摘要
把居民納入保育體系內，是現代保育政策的重要準則之一，但野生動物常因損害農作物而與居民產生衝突，導致居民不願支持保育行動，因此，如何緩解野生動物與居民的衝突，同時適切的管理農業活動，是保育主管機關的重要課題。太魯閣國家公園境內具有農業活動，國家公園管理處曾收購部份耕地使其重新復育回歸自然狀態，在其他仍使用中的農地，農作物常受野生動物侵擾以致影響農民收益。為了保育與永續經營之考量，應了解野生動物危害農作物現況，並尋找減輕人與野生動物間衝突之方法，以增加居民配合與協助國家公園的相關保育政策之意願。並且應了解回收復育耕地及有機農法對生物多樣性之益處，以推動既有農地之轉型，減少農藥與化學肥料對環境之傷害。
本研究分為三個部份，第一個部份以訪查方式，調查西寶、洛韶、薛家場、華祿溪等地農地遭受野生動物損害概況。第二個部份以穿越線調查法、定點調查法、紅外線自動相機、薛爾曼式陷阱四種方法，調查西寶至華祿溪區域之中大型哺乳動物、小型哺乳動物、鳥類、蝶類之動物相，比較慣行農法、有機農法、回收復育耕地之生物多樣性。第三個部份提出各種設備方案，用以防治野生動物入侵農地，並同時進行智慧生態監控。
關於野生動物造成之農損，共收集17份有效問卷，發現多數農民受野生動物損害金額佔其產值的10-30%，高麗菜、番茄、水蜜桃是種植量較多且受損也嚴重的作物，野生動物中以台灣獼猴與野豬造成的損害最嚴重，防治方法中以放鞭炮、養狗最廣泛被採用，但僅有一定程度嚇阻作用，部分受訪者表示近幾年的受損情形較往年更嚴重。
以自動相機調查中大型哺乳動物發現，多數物種的平均出現指數都以回收復育耕地最高，慣行農法耕地又常高於有機農法耕地。小型哺乳類之調查僅在有機農法耕地成功捕獲鼠類(捕獲率2.19 隻/100籠夜)，另外兩種農地均未捕獲。鳥類調查共記錄28科52種鳥類，鳥類的物種數與夏農生物多樣性指數都以有機農法耕地較高，各樣區以雜食者比例最高，但在慣行農法耕地，蟲食者比例低於植食者，另外兩種耕地則相反之。蝶類調查結果顯示，共發現5科90種蝶類，以緣點白粉蝶最優勢，物種數、夏農指數、物種均勻度均以回收復育耕地最高，有機農法與慣行農法耕地之間則互有高低。此外，在蓮花池地區的長期監測(2016-2017)發現，水鹿、野豬、鼬獾等物種的相對豐度有下降趨勢。
我們提出三大類監測裝置供參考，包括：定點偵測裝置、動態偵測裝置、問題動物監測裝置，以電力與網路需求、產品完備度、價格等面向進行分析。另外與兩位農民合作，測試具自動回報功能之紅外線自動相機，用以防治野生動物侵入農田。測試結果顯示該設備於夜間才能發揮效果，故必須尋找更合適的機型，農民有意願繼續使用，但希望由政府補助。
Abstract
Involvement of local residents is emphasized in modern conservation practices. However, conflicts between residents and wildlife often occur due to crop damages. As a result, residents are unwilling to support conservation activities. Therefore, alleviation of human-wildlife conflicts and management of agricultural activities in protected areas are important issues for conservation authorities. Agricultural activity exists in Taroko National Park for several decades. Some of the farms were purchased and restored to nature by the administrators of the national park in recent years, and others are still in use and suffering wildlife damage. For the purpose of conservation and sustainable management, we should understand the status of wildlife damage on crops, and look for ways to mitigate the conflict between humans and wildlife. In addition, we should understand the benefits of farmland restoration and organic farming to biodiversity, so as to promote the transformation of conventional agricultural activity and reduce the use of pesticides and chemical fertilizers.
This study consists of three parts: 1. We interviewed the farmers in Xibao, Luoshao, and Hualuxi area to understand the status of wildlife damage on crops. 2. We applied line transect survey, point count method, camera trap, and Sherman's trap to investigate the faunas of mammal, bird, and butterfly on farms in the Xibao to Huaruxi area, and compare the biodiversity among conventional farms, organic farms, and restored farms. 3. We proposed some approaches which may be contributive to both wildlife damage prevention and ecological monitoring.
We collected 17 valid questionnaires. Approximately 10-30% of crop yield was lost due to wildlife damage. Cabbage, tomatoes, and peaches were the major crops in this area and suffered serious damages. Macaque and wildboar were the two species which caused the most serious damages. For the damage prevention methods, firecrackers and dogs were widely used but not really effective. Some of the respondents indicated that wildlife damage became more serious in recent years.
The results from camera trap survey suggested that medium-large sized mammals occurred most frequently in the restored farms, and usually occurred more frequently in the conventional farms than organic farms. The Sherman’s trap only captured mouse in the organic farms (2.19 mouse/100 trap nights). For the bird fauna survey, we recorded 52 species in 28 families. The species richness and Shannon-Weiner index were both higher in organic farms than conventional farms and restored farms. In all types of farms, the proportions of omnivore were highest. However, herbivores had higher proportions in conventional farms but lower proportions in organic and restored farms than insectivores. For the butterfly fauna survey, we recorded 90 species in 5 families. Pieris canidia is the dominant species in number. Species richness, Shannon-Weiner index, and species evenness were highest in the restored farms, but equally matched between the conventional and organic farms. In addition, our long-term monitoring suggested that the relative abundance of sambar, wildboar, and ferret badger were decreasing in the Lianhua Pond area.
We proposed three types of devices for damage prevention and mammal monitoring, including: fixed-point detection device, agricultural drone, and problem animal detection device. We collected information regarding power and network demand, product completeness, and price for these devices. In addition, we cooperated with two farmers to test a fixed-point detection device using an infrared automatic cameras with an instant report function. The results showed that this device was able to detect and alarm wildlife invasion at night, but not effective in the daytime. Therefore, it is necessary to search for a more suitable model. The farmers were willing to use this device but required a subsidization from the government.