摘要: | 在全球維持生態多樣性的研究中,已認定入侵種是破壞生物多樣性的重要原因之一,因為入侵種會造成生物多樣性、自然棲地、生態平衡的破壞,且對生物多樣性的威脅程度,僅次於棲息地的切割與干擾,因此,台灣入侵種名單中,名列全世界百大入侵植物之一,且嚴重入侵及危害澎湖群島與恆春半島的木本植物-銀合歡,便成為國內相關單位所關心的研究對象。由於銀合歡擴散之景觀生態安全格局研究,澎湖群島已有前人研究報告,故本研究強調在恆春半島之銀合歡擴散的景觀生態安全格局研究,其目的除了探討恆春半島之景觀變遷和景觀生態安全格局之外,主要目的是在評估該二地區銀合歡入侵擴散及景觀生態安全格局之差異性,因此本研究的流程包括:(一)根據二期的國土利用調查成果資料,先萃取恆春半島的森林地區,然後應用1994年和2006年二期的法國SPOT衛星影像進行影像分類,以獲取恆春半島的銀合歡分布圖和土地使用分類圖,供探討恆春半島銀合歡的空間分布特徵、空間結構與配置、及景觀變遷情形,並經由Shannon多樣性t檢定法,探討恆春半島之土地使用變遷情形,以及銀合歡擴散對恆春半島景觀變遷的影響程度,過程中有關景觀指數的量化與分析,包括三種多樣性指數、三種均勻度指數、面積、嵌塊體總數、嵌塊體密度、邊界總長度、邊界密度、平均嵌塊體大小、平均嵌塊體形狀指數、碎形維度指數、散置和並排指數∕分散性指數、聚集度指數等;(二)探討景觀因子與銀合歡空間分布之相關性,包括海拔高度、坡度、坡向、距聚落以及道路的距離;(三)應用景觀生態安全格局方法,分析恆春半島銀合歡的入侵與擴散,並從防治觀點,劃定恆春半島防治銀合歡擴散的戰略點分布圖;(四)最後將恆春半島銀合歡之景觀變遷、景觀生態安全格局、及戰略點類型等結果,跟同樣受到銀合歡嚴重入侵與擴散的澎湖群島做比較,並供評估該二地區銀合歡入侵與擴散之差異性。
研究結果顯示:(一)從1994到2006年的銀合歡分布圖得知,銀合歡的分布面積確實有增長的趨勢,表示銀合歡在12年間有擴散蔓延的現象;從三種多樣性指數證實恆春半島銀合歡因受干擾而造成所占之總景觀比例逐漸增加,而三種均勻度指數亦顯示銀合歡數量呈現增加的趨勢,且分布越趨於均勻,表示銀合歡的蔓延程度已逐漸嚴重;從嵌塊體密度、邊界總長度、邊界密度、平均嵌塊體大小等指數上升的結果得知,銀合歡已從面積較小的嵌塊體,擴散成面積較大的嵌塊體,且因面積、嵌塊體數量及嵌塊體密度的增加,造成嵌塊體連接度的上升,進而增加銀合歡成為純林的機會,並使銀合歡對抗干擾的能力增強;從平均嵌塊體形狀指數、碎形維度指數上升的結果,可得知銀合歡嵌塊體的結構表面趨於複雜、嵌塊體的形狀趨於不規則,造成恆春半島中其他類別的嵌塊體,與銀合歡的接觸機率升高,導致銀合歡入侵的機率增加;從散置和並排指數∕分散性指數、聚集度指數下降的結果,可得知嵌塊體數量增加及擴散範圍增大,導致銀合歡間的嵌塊體聚集度下降、嵌塊體間的分布呈現非常不均勻狀態,且隔離程度也趨於緩和,造成銀合歡的連接程度因而上升,進而合併成為較大面積之銀合歡,抑或成為銀合歡純林;從Shannon多樣性t檢定法得知恆春半島的地景變遷呈顯著性的變化,並證實銀合歡的擴散確實對恆春半島的景觀變遷有顯著性影響;(二)從銀合歡空間分布之分析結果得知,銀合歡主要分布在海拔300公尺以下、坡度0°–20°的範圍內,且分布面積會隨著海拔梯度上升,以及隨著距道路的距離越遠而逐漸減少;(三)從景觀生態安全格局之Logistic模式分析得知,具有顯著影響恆春半島銀合歡入侵擴散的因子包括坡向、距海岸的距離、距道路的距離、以及距銀合歡的距離共四個景觀因子,經利用上述四因子劃定恆春半島防治銀合歡擴散的戰略點分布圖後,將1994年建立的阻力趨勢表面套疊2006年銀合歡分布圖,得知恆春半島的銀合歡已有部分擴散至阻力較高的地區,表示2006年銀合歡的擴散較1994年的嚴重;從戰略點分布圖得知未有銀合歡分布之戰略點有32個,占所有戰略點的82.05%,快要受銀合歡入侵之戰略點有3個,占所有戰略點的7.69%,有銀合歡分布之戰略點有4個,占所有戰略點的10.26%;(四)比較澎湖群島與恆春半島的景觀變遷程度得知,銀合歡的蔓延擴散確實對兩地區的景觀變遷有顯著性的影響;從恆春半島與澎湖群島之最小累積阻力模型所繪製的景觀生態安全格局圖得知,澎湖地區的景觀安全格局單元有34個;而恆春地區有36個,而該二地區之低阻力通道的源間連接形狀,皆類似於樹枝狀的水系類型;從戰略點類型的分析結果得知,恆春半島是網路型阻力趨勢表面的景觀生態戰略點,而澎湖群島亦是網路型阻力趨勢表面的景觀生態戰略點。本研究針對恆春半島銀合歡之景觀變遷、景觀生態安全格局、及戰略點等所得到的結果,以及恆春半島與澎湖群島銀合歡入侵擴散之差異性評估結果,將提供林業相關單位未來防治該二地區銀合歡擴散之經營管理上的參考。
Invasive species have been regardes as one of the important reasons to cause the damage of biodiversity in the study for maintaining ecological diversity of the world. It threats to biodiversity after disturbing habitat, natural habitats, and ecological balance. For these reasons, the related forest units become interesting in the subjects about Leucaena leucocephala that is the list of invasive species in