Catchment-based landscape patterns and divisions in mountain areas: A case study of Qixia City in Shandong Province

This paper used Qixia City (which lies in the center of Jiaodong Peninsula) as a case study of typical mountain areas to investigate catchment-based landscape pattern divisions with cluster analysis of landscape indices generated from digital elevation model (DEM). Firstly, Landsat TM image (2010) covering the whole study area was interpreted and geo-referenced to generate the landscape patterns in Qixia City and the corresponding relief map (1︰25 000) vectorized to build DEM model. Then 300 catchments were divided using hydrology extension support in ArcGIS. Taking into account ecological significance and features of common landscape indices as well as local ground truth, five indices were selected for the analysis. The indices included edge density (ED), Shannon's diversity index (SDI), coefficient of variance of patch size (PSCOV), area-weighted mean patch fractal dimension (AWMPFD) and core area density (CAD). Based on average elevation and spatial distributions, landscape patterns in each catchment were analyzed. The effects of human activity and natural disturbance on regional landscape ecology were also discussed in the paper. All the indices analyzed in each catchment were synthesized using cluster analysis and combined with dominant landscape type in the catchment to build zoning maps of landscape. The results suggested that ED and CAD values increased as altitude gradually increased and significant vertical differentiation characteristics were noted. The SDI values gradually grew smaller with increasing altitude. Then PSCOV values showed inverted U-shaped trend with increasing altitude. The AWMPFD values did not show significant vertical differentiation characteristics. Three landscape ecology classes were regionalized in Qixia City. ClassⅠhad the smallest land area (10.14%) and stable agricultural zones. This implied that the landscape structure and scale were very stable and needed protection. Class Ⅱ had the biggest land area (50.93%). Here, a variety of landscape types had spatially cross distributions with complex patchy shapes. This area belonged to the transition zone of natural landscape and cultural landscape which needed strengthened land use planning and rational distribution of industry. Class Ⅲ had the bigger land area (35.92%) and was mainly concentrated in high altitude areas with main land use types of woodland and grassland. Thus larger reclamation of land in this class was a very difficult task in terms of development and utilization. The study argued that the use of landscape metrics in landscape zoning better reflected prominent landscape features and spatial variation patterns of landscape ecological structure. To a certain extent therefore, the study provided a significant guidance on planning and utilization of land in Qixia City.