This study proposes a framework for constructing and optimizing ecological security patterns (ESP) by integrating patch stability and network connectivity. The ESP of Ningxia, an ecologically vulnerable region, consists of 71 ecological sources covering 10,970.25 km² and 150 ecological corridors covering 3,950.88 km². Ecological sources and corridors along the Yellow River have high patch stability but low network connectivity, while the largest ecological source has high values for both. Removing nodes and edges reduces connectivity robustness, global efficiency, and equivalent connectivity. The variation trends of these indicators under different removal scenarios are inconsistent, indicating a trade-off between patch stability and network connectivity conservation objectives. The study highlights the importance of balancing different conservation objectives in landscape planning. The proposed framework provides guidance for conservation planners to construct and optimize ESP without losing information due to element abstraction in network analysis. The framework uses ecosystem health, human footprint, and circuit model to identify ecological sources and corridors. Land use conflict analysis and node/edge removal methods are used to assess patch stability and network connectivity. The results show that the ESP is composed of 71 ecological sources and 150 ecological corridors, with the largest ecological source having the highest values for both patch stability and network connectivity. The study also identifies the need to consider the internal structure of ESP to ensure effective conservation. The results suggest that the ESP should consider both the contribution to network connectivity and the spatial configuration of ecological sources and corridors. The study concludes that the proposed framework can help optimize ESP by integrating patch stability and network connectivity, providing a more effective solution for ecological security.This study proposes a framework for constructing and optimizing ecological security patterns (ESP) by integrating patch stability and network connectivity. The ESP of Ningxia, an ecologically vulnerable region, consists of 71 ecological sources covering 10,970.25 km² and 150 ecological corridors covering 3,950.88 km². Ecological sources and corridors along the Yellow River have high patch stability but low network connectivity, while the largest ecological source has high values for both. Removing nodes and edges reduces connectivity robustness, global efficiency, and equivalent connectivity. The variation trends of these indicators under different removal scenarios are inconsistent, indicating a trade-off between patch stability and network connectivity conservation objectives. The study highlights the importance of balancing different conservation objectives in landscape planning. The proposed framework provides guidance for conservation planners to construct and optimize ESP without losing information due to element abstraction in network analysis. The framework uses ecosystem health, human footprint, and circuit model to identify ecological sources and corridors. Land use conflict analysis and node/edge removal methods are used to assess patch stability and network connectivity. The results show that the ESP is composed of 71 ecological sources and 150 ecological corridors, with the largest ecological source having the highest values for both patch stability and network connectivity. The study also identifies the need to consider the internal structure of ESP to ensure effective conservation. The results suggest that the ESP should consider both the contribution to network connectivity and the spatial configuration of ecological sources and corridors. The study concludes that the proposed framework can help optimize ESP by integrating patch stability and network connectivity, providing a more effective solution for ecological security.