Taiwan, to be ranked 100 of the World's Worst Invasive Alien Species, and intrusion and harm seriously in the Hengchun Peninsula and the Penghu Islands. Since the research have been reported which landscape security pattern on the dispersal of Leucaena leucocephala in the Penghu Islands, this study emphasizes the area in Hengchun Peninsula. This study focuses not only on the landscape changes and landscape security pattern but also to assess the invasive specie of differences between dispersal and landscape security pattern in two kinds of regions. The research process includes (1) Applying the SPOT–2 image in 1994 and the SPOT–4 image in 2006 to firstly extract the forest area, according to the first and second national land-use survey data produced by the Survey and Mapping Center, and then using remote sensing technique to classify the forest area of Hengchun Peninsula into the categories of Leucaena leucocephala maps and also to generate the land–use classification maps. The result obtained from this process can be a basic map for discussion about the spatial distribution, structure, and configuration and landscape change of Leucaena leucocephala in Hengchun Peninsula. To examine the landscape diversity of Hengchun Peninsula and see if there is a significant landscape change or not by Leucaena leucocephala dispersal through using Shannon diversity t–test. In this study, the quantitative analysis with landscape index includes three landscape diversity indices, three landscape evenness indices, total area, patch number, patch density, edge density, mean patch size, mean shape index, fractal dimension, interspersion and juxtaposition index, and aggregation index.(2)Exploring the relationship between spatial distribution of Leucaena leucocephala and landscape factors such as altitude, slope, aspect, and distance from settlements and roads.(3)Applying landscape security pattern to analyze the invasion and dispersal of Leucaena leucocephala in Hengchun Peninsula, and to establish the strategic points of preventing dispersal of Leucaena leucocephala in Hengchun Peninsula from the control point of view.(4)Finally, to compare the dispersal of Hengchun Peninsula and Penghu Islands that is the same with serious invasion by Leucaena leucocephala through landscape change, landscape security pattern, and strategic point type to assess the difference about invasion and dispersal between these two areas.
