Possible Impacts of Climate Change on Potential Tree Plant Forms of a Mountain Region in Central Taiwan

氣候變遷對台灣中部山區潛在植物型之可能影響

10.6165/tai.2003.48(4).259

關秉宗(Biing-T. Guan)；黃任伯(Zenn-Bor Huang)；鍾年鈞(Nien-June Chung)

生物地理模式 ； Box模式 ； 氣候-植群分類 ； Monte Carlo模擬 ； Biogeographical model ； Box's model ； Climate-vegetation classification ； Monte Carlo simulations

TAIWANIA

48卷4期（2003 / 12 / 01）

259 - 272

英文

本文利用Box's模式探討在不同情境下氣候變遷對台灣中部和社與對高岳山區潛在林木植物型之影響。本文亦利用Monte Carlo模擬方法探討氣候變遷之不確定性對預測結果所可能產生之影響。在模擬研究中，氣候因子增溫的範圍為迄今長期平均溫度加0至5°C，而降水因子其變動範圍則為迄今長期平均降水之±30％。研究結果顯示，研究地區潛在林木植物型於所設定的變動範圍內受降水因子之影響頗小，主要影響將源自於溫度因子之改變。在所設定的氣候變遷情境下，研究地區現今出現之林木植物型其反應可分成：(1)不受影響，(2)逐漸消失（如夏綠型），與(3)逐漸出現（如熱帶樹種）等三大類，而研究地區現今出現之林木植物型大多數屬於第一類。本文所用之方法在缺乏高解析度資料時，仍可有效地用以探討氣候變遷對森林植群相對組成之可能影響。

In this study we used Box's model, which is based on the equilibrium relationships between macroclimate and plant forms, to assess the possible impacts of climate change on the potential tree plant forms of a mountain region in central Taiwan. To account for the effects of uncertainty associated with the projected climatic conditions on the model's predictions, a Monte Carlo study was also carried out. For the temperature variables in the model, they varied between 0 and 5°C above the long-term average; for the precipitation related variables, the projected precipitation varied between ±30% of the long-term average. The responses of tree plant forms to temperature increases could be divided into three categories: (1) those that would not be influenced, (2) those that could disappear gradually, including summergreen tree species, which are more sensitive to temperature and require a lower temperature to exist, and (3) those that could appear gradually, including tropical tree species. Under the projected precipitation conditions, most of the tree plant forms currently present would not be influenced, except for rainforest and raingreen tree plant forms. The necessary data for Box's model could be obtained easily from usual climatic databases, and the results have a higher resolution than Holdridge Life Zone model. When detailed information necessary to run high-resolution models is not available, the approach used in this study could be used as an alternative.