关键词: 生态用地, 景观指数, Logistic回归模型, GIS

Abstract: As one of the most important components of the urban social-economic-natural complex ecosystem, ecological lands provide various critical ecosystem services. As an essential resource for human being, the protection of existing ecological land, the restoration of currently damaged ecological zones, and the return of naturally ecological land are important measures for improving and balancing regionally ecological conditions. These methods are also useful for sustainable development and harmony between human and nature. In view of ecosystem services of ecological land and ecological environment effects of land use and land cover change during rapid urbanization processes, it is necessary to select typical urbanization area to quantify the spatiotemporal characteristics and driving mechanisms of ecological land use change, which plays practical roles in providing spatial strategies for improving regionally ecological security and establishing urban ecological security pattern. The research area of this study is located in Xinzhuang Town of Changshu City, south Jiangsu Province, which experiences the most rapid industrialization across China in the past decades. As time goes by with the acceleration of urbanization, considerable amount of ecological land resource has been consumed with the deteriorated ecological environment and the fragmented landscape. Based on high resolution remote sensing images, integrated with GIS and RS technology, and by utilizing transfer matrix and landscape pattern metrics, this study systematically analyzed the changing characteristics of ecological land in Xinzhuang Town. A logistic regression model has also been employed to analyze the driving forces of the main types of ecological land change, with per capita GDP, per capita gross industrial output, per capita gross agricultural output, per capita income, population density and the distance to nearest rural settlements, major road, river, and village center as possible driving factors. Main conclusions are as follows. During the period of 1991-2009, the spatial structure and area of ecological land have been dramatically changed in Xinzhuang Town, with an accelerated decreasing trend. Among these changes, the area of paddy fields change was obvious that the cumulative reduction was 1 806.61 hm²; while both the aquaculture land and garden plot with higher economic benefits were increased, especially the former, whose average annual growth was 43.01 hm² between 1991 and 2009. The decreased ecological land was mainly transferred to aquaculture land, residential land and industrial land. At class level, the fragmentation and aggregation increased, dominance decreased, and landscape shape was more and more regular for most of the ecological land types. Logistic regression results showed that for the paddy fields coverage change, "distance to nearest river", "distance to nearest road", and "distance to nearest rural settlement" were the most important independent variables in the first period of 1991-2001, while "distance to nearest river", "Per capita GDP", and "Per capita industrial output" were the most important independent variables over the second period of 2001-2009. For aquaculture land coverage change, the main driving factors in the first period were "distance to nearest river" and "distance to nearest road", while "distance to nearest rural settlement" seemed to be more important in the second period. For forest land coverage change, it was found that in the first period, "distance to nearest village center" and "distance to nearest road" have produced greater influence, and in the second period, "distance to nearest road" and "population density" were more important. However, due to the deficiency of quantification and space technology of policy factors, this study did not consider the policy factors, such as land development, management and ecological protection. Further studies are needed to explore how to add policy factors to improve the fitting precision of the model and practical application effects, and thus provide more scientific decision support for urban ecological security and regional sustainable development.

Key words: ecological land, landscape pattern metrics, logistic regression model, GIS