The distribution area of Leucaena Leucocephala based on the land–use map in 1994 to 2006 indicates that the Leucaena leucocephala dispersal continues to expand during 12 years. According to the result of three kinds of landscape diversity indices, the proportion of Leucaena leucocephala to the total landscape is enlarged in the Hengchun Peninsula due to the disturbance. Three landscape evenness indices show that the patch numbers of Leucaena Leucocephala augment and become uniform distribution, it signifies that the dispersal of Leucaena Leucocephala be more serious. From patch density, total edge, edge density, and mean patch size increase, it is known that Leucaena leucocephala already invades from smaller patches into larger patches and becomes more connectivity because of area, patch number, and patch density grow. It makes more opportunity to become pure forest and enhance the ability of Leucaena leucocephala to fight with disturbance. According to mean shape index and fractal dimension, the structure of patches become complex and the shape of patches become irregular. This situation will increase the probability of Leucaena leucocephala to contact with another land–use type of Hengchun Peninsula, and the invasive risk of Leucaena leucocephala. As for the indices of interspersion/juxtaposition and aggregation, the result shows that the aggregation between Leucaena leucocephala patches decreases with the the number of patches and the scope of dispersal and also the distribution of patch is uneven, less isolation and more connection. Furthermore, small patches are merged into a larger area, which will form a pure forest of Leucaena leucocephala. From the result of Shannon diversity t–test, the landscape of Hengchun Peninsula has significant change and Leucaena leucocephala dispersal during 1994–2006 has significant effect on landscape.(2)As for the spatial distribution of Leucaena leucocephala, the result shows that Leucaena leucocephala mainly locates around the altitude of 300 meters below and the slope of 0°–20° range, and increases with the elevation gradient. In addition, it gradually reduces with the farther away from the road.(3)After using the Logistic Model in the method of landscape security pattern, the factor has significant effect on the dispersal of Leucaena leucocephala including slope, the distance from the coast, the distance from the road and the distance from the Leucaena leucocephala. These four factors are then used to produce the map of prevention and control strategic points of Leucaena leucocephala distribution. According to the minimum cumulative resistance in 1994 overlaying the distribution of Leucaena leucocephala in 2006, Leucaena leucocephala has spread to some areas of high resistance. That is, the dispersal of Leucaena leucocephala in 2006 is more severe than 1994. In the result of strategic points, 32 strategic points are not with Leucaena leucocephala, 3 strategic points are almost with Leucaena leucocephala, and 4 strategic points are with Leucaena leucocephala. The proportion to the total strategic points is 82.05%, 7.69%, and 10.26%, respectively. (4)The dispersal of Leucaena leucocephala has significant effect on the landscape of Penghu Islands and Hengchun Peninsula after comparing these two area. According to landscape security pattern from the minimum cumulative resistance model, Penghu Islands exist 34 units and low resistance path is similar to the type of radial water system. Hengchun Peninsula exist 36 units and low resistance path is like the type of dendritic river system. In the result of analysis of strategic point type, both Hengchun Peninsula and Penghu Islands are the network minimum cumulative resistance of landscape security patterns. This study focuses on the landscape change of Leucaena leucocephala, landscape security patterns, strategic points, and a comparison of difference between Hengchun Peninsula and Penghu Islands about the invasive dispersal, the result can be served as a management reference of Leucaena leucocephala in these two area for the forest management units. |