Construction and Optimization of Ecological Security Pattern in Kunming

Insights From Machine Learning Algorithms and Circuit Theory

Bibliographic Data

ID21648905
AuthorsLin Zhang (0000-0001-5966-3827, School of Soil and Water Conservation Beijing Forestry University Beijing China), Yutong Li (0000-0001-8653-6698, Soil and Water Conservation Monitoring Center Taihu Basin Authority of Ministry of Water Resources Shanghai China), Kanli Wei (Shanghai Agricultural Development Promotion Center Shanghai China), Shi Qi (0009-0005-9927-3547, School of Soil and Water Conservation Beijing Forestry University Beijing China, corresponding author)
Year2026
Volume37
Issue8
Pages3594-3608
Publication date2026-05-15
Open AccessYes
TypeARTICLE
VenueLand Degradation and Development (JOURNAL)
Journal identifiersISSN: 1085-3278 • E-ISSN: 1099-145X
PublisherWiley (PUBLISHER • GB)
DOI10.1002/ldr.70330
OpenAlexW4416681199
LanguageEN
References cited43

Constructing and optimizing an ecological security pattern (ESP) is a critical approach to safeguarding ecosystem health and promoting sustainable development, particularly in rapidly developing cities. However, traditional research methods often fail to accurately and effectively construct complex and dynamic ESPs. This limitation poses considerable challenges to the formulation of ecological restoration strategies and the maintenance of ecosystem stability. To address these challenges, the present study innovatively introduces an integrated methodological framework that combines the remote sensing ecological index (RSEI), morphological spatial pattern analysis (MSPA), landscape connectivity index (LCI), machine learning algorithms and circuit theory to construct the ESP of Kunming from 2000 to 2021. The framework is centered on the three elements ‘ecological source’, ‘ecological corridor’ and ‘ecological pinch point’, and further proposes targeted optimization strategies. Results demonstrate the substantial applicability and advantages of machine learning algorithms in ESP construction, and the XGBoost model emerging as the most effective approach for developing ecological resistance surfaces. Between 2000 and 2021, Kunming's ecological source area expanded from 882.90 to 1424.87 km 2 , forming a spatial pattern characterized by clustered distributions in the west and north and scattered distribution in the east. The total length of ecological corridors increased from 1141.28 to 1553.98 km, the number of connected corridors from 39 to 72 and ecological pinch points from 22 to 26. These changes have effectively mitigated ecological fragmentation, facilitated species migration and supported habitat restoration. Additionally, the number of ecological barriers grew from 21 to 31, particularly in the vicinity of Dianchi Lake. Given the obstructive impact of urbanization on the ecological network in this region, enhanced ecological protection and measures to restore network connectivity are imperative. Based on these findings, an optimized ESP framework, referred to as the ‘four zones and three belts’ model, was developed for Kunming city. This framework provides valuable theoretical and methodological insights for the construction and optimization of ESP in urban contexts

Ecological systems theory · Ecosystem · Exploit · Habitat · Safeguarding · Spatial ecology · Sustainable development · Land Use and Ecosystem Services · Water Resources and Sustainability · Wildlife-Road Interactions and Conservation

Tools

Open DOI