucial to achieving sustainable development. Based on panel data from 27 provincial-level administrative regions in China from 2013 to 2023 (Tibet, Qinghai, Gansu, and Ningxia were excluded due to missing data), this paper constructs a fishery economic resilience (FER) evaluation system, and comprehensively employs the entropy-weight TOPSIS method, the modified gravity model, social network analysis, and QAP regression to examine the structural evolution and driving factors of the resilience spatial network. The findings are as follows: (1) The resilience level shows a steady upward trend, but the “high in the east and low in the west” gradient pattern has long been solidified, with the eastern regional mean approximately 1.8 times that of the western regions, and the regional disparity shows no significant sign of convergence (it should be noted that the exclusion of the four western provinces may render the actual disparity more pronounced). (2) The spatial association network has always remained fully connected, with the network hierarchy degree fluctuating from 0.207 down to 0.143. Although the network density and the number of relationships have slightly decreased, the overall evolution exhibits a trend of “hierarchy weakening and structure optimizing,” and a dominant pattern with eastern provinces as the core and central and western regions as the periphery has taken shape; the block model further reveals a clear direction of resilience factor flows converging from the central and western regions toward the east. (3) QAP regression shows that the adjusted R² of the model ranges from 0.056 to 0.092 (all passing the 1% significance test), among which geographic distance is the most stable spatial inhibiting factor (p<0.01), and the influence of the proportion of feed and seedling costs has increased from marginally significant to highly significant (p-value decreasing from 0.073 to 0.007); the influence of aquatic technology promotion has gradually strengthened (p-value decreasing from 0.245 to 0.087); the intensity of fiscal support for agriculture shows a trend of gradual withdrawal (p-value increasing from 0.065 to 0.435); per capita GDP is only significant in the medium term, exhibiting a phased pattern (p-value decreasing from 0.175 to 0.009 and then rising to 0.160). The above findings indicate that enhancing national FER requires moving beyond localized governance and building a regional coordination mechanism anchored in the spatial association network to systematically narrow the spatial resilience gap by strengthening technology spillovers, smoothing factor linkages, and improving data monitoring.