Urbanization Impact on Landscape Patterns in Beijing City, China: A Spatial Heterogeneity Perspective
Huilei Li,Jian Peng,Liu Yanxu,Hu Yi'na
DOI: https://doi.org/10.1016/j.ecolind.2017.06.032
IF: 6.9
2017-01-01
Ecological Indicators
Abstract:The temporal and spatial characteristics of landscape pattern change can reflect the spatial impact of urbanization on the ecological environment. Studying the relationship between urbanization and landscape patterns can provide supports for urban ecological management. Previous studies have examined the quantitative relationship between the social economy and landscape patterns of an entire region, but have not considered the spatial non-stability of this relationship. In this study, we characterized the landscape patterns in Beijing City, China during 2000 and 2010 using four landscape metrics, i.e. patch density (PD), edge density (ED), Shannon's diversity index (SHDI) and the aggregation index (AI). Geographically weighted regression (GWR) was employed to identify the spatial heterogeneity and evolution characteristics of the relationship between the urbanization of population density (POP), gross domestic production (GDP) and nighttime lighting (NTL), and landscape patterns. The evolution of urban landscape patterns indicated a decentralized, aggregated, and fragmented change from the downtown to the suburb and outer suburb. During the 10-year period, the average PD in the downtown increased by 100.6%, and the increase of AI in the suburb was the largest. The PD, ED and SHDI increased by different degrees in the outer suburb. The influences of different urbanization modes on landscape patterns were also different. Inning mode made the landscape patterns more regular and integrated. The landscape was more broken and complicated under the edge-expanding mode, and the leapfrog mode made PD and SHDI increase slightly. In the relationship interpretation, GWR effectively identified the spatial heterogeneity, and improved the explanatory ability compared to ordinary least squares (OLS). The most intense response to urbanization was the forest landscape and the forest-cultivated land ecotone in the northwest of Beijing City, indicating that this region was ecologically fragile. The population density in the urbanization index had a direct effect on landscape patterns, while the PD affected by urbanization was greater than the shape, aggregation and diversity index. Affected by development policy, urban planning and other factors, the explanation degree of social economy to landscape patterns decreased in 2010. GWR is an effective method for quantifying the spatial differentiation characteristics of urbanization impacts on landscape patterns, which can provide more spatial information and decision criteria for the green development of a compact city.