環境因子與動物生活史對陽明山兩棲爬行類動物路殺時空分布的影響

Abstract

道路快速地擴張對野生動物族群造成很多負面影響，包括棲地的流失、破碎化、道路阻隔效應與路殺等等。路殺指的是動物在道路上被行經車輛撞死之事件，會直接增加族群死亡率而影響到族群的存續，並可能間接影響物種多樣性。許多研究發現兩棲爬行類動物的路殺量以及路殺比例較其他脊椎動物類群高，且其路殺在空間與時間上分布並不均勻。導致兩棲爬行類動物路殺在空間分布不均的可能原因包括道路附近的棲地類型及道路本身的結構與交通。道路附近的棲地類型可能影響族群大小，若棲地類型是動物所偏好的，則族群較大，導致動物穿越道路的機率增加而有較高的路殺量。另外，道路的結構與交通狀況可能影響動物在道路上的危險程度，若道路的結構使駕駛者反應能力降低，或交通量高、車速快的情況下，則動物在道路上的危險程度較高而可能有較高的路殺量。導致路殺在時間分布不均的可能原因與動物活動量的時間模式有關。若動物在某些時段活動較高而導致活動範圍擴大，則此時段動物穿越道路的機率增加而有較高的路殺量。暸解路殺時空分布狀況以及相關之重要環境因子有助於建構預測模型以預測高路殺量的區域與時段，藉此設置動物廊道—圍籬系統以及實施交通管制等減緩措施。本研究使用陽明山國家公園持續調查五年的兩棲爬行類動物、盤古蟾蜍、斯文豪氏赤蛙以及黃口攀蜥的路殺資料，利用迴歸模型探討與其路殺空間分布相關之地景、道路特徵因子，以及與其路殺時間分布相關之氣候、生活史因子。結果發現距離生態保護區越近、明渠水溝長度越短的路段會有較高的兩棲類路殺量；附近森林面積越大、人為使用面積越小、路段寬度越窄以及路燈數越少的路段會有較高的爬行類路殺量。此外，「種」分類階層與「綱」分類階層的相關性分析結果比較發現重要的影響因子不同，與物種生態習性有關的因子在「綱」階層的結果中就不見了，顯示在制定減緩措施時，先確定目標對象的重要性。兩棲類的路殺時間分布與氣溫以及相對濕度呈正相關，爬行類的路殺時間分布則只與氣溫呈正相關。此外，「種」分類階層的相關性分析結果顯示除了氣候的影響，在繁殖季時也會有較高的路殺量。依據分析結果，針對兩棲爬行類整體，我建議於附近森林或農地面積大、鄰近生態保護區與水域之路段增設減緩措施。並於路寬窄之路段，移除路旁的雜草與灌叢，避免動物當作微棲地利用。有明渠水溝之路段，應定期清掃以確保其排水功能，避免積水導致兩棲類動物當作微棲地利用。無路燈之路段，建議架設較不會吸引昆蟲之路燈，或架設圍籬引導動物至有路燈路段。若要增設暫時性減緩措施，針對兩棲爬行類整體，則建議於其路殺高峰期氣溫及相對濕度均較高的夜晚實施。最後，在設置措施前後均應搜集路殺資料以評估減緩措施之成效。關鍵字：兩棲爬行動物、路殺、陽明山國家公園、地景分析、氣象

The rapid expansion of roads has negatively impacted wildlife populations in various ways, including loss of habitat, habitat fragmentation, and subdivisions of populations. Among them, vehicle-caused wildlife mortality has a direct impact on the population mortality rate, resulting in declines in population size, and might even present as an extinction threat to some species. Review studies discovered that compared to other vertebrate animals, populations of amphibians and reptiles were most vulnerable to road mortality. In addition, studies have shown that road mortality is aggregated in time and space. Spatial aggregations might be due to landscape composition and road traffic, while landscape composition near roads might influence animal abundance and movement. If a landscape composition is preferred by animals, this would result in larger population sizes, and in turn, increase the number of roadkill. In addition, the traffic and structure of roads would also influence the road mortality rate. If the road structure reduces the driver’s ability to respond to animals on the road or has heavy traffic, it would have higher numbers of roadkill. Temporal aggregations of road mortality might be correlated with animal activity, which is influenced by climate and animal phenology. If the climate condition causes animal activity to increase and expand their movement range during a particular period, this would also positively correlate with the roadkill numbers.To identify high-risk areas and design effective mitigation measures, it is essential to understand the spatial-temporal pattern of road mortality and identify important explanatory factors. Hence, I used roadkill data of amphibians, reptiles, Bufo bankorensis,Odorrana swinhoana, and Diploderma polygonata xanthostoma collected from Yangmingshan National Park between 2016-2020 to study the spatial-temporal pattern of road mortality and its relation with nearby landscape composition, road characteristics, climate, and animal phenology.The results showed that the spatial pattern of amphibian roadkill was associated with distance to the nearest ecological reserve and the length of roadside ditches, while that of reptile roadkill was associated with forest percentage area, artificial land use percentage area, road width and the number of street lights. In addition, important explanatory variables differed for treating amphibians and reptiles as a group from single species. The temporal pattern of amphibian roadkill was associated with temperature and relative humidity, while that of reptile roadkill was only associated with temperature. In addition to climatic variables, Odorrana swinhoana and Diploderma polygonata xanthostoma also had higher roadkill numbers during breeding seasons. Variables related to species traits might get lost when pooling species into groups, indicating the importance of identifying the target of interest before designing mitigation measures.For amphibians and reptiles, mitigation measures should be placed on road segments that have large areas of forest or farmland and are near ecological reserves. Weeds and bushes beside narrow road segments should be removed to prevent herpetofauna from being attracted to roadsides. The drainage function of roadside ditches should be maintained or else it might present ecological traps for anuran populations. In addition, streetlights not attracted to insects should be placed to increase the visibility of drivers at night, or fences should be placed to guide animals to road segments that have streetlights. For temporary mitigation measures, they should be location and time specific, and be implemented at high temperature and relative humidity nights during peaks of roadkill seasonality. Last but not least, it’s important to collect roadkill data both before and after the implementation of roadkill mitigation measures to assess their effectiveness in reducing roadkill numbers.Keywords: Amphibians, Reptiles, Roadkill, Yangmingshan National Park, Landscape analysis, Climate