| A basin is an integrated geographical unit, and the landscape patterns within the basin have usually been formed by the interactions of natural and anthropogenic processess. The changes in landscape patterns within the basin can directly affect the ecological processes of the basin, influencing the ecological integrity and safety of the basin and the areas near the basin.The Irtysh River originates from the Altai Mountains in Xinjiang of northwest China, it flows down via the desert plains on the south foot of the Altai Mountains, goes to the west and out of the border of China, finally flows into the arctic sea through Kazakstan and Russia. Irtysh River Basin in the Altai forest region is composed of two parts: the mountain watershed landscape and the riparian plain landscape, which are connected as an integrative system by the river. The Irtysh River basin nourishes rich biodiversities in Xinjiang, playing important roles in maintaining the ecological integrity and safety of north Xinjiang. Analyzing the landscape pattern and spatial distribution of biodiversity both in mountain watershed and in the riparian plain could help us in understanding the ecological functions of the whole basin landscape, which can provide scientific basis for future reasonable resources use, biodiversity conservation and ecological landscape planning. In this thesis, we selected two typical landscape types in the Irtysh River basin of Xinjiang as the study area, Xiaodonggou representing the mountain watershed landscape and Beitun representing the riparian plain landscape, By combining methods of field investigations, remote sensing image interpretation, GIS analyses and simulation models, the pattern and dynamics of the landscape and spatial distribution of biodiversity in the Irtysh River basin of Xinjiang were explored. The main conclusions are as follows:(1)The landscape pattern and fragmentation in Beitun of Irtysh River Basin and Xiaodongou of Altai Mountains in Xingjian was analyzed. Quantitative indices of landscape pattern were chosen to examine landscape pattern and fragmentation. The landscape in Beitun was classified into 7 types and 10 subtypes, including 5 vegetation types and 8 sub vegetation types. The area percentage of the five vegetation types (farmland, riparian forest, wetland, grassland and desert vegetation ) in the landscape was 16.5%, 2.6%, 1.6%, 10.5%, and 52.2% respectively. The patch maximal perimeter, patch maximal area and mean area indices of riparian forest were lower than the other vegetation types, implying that the riparian forest distributed in smaller patches in the whole landscape. The value of mean patch shape, fractal dimension index, and edge density index of the riparian forest were higher than those of the other vegetation types, indicating that the riparian forest had more complex patch shape. All the pattern indices demonstrated that the riparian forest was the most fragmented type in the Irtysh River Basin. The landscape in Xiaodonggou of Altai Mountains was classified into coniferous forest, broadleaved forest, conifer-broadleaf mixed forest, shrub, grassland. The landscape pattern analysis results showed that forest was the matrix in Xiaodonggou of Altai Mountains, and the broadleaved forest had most patch shape.(2)The landscape pattern and dynamics in Beitun of Irtysh River Basin from 1972 to 2000 were analyzed by using remote sensing image of MSS, TM, and ETM+. Quantitative indices of landscape pattern included patch numbers, mean patch size, mean patch shape index, mean patch fractal dimension, Shannon diversity index, and Shannon evenness index. The results showed that desert was the matrix of the landscape. In the 28 years, the farmland area percent increased 9.62%, the desert area percent decreased 11.75%, residence area increased 0.15%, while the areas of riparian forest and wetland decreased 0.44% and 1.06% respectively. The patch number increased and the mean patch size decreased in the landscape, which indicated that landscape in 2000 was more fragmentized than the landscape in 1972. Both the mean patch shape index and mean patch fractal dimension index had increased from 1972 to 2000, which showed that landscape patch shape were becoming more complex. The Shannon diversity index and Shannon evenness both had a trend of increase from 1972 to 2000, indicating that the landscape heterogeneity decreased and landscape types distributed more evenly.Transition matrix of landscape types from 1972 to 2000 indicated that the riparian forest was the most unstable type, which was usually conversed into grassland or farmland.(3)Predicting the spatial distribution maps of forest stand factors were done in Beitun of Irtysh River based on the field investigation data and GIS. The predicting environmental factors included climate, topography and the distance to river. Residual maps were used to evaluate the accuracy of predicting. Keyran River was one of an important tributaries which located in the north of Irtysh River. The differences of forest communities characteristics were compared between upper, central, and lower part in the Irtysh River and the Keyran River of Beitun respectively. The difference of forest community characteristicss between near, middle, and far riparian zones were also compared in the south vs. north banks of the Irtysh River, and south vs. north banks of the Keyran River of Beitun respectively. The results showed that in the Irtysh River, the order for the mean DBH of the riparian stands was central reach >lower reach >upper reach. The standing volume and canopy density order was central reach >upper reach >lower reach. Tree mean height, species richness, and Fisher diversity index order was upper reach >central reach >lower reach. In the Keyran River, the tree mean height order was upper reach >central reach >lower reach, the order for the other five stand factors was upper reach Keyran River. The changes of the stand factors were affected by the shape and topography of the Irthsh River and the Keyran River in Beitun.