目前，我国正处于快速城市化阶段，城市规模和数量激增，自然生态用地不断被侵占。同时，自然灾害也是造成生态斑块破碎化、生态环境受损的主要原因之一。强烈的人为活动和频繁的自然灾害导致大型生态斑块数量减少、质量下降，生态系统稳定性降低，生态安全风险与日俱增，对社会的可持续发展造成巨大阻碍。十九大报告明确指出我国要“加大生态系统保护力度。构建生态廊道和生物多样性保护网络”。作为生态系统保护和修复的有力手段，生态网络建设对于提升生态系统质量和稳定性，适应和减轻生态安全风险，提高区域生态减灾能力，建设生态文明具有非常重要的意义。 本研究选择城市化进程迅速且自然灾害频繁的四川省作为研究区域。首先，本文以景观阻力最小化和网络连通性最大化为目标，提出了基于最小耗费距离模型和Prim算法的ESOM模型，构建了累计景观阻力最低且斑块全连通的四川省最小生态网络。在此基础上，采用重力模型和最大潜在损失模型识别最小生态网络中的关键廊道，对比分析两组关键廊道的一致性和差异性，探讨两种模型的适用情景。最后，从斑块自身属性和网络结构影响两个视角出发，引入斑块形态学指标和生态网络结构指标，以构成斑块稳定性指标，并基于帕累托最优进行斑块稳定性排序，定量评价了生态斑块的稳定性。主要结论如下： 1）四川省生态斑块空间分布不均，主要集中在西北部和西南部；景观阻力面的分布较均匀，最高值出现在成都市，低值区域集中在北部高原和山地的阿坝州和绵阳市；四川省最小生态网络呈“五心放射状”；东部生态网络有待完善。 2）四川省最小生态网络是一个无标度网络，网络连通性对于斑块随机破坏具有很高的鲁棒性，但在蓄意破坏下，网络的脆弱性很高。 3）重力模型和最大潜在损失模型分别识别出的两条关键廊道既有一致处，也有差异处。结果一致表明，四川省西北部分布有关键廊道，东部不存在关键廊道；有四条共同的关键走廊，包括最关键廊道都是连接阿坝州和雅安市的生态廊道；最核心生态斑块是0号斑块，位于阿坝州。不同之处在于，一组关键廊道分布在四川省西北部，连接了大斑块；另一关键廊道分布在整个西部，不仅连接了网络核心区，也连接了网络边缘区。 4）在生态规划和保护中，重力模型更适用于核心生态区的识别和保护，而最大潜在损失模型则是维护和加强与整个区域连通性的网络功能的更好选择。 5）本文提出的斑块稳定性指标不仅考虑了斑块自身属性，也增加考虑了生态网络对于斑块稳定性的影响。四川省最小生态网络中的生态斑块的稳定性呈现出西高东低的空间格局，且南部斑块更容易受到人为干扰和物种寄生。 6）川西地区自然灾害发生频繁。灾前重点保护，灾后优先恢复稳定斑块，尤其是稳定性最高的第一等级稳定斑块，可有力维持和修复区域生态系统稳定性，增强生态系统服务功能，保护生物多样性。 本研究创新了生态网络的构建方法，从实践上为区域生态网络构建提供指导，进行了关键廊道识别和斑块稳定性评价的有益尝试，为生态规划者和决策者提供了一个耗费最少的生态成本发挥最大的生态效用的生态网络构建工具ESOM模型，助力区域生态系统可持续发展。

Natural hazards and human activities are the primary causes of large-scale habitat loss, fragmentation, and degradation. Identifying and optimizing ecological networks can maintain landscape integrity and enable connections between fragmented patches to be created or restored. As a powerful means to alleviate the contradiction between urban construction and ecological protection, ecological network construction is of great significance for improving the quality and stability of ecosystems, achieving sustainable development and building ecological civilization.This study is aimed to develop an innovative quantitative approach to construct a regional ecological network and identify critical corridors within the network, in order to provide scientific support for ecological planning. In this study, Sichuan Province was selected as the research area. Firstly, aiming at minimizing landscape resistance and maximizing network connectivity, we proposed the ecological network structure optimization model based on the least cost path model and Prim algorithm, and constructed the minimal ecological network of Sichuan. Then, the gravity model and the maximum potential loss model are used to identify the critical corridors in this network. Finally, from the perspectives of patch attributes and network structure, patch morphology index and ecological network structure index are introduced to constitute the stability quantitative index of patches, and patch stability is ranked based on Pareto optimum. The main conclusions are as follows: 1) The spatial distribution of patches in Sichuan is uneven, mainly concentrated in the northwest and southwest; the distribution of landscape resistance surface is relatively uniform; the smallest ecological network is "five-center radial"; the eastern ecological network needs to be improved. 2) The minimum ecological network in Sichuan Province is a scale-free network. Its connectivity is robust to random damage, but it is vulnerable to intentional damage. 3) The gravity model and the maximum potential loss model identify two group of critical corridors which have similarities and differences. The results show that the critical corridors is in the northwest and not in the east. There are four common critical corridors, including the most critical corridor connecting Aba and Ya'an; the Patch 0 (Aba) is core. The difference is that corridors I are distributed in the northwest, connecting large patches; the corridors II are throughout the western region, connecting not only the core area of the network, but also the edge area of the network. 4) In ecological planning and protection, gravity model is more suitable for the identification and protection of core ecological areas, while maximum potential loss model is a better choice to maintain and strengthen the network function of connectivity with the whole region. 5) The proposed patch stability index not only consider the patch's own properties, but also add the impact of ecological network on patch stability. The stability of patches shows a spatial pattern of high in the West and low in the east, and the patches in the south are more susceptible to human disturbance and species parasitism. 6) Natural disasters occur frequently in western Sichuan. Priority should be given to the reliable patches during pre-disaster protection and post-disaster restoration. This study innovates the construction method of ecological network, provides guidance for the construction of regional ecological network in practice, manage to identify critical corridors and evaluate the stability of patches. It provides ecological planners and decision makers with an ecological network construction tool that consumes the least ecological cost and maximizes the ecological utility.