墾丁地區台灣梅花鹿（Cervus nippon taiouanus）農損分布與防治圍欄之評估

Spatial Distribution of Crop Damage by the Formosan Sika Deer (Cervus nippon taiouanus) and Evaluation of the Deer-Exclusion Fences in Kenting, Taiwan

鹿科動物造成的農業經濟損失是目前全世界關注的議題之一。台灣梅花鹿（Cervus nippon taiouanus）自1984年於墾丁國家公園復育至今，野外族群數量快速增長並擴散，並已造成部分墾丁地區發生農損。本研究探討墾丁地區梅花鹿造成農損的樣態及分布，並評估適用於台灣梅花鹿的危害防治圍欄設計。本研究共訪問22位農損戶，結果顯示梅花鹿農損的歷史與分布符合梅花鹿族群的成長和擴散歷程。受損的作物以火龍果（Hylocereus undatus）及黑豆（Glysine max）為主。農損有明顯的區域性，絕大多數發生在港口村和永靖村，與梅花鹿主要分布區和作物種植密集區的重疊有關。在評估適用於台灣梅花鹿的危害防治圍欄方面，本研究測試了防治農損的10 cm×10 cm網目菱形網圍欄及電圍欄。結果顯示，在圈養環境下，菱形網圍欄高度150 cm時可完全阻隔梅花鹿，而電圍欄在135 cm高、16條網線時可有效阻卻鹿隻的穿越|然而，田間的電圍欄測試則顯示在90 cm高、3條網線密度時，就會產生阻隔效果，惟恐觀察紀錄不足，暫時建議以120 cm高、13網線的樣式作為田間使用的基準，而當地自然食物資源與造成鹿隻緊迫因子的多寡，也應納入考慮。本研究建議使用菱形網長期防護多年生、全年性、高經濟的作物|而使用電圍欄作為短期性、季節性和大面積的防護措施。最後，也建議盡量區域性設置大面積的圍欄，以降低成本、增加效益。

Crop damage caused by deer has been a major concern worldwide. The Formosan sika deer (Cervus nippon taiouanus) restoration program was initiated in Kenting National Park, Taiwan, in 1984. Following the rapid growth in number and expansion in distribution of the deer population, crop damages caused by the deer also occurred in parts of the Kenting area. The present study aimed to understand patterns and spatial distribution of crop damage, as well as to evaluate the effectiveness of deer-exclusion fencing for damage control. We interviewed 22 farmers experiencing deer-inflicted crop damage. The results showed that the history and spatial distribution of the crop damage was consistent with the growth and expansion of the deer population. Pitayas (Hylocereus undatus) and black beans (Glysine max) were the most damaged crops. The majority of the crop damage occurred in a restricted area between Gang-Kou and Yong-Jing villages, where intensive agricultural practices overlapped with an area of concentrated deer population. Regarding the effectiveness of various deer-exclusion fence designs, results showed that, under captive conditions, a 150 cm-high conventional chain-link fence or 135cm-high, 16-strand electric fence both could effectively stop the deer. Our results also demonstrated that a 90 cm-high, three-strand electric fence could exclude the deer efficiently in the field. However, due to the short period of observation, we tentatively recommend 120 cm-high, 13-strand electric fences for use in the field. When determining the type and height of the fence, the availability of food resources and the level of stress inflicted on the deer should also be taken into account. We also recommend that chain-link fences be used for permanent protection of perennial or year-round growth and high value products| while electric fences be utilized to protect short or seasonal growth products, or in large-area fields. Lastly, we suggest that regional fencing of a large area should be prioritized for cost reduction and optimal beneficial results.