台灣綠蠵龜於覓食棲地之族群生態學及保育

Abstract

綠蠵龜（Chelonia mydas）被IUCN列為瀕危物種，主要分布於亞熱帶與熱帶海域，也是台灣最常見的海龜種類。過往研究多集中於產卵族群，對近岸覓食族群之瞭解相對有限。然而，海龜生命週期中絕大多數的時間都在覓食棲息地發育、生長並累積能量準備繁殖，若能更瞭解覓食族群和棲息地概況，有助於制定保育及管理策略。本研究結合公眾參與及系統性調查，探討臺灣地區覓食綠蠵龜之族群生態。首先，透過與公民科學家合作建立「海龜點點名」計畫，收集水下海龜目擊資料，並藉由個體辨識，建構台灣沿岸海龜密度地圖、確認覓食熱區與瞭解定居模式。小琉球、恆春和綠島為重要覓食熱點。定居模式估算出約43.4%的個體會在同一地區停留超過一年，而且成龜的定居時間和目擊次比青年龜和亞成龜來得更久。此外，約有10%的目擊紀錄，與受傷個體有關，包括漁線纏繞、背甲受傷或其他外觀異常，顯示出人為活動造成的威脅。其次，我們在覓食熱區小琉球進行兩年的系統性浮潛調查，驗證調查設計對資料有效性的影響，並進一步評估小琉球覓食族群的棲地忠誠度。潮汐狀態和調查設計會影響海龜個體偵測機率：漲潮和滿潮時海龜數量最多，而漲潮和退潮時則有較多不同個體；重覆調查較傳統單次調查記錄到八倍多的目擊數量和快兩倍多的個體及重複目擊率。我們用標準化棲地忠誠指標（SSFI）將海龜個體的定居時間和目擊頻率進行量化，並用階層式群集分析辨識出兩種忠誠度組別，低與高忠誠度約各佔一半，並其他資料集和不同地區綠蠵龜進行棲地忠誠度的比較，發現小琉球海龜與菲律賓Oslob相似，且比公民科學資料提供更好資料解析度。最後，透過整合公民科學與系統性調查的長期資料，發現在小琉球約13.3%野外族群有外傷或疤痕，但80%的個體都觀察到已復原。本研究觀察到部分受傷個體從一開始受傷到逐漸恢復的完整過程。受傷成因包括螺旋槳擊傷、魚線纏繞和不同原因，復原時間平均皆需一年半。就野外觀察，嚴重傷口平均需約600天癒合，而輕微傷口則約491天。部分嚴重個案，如因魚線纏繞進而潰爛斷肢，和螺旋槳擊傷造成背甲變形，康復後仍持續在小琉球被目擊達106至2,264天，顯示其強大的野外康復能力和適應力。但海龜重複受傷的現象和受傷個體大多出沒於港口附近，都顯示設立限速區和娛樂釣魚規範的迫切和重要性。本研究透過公民科學加深對台灣地區海龜覓食族群生態的瞭解，並針對調查標準化、棲地忠誠度評估與人為威脅提出建議，對臺灣及東印度洋和東南亞區域的海龜族群之保育與管理具有重要參考價值。The green turtle (Chelonia mydas), listed as Endangered by the IUCN, is primarily distributed in subtropical and tropical waters and is the most commonly sighted marine turtle species in Taiwan. Research has historically focused on the nesting population and stranded individuals, leaving the ecology of foraging populations understudied. Marine turtles spend the majority of their lifetime in foraging grounds, where they seek shelter, grow, and gain energy for breeding and migration. Understanding their population ecology and residency patterns in foraging grounds is crucial for developing effective conservation and management strategies.This Ph.D. study integrates a citizen science approach and standardized field surveys to understand the population ecology of green turtles in Taiwan. First, a citizen science initiative, TurtleSpot Taiwan, was co-found with citizen scientists to collect sightings of marine turtles and identify individuals through photo identification according to their unique facial scute patterns. Based on five years of nationwide data, several main foraging grounds (Liuqiu Island, Kenting, and Green Island) with a substantial number of foraging aggregations have been identified. Approximately 43.3% of turtles remained in the same area for over a year, with adult-sized turtles having more repetitive sightings and longer residency durations than juveniles and subadults. Approximately 10% of sightings involved turtles with external injuries or anthropogenic threats, including entanglement in fishing lines or hooks, carapace injuries, or missing flippers.Second, a two-year, multiple-event snorkel-based survey was conducted at a key foraging ground, Liuqiu Island, to validate the survey design and assess the site fidelity of the local aggregation of green turtles. Turtle detection probability was influenced by tidal phase and survey design. More turtles were observed during high and flood tides, and a higher number of unique individuals were observed during flood and ebb tides. Multiple-event surveys recorded over eight times more sightings and nearly twice as many individual turtles and resighting rate as single-event surveys. We used the Standardized Site Fidelity Index (SSFI) to measure site fidelity for each individual, and clustering analysis revealed two distinct fidelity groups (Low Fidelity: 49%; High Fidelity: 51%) at sites. SSFI also allows for comparisons with other regions using open-access photo-ID datasets and found that the site fidelity of green turtles on Liuqiu Island was similar to those at Oslob, the Philippines, and offers a better resolution than citizen science data on Liuqiu Island.Finally, long-term sightings from citizen science and survey data revealed rare but valuable records of injury recovery in the wild, offering new insights into the resilience and vulnerability of foraging turtles. The prevalence of external injuries among marine turtles at Liuqiu Island is approximately 13.3%, with most exhibiting signs of healing. Regardless of the injury causes (i.e., propeller strike, fishing line entanglement, and unidentifiable cause), the average healing times were about 1.5 years. Severe injuries typically take about 600 days to heal, while minor injuries generally take around 491 days to recover fully. Certain animals recoverfrom severe injuries, such as flipper amputation caused by fishing line entanglement, and propeller strikes leading to carapace deformity permanently; however, the long-term observations after their recoveries for at least 106 to 2,264 days demonstrate remarkable resilience. Individuals with recurring injuries were observed, with tracked individuals showing frequent use of nearshore areas near ports, emphasizing the need for targeted conservation measures specific to port areas and by extension regions with high boat traffic. These include go-slow zones, recreational fishing regulations, and increases in public outreach.This study demonstrates the value of combining citizen science with systematic monitoring to assess turtle distribution, habitat use, and responses to human impacts. It contributes practical solutions for the foraging population study and marine turtle conservation in Taiwan and can be applied to the broader East Indian and Southeast Asian Regional Management Unit.