- 學位論文
台灣淺山地區石虎 (Prionailurus bengalensis) 的空間生態學
Spatial ecology of leopard cats (Prionailurus bengalensis) in a fragmented rural landscape in northwestern Taiwan
摘要
石虎(Prionailurus bengalensis)在台灣屬於I級保育類野生動物，過去並未有針對此物種的生態學研究。本研究以自動相機和無線電追蹤進行石虎的生態研究，以了解石虎的活動範圍、核心區、日移動模式、日活動模式，以及不同尺度下對於棲地利用的情形。2005年至2008年，於新竹縣和苗栗縣1000公尺以下的淺山地區，共架設253台自動相機，並於通霄地區對6隻石虎配戴發報器進行無線電追蹤。 無線電個體追蹤結果顯示，平均活動範圍(100% MCP)和核心區(50% MCP)分別為5.0和0.7平方公里；雄性石虎的活動範圍和核心區大於雌性石虎。雄性石虎於濕季時有較大的活動範圍和核心區。雖然，石虎有領域行為，然而只有核心區不重疊，個體間的活動範圍有重疊現象。另外，雄性石虎有較大的日移動距離和範圍，而雌性石虎則對於活動範圍有較集中和有效地利用。24小時活動資料顯示，石虎為夜行性動物，晨昏各有活動高峰期，偶爾白天也會活動。 石虎棲地利用分為3個尺度分析，在活動範圍尺度(third-order scale)，石虎對於休息地點有所偏好或迴避，每隻個體偏好或迴避有所差異；然而，活動地點的選擇，只有雄性石虎有偏好或迴避的棲地類型。在追蹤樣區尺度(second-order scale)上，石虎對於棲地類型的選擇也有所偏好，活動範圍的偏好依序為天然林>非天然林>草生-農墾地；核心區的偏好依序為非天然林>天然林>草生-農墾地。在大樣區尺度 (first-order scale)上，邏輯斯迴歸(logistic regression)分析結果顯示，與石虎出現有顯著相關之地景層級因子為綴塊密度(Patch density)、地景歧異度 (Shannon's diversity index)和地景形狀指數(Landscape shape index)；綴塊類型層級的因子為竹林綴塊聚集度 (Patch Cohesion Index)、草生地綴塊形狀指數(Landscape shape index)、農墾地綴塊形狀指數、人工林綴塊形狀指數(SHAPE_AM)，和離溪流距離；微棲地因子層級則是鼠類出現頻度、DBH<10cm木本植物密度和地表石頭覆蓋度有顯著相關。 本研究結果顯示對於石虎此種瀕臨絕種之保育類動物的保育和經營對策，必須包含不同尺度的考量。同時，非法毒殺、獵捕、道路阻隔和路殺，對於石虎的生存和族群存續有極大影響，破碎地景所形成的鑲嵌異質環境，雖然可提供有利於石虎的邊際效應，卻也形成所謂的”ecological trap”，使石虎面臨更多的人為干擾和影響。
並列摘要
In Taiwan, the leopard cat is listed as Endangered under the “Wildlife Conservation Act”. The ecology of leopard cats (Prionailurus bengalensis) was investigated in a fragmented rural area in northwestern Taiwan. Emphasis was placed on determining home range sizes, core areas sizes, movements, activity patterns, and its relationship with habitat use at multi-scales. This study is the first detailed, ecological study on the leopard cat in Taiwan. During 2005 to 2008 six leopard cats (2 females, 4 males) were trapped and radio-tracked, as well as total 253 camera trapping were set and collected data in the study areas. Average minimum convex polygon home range size (MCP100) and core area size (MCP50) were 5.0 and 0.7 km2, respectively. Males had much larger home ranges than females, but only males reduced their ranges significantly in the dry season. Home ranges overlapped extensively, but their core areas were not overlapped suggesting territoriality. Movement analyses showed that males were more mobile than females, but females use their home range in a more concentrated and effective manner. Leopard cats were nocturnal and exhibited crepuscular activity patterns in the wet season, but were arrhythmic in the dry season. Nocturnal activity was low during the dry season. Overall, broadleaf forest (29.3±19.6%) and plantation forest (24.7±4.0%) were the prominent habitat types within home ranges (FK95) of the four radio-tracked leopard cats, while man-made construction (4.2±1.6%) was the least used habitat type. Similarly, plantation forest (2.9±10.6%) and broadleaf forest (25.1±16.1%) were the prominent habitat types within core areas (FK50) of the four radio-tracked leopard cats. However, percentage of each habitat type within the home ranges and core areas varied by individuals. At the third-order scale, the use of different habitat types for resting significantly differed from the relative availability for each individual cat, whereas the use for activity significantly differed from the relative availability for both of males. At the second-order scale, leopard cats do not establish home ranges at random with the order: natural forests > unnatural forests > grass-agriculture lands, whereas they do not establish core areas at random with the order: unnatural forests > natural forests > grasslands-agriculture lands. At the first-order scale, the logistic regression models showed that leopard cats preferred areas with greater patch density (PD), higher Shannon's diversity index (SHDI), and higher landscape shape index (LSI) which indicated the mosaic landscape characteristics. Leopard cats also preferred the areas with less aggregated bamboo patches (COHESION_BB), lower landscape shape index of grassland (LSI_GL), as well as higher area weighted mean shape index of PF (SHAPE_AM_PF), higher landscape shape index of agriculture land (LSI_AL), and longer distance to streams (D_Stream). Occurrence index of murids, tree density (DBH<10cm) and percentage of rock cover were those microhabitat factors which influence the presence of leopard cats. The information provided here suggested that efficient managements and conservation strategies for leopard cats should be conducted at multi-scales in the mosaic rural habitat. In addition, this study found that illegal poisoning, trapping, road impediment, as well as road-kill incidents have the potential to affect the survival of the local leopard cat population. Hence, fragmented landscape and heterogeneous habitat which associated with forest edges could benefit leopard cats for supplementary resources. On the other hand, it could be an “ecological trap” for this endangered species, because the leopard cats may be more exposed to human persecution.