The construction of ecological security patterns has become a crucial approach to assessing ecosystem health and integrity and is essential for achieving sustainable and high-quality development in both urban and rural areas. Using Lanzhou City as an example, our study employed the InVEST model, ecological service importance evaluation, and hotspot analysis to classify ecological sources. Additionally, ecological corridors were identified and optimized using the least resistance model and circuit theory. The identified corridor pattern was further analyzed using space syntax and neural networks to determine the influences of various factors. This framework can be applied to the circular construction of corridors. Our findings revealed a three-stage differentiation trend in the importance of ecosystem services. Ecological source areas and corridors were densely distributed in the northwest of Lanzhou. The optimized ecological source area increased from 2914.1 km² to 4542.5 km², raising its proportion in the study area from 22.2% to 34.7%. The total number of corridors after optimization was 217, spanning a 2657.3 km length. The Gaolan Mountain area had the highest current density, whereas the ecological barrier area was mainly distributed in the northwest of Yongdeng County and the north of Yuzhong County. The spatial syntax index indicated significant potential reachability between the Honggu area and the northwest area. Finally, using neural network perceptrons to simulate ecosystem service functions, our findings revealed that habitat quality showed the best fit under single-dependent-variable prediction, followed by water yield, with soil conservation showing a poor fit. Under three-dependent-variable prediction conditions, population factors had the greatest impact on ecosystem services, while slope had the least impact. Therefore, it is important to promote the construction of green infrastructure in the northwest and southeast, improve the connectivity of ecological corridors in Honggu District, and adopt corresponding spatial corridor optimization strategies according to different ecological needs. Collectively, our findings provide a theoretical and practical basis for the construction and optimization of urban and rural ecological security patterns in the semi-arid region of Lanzhou.

Land management strategies play a pivotal role in the sustainable development of a region. Integrating space syntax into the ecological–social perspective to assess habitat services and optimize multi-scenario simulations and evaluations is crucial for developing resilient strategies for the future. This study takes Lanzhou, a semi-arid region, as a case study, combining multi-model analysis to explore the relationship between habitat quality and spatial accessibility and to conduct habitat service zoning. The findings indicate that under four development scenarios, the ecological network generally shows a three-segment distribution. The factors that have the most significant impacts on cultivated land, forests, shrubs, construction land, and bare land are GDP, precipitation, temperature, population density, and NDVI, respectively. The ecological priority scenario features the most corridors, while the cultivated land protection scenario incurs the lowest construction costs. Across various analysis radii of space syntax, except for MED at a 6000 m radius, the ecological priority scenario exhibits excellent network accessibility. The coupling coordination degree of the four scenarios generally lies within a mild imbalance level, with a spatial distribution pattern characterized by “high in the west and low in the east”. Based on 10 types of habitat services, a priority management sequence for land and key governance towns was established, leading to the proposal of a “dual coordination” multi-center compact network layout model. This research not only enriches the theory of land ecology but also overcomes the shortcomings in land spatial planning, addresses the practical problems of land development transformation in Lanzhou, and offers new data support and ideas for the construction of ecological cities in semi-arid regions.