Reconsidering the optimization of watershed landscape pattern based on the enhancement of ecological network resilience.

River basins face severe landscape fragmentation from urban expansion, necessitating highly resilient ecological networks (ENs) that coordinate ecosystems across administrative scales. However, current EN resilience assessments rely heavily on static, one-dimensional indicators and lack optimization validation. Using the Qiantang River Basin, this study develops a systematic "identification-assessment-optimization-validation" framework. We integrate structural and functional indicators for static resilience and employ complex network modeling to simulate dynamic resilience under random and targeted attacks. Static resilience characterizes the inherent structural and functional condition of the network at each time point, whereas dynamic resilience is evaluated through changes in network performance during progressive node removal. The results show that static resilience continuously declined from 2000 to 2020, while the network collapse threshold under random attacks increased from 20 to 26%. This divergence indicates that deterioration in the baseline network condition can coexist with increased tolerance to stochastic node loss, highlighting the complementary value of static and dynamic resilience assessments. Targeted attacks further identified critical ecological sources that play important roles in maintaining network resilience. After ecological stepping stones were added and differentiated buffer zones were established, reassessment showed that the optimized network exhibited improved resilience. By linking ecological network identification, resilience diagnosis, spatial optimization, and effectiveness validation, the proposed framework provides a practical basis for watershed ecological conservation and spatial management.