| Beijing mountain area is an important part of Beijing suburb and it is a natural barrier surrounding Beijing city. The forest quality has a direct impact on the balance and stability of Beijing ecological system. Based on the quantitative description and analysis of the effect scale of forest landscape composition, structure, and spatial pattern in Beijing mountain area, the thesis clarifies the mountain forest resources, reveals the spatial structure of forest types as well as the regular pattern and scale effect of spatial changes from the perspective of the landscape. It provides a theoretical reference for forest landscape planning and forest landscape ecosystems optimization. Meanwhile, it has an important significance to improve the forest structure of Beijing mountain area, realize the optimal allocation of forest spatial layout, and develop the ecological function of urban forestry.On the basis of a great deal of domestic and international documents, combining the characteristics of the study area, selecting forest landscape spatial pattern in Beijing mountain area, forest type terrain differentiation under the landscape scale, and forest landscape scale effect these three main aspects as the research content, taking landscape ecology as the theoretical basis, using the multi-remote sensing images, together with historical data (all types of forest survey data) and plots measured data, and making full use of a variety of methods, such as remote sensing technology, geographic information system technology, quantifying pattern analysis technology, space geographic statistical technology, the thesis systematically studies the forest landscape pattern and the scale effect in Beijing mountain area. Main conclusions are as follows:(1) With indices calculation of forest landscape spatial pattern in Beijing mountain area, and through the researches of patches type level and landscape level, it is found that: The composition,structure and spatial distribution pattern of in Beijing mountain area is extremely uneven, and the spatial heterogeneity characteristics is significantly. These four types of forest landscape are coniferous forest, broad-leaves forest, mixed broadleaf-conifer forest and shrub forest.(2) Through the research of terrain factor differentiation of forest landscape pattern in Beijing mountain area, it is found that: With the changes of altitude and local topographic features (aspect, slope), forest vegetation is also experiencing a regular changes, and the impacts are various from different terrain-grade level, which presents a significant distribution.(3) Based on line pattern sampling, by the calculation of information entropy indices in Huairou and Mentougou District, it is found that: All types of forest landscape in different line transects reflect the same general trend with the changes of scale. Namely, along the changes of sample plots from small to large, the information entropy indexes were from small value gradually became larger and ultimately achieved their maximum; every landscape elements had its maxima expression scale of heterogeneous. Under the scale, elements were uneven in the distribution of line transects, which showed certain features of macroheterogeneity; exceed the scale, this uneven features were concealed as a detail, elements were even in the distribution of line transects, which showed microheterogeneity feature.(4) Based on grid pattern sampling, by the analysis of lacunarity indices in Huairou and Mentougou District, it is found that: In different scales, the higher lacunarity indices are, the greater heterogeneity of spatial distribution for forest landscape elements is. This performance can be considered as macroheterogeneity characteristics. Oppositely, the lower lacunarity indices values are, the smaller heterogeneity of spatial distribution for forest landscape elements are. This performance can be considered as microheterogeneity characteristics.The order of lacunarity index is contrary to the order of coverage rate of the forest landscape elements in the same scale, consequently, lacunarity indices values in experiments can help to study the coverage rate of forest landscape elements changing with the spatial scale. With gradually larger scale, the precision that lacunarity indices reflect the landscape coverage. rate is poorer.With gradually increasing analysis scale, all standard lacunarity indices of forest landscape types show the decreasing trend. However, the changed speed of standards lacunarity indices for different landscape elements is different along scale changes, which reflect the sensitivity difference of scaleeffect to different types of forest landscape in different studies areas.(5) By the study of spatial heterogeneity in different scales of the forest landscape pattern and Kriging interpolation, it is found that: The spatial distribution of forest landscape diversity in experimental area was of anisotropy, its spatial dependence and the level of spatial heterogeneity are significant in different directions: With the increased extents, the lag interval of anisotropy gradually increased; also showed a similar trend under different ranges. Diversity index interpolation of forest landscape in different scale can reflect the spatial distribution pattern of landscape diversity; however the results are varied with different details and accuracies. In small-scale, estimated diversity indices of different forest landscape, the patches are broken. With the increased scale, the complexity of diversity pattern of forest landscape became simple, spatial variation is reduced; regional connectivity with the same estimated value of diversity index gradually strengthened. It showed that the details in small-scale is being gradually inundated in large-scale with the increasing observed scale, tended to be more macroheterogeneity.There are three innovations in the paper: (1) Analyze and compare forest landscape pattern and spatial distribution in Beijing mountain area by the application of pattern quantification analysis theory. (2) Study the different scale effect of forest landscape pattern in Beijing mountain area by the sampling of the line pattern, grid pattern and polygon pattern interpolated from point pattern, with the help of information entropy, lacunarity index and geostatistical analysis. (3) Considered forest landscape diversity index as a regionalized variables, it is carried out exploratory study of spatial heterogeneity and the scale effect by geostatistical methods, which has certain theoretical and practical significance. |
PDF Full Text Request