玉山國家公園景觀變遷的驅動機制剖析

玉山國家公園的複雜、破碎之森林生態系統及其呈現的動態、不規則之土地紋理，為因應全球氣候暖化的趨勢，宜解析其景觀格局與變遷背後的驅動機制，以有效掌握其土地使用之影響衝擊，方能提出更合宜的規劃與經營策略。因此，本研究假設玉山國家公園森林生態環境係造成其景觀格局與變遷的顯著因子，採用遙感探測、地理資訊系統與景觀指數的整合性研究方法，從玉山國家公園森林景觀格局在時間和空間的結構與量之變化及趨勢，剖析其景觀變遷的驅動機制。研究過程係以植被多樣性、破碎性、網絡與廊道、踏腳石等整體景觀結構特性，及綴塊尺寸、數量、位置、邊緣結構、形狀等植群景觀特性，建立景觀指數分析之研究架構；以1972、1979、1990、2000及2010年遙感探測之影像判讀資料，呈現其景觀格局特徵與動態；再以景觀指數解析其景觀格局變遷的驅動機制與意涵。研究中發現整體景觀結構的植群多樣性，呈現不均勻分佈，聚集程度降低等特性；植群破碎性呈現內部棲地減少等特性；植群網絡與廊道呈現連結性低之特性；踏腳石呈現接近性增加之特性；植群景觀特性的綴塊尺寸，呈現邊緣棲地比例增加，重新連結機率增加等特性；綴塊數量呈現綴塊密度增加之特性；綴塊位置呈現綴塊孤立作用降低，重新拓殖性增加，距離穩定等特性；邊緣結構呈現綴塊形狀不規則、複雜化等特性；綴塊形狀呈現綴塊伸展程度漸增，惟變化幅度不大之特性。最後，整體研究結論表明，玉山國家公園的整體景觀結構，“粗細交錯、疏密有秩的景觀結構”是生態物種與能量流動的良好運作基礎，關鍵是“物種與能量的流動”；植群景觀特性上，“婉轉有形、聚散合宜的格局韻律”是生態位置與型態組合的健全運作機制，關鍵是“位置與型態的組合”，而這物種與能量運作基礎及位置與型態操作機制的組合架構，即玉山國家公園景觀變遷的驅動機制之雛形，也就是本研究所設定的目的與成果，除提供作為玉山國家公園永續經營的具體環境資訊，後續將進一步建立台灣高山國家公園之永續規劃架構與空間分析技術。

關鍵字

景觀格局；景觀變遷；景觀指數

並列摘要

For the complex, fragmented ecosystem and dynamic, irregular landscape texture of Yushan National Park, the park authority should analyze the driving mechanism behind the landscape pattern and changes, for responding to the global warming trend, to effectively grasp the impact land use change and suitably propose the planning and managing strategy. Therefore, assuming that the forest ecological environment of Yushan National Park is the significant factor causing its landscape pattern and changes, this research adopts the integrated research method of remote sensing, geographic information systems and landscape indices, studying the changes and trends of forest landscape pattern in temporal and spatial structure, to analyze the driving mechanism of its landscape change. This research first established the landscape indices analysis framework by the overall landscape structure characteristics of vegetation diversity, fragmentation, networks and corridors, stepping stone, and the vegetation landscape characteristics of patch size, number, location, edge structure, shape, then, presented the landscape pattern characteristics and dynamic by the 1972, 1979, 1990, 2000 and 2010 remote sensing interpretation data, finally, analyzed the driving mechanism and implications of its landscape pattern and changes by the landscape indices. The research found that the vegetation diversity in the overall landscape structure presents the nonuniform distribution and aggregation reduction characteristics; vegetation fragmentation presents internal habitat reduction of characteristics; vegetation networks and corridors presents low connectivity characteristics; stepping stone presents proximity increased characteristics; the patch size in vegetation landscape characteristics presents the edge habitat proportion increased and relink probability increased characteristics; patch amount presents patch density increased characteristics; patch location presents isolation reduction, re-colonization increased and distance stability characteristics; edge structure presents irregular patch shape and complexity characteristics; patch shape presents patch stretching increased characteristics. Finally, the research conclusion first indicates that the thick and thin, sparse and dense staggered landscape structure is the well-functioning basis of ecology species and energy flow, for the overall landscape structure of Yushan National Park. The key issue is the flow of species and energy. Then, the circumlocutory pattern with gather, dispersed relationship and rhythm landscape pattern is the perfect operation mechanism of ecological location and pattern combination, for the landscape ecological characteristic of Yushan National Park. The key issue is the combination of location and pattern. Therefore, the combination framework of species-energy operation basis and location-pattern operation mechanism should be the driving mechanism of landscape pattern and change of Yushan National Park, which is the prospective study result of this research. This research excepted provides the specific environment information for the sustainable management of Yushan National Park, further establishing the sustainable planning framework and spatial analysis techniques for Taiwan Alpine National Park in future.