An ex-post evaluation approach to assess the impacts of accomplished urban structure shift on landscape connectivity.

Many studies have been conducted to evaluate the effects of different urban structures on landscape connectivity, and most of them rely on the comparison approach or ex-ante scenario analysis. However, we still lack an ex-post method to capture the consequences of accomplished urban structure shift (from monocentric to polycentric), which is guided by the land use planning. To fill this gap, we develop an ex-post evaluation approach which integrates counterfactual analysis and landscape graphs. Counterfactual analysis is combined with cellular automata simulation model, to uncover what the city might look like, if it had continued to expand in a monocentric structure; and the landscape graphs enable us to reveal the possible landscape connectivity in actual and counterfactual scenarios. We select Nanjing city as the study area and 4 target species, to delve into the varying impacts of the urban structure shift on different taxonomic groups. Our case study demonstrates that: (1) the impact of urban structure shift is more relevant to the long disperser; (2) the actual landscape (polycentric) would facilitate the bird's dispersal, while (3) forest mammals have higher connectivity in the counterfactual scenario (monocentric), and the possible reasons are discussed. Finally, we demonstrate that the urban structure shift might not necessarily cause the connectivity decline, on condition that the key connectivity providers are identified by integrating ecological network analysis into the land use planning, and well preserved in the shift.