基于网络效能分析的生境网络构建与优化——以苏锡常地区白鹭为例

A network efficiency analysis approach to habitat network optimization:a case study of the little egret(Egretta garzetta) in the Su-Xi-Chang area,China

从网络效能分析出发研究生境网络优化具有重要意义。以快速城市化苏锡常地区为小尺度研究区、苏锡常及周边15个城市为大尺度泛研究区、白鹭为焦点物种,通过采用引入并构建网络效能概念框架、识别和构建不同尺度生境网络、调整网络结构和转换网络结构特性相结合的方法,实现了跨尺度下白鹭生境网络优化。结果表明:1)依据大尺度迁移网络完善小尺度生境网络得到的方案一和改变方案一网络特性得到的方案二均是现状生境网络的优化方案;2)方案二与方案一相比,在平均路径长度值增加、聚类系数值减少、节点度概率分布幂次关系拟合度提高,即小世界特性减弱、无标度特性增强的基础上,测度网络连接度的α、β、γ指数值得到增加,同时新增生境节点和迁移廊道数量更少,具有更好的优化成效;3)从网络效能分析出发,有益于提高生境网络跨尺度下优化方案的可行性,其分析思路既丰富了网络优化的方法论研究,也为提高上位与下位规划空间契合度的实践提供了理论支持。但从网络效能概念框架看,如网络层级分析、其他尺度情景研究、网络质量测度指标等方面还亟待丰富与完善。

The network efficiency analysis method is used to study network systematization and integrity, which are important for the assessment of ecological networks. However, this method has rarely been applied in habitat network optimization studies. The objective of the present study was to use the network efficiency analysis method to develop a new approach to optimize habitat network analysis. A small-scale area of 1.75×10^5 km^2 was selected in the Su-Xi-Chang region as the case study; and 12 neighboring cities, together with the Su-Xi-Chang area, covering 11.91×10^5 km^2, were selected as the large-scale study area. The little egret （Egretta garzetta） was chosen as a regionally representative species in the study area. The method included three steps： 1） a conceptual framework for the network efficiency analysis was built to optimize the habitat network, in which the small-world and scale-free network concepts were introduced, with their network structural characteristics, which included average path length, cluster coefficients, and the probability distributions of node degrees. These structural characteristics were applied in step 3 as the assessment indicators; 2） the little egret habitat patches and small- and large-scale habitat networks were identified and simulated. The present status of the small-scale habitat network was obtained from previous research results. In the large-scale area analysis, the observation sites were obtained from a dataset for 2003-2015 on the China Bird Report website. These were regarded as the habitats and those sites where 〉 20 little egrets were observed simultaneously were identified and transferred to habitat patches. Then, a large-scale habitat network was simulated using the same method used in the previous research; and 3） the present status of the small-scale habitat network was optimized using the large-scale habitat network and was designated as network A. Network B was generated by adjusting the network A structure and transferring its network structural characteristics. The network connectivity indexes α, β, and γ of the present status small-scale habitat network were used to calculate and compare networks A and B. The results showed that 1） the A and B networks represented the optimized results of the present status small-scale habitat network; 2） the increased average path length value, decreased cluster coefficient value, and increased goodness-of-fit of probability distribution of node degree for network B compared to that of network A showed that the network B small-world network characteristics had decreased, but its scale-free network characteristics had increased. This suggests that network B has a better fit to the large-scale habitat network than network A; and 3） the number of newly-added habitat nodes and the migration corridor for network B were smaller than network A, but its corresponding values for the network connectivity indexes α, β, and γ were greater. This suggests that network B results were better than those for network A, and therefore are the recommended, optimized results. The present study indicated that optimizing habitat networks using the network efficiency analysis method improved effectiveness, enriched network structure optimization methodology, and provided a theoretical basis for improved goodness-of-fit when undertaking spatial planning among upper- and lower-levels. However, further studies are required to improve conceptual frameworks, including aspects of network hierarchical analysis, scaling, and network quality assessment indicators.