Rapid urbanization and economic development in China has resulted in urban sprawl and has also led to a reduction of natural resources and habitat fragmentation in urban areas. Habitat fragmentation and the resultant decrease in habitat connectivity has become a serious challenge for biodiversity conservation. Green spaces are important habitat resources in the city. The development of green space ecological networks can be used to increase the connectivity between fragmented habitat patches, which promotes urban biological diversity conservation. Taking Bazhong western new district, a rapidly urbanizing area in Sichuan, China, as a case study, a novel method is proposed using Guidos, GIS and the gravity model to develop green space ecological networks.
First, the morphological spatial pattern analysis (MSPA) model in the software Guidos was used to identify the main types and structure of green spaces in the landscape. The "cores" and "bridges" were extracted, which will make the foundation of the green space network. Second, using the software Conefor, the integral index of connectivity (IIC), probability of connectivity (PC) and the percentage of importance (dI) of each individual element were used to evaluate the relative importance of these cores and bridges to landscape connectivity. Accordingly, the top ten core patches acting as ecological sources were identified. The remaining cores and bridges were ranked based on their importance in maintaining ecological network connectivity. Finally, using GIS, potential ecological networks were developed using the least-cost path method, and the important corridors were identified using the Gravity model. A plan to optimize the ecological network in the study area was put forward. This research shows that the structural elements, the cores and bridges of ecological networks can be identified via the MSPA method and that landscape connectivity measurements (IIC, PC and dI) provide good quantitative measures of the relative importance of landscape patches, especially when considering landscape patch spatial distributions. The evaluation of landscape patches can help to identify and extract important protected landscape elements. Least-cost path analysis also provides important reference information for evaluating landscape patch habitat quality, which most previous researches have evaluated by subjective assessments. The research results show that the combination of MSPA with measures of landscape connectivity and least-cost path analysis are a novel and useful method for developing green space ecological networks in urban and other fragmented areas. The results will help to guide the development and management of green spaces for biodiversity conservation.