(4)Based on the field investigation plots data and environmental factors, cluster analysis and ordination were used to analyze forest vegetations in Xiaodonggou. The cluster analysis identified 6 woody community types, including 5 tree communities and 1 shrub community. The 6 community types included: shrub, Populus tremula,, Populus laurifolia, Betula pendula, Larix sibirica, and Abies sibirica + Larix sibirica+ Picea obovata dominated plant communities, which had different distribution patterns along the major environmental gradients in the Altai mountains. The DCCA ordination showed that the topographical factors and soil factors were the dominant environmental factors affecting plant species distribution in the Xiaodonggou area. The distribution of the vegetations in the area was mainly controlled by the interactions of topographical and soil factors.(5)Relationship between the vegetation distribution and topographic factors were explored based on overlays of remote sensing classification map with the maps of slope, aspect, profile curvature and altitude which derived from DEM. The imagery interpreted vegetation types included coniferous forest, broadleaved forest, conifer-broadleaf mixed forest, shrub, and grassland in Xiaodonggou. The optimal habitat combination of topography for these vegetation types in Xiaodonggou were as follows: coniferous forest was mainly distributed in slope from 15°to 35°, west-north aspect, the profile curvature ranged from 0°to 3°and altitude ranged from 1 800m to 2 000m. Broadleaved forest was mainly distributed in slope from 15°to 35°, west-north aspect, the profile curvature ranged from 5°to 10°and altitude ranged from 1 400m to 1600m. Conifer-broadleaf mixed forest was mainly distributed in slope from 15°to 35°, north aspect, the profile curvature ranged from 5°to 10°and altitude ranged from 1 600m to 1 800m. Shrub was mainly distributed in slope from 15°to 35°, west aspect, the profile curvature ranged from 5°to 10°and altitude ranged from 1 400m to 1 600m. Grassland was mainly distributed in slope from 15°to 35°, south aspect, the profile curvature ranged from 5°to 10°and altitude ranged from 1 200m to 1 400m.(6)Patterns of species diversity along altitude in northern slope of Altai Mountain were studied based on the data of vegetation plots in Xiaodonggou. Species richness and Shannon-Wiener diversity index were chosen for the analysis. Diversity order of the three life forms was: trees < shrubs < herbaceous. Diversity indices along altitudinal gradient of the three life forms exhibited different trends. Tree diversity index exhibited an unimodal pattern with the increase of elevation, peaking at the altitude range of 1 600 m to 1 700 m; The shrubs showed a similar altitudinal gradient pattern. The herbaceous plant diversity index exhibited a different trend, which increased slightly with increasing elevation, but the change was not significant. The total diversity of trees, shrubs, and herbaceous showed an unimodal pattern with the increase of elevation. Altitude from 1 600m to 1 700m was an ecotone between conifer and broad-leaved stands, where maximum diversities occurred.(7)The predicting model for tree species richness was constructed on basis of field plot investigation and environmental variables in Xiaodonggou. The environmental predictor variables included climate, topography, and NDVI. The residual map was used to evaluate the validity of the model. The predicted spatial distribution map of species richness was overlaid with the slope, aspect and elevation maps respectively. The tree species richness of most areas in Xiodonggou ranged between 3 and 4, which occupied 70.28% of the total study area. In slopes from 0°to 5°, tree species richness value of 3 had the highest presence frequency, while in other range of slopes, the species richness value of 4 had the highest presence frequency. In aspects of west and west-north, the tree species richness of 3 had the highest presence frequency, while in the others aspect the tree species richness value of 4 had the highest presence frequency. Along the altitudinal gradient, the tree species richness showed an unimodal pattern of distribution. The residual type results showed that strongly predicting area and moderately predicting area reached 94.62% of the total area, which implied that the predicting model was accurate enough.(8)Using ETM+ data and the biomass by typical sample investigation plots, the Pearson correlation between RVI,NDVI,SAVI,DVI,IPVI and plots biomass was made respectively, the linear regression models for aboveground biomass in Xiaodonggou were constructed. Residual map was used to evaluate the accuracy of predicting accuracy. The correlation between NDVI and biomass was much better among the chosen vegetation indices. The residual type results showed that strongly predicting area and moderately predicting area reached 66.60% and 30.31% respectively, which implied the predicting model was accurate. The map of tree aboveground biomass spatial distribution was overlaid with the slope, aspect and elevation maps respectively. The results showed that the tree aboveground biomass order was inclined/steep slope >flat/gentle slope> rapid/dangerous slope. In aspect, the east-south and south aspect had the lower biomass than the other aspects. In elevation, the biomass order was mid elevation > high elevation > low elevation, from low elevation to high elevation, the biomass exhibited an unimodal pattern.
|
PDF Full Text Request