Optimization of ecological network function and structure by coupling spatial operators and biomimetic intelligent algorithm

Ecological networks (ENs) can maintain regional ecological security and reconcile the contradiction between development and conservation. Optimization of ENs is crucial for promoting sustainable development and human well-being. Although various valuable methods have been explored to optimize ENs, collaborative optimization research on the EN function and structure through a quantitative simulation at the patch level is yet to be conducted. This study integrated spatial function and structural operators into a multi-type ant colony optimization model and constructed a high-performance land-use conversion framework for EN optimization. Results showed that: (1) the optimized ecological source exhibits heightened functionality and improved connectivity, resulting in a more stable and rational network structure; (2) the proposed ecological nodes emergence mechanism prevents fragmentation of local ENs and improves the overall structural integrity and resilience; and (3) the introduction of graphics processing unit (GPU) parallelization improves the efficiency of the model at city level by 13.49 times, resolving the paradox between the computational efficiency of high-precision land-use simulation and the construction of large-scale ENs. This study provides methodological support for patch-level landscape planning and reference for formulating reasonable ecological restoration strategies. [Display omitted] •A spatial-operator based MACO model for EN optimization was proposed.•Spatial operators were designed to support patch-level optimization.•A global ecological node emergence mechanism was developed.•GPU parallelization was used to improve computing performance.