Islanding Landscape In Thousand-Island Lake Region And Its Effect On Plant Diversity

Landscape fragmentation is defined as the phenomenon that functional diversity relationships are ruptured, or that connectivity is reduced among ecosystems within the landscape. It can also be seen as a process of landscape from simpleness to complexity caused by disturbance of natural or human factors. Landscape fragmentation brings lose or degradation of habitat and accordingly induces the loss of biodiversity. Landscape fragmentation can affect species, heredity and ecosystem diversity. Effects of landscape fragmentation on biodiversity is one of a focus fields of landscape ecology research. Thousand-island Lake region is a fragmental landscape caused by strongly factitious disturbing.Study of this thesis focuses on a key aspect which is dynamics of fragmental Landscape in Thousand-Island Lake region and its effect on plant diversity. We analyzed the changes in land use and land cover pattern in this key region for a period of 39 years (1964-2003), studied the effects of landscape fragmatation on plant diversity, edge effect of shrub layer and species-area relationship on small islands. The major results can be summarized as follows:1. Our results showed that the total number of patches and the degree of fragmentation basically remained the same, whereas the patch number, patch area, and dominance of each patch type changed significantly in Xiaojinshan of Thousand-Island Lake region. The most dominant patch type changed from treeless barrens (47.44% of the terrestrial land) in 1964 to the Pinus massoniana forest (52.81% of the terrestrial land) in 2003. We also found that the value of landscape diversity decreased slightly. It is important to note that, because landscape diversity is a synoptic measure of the richness and evenness of all land use and land cover types, it is not always positively related to landscape functioning. The slight decrease in landscape diversity in our study was a result of dynamics of all existing patches in the study region. Our results suggested that the forest vegetation of the area had improved in terms of coverage since 1964 (i.e., 5 years after the dam construction). These changes may be attributed to the many measures of ecological protection in thisregion which allowed natural succession to take its cause. The improved landscape pattern enhanced ecological functioning of the island ecosystems, including restoration of habitats, increased biodiversity, and reduction in soil erosion. For example, some severely endangered species such as Phoebe bournei and Syrmaticus ellioti were able to survive and successfully reproduce. While the forest vegetation has been recovering, the increasing dominance by the structurally simple Pinus massoniana forest in this region should be considered in future efforts of biodiversity conservation, ecological planning, and sustainable development.2. 26 samples on 18 islands in which small, medium and large islands were included were selected. Species and number of trees and shrubs were investigated. The results of one-way variance analysis (ANOVA) of species richness of trees showed that the species richness of trees among all kinds of islands were differed significantly (F=l3.055, P=0.000). We found the species richness of trees in large islands was significant higher than small and medium islands and was not differed significantly to land by multi-ply compare. Analysis of shrubs species showed that species richness of shrubs were not differed significantly in small, medium and large islands and land samples. Our further analysis using Spearman correlation showed that the species richness of trees was significantly correlated with islands area and the specie richness increased with island area, but the correlation between specie richness of shrubs and islands area was not significant. The analysis using Shannon-Wiener Index (H) showed that the diversity both trees and shrubs on large islands was largest and next was land samples, but the diversity in small islands was larger than medium islands. We also analyzed the dominance and evenness using Simpson index and Pielou index. To tree, the species evenness was highest and dominance was lowest on large islands but it was small islands to shrub.3. On 10 less-disturbed and more isolated islands, we investigated the composition of shrub-layer plant species in response to the landscape fragmentation induced by damming. The results of one-way variance analysis (ANOVA) showed that the abundance and diversity of species differed significantly between the edge versus interior environment (F=23.386, P=0.000<0.01). Specifically, both plantabundance and species richness in edge environment were higher than those in interior environment. We also quantified the plant species diversity for the two environments using Shannon-Wiener Index (H), Simpson Index (D), and Pielou Evenness Index (E). For the edge, H=2.304, D=0.883, and E=0.880; for the interior, H=1.716, D=0.751, and E=0.755. Clearly, plant species diversity was higher at the edge than the interior of the islands. Also, several species, such as Vitex negundo L.var. ccmnabifolia, and Rubus swinhoei Hance, were only found in the edge environment. The measure of species similarity between the edge and interior varied greatly (B=[0.30, 0.800]). This indicates that the plant species diversity of at least some of the islands has been significantly affected by landscape fragmentation. Our further analysis using Spearman correlation showed that the shrub-layer plant species richness was not significantly correlated with island area in either the edge or interior environment.4. We selected 22 small land-bridge islands which areas are less than 2 hm2 as sample of our study. We studied the species number, quantity of tree and shrub of these islands in a quantitative method and analyzed the relationship between the species number, quantity of tree and brush and islands area. The power curve was not the best one which was used in traditional island biodiversity theory when we showed the Species-area relationship in our study. The logarithm curve was showed the relationship between tree species number and islands area rather than power curve, as well as the relationship between non-herbage (tree and shrub) species number and islands area. S curve can be showed the relationship between the shrub species number and islands area more exactly. The sensitivity analysis of tree and shrub species of the islands showed that the number of sensitive species took on declining trend with the reducing of island area while. However, the number of non-sensitive species kept stabilization.