| Summary: | 碩士 === 國立中山大學 === 生物科學系研究所 === 104 === Alpine vegetation studies typically include vegetation classification and its relation to environment. In Taiwan, there were very few such studies because of the dangerous terrain and the harsh climate, so the aim of this study is to build a classification for Taiwan alpine vegetation from field survey and literature data.
My study took place in the alpine vegetation areas above the treeline of the Zhongyang Range, the Xueshan Range and the Yushan Range. Relevé data, including species cover and environmental information, were collected following the Braun-Blanquet methods. The collected data were analyzed in two steps. All the relevés were first grouped into composite samples by means of two-way indicator species analysis and tabular comparison method. Second, the composite samples were classified into vegetation groups by cluster analysis. Then, these vegetation groups were combined and ranked by means of discriminant analysis, detrended correspondence analysis and canonical correspondence analysis. The vegetation types classified were Lycopodium obscurum vegetation type, Carex satzumensis vegetation type, Carex oxyandra vegetation type, Carex nubigena vegetation type, Picris hieracioides subsp. ohwiana vegetation type, Pedicularis verticillata vegetation type, and Potentilla leuconota vegetation type. The Lycopodium obscurum vegetation type were further divided into Gentiana scabrida var. scabrida subvegetation type, Carex breviculmis subvegetation type, Lycopodium clavatum subvegetation type, Veratrum formosanum subvegetation type, Spiraea hayatana subvegetation type, and Anaphalis morrisonicola subvegetation type.
The vegetation research results from the past can be incorporated into the current classification very well. When the original Braun-Blanquet data is available, the diagnostic species were used to determine the vegetation type. The discriminant analysis then used to check the correct classified percentage. Only very few vegetation releves were misclassified, resulted from the different data used in the table comparision method and the discriminant analysis. This indicated the classification in this study may be applied to all Taiwan alpine area. When the original Braun-Blanquet data is not available, the diagnostic species listed are used to determine the possible vegetation type. There were also a very few vegetation types from the literature cannot be found its proper position in the system. These may be resulted from different sampling methods, vegetation changed or microhabitat not surveyed in the current study
|
|---|
