Abstract: Green space ecological network is important for improving the fragmentation of regional ecological space, biodiversity loss, imbalance of supply and demand in ecosystem services, and ensuring regional ecology. With Dongting Lake region as an example, the ecological sources were evaluated and identified, and basic ecological resistance of grid element was calculated from the perspectives of comprehensive ecosystem service function evaluation, potential biodiversity comprehensive index evaluation and morphological spatial pattern analysis with the support of 3S technology. We corrected the basic ecological resistance by night light index, used the minimum cumulative resistance model to identify ecological corridors, and constructed a weighted evaluation model, eva-luated the degree of aggregation and dispersion of the sources, as well as the contribution of ecological connection of the corridors. Furthermore, we compared and evaluated the structural characteristic index of the integrated networks, the potential “source sink” networks and the planning networks. The results showed that the spatial distribution of the sources was uneven, with forest land, shrubland and wetland acounting for 95.9% of the total area. The Dongting Lake wetland located in the middle of the region was at high ecological risk. The closer the source was to the center of the ecological network system and the smaller the average minimum cumulative resistance was to other sources, the stronger the advantages of aggregation and dispersion. The denser the distribution of medium and high ecological quality sources was around the high ecological quality sources, the higher the aggregation and dispersion. The closer the corridor was to the high ecological quality sources, the greater contribution of ecological connection. Forest land, shrubland, especially river played an important role in connecting natural ecosystem and human social system. The “source-sink” planning greenways formed a supplement to the “source-sink” potential ecological corridors. Compared with the “source-sink” potential network, the α, β, γ and ρ index of the integrated network was increased by 123.1%, 25.8%, 26.2% and 74.6% respectively. Compared with the “source-sink” planning network, the α, β, γ and ρ index was increased by 190.0%, 31.1%, 32.5% and 114.6% respectively. The results could provide reference for the construction of green space ecological network and planning of land use in Dongting Lake region.

Key words: green space ecological network, source-sink, potential corridor, planning greenway, comprehensive evaluation method, Dongting Lake region