Two-step forest vegetation analysis based on large data sets in northeastern Taiwan

本研究介紹植群分類學中龐大植物社會資料組之二階段分析方法。資料組之分層建議 有兩種方式：在較大範圍及異質地理區域內，可採用局部地點的資料先分出若干子集;如資料 組包含一地區所有植群型，則按植群型分出子集。第一階段首先在每一資料子集進行基本的群 團分析或分類工作，然後在形成之初級群團中計算組成物種之覆蓋－豐富度整合值，所有基本 群團都經由相同之處理程序。第二階段是以初級的群團為單位，再度進行群團分析與分類，以 劃分社會型。本文建議的整合值可以反映恆存度及覆蓋－豐富度，但特別強調物種須擁有高恆 存度才可以當作鑑別種。 本方法以TWINSPAN 程式處理台灣東北氣候區496 個樣區及374 種木本植物之植群分類，並 以層級理論及分類理論來探討一些有關問題及未來發展方向。A two-step method for the classification of very large phytosociological data sets in northeastern Taiwan is presented. Stratification of the data set is suggested either by subdivided areas in the case of a large and geographically heterogeneous region, or by vegetation types in the case of a set covering all the plant communities in the region. First, cluster analysis or any classification pro-gram is processed on each subset. The resultant basic clusters are summarized by calculating a synop-tic cover-abundance value for each species in each cluster. All basic clusters are then accorded to the same procedure of classification. Second order clusters are explained as community types. The syn-optic value proposed for each plant reflects both constancy and average cover-abundance. It is em-phasized that a species should have a high constancy to be used as a diagnostic species. The method is interpreted with a data set of 496 relevés and 374 species of forest vegetation in northeastern Taiwan. The program TWINSPAN is adopted in each stage of classification. Some problems and perspectives of the approach are discussed in the light of hierarchy theory and classifi-cation theory