臺灣木本植物多樣性影響因子之探討

Abstract

了解物種多樣性沿環境梯度（如地形、氣候等）的變化以及影響它們分佈趨勢的生態因子為近年來研究多樣性的重要議題之一，同時也是未來保育經營管理的基礎。然而，探討生物因子〔如推移帶效應（ecotone effects）、質量效應（mass effects）等〕對多樣性的影響及在東亞的研究仍然闕如。本研究利用「國家植群多樣性調查及製圖計劃」之樣區資料，以廣義線性模式（GLM）探討台灣島木本植物多樣性之影響因子，這些因子包含地形、氣候及植群等解釋變數共三十項。研究方法先以相關性矩陣剔除高相關變數，接著採用手動操作之正向逐步選取法（forward stepwise）依次選出解釋變量改變最多的因子，並以卡方檢定檢驗各項的顯著性。研究結果顯示，最佳模式由九項變數組成，其中包含年平均溫度的二次及三次式，共解釋53.3%的變異。木本植物在較高的年平均溫度及最大之月降雨量、潛在推移帶、低岩石地比例、少植群型類型及較大的年溫差下有較高的多樣性。此結果可由物種—能量假說（species-energy hypothesis）、環境適合性假說（environmental favorableness hypothesis）及推移帶效應解釋；環境壓力假說（environmental stress hypothesis）、環境穩定性假說（environmental stability hypothesis）、環境異質性假說（environmental heterogeneity hypothesis）及質量效應不適合解釋台灣木本植物多樣性之趨勢。研究結論：影響大尺度木本植物多樣性分佈趨勢最重要的因子為氣候及推移帶，地形變化的效應不明顯；各類型的環境因子及高次項變數為建構模式的必要元素，而手動操作之正向逐步選取法為配適多樣性—環境因子模式的可行方法。

Macroecologists have often suggested that environmental factors such as climate and topography exert great influences on woody plant species richness. However, studies which have empirically documented the link between biological determinants and species richness are scant, e.g. the ecotone effect and the mass effect. Few studies have reported on an East Asian landscape too. Therefore, the aim of this article attempts to explore how species richness and independent environmental descriptors are related in the scale of Taiwan Island. Generalized linear models (GLM) were used to correlate species richness of woody plants per quadrat (400 m2) with three sets of variables, which are topographic, climatic and vegetation variables. Regression models were constructed by a manually operated forward stepwise procedure as the following process: the reduction of collinearity among variables, the model selection based on the percentage of deviance change (D2) and a Chi-squared significant test. The best explanatory model consists of nine variables. Including quadratic and cubic forms of temperature, the model explains about 53% of the deviance. Maximum woody plant species richness occurs at higher annual average temperatures with larger temperature ranges, in a potential ecotone, with higher maximum monthly rainfall, with less rock outcrops and fewer vegetation formations. Our results could be explained by the species-energy hypothesis, the environmental favorableness hypothesis and the ecotone effect. Stressful conditions, stability of climate, environmental heterogeneity and the mass effect proved to be of minor importance for determining regional patterns of species richness. We conclude that climate-driven process and ecotones are primary determinants of woody plant species richness in Taiwan Island. It is recommended using several sets of predictors and the methods that we proposed to capture the underlying complexity of biodiversity.