An optimized evaluation method of an urban ecological network: The case of the Minhang District of Shanghai

•The framework was aimed at optimizing the evaluation methods of an urban ecological network.•MSPA methods and graph theory were combined in order to define the connectivity of the ecological network.•The high-quality ecological space structure was analysed, and classification of the urban land was detailed.•The urban ecological network was constructed for species with level 4 dispersal capacities.•We identified the patch locations that were most important for overall connectivity in cities. Obtaining the characteristics of an urban ecological network structure and connectivity plays an important guiding role in the concrete implementation of ecological construction in a space. Resistance assignment is a key step in ecological network construction. However, most research ignores the diversity of land use in an urbanization environment and the difference between ecological space centres and boundaries. Taking Minhang District of Shanghai as an example, this paper puts forward an optimal evaluation method for an urban ecological network. By combining the morphological spatial pattern analysis method and graph theory, the resistance value of different urban land uses was evaluated, and the ecological network was constructed by comprehensively considering the land use types, land area, forms, and positions. The connectivity and the importance of individual nodes at four different thresholds were quantified. The results revealed that the core and the edge areas of the morphological spatial pattern analysis occupied the highest amount of space, 33.34 % and 21.92 %, respectively. In addition, there were relatively few large and high-quality ecological spaces, and the green space was seriously fragmented in the study area. The degree of structural connectivity was related not only to the area proportion of the ecological space but also to the area of connected patches. The contribution value of each node to the overall network connectivity was affected by the mobility of the species, the node area, the spatial location, and the surrounding matrix environment. This study provided a framework designed to optimize the evaluation method of an urban ecological network. The influence of the type and structure of urban land use on the resistance value was analysed in a more detailed way, and the results of the connectivity and the important nodes were obtained more accurately.