透過您的圖書館登入
IP:13.58.121.131
  • 期刊

梅花鹿再引入後的空間現象:墾丁礁林與次生林的覓食棲地選擇

Impacts of Reintroduced Sika Deer in Kenting: Spatial Pattern of Foraging Habitat Choice in Reef and Secondary Forests

摘要


臺灣梅花鹿野外族群曾於1969年滅絕於野外,幸賴台北市立動物園所保存之種源,於1990年代逐步野化,讓梅花鹿引入墾丁國家公園社頂地區,放養於半開放的森林地帶。2000年以後復育計畫開始讓繁殖族群擴散至籠仔埔草原,結果發現局部鹿群會侵入墾丁高位珊瑚礁保留區,並啃食當地植被造成損害。藉由重新檢驗1993-1994年鹿群覓食在半開放珊瑚礁森林的逐月紀錄,本研究目的為:1.比較乾濕季梅花鹿之覓食與休息所在地的植被組成特色;2.探討覓食與休息之空間分布是否有群聚現象;3.根據當年送測的鹿食植物之營養成分,能否進一步理解梅花鹿於乾溼季間棲地選擇的差異。當地23種鹿食植物之營養成分分析中,以原生植物月橘及相思樹所含熱能與含氮量俱高;熱能次高之食草為次生林中的馬纓丹和芭樂,但只限在乾季時才啃食這些外來種的幼葉;至於「珊瑚礁林」苗木的營養價值並無突出。倒是外來入侵種銀合歡含氮量極高,全年為鹿隻於林下廣泛搜尋。在鹿群行為的空間近鄰分析中,發現休息地點為典型的廣泛分布而呈離散;濕季時鹿隻跑動明顯,離散程度次之。相反地,乾季覓食困難而叢集現象明顯,鹿群會特別集中於礁林原生苗木處形成啃食壓力。運用空間自相關分析技術,空間延遲模型(SLM)由於引入相鄰小區覓食的表現,模型解釋變異程度顯著提高。相較之下,傳統廻歸(OLS)不採計空間群集現象,將誤判鹿群覓食之生態衝擊。

並列摘要


The Formosan sika deer, which were extirpated in the wilderness of Taiwan in 1969, have been reintroduced into Sheting protected area of Kenting National Park since the 1990s. Recovered deer populations gradually expanded into primary reef forests after the 2000s, and even crossed widely over to local grasslands and the high-reef primary forest. This study aids at: 1. Comparing vegetative composition between resting and foraging sites; 2. Examining spatial characters of clustering habitats; 3. Explanation by nutrients and spatial autocorrelation. An analysis of nutrient contents of 23 local deer food plants was conducted. The heat energy and nitrogen content of Murraya paniculata and Acacia confusa were highest; the following highest energy contents include Lantana camara and Psidium guajava of the secondary forests. The nutritional value of the invasive alien species, Leucaena leucocephala, had an extremely high level of nitrogen, resulting in an intensive use of deer in all kinds of forest. In the nearest neighbor analysis, the resting sites were evenly distributed and scattered; while in the dry season reached a minimum, indicating a clumped feeding on the reef forests. In the dry season, deer fed on reef forests due to scarce resources. In this study, the spatial autocorrelation analysis was adopted to model the average foraging time on dominant plants. Due to the introduction of the spatial lag model (SLM), the correlation increased, showing a significant improvement when considering the clumped vegetation in the area. In contrast, under the classic regression indicated that the ecological impacts of seasonal feeding might be misjudged if using the method of ordinary least squares (OLS), assuming independent and homogeneous spatial units.

延伸閱讀