Ecological security pattern construction provides a systematic framework for coordinating ecological conservation, ecosystem restoration, territorial spatial governance, and sustainable regional development. Under the combined pressures of global environmental change and increasing human activities, ecosystems in arid and semi-arid regions have become increasingly vulnerable, resulting in ecological degradation, landscape fragmentation, biodiversity loss, and declining ecosystem resilience. Establishing a scientifically sound ecological security pattern has therefore become an essential prerequisite for maintaining regional ecological processes and enhancing ecosystem stability. However, existing studies generally identify ecological sources based solely on land cover types or ecological management units, which are often subjective and fail to adequately reflect the spatial heterogeneity of ecosystem functions and ecological background conditions. To overcome these limitations, this study takes the Ningxia Hui Autonomous Region, a representative arid and semi-arid region in northwestern China, as the study area and develops an ecological security pattern by integrating ecosystem service assessments with ecological management units. The objective is to improve both the scientific reliability and practical applicability of ecological source identification while providing technical support for ecological conservation, ecological restoration, and territorial spatial optimization.

Four ecosystem services closely associated with regional ecological security, namely soil conservation, windbreak and sand fixation, habitat maintenance, and carbon sequestration, were selected to comprehensively evaluate ecosystem service capacity. Soil conservation, habitat maintenance, and carbon sequestration were quantified using the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model, whereas windbreak and sand fixation were evaluated using the Revised Wind Erosion Equation (RWEQ). The outputs of the four ecosystem services were normalized, assigned equal weights, and classified into five levels using the natural breaks method. Ecological source areas were subsequently identified by integrating areas with medium or higher ecosystem service levels with existing nature reserves and ecological protection redlines, while small and isolated patches were removed to improve ecological integrity and spatial continuity. To characterize the resistance of ecological flows, an ecological resistance surface was established by incorporating both natural environmental conditions and anthropogenic disturbance factors. Seven resistance indicators were selected, including normalized difference vegetation index (NDVI), digital elevation model (DEM), slope, land cover type, population density, distance to roads, and nighttime light intensity, representing vegetation conditions, terrain constraints, landscape characteristics, and human disturbance intensity. The analytic hierarchy process (AHP) was employed to determine the relative importance of each resistance factor. All datasets were standardized and resampled to a spatial resolution of 1 km before integration to ensure spatial consistency. The minimum cumulative resistance (MCR) model was then used to generate the ecological resistance surface. Based on the identified ecological sources and resistance surface, ecological corridors, pinch points, and barriers were extracted using circuit theory implemented in the Linkage Mapper tool. Compared with conventional least-cost path methods, circuit theory simulates ecological flows as electrical currents moving through resistance networks, allowing multiple potential movement pathways to be identified simultaneously and providing a more realistic representation of regional ecological connectivity. Ecological corridors represent key pathways for species migration and ecological flows, pinch points indicate critical locations with concentrated ecological flows that are highly sensitive to external disturbances, and ecological barriers denote areas where ecological processes are substantially impeded and ecological restoration should be prioritized.

A total of 66 ecological source areas were identified, covering 6.90×10³ km² and accounting for 10.4% of the total study area. These ecological sources were mainly distributed in the Helan Mountains, Liupan Mountains, and other mountainous regions, exhibiting an obvious spatial pattern characterized by “more in peripheral areas and fewer in central areas.” A total of 169 ecological corridors with a combined length of 4.32×10³ k km were extracted. Among them, key ecological corridors accounted for 24.4% of the total corridor length and played an essential role in maintaining regional ecological connectivity and facilitating ecological flows. In addition, 75 ecological pinch points were identified, primarily surrounding mountainous ecological sources, indicating areas where ecological connectivity was highly vulnerable to external disturbances. Ecological barriers covered an area of 1.89×10⁴ km² and were widely distributed along ecological corridors, suggesting priority areas for ecological restoration and connectivity enhancement. By integrating ecological sources, ecological corridors, pinch points, and ecological barriers, an ecological security pattern characterized as “one axis, three belts, and three zones” was established. Specifically, the “one axis” refers to the ecological security axis along the Yellow River, which connects major ecological patches and promotes ecological flows; the “three belts” represent important ecological barrier zones formed by the Helan Mountains, Liupan Mountains, and central ecological transition areas; and the “three zones” comprise ecological conservation zones, ecological restoration zones, and ecological regulation zones with different ecological functions, management objectives, and restoration priorities.

Integrating ecosystem service assessments with ecological management units effectively improves the scientific reliability, ecological rationality, and practical applicability of ecological source identification by simultaneously considering ecosystem functions and existing ecological conservation policies. Compared with conventional methods based solely on land cover or management units, the proposed approach effectively avoids including ecologically degraded or low ecosystem service areas as ecological sources, thereby improving the accuracy and robustness of ecological security pattern construction. The ecological security pattern established in this study provides a comprehensive framework for identifying ecological conservation priorities, optimizing ecological networks, enhancing landscape connectivity, and supporting ecological restoration planning. The proposed methodology is readily transferable to other ecologically fragile regions, particularly arid and semi-arid areas experiencing similar environmental pressures, and provides valuable scientific support for territorial spatial planning, ecosystem management, biodiversity conservation, and sustainable regional development under the context of global environmental change.