|
Forests, as an important component of the landscape in Fenhe watershed, play a significant role in the landscape diversity of basin and the exchange of material and energy. Guandishan forest area, represented the watershed forests, is not only representative of the natural secondary forest in North China mountain area, but also significant in the landscape pattern analysis research.Under the guidance of landscape ecologic theory, using, based on the Landsat MSS images in 1973 and the Landsat TM images in 1990 and 2007, we used the RS and GIS technology, combined with the collected field data, firstly,to make an image process on three periods of remote sensing images in the study area, and then extracted the landscape elements.Then, In the support of FRAGSTATS, a kind of landscape pattern software, we selected the landscape pattern indexes which can fully reflect the characteristics of study area on class level and landscape level respectively. Taking landscape pattern in 2007 as an example,we made the index conducted analysis on the aspects of patch formation, patch feature, shape index, fragmentation index, heterogeneity and spatial distribution, on all kind of landscape element in study area. We then made a comparison analysis on three periods of landscape elements to obtain the significant changes of all kinds of landscape elements in the thirty years. Finally, on the basis of dynamic analysis of landscape pattern, we applied the Markov forecast model to forecast the development of the future landscape element in study area,and analyzed the driving forces of the changes. Analysis results show that:(1) In the landscape element formation in the study area, the forest landscape has the biggest advantaged, which is the matrix of the whole landscape, the grass landscape also has quite a proportion, while the cultivated land, resident and mine land, and the river take up a smaller proportion of landscape. In the thirty years, the dynamic changes of forest area show that: 1973 > 2007 > 1990, and the changes of forestland on area feature indexes, shape indexes, fractal dimension and fragmentation degree is dynamic consistent with the forestland area that indicates the forest interfered a lot. The increasing degree of landscape gathering and heterogeneity also lead to the increase of the landscape element diversity. From the spatial distribution of landscape, three periods of landscape elements are consistent on the distribution and variation of the levels on altitude, slope and aspect, represented a certain regularity.(2)Forest landscape pattern analysis of first-class is relatively complex, which is mainly represented that increasing the coniferous forest area yearly, and the broad leaved forest and shrubbery are in a dynamic change, corresponding indexes of area features, shape,fragmentation and heterogeneity had such a change in the thirty years and exhibit a certain regularity.(3)The forest landscape situation of second-class has more complicated. Different woodland compositions have distinctive change ranges in various years: the indexes of coniferous dense crop, coniferous open forest and broad leaved dense forest change a lot, the landscape heterogeneity is rise and reached the highest in 1990, the whole landscape diversity increased year by year, and the evenness is in change. Meanwhile, three periods of secondary forest landscape have the consistent performance in different levels of terrain factors.
|