基于兴趣点数据和最大熵模型的苏州市鸟类多样性研究
Study on Bird Diversity in Suzhou Based on POI Data and MaxEnt Model
-
摘要: 鸟类群落对于城市环境中的食物及栖息地有较强的依赖性, 导致城市鸟类与环境因子的关系非常密切。以苏州市为研究区域, 提取影响城市鸟类生存繁衍的关键环境因子, 探索性地引入兴趣点(point of interest, POI)数据并构建多类别POI核密度指数, 揭示苏州市不同城市空间布局对鸟类分布的影响。通过收集苏州市2020-2022年鸟类分布数据, 利用最大熵模型(maximum entropy model, MaxEnt), 对205种鸟类在苏州范围内的潜在适生区进行模拟, 分析了苏州市鸟类多样性的空间分布格局及热点地区。结果显示, MaxEnt模型预测结果可信度较高(AUC平均值为0.854);土地利用类型、归一化植被指数和距离绿地距离是影响苏州市鸟类空间分布的主要环境变量, 林鸟与水鸟的差异影响因子主要为冠层高度和距离水源距离; 结合偏相关分析得出, 在12类POI核密度因子中, 乡村住宅用地类型对于鸟类的分布影响最大, 这表明在城市发展的同时也需要注重维持乡村生态平衡。Abstract: Bird communities in urban environments exhibit a robust dependence on food availability and habitat suitability, forging intricate connections between urban avifauna and environmental dynamics. Taking Suzhou City as the study area, the study extracted key environmental factors that affecting the survival and reproduction of urban birds, used point of interest (POI) data to construct a multi-category POI kernel density index, and revealed the influence of different urban spatial layouts on bird distribution. By collecting bird distribution data from 2020 to 2022, using the Maximum Entropy (MaxEnt) model, the potential suitable habitats for 205 bird species within the Suzhou area is simulated, and the spatial distribution patterns and hotspot areas of bird diversity are analyzed. The results show that the MaxEnt model has a high level of credibility (with an average AUC value of 0.854). Land use type, normalized difference vegetation index, and distance to green spaces are the main environmental variables influencing the spatial distribution of birds in Suzhou. The differences between forest birds and water birds are mainly influenced by canopy height and distance to water sources. Based on partial correlation analysis, among the 12 POI kernel density factors, rural residential land use has the greatest effect on bird distribution, indicating that while urban development is important, it is also necessary to focus on maintaining ecological balance in rural areas.