生态学报2026，Vol.46Issue(8)：3999-4016,18.DOI:10.20103/j.stxb.202507161869
大熊猫国家公园四川片区生态质量时空演变及模拟预测
Spatiotemporal evolution and simulation prediction of ecological quality in the Sichuan Area of the Giant Panda National Park
摘要
Abstract
The Giant Panda National Park serves as an ecological security barrier for biodiversity conservation in the upper reaches of the Yangtze River in China.A scientific assessment of its current ecological status,along with an analysis of its evolutionary patterns and future trends,is crucial for achieving refined management and sustainable development of the national park.This study utilized multi-source remote sensing,meteorological,and socio-economic data from 2000 to 2023 to establish an ecological quality evaluation index.Spatial autocorrelation and geographical detector models were employed to analyze the spatiotemporal dynamics of ecological quality and the driving mechanisms behind these changes.Additionally,the performance of four models—Bayesian Neural Network,BP Neural Network,GA-BP Neural Network,and LM Neural Network—was compared to assess their ability to simulate ecological quality.The study also included simulations for 2030,forecasting ecological quality under different scenarios,providing insights into the future ecological trends and their implications.The performance of four models,including the Bayesian neural network,was compared,and ecological quality under different scenarios for 2030 was simulated.The results indicated that:(1)From 2000 to 2023,the overall ecological quality of the study area showed significant improvement,with the proportion of areas rated as good or excellent increasing from 77.49%to 87.78%.Spatially,it exhibited a strong clustering pattern characterized by"lower values in the northwest and higher values in the southeast."(2)Land use ecological index,elevation,and temperature were identified as the core factors driving the spatial differentiation of ecological quality.All factor interactions demonstrated enhancing effects,particularly when socio-economic factors such as population and GDP interacted with natural factors like elevation,which significantly amplified their influence.(3)The LM neural network model demonstrated more robust predictive performance.Based on this model,under an ecological protection priority scenario,the area of excellent ecological zones was projected to increase by 5.34%by 2030;under a natural development scenario,a slight risk of degradation appeared,with the area of excellent zones decreasing by 1.86%;while under an economic development priority scenario,the area of excellent zones showed a sharp decrease of 53.24%,indicating severe ecological degradation.The study concluded that while the ecological foundation of the Giant Panda National Park was determined by its natural geographical patterns,human activities influenced the ecosystem through interactions with these natural conditions.These findings provided a scientific basis for formulating refined management and adaptive conservation strategies for the national park.关键词
生态质量/驱动机制/机器学习/情景模拟/大熊猫国家公园Key words
ecological quality/driving mechanism/machine learning/scenario simulation/Giant Panda National Park引用本文复制引用
王瑶,向明顺,邬真妮,李建华..大熊猫国家公园四川片区生态质量时空演变及模拟预测[J].生态学报,2026,46(8):3999-4016,18.基金项目
国家自然科学基金项目(42471229) （42471229）
