利用BAG模式建立台灣中部沙里仙地區植物功能型-氣候關係之研究

Abstract

氣候變遷的研究均指出氣象因子的變化不僅有區域性的差異，甚至不同的時節反應也不一致。台灣以往的植群-氣候關係模式多著眼在年平均氣象參數的研究，而忽略季節性氣候的差異，而以全球性的植群分布氣象範圍套用台灣的植群-氣候關係研究，可能忽略小尺度地域的特殊演化造成的氣象適應力不同。本研究針對中部沙里仙地區植群分布情形，嘗試以包含逐月氣象參數做模擬的BAG model，建立較小尺度地域的植群-氣候關係，以供氣候變遷中植群動態研究的參考。 在三個植物功能型下的BAG分群結果，以複相關係數R-square值0.7為判別值時，可將常綠針葉樹型分成6群、常綠闊葉樹型11群及落葉性闊葉樹型8群共計25群BAG。本研究中所用的降雨參數變化幅度小，相對於計量單位而言較不具意義，分群區別主要表現在溫度以及光空域的參數項。氣象條件範圍以常綠針葉樹型的分群區別最佳，而範圍相近的分群可藉由潛力高值在海拔及WLS梯度的集中趨勢加以區別。研究顯示樹種實際樣區數會減弱潛力樣區的判定，而若干極端性的樣區會擴張分群的潛力分布，因此減弱氣象範圍界定的準確性，並使分群區別變得不易。 本研究受限於氣象紀錄的缺乏，代用的參數項資料較為粗放，且在小尺度的研究中，逐月的氣象資料間容易具高相關性，但結果仍可區別出分群的氣象條件範圍，若能進一步改進選用的氣象參數，應可對植群-氣候分群關係能有更詳細完善的解釋。

Climate change studies have pointed out that meteorological changes have both regional and seasonal differences. Previous research on the possible impacts of climate change on vegetation in Taiwan mainly focused on the changes in annual mean climatic parameters, therefore neglected the seasonal differences. In addition, using climatic envelopes derived globally to model vegetation-climate relations in Taiwan also likely obscured the small-scale vegetation and climate evolutional relationships. The main objective of this study was to develop vegetation-climate relationships based on monthly meteorology parameters for tree species in Salisen area, a mountainous region with elevation ranging from 1200 m to 3952 m, of central Taiwan. Monthly climatic parameters for the study area were first derived from long-term meteorological data using a generalized additive model (GAM) approach, with altitude and spatial coordinates as the explanatory variables. Then a modified Bioclimatic Affinity Groups (BAG) model was used to construct BAGs for three dominant plant functional types, namely, evergreen needle-leaved, evergreen broad-leaved trees, and deciduous broad-leaved trees, in the study area. Tree species distributions within the study area were based on 72 composite plots from a previous study. Using R-square value 0.7 as the threshold value, 6, 11, and 8 BAGs were derived for evergreen needle-leaved, evergreen broad-leaved, and deciduous broad-leaved trees, respectively. In this study, due to limited variations, precipitation parameters had little contribution to explain the differences among BAGs. Temperature and light sky space parameters were the main factors in delineating BAGs, especially in evergreen needle-leaved tree type. In general, temperature was the main factor in differentiating BAGs. BAGS with similar temperature requirements could be further distinguished by whole light sky space (WLS). The main difficulty encountered in this study was that some species were only present in a limited number of plots. For those species, the derived BAGs overextended their distribution ranges. The second difficulty encountered was that in a small region monthly temperature parameters were highly correlated. Despite these difficulties, the BAGs derived in this study could still be mapped directly to the main vegetation types in the study area, suggesting that the derived BAGs were in agreement with reality. The results of this study would enhance future modeling efforts on possible impacts of climate changes on the study area’s vegetation dynamics.