Highly urbanized areas are not absolute barriers: Assessing avian habitat connectivity in a megacity via the omnidirectional connectivity model

Maintaining habitat connectivity for birds has become a central challenge in urban biodiversity conservation. Taking the Shanghai metropolitan area as a representative case, this study integrates citizen science observations with field surveys to select 25 surrogate bird species. Suitable habitats were identified using the MaxEnt model in combination with the human footprint index, and the omnidirectional connectivity model (OCM) was applied to evaluate avian habitat connectivity across the city. Furthermore, an integrated XGBoost-SHAP-MGWR analytical framework was employed to investigate the mechanisms and spatial heterogeneity through which urban environmental factors influence habitat connectivity. The results indicated: (1) a total of 418 ecological sources were identified, covering 921.84 km², primarily concentrated in the western and southeastern regions, while the central urban districts exhibited severe fragmentation; (2) overall connectivity in the study area remains relatively high, with peripheral suburban zones forming well-connected ecological networks, whereas the central city and peripheral industrial areas demonstrated poor connectivity; and (3) environmental factors exert heterogeneous effects on avian connectivity, with habitat-related variables playing a decisive role. In particular, increasing the proportion of blue infrastructure (PB > 4%) and enhancing ecological quality (RSEI > 0.57) significantly improve habitat connectivity. Overall, this study underscores the necessity of integrating landscape connectivity research into urban planning. The findings not only elucidate the complex interactions between urban environmental characteristics and avian connectivity but also provide a transferable methodological pathway to support biodiversity conservation and tailored policy interventions in other high-density metropolitan contexts.