- 學位論文
影響淺山道路兩棲爬蟲動物路殺熱點分佈之主因 以新北市土城區承天路與龍泉路為例
Factors influencing the distribution of roadkill hotspots for amphibians and reptiles in suburban hills: A case study of Tucheng District, New Taipei City
摘要
隨著人類社會與經濟的發展、都市化進程加速，道路的建設和擴張與日俱增。然而，道路的建設會切割原有的自然地景，對野生動物的棲息地產生負面影響，包括棲息地的破碎化、外來物種沿著道路入侵、人類活動干擾和車輛碰撞等，這些現象對野生動物族群的存續和生物多樣性造成威脅。而在這些影響中，野生動物與車輛的碰撞是對生態系統影響最直接且顯著的，尤其兩棲爬蟲動物因其生活史特性、體型較小等原因，使其更容易受到路殺事件的影響。 許多研究指出，道路環境、交通流量和氣候條件等因素對兩棲爬蟲動物的路殺風險有顯著影響。路殺事件不僅會對生物個體造成威脅，還可能對整個生態系統產生負面效應，甚至導致物種滅絕。為了降低道路對野生動物的影響，研究人員和相關團體已開始尋求有效的保護措施。 本研究旨在探討影響淺山道路兩棲爬蟲動物路殺熱點分佈的主要因素。本研究選擇新北市土城區的承天路與龍泉路作為研究區域，這些地區擁有豐富的兩棲爬蟲動物資源。延續前人與新北市農業局合作的調查計劃，本研究進行了為期一年的現地調查和數據收集。通過統計和分析所收集的數據，本研究試圖找出影響路殺熱點分佈的主要因素，並提出相應的保護建議，減輕路殺對野生動物的負面影響。 研究結果顯示，新北市土城區承天路與龍泉路的路殺物種以兩棲類大宗，爬蟲類次之，其路殺高峰期為春夏兩季。本研究以核密度估計來判斷路殺熱點位置分佈，並使用羅吉斯回歸分析來找出哪些影響因子對於路殺熱點的產生有較高影響力與解釋力。本研究發現路殺熱點的形成與道路兩側之明渠排水溝分佈、道路受植物遮蔽邊界情形、植生邊坡分佈、漿砌卵石擋土牆分佈以及道路坡度有一定的關聯性。同時本研究將路殺熱點之分佈與歷史資料做比對，發現熱點位置與過往分佈區域有所出入，影響原因為新北市農業局有針對過往之路殺熱點區域進行路殺改善工程，其工程結果有效的降低了路殺事件的發生。 本研究之調查資料及分析結果可以提供未來擁有類似道路條件之路段進行參考，作為路殺調查之監測基礎以及路殺改善工程之參酌。但因本調查之時間尺度僅有一個年度，若需要更詳盡的調查資料，則需要更深入且為期更長的調查。
並列摘要
With the development of human society and economy, along with the acceleration of urbanization, the construction and expansion of roads have been increasing rapidly. However, road construction often results in the fragmentation of natural landscapes, negatively impacting wildlife habitats. This includes habitat fragmentation, the invasion of alien species along roadsides, disturbance from human activities, and vehicle collisions. These phenomena pose threats to the survival of wildlife populations and the biodiversity of ecosystems. Among these impacts, wildlife-vehicle collisions have the most direct and significant influence on the ecosystem, particularly for amphibians and reptiles who are more susceptible to roadkill due to their life history characteristics and smaller body size. Numerous studies have indicated that road conditions, traffic flow, and climate factors significantly influence the risk of roadkill for amphibians and reptiles. Roadkill incidents not only pose threats to individual organisms but can also have negative effects on the entire ecosystem, potentially leading to species extinction. In order to mitigate the impact of roads on wildlife, researchers and relevant organizations have begun to seek effective conservation measures. The aim of this study is to investigate the main factors influencing the distribution of roadkill hotspots for amphibians and reptiles on shallow mountain roads. We selected the areas of Chengtian Road and Longquan Road in Tucheng District, New Taipei City as our research area, as these locations possess abundant amphibian and reptile resources. Building upon previous research that collaborated with the New Taipei City Agriculture Bureau, we conducted a one-year field survey and data collection. Through statistical analysis of the collected data, our objective is to identify the primary factors influencing the distribution of roadkill hotspots and propose corresponding conservation recommendations to mitigate the negative impact of roadkill on wildlife. The results of the study reveal that the majority of roadkill species on Chengtian Road and Longquan Road in Tucheng District, New Taipei City are amphibians, followed by reptiles, with the peak period of roadkill occurring in spring and summer seasons. We utilized kernel density estimation to determine the distribution of roadkill hotspots and employed logistic regression analysis to identify influential factors with higher explanatory power in hotspot formation. Our findings indicate a correlation between the formation of roadkill hotspots and the distribution of open drainage ditches, crown coverage, retaining walls, vegetation slopes, and road gradients. Furthermore, comparing the distribution of roadkill hotspots with historical data, we observed differences in hotspot locations, which can be attributed to roadkill improvement projects implemented by the New Taipei City Agriculture Bureau in previously identified hotspot areas. The results of these engineering interventions have effectively reduced the occurrence of roadkill incidents. The survey data and analysis results of this study can serve as a reference for future road sections with similar conditions, providing a basis for roadkill monitoring and guidance for roadkill improvement projects. However, it should be noted that the timeframe of this investigation was limited to a single year. For more comprehensive data, a more in-depth and longer-term survey would be necessary.